Pneumonia Due to Mycoplasma in Gnotobiotic Mice

JOURNAL OF BACTERIOLOGY, Oct., 1966Copyright i) 1966 American Society for Microbiology

Vol. 92, No. 4Printed in U.S.A.

Pneumonia Due to Mycoplasma in Gnotobiotic MiceI. Pathogenicity of Mycoplasma pneumoniae, Mycoplasma salivarium,

and Mycoplasma pulmonis for the Lungs of Conventionaland Gnotobiotic Mice

IRVING I. LUTSKY AND AVRUM B. ORGANICKAllen-Bradley Medical Science Laboratory and Department of Medicine, Marquette University School

of Medicine, Milwaukee, Wisconsin

Received for publication 18 June 1966

ABSTRACTSLUTSKY, IRVING I. (Marquette University School of Medicine, Milwaukee, Wis.),

AND AVRUM B. ORGANICK. Pneumonia due to mycoplasma in gnotobiotic mice. I.

Pathogenicity of Mycoplasma pneumoniae, Mycoplasma salivarium, and Mycoplasmapulmonis for the lungs of conventional and gnotobiotic mice. J. Bacteriol. 92:1154-1163. 1966.-Two species of mycoplasma of human origin, Mycoplasma pneumoniaeand M. salivarium, were tested for their ability to produce respiratory disease in theHa/ICR mouse when inoculated by the intranasal route. The mouse pathogen M.pulmonis was studied as a positive control. Conventional and gnotobiotic Ha/ICRmice were employed, the latter to provide a system free from indigenous mycoplasmaand bacteria. Pneumonia from which mycoplasma were isolated was produced in allgroups of the conventional Ha/ICR mice, including those inoculated with sterilebroth. Only M. pulmonis produced disease when inoculated intranasally into thegnotobiotic mice, and the gross and microscopic lesions resembled those describedin conventional mice. The gnotobiotic mouse provided a tool to study the patho-genicity of different mycoplasma species, and indicated marked differences in hostspecificity that could not be clearly seen when conventional mice were used.

The demonstration that the agent of atypicalpneumonia in man is a mycoplasma, Mycoplasinapneumoniae (1, 2), suggested that other myco-plasma species commonly found in the upperrespiratory tract, e.g., M. salivarium, might becapable of causing respiratory disease. Studies ofexperimental infection with mycoplasma ofhuman origin in laboratory animals (3, 4, 7)have the disadvantage of altering the factor ofhost-parasite specificity, but they afford theopportunity to examine details of the infectiousprocess not readily available in human studies,such as the evolution of lesions and the localiza-tion of microorganisms. Intranasal inoculationsin conventional rodents are often complicatedby pneumonias due to indigenous murine path-ogens (9). In the animal studies to be reported,gnotobiotic mice provided a system free fromindigenous murine mycoplasmata and bacteria.The mouse pathogen, M. pulmonis, was used asa positive control with which the results of intra-nasal inoculations of mycoplasma of human ori-gin, M. pneumoniae and M. salivarium, might becompared.

MATERIALS AND METHODS

Conventional and gnotobiotic Ha/ICR mice, 4 to6 weeks of age and of both sexes, were used through-out the study. The gnotobiotic Ha/ICR mice (sup-plied by A. R. Schmidt Co., Madison, Wis.) werederived originally from Ha/ICR mice by use of hys-terectomy, and foster-nursed on gnotobiotic ND-1mice. Preliminary screening in our laboratory re-vealed that the lungs, liver, spleen, and kidneys ofthese gnotobiotic mice were free from mycoplasmaand bacteria.

Mice were housed in clear, autoclavable plasticcages within flexible film isolators. With conven-tional animals, the isolators were used to preventaerosol cross-contamination in concurrent experi-ments in which different mycoplasma species wereused. Three species of mycoplasma were used in theexperiments: M. pneumoniae (M-52 strain), M.salivarium, and a murine mycoplasma species whichproduces infectious catarrh in mice, M. pulmonis (8).The M-52 strain of M. pnteumoniae (ATCC 15293;

obtained from York E. Crawford, Naval MedicalResearch Unit No. 4, Great Lakes, Ill.) was passedfour times in 20% horse serum-enriched PPLO broth(2) containing 25% yeast extract, 1,000 jug/ml ofpenicillin, 1:2,000 thallium acetate, and 5 jug/ml of

