APPLIED MICROBIOLOGY, Sept. 1969, p. 427-432 Vol. 18, No. 3Copyright ( 1969 American Society for Microbiology Printed in U.S.A.

Comparative Pathogenicity of Actinomycesnaeslundii and Actinomyces israelii

R. MARIE COLEMAN AND LUCILLE K. GEORGHealth Services and Mental Health Administration, National Commuwicable Disease Center, Atlanta,

Georgia 30333

Received for publication July 8 1969

Typical actinomycosis has been produced in mice following single intraperitonealinjections of saline suspensions of Actinomyces israelfi and A. naeslundii. A. israeliiproduced infections in 95.8% of the animals inoculated. A. naeslundii, generallyconsidered to be a saprophytic organism, produced lesions in 89.7% of the inocu-lated animals. The finding that A. naeslundii produced lesions in mice similar tothose produced by A. israelfi suggests that A. naeslundii has similar pathogenicpotential for man. The isolation of A. naeslundii from suppurative lesions of manalso supports this conclusion.

The production of actinomycosis in experi-mental animals with pure cultures of Actinomycesisraelii was first clearly described in 1891 byWolff and Israel (14). They were able to producetypical actinomycotic lesions in 18 rabbits andthree guinea pigs. Typical grains with mycelialfilaments and clubbing were present in mostlesions. Since that time numerous workers (1, 4,5, 8, 11, 15), have obtained similar results withvarious animals. Irregular or poor results havebeen obtained by other workers (3, 10, 12).Although numerous attempts utilizing pure

cultures of A. isrealii have been made to produceactinomycosis in the experimental animal, fewattempts have been made to produce infectionswith A. naeslundii. Howell et al. (6) used threestrains of A. naeslundii to inject 28 (21- to 25-day-old) hamsters intraperitoneally to test for patho-genicity. In two animals, one sacrificed after 8days and the other after 14 days, a solitary noduleapproximately 1 mm in diameter was found on theperitoneal lining of the body wall. Actinomyceteswere demonstrated in smears from each lesion.Using white mice as the experimental animals andone strain of A. naeslundii, Buchanan and Pine(1) were able to produce infections in seven of16 mice inoculated intraperitoneally Four of theseven infected mice had only small lesions on theperitoneal wall or at the point of inoculation. Theother three infected mice had lesions that weremuch more extensive. Two mice had encapsulatedabscesses with diameters greater than 2 cm, oneon the intestine and the other on the liver, and anumber of other small abscesses. The third

mouse had a large intestinal abscess and alsosmall abscesses on the liver and mesentery.During the past several years at the National

Communicable Disease Center (NCDC, Atlanta,Ga.), A. naeslundii has been isolated from humanclinical materials other than the human mouthwhich is considered its normal habitat (9, 13).These isolations have led to the question of thepossible role of A. naeslundii as an agent ofhumanactinomycosis. In a previous report (2), fiveisolates of A. naeslundii from the normal mouthand seven from pathological clinical materialshave been characterized morphologically, bio-chemically, and serologically, and were found tobe essentially similar. The present study wasundertaken to investigate the pathogenic potenti-alities of these same 12 isolates of A. naeslundiiand to compare the pathogenicity of these orga-nisms for mice to 12 isolates of A. israelii.

MATERIALS AND METHODSTwelve strains of A. israelii and 12 strains of A.

naeslundii were studied. A list of the A. israeli strains,sources, and clinical data are given in Table 1. Thesources of the A. naeslundii strains are given in the pre-vious study (2). All cultures were routinely grown inenriched thioglycollate broth (1.5 of dehydratedTrypticase Soy Broth and 1.25 g of Tryptose Brothper liter of Thioglycollate Broth, BBL).

