General Characteristics

The genera described in this chapter are all catalase-negative, non-spore-forming, gram-positive rods; some may exhibit rudimentary branching. Erysipelothrix rhusiopathiae is one of three species in the genus, but it is considered the only human pathogen. E. rhusiopathiae consists of several serovars based on peptidoglycan structure. The serovars most commonly associated with human infection include serovars 1 and 2. Arcanobacterium spp. demonstrate irregular, gram-positive rods on Gram stain. Gardnerella sp. fermentation byproducts include acetic and lactic acid. The cell wall of Gardnerella sp. is significantly thinner and contains less peptidoglycan than the typically gram-positive bacteria. Weissella confusa, formerly classified as Lactobacillus confusus, is included in Tables 18-3 and 18-4 because it is easily confused on culture media with the organisms included in this chapter, and in rare cases it has been isolated associated with bacteremia and endocarditis.

Epidemiology

Erysipelothrix spp. are found worldwide in a variety of vertebrate and invertebrate animals, including mammals, birds, and fish. Other domestic animals that may be infected include sheep, rabbits, cattle, and turkeys. The organism may be transmitted through direct contact or ingestion of contaminated water or meat. Arcanobacterium spp. are normal inhabitants of the mucosal membranes of cattle, sheep, dogs, cats, and pigs. The organisms listed in Table 18-1 include those that are closely associated with animals and are contracted by humans through animal exposure (e.g., E. rhusiopathiae and Arcanobacterium pyogenes) and those that are part of the normal human flora (e.g., Lactobacillus spp. and Gardnerella vaginalis).

Pathogenesis and Spectrum of Disease

G. vaginalis and Lactobacillus spp. (Table 18-2) are natural inhabitants of the human vagina. Vaginal infections with G. vaginalis are often found in association with a variety of mixed anaerobic flora. Extravaginal infections are uncommon but have been identified associated with postpartum endometritis, septic abortion, and cesarean birth.

Generally, no special considerations are required for specimen collection and transport of the organisms discussed in this chapter. Of note, skin lesions for Erysipelothrix should be collected by biopsy of the full thickness of skin at the leading edge of the discolored area. Refer to Table 5-1 for other general information on specimen collection and transport.

Direct Detection Methods

Gram staining of Arcanobacterium spp. demonstrates delicate, curved, gram-positive rods with pointed ends and occasional rudimentary branching. This branching is more pronounced after these organisms have been cultured anaerobically. Arcanobacterium spp. stain unevenly after 48 hours of growth on solid media and also exhibit coccal forms.

Lactobacillus is highly pleomorphic, occurring in long chaining rods and in coccobacilli and spiral forms (Figure 18-1).

E. rhusiopathiae stains as both short rods and long filaments. These morphologies correspond to two colonial types: (1) rough colonies that contain slender, filamentous, gram-positive rods with a tendency to overdecolorize and appear gram negative and (2) smooth colonies that contain small, slender rods. This variability in staining and colonial morphology may be mistaken for a polymicrobial infection both on direct examination and culture.

Gardnerella organisms are small, pleomorphic gram-variable or gram-negative coccobacilli and short rods. Wet mount and Gram staining of vaginal secretions are key tests for diagnosing bacterial vaginosis caused by G. vaginalis. A wet mount prepared in saline reveals the characteristic “clue cells,” which are large, squamous epithelial cells with numerous attached small rods. A Gram-stained smear of the discharge shows the attached organisms to be gram-variable coccobacilli. In bacterial vaginosis, clue cells are typically present, and large numbers of other gram-positive rods (i.e., lactobacilli), representing normal vaginal flora, are absent or few in number. In addition, the BDaffirm vaginal DNA probe (VDP) may be used for direct detection from genital specimens. Special vials containing transport reagent are used to stabilize the organism’s nucleic acids prior to testing (Becton, Dickinson and Company Franklin Lakes, NJ).

Cultivation

Media of Choice.

All the genera described in this chapter grow on 5% sheep blood and chocolate agars. They do not grow on MacConkey agar but do grow on Columbia colistin-nalidixic acid (CNA) agar. CNA agar is a nutritional base that may include 5% sheep blood to enhance the growth of fastidious organisms. The antibiotics colistin and nalidixic acid prevent the overgrowth of gram-negative organisms. All genera except Gardnerella sp. grow in commercially available blood culture broths. Gardnerella organisms are inhibited by sodium polyanetholsulfonate (SPS), which currently is used as an anticoagulant in most commercial blood culture media. An SPS-free medium or a medium with SPS that is supplemented with gelatin should be used when G. vaginalis sepsis is suspected.