1154

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

VOL. 92, 1966 PNEUMONIA DUE TO MYCOPLASMA IN GNOTOBIOTIC MICE. I

amphotericin B, hereafter referred to as mycoplasmabroth. A 7-day growth in mycoplasma broth wasdivided into 2-ml portions, distributed into sterileglass ampoules, sealed, quick-frozen in an alcohol-Dry Ice mixture, and stored at -65 C. After 10 days,colony counts performed in triplicate after thawingshowed 1.2 X 107 colony-forming units (CFU) permilliliter.The M. salivarium strain was isolated from the

pharynx of a normal human subject, and samples ofa 24-hr growth in mycoplasma broth contained 2.8X 107 CFU/ml.The M. pulmonis strain (kindly provided by John

B. Nelson, Rockefeller University, New York, N.Y.)was derived from mice inoculated with a 30th mouselung passage. Lungs of these mice were ground in atissue grinder and passed five times in mycoplasmabroth. Samples ofa 5-day growth in mycoplasma brothcontained 1.8 X 106 CFU after a single freeze-thawcycle. The inoculum in control animals consisted ofsterile mycoplasma broth.

Intranasal inoculation of the mice with 0.05 ml ofthe inoculum was made with a Microtiter (CookeEngineering, Alexandria, Va.) delivery pipette (de-livering 0.025 ml per drop) to mice in the isolatorsafter ether anesthesia.

In the studies involving the gnotobiotic mice,samplings of feces, bedding, and water for bacteria,mycoplasma, and fungi were made according tostandard isolator techniques (12). The clinical ap-pearance of the animals was observed several timesdaily through the clear cages, and all animals founddead were removed as soon as possible. Mice to besacrificed were killed by traction of the neck. Sterilepaper bags were used to transfer gnotobiotic micetrom the isolator to a sterile field within a hood, wherethe mice were necropsied. Impressions of cut surfacesof lungs were cultured on mycoplasma agar and onblood-agar, and portions of lungs were cultured inthioglycolate broth. Portions of lung were fixed in10% formaldehyde, Bouin's or Helle's fixative, andhistological sections were stained with hematoxylinand eosin.

RESULTSIntranasal inoculation of mycoplasma in con-

ventional mice. To serve as a basis for com-parison with the studies in gnotobiotic mice,151 conventional mice were used in an initialstudy. The results are summarized in Table 1.Pneumonia was produced in all groups of con-ventional mice receiving intranasal inoculations,even those receiving sterile broth. The percentageof animals that developed pneumonia after inocu-lation with M. pulmonis was significantly greaterthan the controls (x2 = 16.64). M. pulmoniswas also more pathogenic than the two myco-plasma species of human origin, M. pneumoniaeand M. salivarium. Although the pathogenicityof these latter species was indistinguishable interse, they were more pathogenic than sterile broth(x2 = 9.48). Mycoplasma indigenous to the mice

TABLE 1. Results of intranasal inioculationt withsterile broth and with three mycoplasma species

in 151 coonventional mice

Dosei No of GrossInoculum colony-forming mice pneu- cent

monia

Sterile broth(control) . 28a 5/27 18.5

M.pneumoniae 6.0 X 105 64b 32/60 53.3M. salivarium... 1.4 X 10i 30 17/30 56.6M. pulmonis 9.0 X 104 29C 22/28 78.5

a One mouse wascropsied.bFour mice were

cropsied.c One mouse was

cropsied.

cannibalized and not ne-



were indeed present, as demonstrated by thepresence of pneumonia and isolation of myco-plasma from animals receiving sterile broth(Table 2). It is possible that these indigenousmycoplasma contributed to some extent to thedisease produced by inoculation with the othermycoplasma species.

Bacteria indigenous to the conventional micealso might have played a role (Table 3). Largenumbers of bacteria (heavy growth of bacteriain impressions on blood-agar) were present in24 of 71 animals with pneumonia inoculatedintranasally with material containing onlymycoplasma. Most of these isolations (20 of 24)were from pneumonic lungs which also containedmycoplasma. Bacteria were isolated only oncefrom 47 normal lungs, but mycoplasma werealso isolated from this specimen.