For animal inoculations, transfers were made froman actively growing thioglycollate broth culture toslants of Brain Heart Infusion Agar and incubatedunder anaerobic seals, as described in the previousstudy (2). Cultures were incubated at 37 C until maxi-mum growth was obtained, usually 4 to 5 days. Theinoculum was then prepared by washing the growthfrom the slants with sterile saline. Where clumps of

427

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

COLEMAN AND GEORG

TABLE 1. A. israelii cultures studied

sNCDCn NCDC diagnostic no. Other identification Clinical source

W855 Howell, NIHb, 277, ATCCC Brain abscess12102 (A. israelii, type strain)

X695 45-395-64 (received 4/30/64) V.A. Hospital, 6394, Atlanta, RetroperitonealGa. abscess

W726 Howell, NIH, 287, ATCC, 12103 Cervico-facial abscessW749 45-1153-64 (received 12/31/64) Weed, Mayo Clinic, Rochester, Liver abscess

Minn. (strain, Kramer)W750 45-1154-64 (received 12/31/64) Weed, Mayo Clinic, Rochester, Draining sinus on

Minn., strain, Tobeck torsoW796 44-282-65 (received 8/19/65) Wash. State Dept. of Health, Parotid gland abscess

CDS, 195W825 Emmons, NIH, 1829, ATCC, Pleural fluid

10048W838 44-12-66 (received 2/16/66) Weed, Mayo Clinic, Rochester, Parotid gland abscess

Minn. # 2-606-032W850 44-48-66 (received 3/21/66) Mass. General Hospital, F582, Facial abscess

Boston, Mass.W946 44-202-66 (received 10/12/66) V.A. Hospital, Seattle, Wash. Pus from molar ex-

(strain, Greenberg) tractionW954 44-218-66 (received 11/18/66) Kobayashi, Washington Univ. Draining sinus on

School of Medicine, St. torsoLouis, Mo. (strain, Holmes)

W1045 44-148-67 (received 7/20/67) V.A. Hospital, 8988, Atlanta, Peritoneal abscessGa.

a NCDC, National Communicable Disease Center, Atlanta, Ga.b NIH, National Institutes of Health, Bethesda, Md.c ATCC, American Type Culture Collection, Rockville, Md.

organisms occurred, the growth was homogenized byrepeatedly drawing up the suspension in a syringe andexpelling it. The suspension was then adjusted tomatch the turbidity of tube 8 of the McFarland nephe-lometer standard (McFarland Engineering & PumpCo., Inc., Houston, Tex.). With a minimal loss of time,0.5 ml of the inocula was injected intraperitoneallyinto 15- to 30-g, white, male mice.The mice were sacrificed routinely 10 to 15 days

after inoculation. A. israel/i strains W749, W750, andW825 and A. naeslundii strains X569 and W752 weremaintained up to eight weeks before autopsy to deter-mine whether infection would kill the animals andwhether the lesions would persist.

Smears for Gram strains and fluorescent antibody(FA) studies were made of purulent material found inthe lesions. In some instances, tissue sections weremade for histo-pathological studies. The extent of theinfection was graded according to the following scale:0, no lesions found; 1+, abscess at point of inocula-tion only; 2+, abscesses on peritoneal surfaces and inmesenteries; 3+, above with abscesses 10 mm or morein diam in the area of spleen and stomach; 4+, abovewith lesions on the liver, diaphragm, or both; 5+,above with lesions in the pleural cavity.

Only abscesses containing gram-positive, morpho-logically typical organisms (Fig. 1) which could beidentified by specific FA reactions were considered tobe actinomycotic lesions. FA tests were performed by FIG. 1. A. nathe direct staining technique described by Lambert tally infected net al. (7). stain. X 1,125.