Isolation of G. vaginalis from female genital tract specimens is best accomplished using the selective medium human blood bilayer Tween agar (HBT). HBT is CNA agar with amphotericin B added to prevent the growth of yeasts and filamentous fungi. Human blood is layered over the top to enhance the beta-hemolytic pattern of G. vaginalis.

Incubation Conditions and Duration

Detectable growth of these organisms should occur on 5% sheep blood and chocolate agars, CNA, and HBT incubated at 35°C in 5% to 10% carbon dioxide (CO₂) within 48 hours of inoculation.

Colonial Appearance

Table 18-3 describes the colonial appearance and other distinguishing characteristics (e.g., hemolysis) of each genus on sheep blood agar. G. vaginalis produces small, gray, opaque colonies surrounded by a diffuse zone of beta-hemolysis on HBT agar (Figure 18-2).

TABLE 18-3

Colonial Appearance on 5% Sheep Blood Agar and Other Characteristics

Organism	Appearance
Arcanobacterium spp.	Small to large colonies with various appearances, including smooth, mucoid, and white and dry, friable, and gray; may be surrounded by narrow zone of beta-hemolysis
Erysipelothrix rhusiopathiae	Two colony types: large and rough or small, smooth, and translucent; shows alpha-hemolysis after prolonged incubation
Gardnerella vaginalis	Pinpoint; nonhemolytic
Lactobacillus spp.	Multiple colonial morphologies, ranging from pinpoint, alpha-hemolytic colonies resembling streptococci to rough, gray colonies
Weissella confusa	Pinpoint; alpha-hemolytic and may be confused with organisms presented in this chapter.

Approach to Identification

The identification of the four genera described in this chapter must be considered along with that of Actinomyces, Bifidobacterium, and Propionibacterium spp., which are discussed in Chapter 42. Although the latter genera are usually considered with the anaerobic bacteria, they grow on routine laboratory media in 5% to 10% CO₂. Some are catalase negative. Therefore, as shown in Table 18-4, these organisms must be considered together when a laboratory encounters catalase-negative, gram-positive, non-spore-forming rods.

TABLE 18-4

Biochemical and Physiologic Characteristics of Catalase-Negative, Gram-Positive, Aerotolerant, Non–Spore-Forming Rods

	Urease	Nitrate Reduction	Beta-Hemolysisb	FERMENTATIONa OF:							Other Comments
				Glucose	Maltose	Mannitol	Sucrose	Xylose	CAMPc	GLCd
Actinomyces israelii	−	v	−	+	+	v	+	+	−	A, L, S
A. odontolyticus	−	+	−e	+	v	−	+	v	−	A, S	Red pigment produced after 1 week on SBA
A. naeslundii	+	+	−	+	+	v	+	v	−	A, L, S
A. radingae	−	−	−w	+	+	−	+	+	−	S?	Pyrazinamidase, beta-galactosidase−positive and esculin-positive
A. turicensis	−	−	−w	+	v	−	+	+	−	NT	Pyrazinamidase, beta-galactosidase−negative and esculin-negative
A. graevenitzii	−	−	−	+	+	−	+	−	ND	L > S
Actinobaculum schaalii	−	−	−	+	+	−	v	+	+w	A, s	Beta-galactosidase−negative
Arcanobacterium haemolyticum	−	−	+	+	+	−	v	−	Reverse +f	A, L, S	Gelatin-negative at 48 hr; beta-hemolysis is stronger on agar containing human or rabbit blood
A. pyogenes	−	−	+g	+	v	v	v	+	−	A, L, S	Gelatin-positive at 48 hr; casein-positive
A. bernardiae	−	−	−	+	+	−	−	−	−	A, L, S
Bifidobacterium adolescentis	−	−	−	+	+	−	+	+	ND	A > L (s)
Erysipelothrix sp.	−	−	−	+h	−	−	−	−		A, L, S	H2S-positive in TSI butt; vancomycin-resistant; alpha-hemolytic
Lactobacillus spp.	−	−	−	+	+	v	+	ND	ND	L (a s)	Some strains vancomycin-resistant; alpha-hemolytic
Propionibacterium acnes	−	+	−	+	−	−	−	−	+	A, P (iv L s)	Indole-positive; may show beta-hemolysis on rabbit blood agar
P. propionicumj	−	+	−	+	+	+	+	−	ND	A, P, S, (L)	Colony may show red fluorescence under long-wavelength UV light
Gardnerella vaginalis	−	−	−	+	+	−	v	−k	ND	A (l s)	Beta-hemolysis on HBT; usually hydrolyses hippurate
Weissella spp.	NT	−	−	+	+	−	+	v	NT	L (as)	Vancomycin-resistant, small, short rods; produces gas from MRS broth; alpha-hemolytic; esculin-positive; arginine-positive