General observations on intranasal inoculationof mycoplasma in gnotobiotic mice. Althoughstudies in which conventional mice were used(Table 1) indicated that pneumonia was oftenassociated with indigenous mycoplasma andbacteria, the studies with gnotobiotic miceshowed clear-cut results (Table 4). Intranasalinoculation of sterile broth in 66 gnotobiotic micedid not produce a single instance of pneumonia.Furthermore, at no time were mycoplasma orbacteria ever isolated from the lungs of thesecontrol animals. Gross pneumonia was evidentonly in mice inoculated with M. pulmonis, butnot in mice inoculated with the two mycoplasmataof human respiratory-tract origin, M. pneumoniaeand M. salivarium.

Results of intranasal inoculcation with M. pul-monis in gnotobiotic mice. The association ofgross lesions of pneumonia with the isolation ofmycoplasma is seen in Table 5. Mycoplasma were

1155

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

LUTSKY AND ORGANICK

TABLE 2. Results of intranasal inoculation withsterile broth in 28 conventional mice

Anatomical findings

Microbiological findingsGross Grossly Totala

pneumonia normal lungs

Mycoplasma .......... 4 2 6Mycoplasma and

bacteria............. 0 0 0Bacteria .............. 0 0 0No growth....... 1 20 21

a One mouse was cannibalized, and thereforenot tested.

TABLE 3. Results of intranasal inoculation withbacteriologically sterile broth cultures ofMyco-plasma pneumoniae, M. salivarium, and M.pulmonis in 123 conventional mice

Anatomical findings

Microbiological findingsGross Grossly Totalsa

pneumonia normal lungs

Mycoplasma.43 11 54Mycoplasma andbacteria... 20 1 21

Bacteria .............. 4 0 4No growth............ 4 35 39

Five mice were cannibalized, and thereforenot tested.

TABLE 4. Results of intranasal inoculation withsterile broth and with three mycoplasma species

in 214 gnotobiotic mice

Inoculum Dose in colony- No. of Gross findingsforming units mice of pneumonia

Sterile broth(control)..66 0

M. pneumoniae 6.0 X 105 34 0M. salivarium .1.4 X 106 35 0M. pulmonis.... 9.0 X 104 79a 50/73

a Six mice were cannibalized and not necropsied.

isolated in 50 of 51 lungs when pneumonia wasevident grossly. Mycoplasma were also isolatedin 15 of 22 instances where the lungs were grosslynormal. Microscopic lesions, however, werepresent in 6 of these 15 lung specimens. Thesefindings, and the absence of mycoplasma fromthe lungs of control gnotobiotic animals inocu-lated with sterile broth, confirm the sensitivity ofthe culture techniques employed.The morphological features of mycoplasma

colonies isolated from the lungs of mice in this

TABLE 5. Results of intranasal inoculation withMycoplasma pulmonis in 79 gnotobiotic mice

Anatomical findings

Microbiological findingsGross Grossly nor- Totalsa

pneumonia mal lungs

Mycoplasma. SOb l5C 65Mycoplasma andbacteria. 0 0 0

Bacteria.0 0 0No growth.1 7d 8

a Six mice were cannibalized, and therefore nottested.

b Includes one mouse sacrificed at 2 days.¢ Includes two mice sacrificed 2 hr after inocu-

lation, and four mice sacrificed 2 days afterinoculation.

d Includes two mice sacrificed immediately afterinoculation, and two mice sacrificed 2 hr afterinoculation.

portion of the experiment were similar to thoseof the inoculum. Colonies were characterizedby a highly vacuolated peripheral zone and byirregular and indistinct central portions (11).The periphery of the colonies was often irregularon the primary isolations from cut surfaces oflungs on mycoplasma agar, but was usuallysmoother on subculture (Fig. 1).The clinical course of pneumonia in mice

inoculated with M. pulmonis was observed andcompared with the appearance and behavior ofmice inoculated with sterile broth. Most animalsrecovered quickly from the effects of anesthesiaand of the intranasal inoculation. On the dayafter the intranasal inoculation and during thesubsequent 48 to 72 hr, all mice appeared normal.By the 4th day, however, a number of micethat had received M. pulmonis appeared ill.Labored respirations, ruffled fur, and slownessof movement were evident. Cyanosis was de-tected in the skin of the tail and feet and in adarker than normal red reflex from the eyes.Disease progressed rapidly in some animals,and the first deaths occurred on the 5th day. Be-ween the 5th and 12th days after inoculation withM. pulmonis, most of the mice appeared ill. Ofthe 18 deaths, 16 occurred during this period(Fig. 2). Cannibalization of dead mice wasfrequent (in 6 of the 18). When necropsies wereperformed, total or nearly total involvement ofall lobes with firm, uniform red to reddish-grayconsolidation was seen, and mycoplasma wereisolated from such lungs in every instance. Afterthe 12th day, surviving mice appeared to im-prove somewhat, although they never appearedas healthy as control animals. Their food and