zeslundii W826 in pus from experimen-nouse 15 days postinoculation; Gram

428 APPL. MICROBIOL.

, , - 7 - - - -

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

VOL. 18, 1969 PATHOGENICITY OF A. NAESLUNDII VERSUS A. ISRAELII

RESULTS AND DISCUSSIONActinomyces israelii. Of 95 mice inoculated with

strains of A. israelii, 91 (95.8%) demonstratedlesions (Table 2); of these, 71 (78%) were of3 to 4+ severity. Organswhichweremostfrequentlyinfected were the peritoneal surfaces, mesenteries,diaphragm, and liver. In many instances verylarge abscesses, up to 20 mm in diam, were foundattached to the base of the stomach and theposterior surface of the spleen. Also largeabscesses were frequently found in the lowerabdomen. In no instances were abscesses foundin the pleural cavity.The majority of the mice showed no apparent

ill effects from the infection. Only one death wasattributed to the infection and this occurred inthe group injected with strain W749. The mousewas found dead on the 10th day and had 3 +lesions.Of 16 mice held for 8 weeks, 14 (88%) showed

2-4+ lesions. This indicates that these infectionswere not transient in nature. Retrocultures of A.israelii were obtained from some of these animals.Both granules with and without clubbing were

readily found in exudates from lesions and also in

histologic sections. Positive Gram stains and FAtests were obtained from all infected animals.Actinomyces naeslundii. Of the 117 mice inoc-

ulated with A. naeslundii strains, 105 (89.7%)demonstrated lesions (Table 3). Of these, 44(41.9%) were of 3-5+ severity. Figure 2 illus-trates 4+ lesions on the liver and spleen of amouse inoculated with A. naeslundii and founddead on the 11th day after inoculation. The dis-tribution of lesions was similar to that producedby A. israelii, except that in one mouse inoculatedwith A. naeslundii W821, a large lesion developedin the pleural cavity in addition to those in theabdominal area. The pleural lesion consisted ofan abscess that completely encompassed the upperhalf of the right lung (Fig. 3). Upon histologicalexamination, very little of the lung tissue re-mained. The lesion was made up of a well walled-off abscess that contained fibrin, pus, and largenumbers of gram-positive, filamentous, branchedorganisms. Retroculture of A. naeslundii was

obtained.Three other deaths were attributed to this

organism. One death occurred in the group in-jected with strain ATCC 12104 (American Type

TABLE 2. Experimental pathology in mice following a single intraperitoneal injection of A. israelii

Mice showing Extent ofStrain Human clinical source macroscopic lesions (15 Remarks

W855 Brain abscess (type 6/6 4+ 6 of 6 mice with 4+ lesions(ATCC, strain)12102)

X695 Retroperitoneal 10/10 3-4+ 10 of 10 mice with 3-4+ lesionsabscess

W726 Cervico-facial 5/6 0-4+ 1 mouse not infected; 5 mice with 4+ lesionsabscess

W749 Liver abscess 10/12 2-4+ (1 3 of 6 mice by 6th wk with 3-4+ lesions; 6dead at mice held 8 wk; 2 negative and 4 with10 days, 2-3+ lesions3+)

W750 Draining sinus on 12/12 1-4+ 3 of 6 mice by 6th wk with 3-4+ lesions; 6torso mice held 8 wk: 4 with 2+, and 2 with

3+ lesionsW796 Parotid gland 6/6 3-4+ 5 of 6 mice with 4+ lesions

abscessW825 Pleural fluid 6/6 3+ 5 mice held 8 wk: 3 with 2+, 1 with 3+, and

1 with 4+ lesionsW838 Parotid gland 5/5 2-4+ 4 of 5 mice with 3-4+ lesions

abscessW850 Facial abscess 6/6 1-4+ 3 of 6 mice with 4+ lesionsW946 Pus from molar 8/8 1-4+ 1 mouse with sinus tract at point of inocula-

extraction tion, only 1+ lesions; 7 mice with 3-4+lesions

W954 Draining sinus on 7/7 1-4+ 1 mouse with sinus tract at point of inocula-torso tion, only 1+ lesions; 6 mice with 3-4+

lesionsW1045 Peritoneal abscess 10/11 3-4+ 9 of 11 mice with 3-4+ lesions

aNumerator, number of mice infected; denominator, number of mice inoculated. Total numberinfected, 91/95; percentage infected, 95.8%.

429

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

COLEMAN AND GEORG

TABLE 3. Experimental pathology in mice following a single intraperitoneal injection of A. naeslundii

Strain Human clinical source Mice showing Extent ofno. ain u macroscopic lesions (15 Remarks

lesionsa days)

W826 Sinus tract (type 9/10 0-3+ (1 1 mouse not infected; 1 mouse with lesion(ATCC, strain) dead at only at inoculation point; 4 of 10 mice12104) 10th day, with 3+ lesions

3+)X569 Blood stream 16/19 1-4+ 5 mice held 8 wk: 3 negative, one 1+, and 1

with 2+ lesionsX600 Dental calculus 8/8 2-4+ 1 mouse had extensive 4+ lesionsW752 Postoperative 9/10 2+ 5 mice held 8 wk: 3 negative, 2 with 1+

wound lesionsW821 Fluid from ankle 13/15 1-5+ (3 2 mice with lesion only at inoculation point,

lesion dead by 1+; 6 mice with 3-5+ lesions15 days,2-5+)

W833 Apex of tooth 9/12 1+W869 Dental calculus 3/4 1-2+ 1 mouse with lesion only at inoculation

point, 1+W953 Sputum 5/5 1-2+ 3 mice with lesions only at inoculation

point, 1+W1003 Blood stream 7/8 2-3+W1048 Abdominal wound 10/10 2-3+W1096 Gall bladder 8/8 2-4+

empyemaW1105 Abdominal wound 8/8 2-3+

iNumerator, number ou mice intecteu; uerinfected, 105/117; percentage infected, 89.7%.

nominator, number oi mice inocuiatea. 1iotai number

FIG. 2. Large abscesses on posterior surfaces of FIG. 3. Right lung (organ on far right) with ab-liver lobes and spleen due to A. naeslundii W821, 11 scessed upper lobe. Organ on far left is abscesseddays postinoculation. spleen from same animal. A spleen from a normal

mouse is placed in the center for comparison. Experi-Culture Collection, Rockville, Md.), and two mental infection due to A. naeslundii W821, 11 daysJthIr dpnthke niJrInA-l i;nthuLa1 r^n ;nitAij Ax6Lu postinoculation.

strains W821. All three of these mice had 2-4+lesions. The remainder of the mice showed noapparent ill effects from the injections.Of 10 mice held for eight weeks, six showed no

lesions, three had 1+ lesions and one had 2+lesions. This suggested that in animals surviving

this period of time, the lesions were spontaneouslyregressing.

Granules without clubbing were found in someexudates and histologic sections (Fig. 4a and b).Positive Gram stains and FA tests were obtained

430 APPL. MICROBIOL.

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

PATHOGENICITY OF A. NAESLUNDII VERSUS A. ISRAELII

FIG. 4. (a) Experimental splenic lesion due to A. naesluwdii W821. Shows granules surrounded by mass ofleucocytes. Hematoxylin and eosin stain. X 50. (b) Individual granule in A. naeslundii W821 abscess showingbranchedfilamentous organisms; Brown and Brenn stain. X 1,150.

TABLE 4. Comparison of lesions produced inmice by A. israelii with those produced

by A. naeslundii

A. israelii A. naeslundii

Extent ofinfectionsa No. of Pecetage No. of Percentage

minfection~cefof total no. micef of total no.inoculated iouae

No lesions 4 4.2 12 10.31+ lesions 5 5.3 5 4.272+ lesions 15 15.8 56 47.93+ lesions 35 36.8 37 31.64+ lesions 36 37.9 6 5.15+ lesions 0 0 1 0.9Death I 1.1 4 3.4Total infected 91 95.8 105 89.7

a 1+, abscess at point of inoculation only; 2+,abscess on peritoneal surfaces and in mesenteries;3+, above with abscesses 10 mm or more in diamin area of spleen and stomach; 4+, above with lesionson the liver, diaphragm, or both; 5+, above withlesions in pleural cavity.