1156 J. BAmrRioL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.


4

e..

V.

t~~~~~~~~~~~~~~~~~~~'S:FIG. 1. Colonies of Mycoplasma pulmonis in subculture on mycoplasma agar (X 100) The larger colonies have

a granular peripheral zone, and the central zone is not clearly demarcated.

water consumption remained diminished, andtheir growth rate was retarded. After the initialpeak in mortality, deaths occurred in two animalsduring the 4th week.The gross and microscopic lesions in gnoto-

biotic mice were similar to those described inconventional mice (5, 8, 10). Early pulmonarylesions (4 to 7 days) were grossly red to reddish-gray, and involved segments or entire lobesusually on the dorsal aspect of the lungs. Micro-scopically, the early lesions showed capillaryengorgement, edema, hemorrhage into thealveolar spaces, and a brisk polymorphonuclearleukocytic response in the alveoli and bronchi,the lumen of bronchi often filled with exudateof these leukocytes (Fig. 3). Small numbers ofround cells, principally small lymphocytes,began to appear in the interstitial spaces sur-rounding medium-sized blood vessels betweenthe 4th and 7th days (Fig. 4). At 2, 3, and 4weeks, gross lesions were gray and translucentinstead of red. Infiltration by lymphocytes be-came more conspicuous first in the perivascularand then in the peribronchial spaces. The peri-bronchial and perivascular infiltrate with largeand small lymphocytes and with plasma cellsbecame the most conspicuous microscopic fea-ture at 3 and 4 weeks (Fig. 5 and 6). In the alveoli,large pink-staining foamy macrophages were thepredominant cell type, although polymorpho-

nuclear leukocytes persisted in some pneumonicareas. The exudate in the lumen of bronchi re-mained predominantly polymorphonuclear intype.

In spite of the vigorous cellular response duringthe acute stages, necrosis of bronchial epithelialcells, abscess formation, and empyema werenever seen among gnotobiotes inoculated with M.pulmonis. Purulent exudate was expressed fromthe nasal passages in a small number of instances.The upper respiratory tract and middle earswere not routinely examined. No gross or micro-scopic lesions were noted in the spleen, liver,kidneys, or small intestine. Mycoplasma wereisolated in small numbers from 4 of 5 normal-appearing spleens 4 and 7 days after inoculation.

Results of intranasal inoculations with M.pneumoniae and M. salivarium in gnotobioticmice. No deaths, clinical illness, or pneumonia ongross or microscopic examination of the lungswas seen in any of 34 gnotobiotic mice inoculatedwith M. pneumoniae. Mycoplasma colonies wereisolated from the lungs of 3 of 17 mice sacrificed2 weeks after intranasal inoculation, but fromnone of 17 mice sacrificed 4 weeks after intra-nasal inoculation (Table 6). The 35 gnotobioticmice inoculated intranasally with M. salivariumremained healthy throughout the 30-day ob-servation period. The lungs were grossly andmicroscopically normal, and no mycoplasma were

1157

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

LUTSKY AND ORGANICK

TIME COURSE OF DISEASE IN 365CONVENTIONAL AND GNOTOBIOTIC MICE

DAY OF DEATH IN RELATION TO INOCULATION

CONVENTIONAL MICE

M. pneumaonioe (64 mice)

Deaths g5-

Deaths

GNOTOBIOTIC MICE

M pneumonioe (34 mice)

uir (30 mice)

DeothsoI * [

M pulmonis (29 mice)

5.

Control (28 mice)5

0. VA0 5 10 15 20 25

DAYS

Deoths

Deaths0l

vU

M solivarium (35 mice)5.

M _pulmons (79 mice)

Mpulmanis:1Control (66 mice)

0 5 l0 15DAYS

FIG. 2. Occurrence of deaths in 365 conventional and gnotobiotic mice in relation to the time of intranasal in-oculation.

isolated from the lungs at sacrifice 2 and 4 weeksafter intranasal inoculation.

General observations on comparison of diseasein gnotobiotic and conventional mice. The timecourse of the disease is seen in Fig. 2, whichshows the day of death in relation to the date ofinoculation in 365 inoculated mice. A peak inmortality between the 5th and 12th day in all ofthe conventional animals corresponded to thepeak in mortality in gnotobiotic mice inoculatedwith M. pulimonis. These data strongly suggestthat pathogenic mycoplasma of murine originwere primarily responsible for disease and deathsin conventional mice. Isolation of mycoplasmafrom the lungs of mice inoculated with sterilebroth (Table 2) confirmed that conventionalmice harbor indigenous murine mycoplasmata,and supported the view that disease in con-ventional animals is due to these murine myco-plasma.The microscopic pathology of the lung lesions

and the disease course through the acute andchronic stages were the same in the conventionalmice as in gnotobiotic mice inoculated with M.

pulmonis, with but three exceptions. In thesethree cases, abscesses were present on gross andmicroscopic examination of the lungs, and signifi-cant numbers of bacterial colonies were presentin impressions of the lungs on blood-agar; thus,the role of bacterial infection in these three con-ventional animals was unmistakable (Fig. 7).The role of bacterial infection in the remaininganimals can only be surmised, because bacterialisolations were common in lungs showing grossevidence of pneumonia, and were rare in lungsnot showing pneumonia (Table 3).An additional difference between conventional

and gnotobiotic mice was the striking absenceof round-cell infiltration in the perivascular andperibronchial interstitial spaces in gnotobioticmice inoculated with sterile broth. This findingpresented a contrast to the regular appearanceof small numbers of round cells in the interstitialspaces in conventional mice inoculated intra-nasally with sterile broth. The lungs of thesegnotobiotic mice were not totally devoid oflymphocytes, for lymphocytes were seen in hilarlymph nodes and in small lymphatic vessels.

M. salivor

Deaths M

Deoths

Deaths

20 25

158 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

FIG. 3. Microscopic appearance of the early (2nd day) pulmonary lesion in the gnotobiotic mouse (X 125).Hematoxylin and eosin stain. Areas of consolidation have a uniform appearance; the alveolar exudate consists ofpink-staining fluid and polymorphonuclear leukocytes. Small vessels and bronchi are visible within the pneumoniczone, but they have no distinguishing features.

FIG. 4. Pneumonia in the gnotobiotic mouse 4 days after intranasal inoculation with Mycoplasma pulmonis(X 125). Hematoxylin and eosin stain. A zone of darkly staining lymphocytes is visible surrounding a small pera-bronchial blood vessel. The adjacent bronchus contains an exudate ofpolymorphonuclear leukocytes, but there isno peribronchial infiltration bYv lymphocytes.

1159

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

FIG. 5. Features ofpneumonia in the gnotobiotic mouse 3 weeks after intranasal inoculation with Mycoplasmapulmonis are seen (X 125). Hematoxylin and eosin stain. A large bronchus is nearly occluded by polymorphonuclearexudate. Broad zones ofdarkly staining lymphocytes and plasma cells surround bronchi and vessels. Alveoli containlarge mononuclear cells.

s* 0

4a

e *~I o

_

FIG. 6. Perivascular zone in the lung of a gnotobiotic mouse 3 weeks after intranasal inoculation with Myco-plasma pulmonis (X 550). Hematoxylin and eosin stain. A nodule ofsmall, darkly staining lymphocytes and denselypacked, polyhedral plasma cells can be seen. Adjacent alveoli contain large, pale-staining mononuclear cells withabundant cytoplasm.