from all infected animals. Retrocultures fromsome of the mice inoculated with strains X569,W821, W1003, W1048, and W1096 were alsoobtained. This indicates that the organisms were

viable and actively multiply in the infectedanimals.As seen in Table 4, the lesions were somewhat

more extensive with A. israelii, the majority ofmice showing lesions of 3-4+ severity. With A.naeslundii, the majority of mice showed lesionsof 2-3 + severity, only 6.0% showing 4-5 +lesions as compared to 37.9% with A. israelii. Thelesions produced by these two species did not differgreatly. The only significant difference was thatthe A. naeslundii strains tested produced granuleswithout clubs, whereas with many of the A.israelii strains, typical granules with clubbing werecommonly seen.

ACKNOWLEDGMENTWe wich to acknowledge John P. Gust of the National Medical

Audiovisual Center, Atlanta, Ga. for the preparation of thephotographs.

LITERATURE CITED

1. Buchanan, B. B., and L. Pine. 1962. Characterization of apropionic acid producing actinomycete, Actinomyces propi-onicus. Sp. nov. J. Gen. Microbiol. 28:305-323.

2. Coleman, R. M., L. K. Georg, and A. R. Rozzell. 1969.Actinomyces naeslundii as an agent of human actinomycosis.Appl. Microbiol. 18:420-426.

3. Emmons, C. W. 1939. The isolation of Actinomyces bovis fromtonsillar granules. Pub. Health Rep. 53:1967-1975.

VOL. 18, 1969 431

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from

432 COLEMAN AND GEORG

4. Gale, D., and C. A. Waldron. 1955. Experimental actinomy-cosis with Actinomyces israelii. The development of visceralactinomycosis. J. Infec. Dis. 97:251-261.

5. Hazen, E. L., G. N. Little, and H. Resnick. 1952. The hamsteras a vehicle for the demonstration of pathogenicity ofActinomyces bovis. J. Lab. Clin. Med. 40:914-918.

6. Howell, A., Jr., W. C. Murphy, III, F. Paul, and R. M.Stephan. 1959. Oral strains of Actinomyces. J. Bacteriol.78:82-95.

7. Lambert, F. W., Jr., J. M. Brown, and L. K. Georg. 1967.Identification of Actinomyces israelii and Actinomycesnaeslundii by fluorescent-antibody and agar-gel diffusiontechniques. J. Bacteriol. 94:1287-1295.

8. Magnusson, H. 1928. (Suppl.) The commonest forms ofactinomycosis in domestic animals and their etiology. ActaPathol. Microbiol. Scand. 5:170-245.

9. Naeslund, C. 1925. Studies of Actinomyces from the oralcavity. Acta Pathol. Microbiol. Scand. 2:110-140.

APPL. MIcRoBIoL.

10. Rosebury, T., L. J. Epps, and A. R. Clark. 1944. A study ofthe isolation, cultivation and pathogenicity of Actinomycesisraelii recovered from the human mouth and from ac-tinomycosis in man. J. Infec. Dis. 74:131-149.

11. Slack, J. 1942. The source of infection in actinomycosis. J.Bacteriol. 43:193-209.

12. Sullivan, H. R., and N. E. Goldsworthy. 1940. A comparativestudy of anaerobic strains of Actinomyces from clinicallynormal mouths and from actinomycotic lesions. J. Pathol.(Australia) 51:253-261.

13. Thompson, L., and S. A. Lovestedt. 1951. An Actinomyces-like organism obtained from the human mouth. Proc. MayoClinic 26:169-175.

14. Wolff, M., and J. Israel. 1891. Ueber Reincultur des Actino-myces und seine Uebertragbarkeit auf Thiere. VirchowsArch. Pathol. 126:11-59.

15. Wright, J. H. 1905. The biology of the microorganism ofactinomycosis. J. Med. Res. 13:349-404.

on Novem

ber 8, 2020 by guesthttp://aem

.asm.org/

Dow

nloaded from