1160

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

VOL. 92,1966 PNEUMONIA DUE TO MYCOPLASMA IN GNOTOBIOTIC MICE. I

Also, the gnotobiotic mouse was found capableof a round-cell response and of proliferation ofthese cells, since round cells made an appearancein the perivascular interstitial tissue within 4 to6 days after intranasal inoculation with M. pl-monis and proliferated rapidly in these locations.By 2 to 4 weeks, the multiplication of round cellsin the interstitial spaces was so massive that, inthe infected animals, gnotobiotes as well as con-ventionals, round-cell proliferation representedthe outstanding histological feature.The final difference noted was the absence of

spontaneous pulmonary neoplasms in the lungs

TABLE 6. Results of intranasal inoculation withMycoplasma pneumoniae in 34 gnotobiotic mice

Anatomical findings

Microbiological findingsGross Grossly nor- Total

pneumonia mal lungs

Mycoplasma. 0 3a 3Mycoplasma andbacteria.0 0 0Bacteria.0 0 0No growth . 31 31

a Colonies morphologically consistent with M.pneumoniae were isolated in 3 of 17 animals sacri-ficed 2 weeks after inoculation.

of these young gnotobiotic mice. On microscopicexamination, neoplasms were found in 17 of 102conventional mice, but not once in the lungs of179 gnotobiotic mice of the same strain (Ha/ICR)and age.

DIscussioN

The selection of gnotobiotic mice for this studyrepresents a logical extension of the work ofprevious investigators (5, 8, 10), who recognizedthe need for stocks of mice free from catarrh forexperiments involving intranasal inoculations.A primary concern in 1940 was the validity ofexperiments involving intranasal inoculations ofviruses: "For virus research, the ideal would bestocks of animals as far as possible free from alllatent infections, just as uncontaminated nutrientmedia are essential for the study of bacteria"(5). This statement is no less true for researchinvolving mycoplasma administered by the intra-nasal route. Technical advances in gnotobioticmethods since 1950 have made available stocksof animals which, in the main, conform to theideal set by Edward in 1940. The lungs of thecontrol groups in the gnotobiotic mice used inour studies were consistently negative for myco-plasma and bacteria after the stress of intranasalinoculation of sterile broth.The factor of host specificity is of extreme im-

FIG. 7. Portion of a large abscess in the lung of a convenional mouse inoculated with Mycoplasma pulmonis(X 125). Hematoxylin and eosin stain.

1 161

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

LUTSKY AND ORGANICK

portance and should be discussed in its properperspective. The host specificity in the gnoto-biotic Ha/ICR mouse is very marked, for notonly did M. salivarium, a commensal organismpresent in the pharynx of approximately 50%of normal human subjects, fail to produce anymeasurable effect, but even M. pneumoniae, apathogenic species of human origin which pro-duces bronchitis and pneumonia (the "atypicalpneumonia" associated with cold agglutinins)in man, failed to produce a single instance ofillness or gross pneumonia. M. pulmonis, on theother hand, a pathogenic mycoplasma of mouseorigin, produced illness and pneumonia in ahigh proportion of animals.The conventional Ha/ICR mouse represents

a different host for which M. salivarium andM. pneumoniae were distinctly more pathogenicthan was sterile broth. It was not possible todistinguish the pathogenicity of the two speciesof human origin, M. salivarium and M. pneu-moniae inter se, but both were less pathogenicthan M. pulmonis. The results in the conventionalanimal are difficult to interpret. Synergisticaction of M. salivarium or M. pneumoniae withindigenous murine mycoplasma in the conven-tional Ha/ICR stock and with indigenous bac-teria could be an important factor. Gnotobioticanimals may be more, rather than less, resistantto inoculation with large numbers of a singlemicrobial species introduced from the outsidethan conventional animals. It has been founddifficult to establish certain single bacterialspecies in germ-free mice (6). The complete dis-appearance of M. salivarium from the lungs of allof the inoculated gnotobiotic mice by 2 weeksshows impressive ability of the mammalianorganism to clear itself of certain foreign mi-crobes.The pathogenicity of M. pulmonis for the

gnotobiotic Ha/ICR mouse was very striking.To the authors' knowledge, this is the first de-scription of the use of the gnotobiotic mouse inpathogenicity studies involving intranasal inocu-lations of mycoplasma. Compared with theslower onset of disease in conventional micedescribed by Nelson (8) and Edward (5), theclinical course that we observed was more rapid.Sullivan and Dienes (10) also observed deathswithin a few days after intranasal inoculation of asuspension of lungs with almost total pulmonaryconsolidation, but did not describe the histo-pathology of the acute stage. The difference intime course of the disease may be related todifferences in the dose of organism or to thestrain of mouse used.The round-cell infiltration of the peribronchial

space of the conventional animals would seem

to represent morphological evidence of an im-mune response in animals living in a normal en-vironment to microbes which impinge upon therespiratory passageways. The absence of round-cell infiltration from the respiratory tract of thecontrol gnotobiotic mice would seem to representan equivalent to the conspicuous diminution orabsence of round cell infiltration in the sub-mucosal tissues of the gastrointestinal tract ofgnotobiotes in the absence of viable microbes inthe fecal stream.Our observation on the absence of pulmonary

neoplasms in gnotobiotes is of interest, becauseit does not confirm previous reports (Ward,Federation Proc., 20:150, 1961) that spontaneoustumors have an approximately equal incidencein gnotobiotic and conventional animals. Thediscrepancy in findings may be due to the factthat all of the mice in our study were 10 weeksof age or less.

ACKNOWLEDGMENTS

This investigation was supported by grantsfrom The American Thoracic Society, PublicHealth Service general research support grant 5SO1FR 5434, training grant 5T1 HE 5538, andPublic Health Service grant AM4150-05 from theNational Institute of Arthritis and MetabolicDiseases.The technical assistance of Patricia Goggans is

gratefully acknowledged. We also wish to ac-knowledge the help of the Staff of the VeteransAdministration Biostatistics Research SupportCenter, Hines, Ill., in the statistical analysis.

LITERATURE CITED1. CHANOCK, R. M., L. DIENES, M. D. EATON,

D. G. FF. EDWARD, E. A. FREUNDT, L. HAY-FLICK, J. F. PH. HERS, K. E. JENSEN, C. Liu,B. P. MARMION, H. E. MORTON, M. A. MUF-SON, P. F. SMITH, N. L. SOMERSON, AND D.TAYLOR-ROBINSON. 1963. Mycoplasma pneu-moniae: proposed nomenclature for atypicalpneumonia organism (Eaton agent). Science140:662.

2. CHANOCK, R. M., L. HAYFLICK, AND M. F.BARILE. 1962. Growth on artificial medium ofan agent associated with atypical pneumoniaand its identification as a PPLO. Proc. Natl.Acad. Sci. U.S. 48:41-49.

3. DAJANI, A. S., W. A. CLYDE, AND F. W. DENNY.1965. Experimental infection with Mycoplasmapneumoniae (Eaton's agent). J. Exptl. Med.121:1071-1086.

4. EATON, M. D., G. MEIKLEJOHN, AND W. VANHERICK. 1944. Studies on the etiology of pri-mary atypical pneumonia. A filterable agenttransmissible to cotton rats, hamsters, andchick embryos. J. Exptl. Med. 79:649-668.

5. EDWARD, D. G. FF. 1940. The occurrence in

1162 J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.


normal mice of pleuropneumonia-like organ-isms capable of producing pneumonia. J.Pathol. Bacteriol. 50:409-418.

6. GIBBONS, R. J., S. S. SOCRANSKY, AND B. KAP-SIMALIS. 1964. Establishment of human in-digenous bacteria in germ-free mice. J. Bac-teriol. 88:1316-1323.

7. GOODBURN, G., AND B. P. MARMION. 1962. Astudy of the properties of Eaton's primaryatypigal pneumonia organism. J. Gen. Micro-biol. 29:271-290.

8. NELSON, J. B. 1937. Infectious catarrh of mice.J. Exptl. Med. 65:833-860.

9. NELSON, J. B. 1958. Infection in laboratory ani-mals with pleuropneumonia-like organisms. J.Natl. Cancer Inst. 20:911-915.

10. SULLIVAN, E. R., AND L. DIENEs. 1939. Pneu-monia in white mice produced by a pleuro-pneumonia-like micro-organism. Proc. Soc.Exptl. Biol. Med. 41:620-622.

11. TULLY, J. G. 1965. Biochemical, morphological,and serological characterization of mycoplasmaof murine origin. J. Infect. Diseases 115:171-185.

12. WAGNER, M. 1959. Determination of germ-freestatus. Ann. N.Y. Acad. Sci. 78:89-100.

1163

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

28

Dec

embe

r 20

21 b

y 20

1.93

.118

.222

.

Documents

Pneumonia Due to Mycoplasma in Gnotobiotic Mice