HBT, Human blood bilayer Tween agar; iv, isovaleric acid; ND; not done, NT; not tested; SBA, 5% sheep blood agar; TSI, triple sugar iron agar; v, variable; w, weak; +, ≥90% of strains positive; −, ≥90% of strains negative.

ⁱSome strains are catalase negative.

^aFermentation is detected in peptone base with Andrade’s indicator.

^bOn sheep blood agar.

^cCAMP test using a beta-lysin–producing strain of Staphylococcus aureus.

^dEnd products of glucose metabolism: A, Acetic acid; L, lactic acid; P, propionic acid; S, succinic acid; ( ), may or may not produce acid end product.

^eMay show beta-hemolysis on brain-heart infusion agar with sheep or human blood.

^fReverse CAMP test; Staphylococcus aureus beta-lysins are inhibited by a diffusible substance produced by A. haemolyticum (Figure 18-3).

^gMay also show beta-hemolysis on brain-heart infusion agar with human blood.

^hReaction may be weak or delayed.

^jFormerly Arachnia propionica.

^kGardnerella vaginalis–like organisms ferment xylose.

Several commercial systems for fastidious gram-negative bacterial identifications will adequately identify Gardnerella. The HNID panel (Haemophilus-Neisseria identification panel, Dade MicroScan, West Sacramento, California) works particularly well. However, rapid identification panels usually are used for isolates from extragenital sources (e.g., blood).

Comments Regarding Specific Organisms

A presumptive identification of G. vaginalis is sufficient for genital isolates, based on typical appearance on Gram stain, beta-hemolysis on HBT agar, and negative tests for oxidase and catalase. Corynebacterium lipophiloflavum, a bacteria isolated from females with bacterial vaginosis, is catalase positive.

The beta-hemolytic Arcanobacterium spp. resemble the beta-hemolytic streptococci but can be differentiated from them by Gram stain morphology. A. haemolyticum and A. pyogenes can be differentiated based on liquefaction of gelatin; A. pyogenes is positive and A. haemolyticum is negative. A. bernardiae is nonhemolytic.

Erysipelothrix sp. is the only catalase-negative, gram-positive non-spore-forming rod that produces hydrogen sulfide (H₂S) when inoculated into triple sugar iron (TSI) agar (Figure 18-4). Some Bacillus spp. also blacken the butt of TSI, but they are catalase positive and produce spores. Automated identification with the Vitek2 and Phoenix systems and the API ID system is reliable for identification.

Lactobacillus spp. are usually identified based on colony and Gram stain morphologies and catalase reaction (negative). Differentiation from viridans streptococci may be difficult, but the formation of rods in chains rather than cocci in thioglycollate broth is helpful. Alternatively, a Gram stain of growth just outside the zone of inhibition surrounding the 10-U penicillin disk placed on a blood agar plate inoculated with a lawn of the organism should show long bacilli rather than coccoid forms if the organism is Lactobacillus spp.

Antimicrobial Susceptibility Testing and Therapy

The rarity with which most of these organisms are encountered as the cause of infection has made the development of validated in vitro susceptibility testing methods difficult (Table 18-5). However, most of the organisms are susceptible to the agents used to eradicate them, therefore in vitro testing is not usually necessary to guide therapy. Lactobacillus spp. can be resistant to various antimicrobial agents. Fortunately, these organisms are rarely implicated in infections. When they are encountered in specimens from normally sterile sites, careful evaluation of their clinical significance is warranted before any attempt is made at performing a nonstandardized susceptibility test.

Although some of these organisms may grow on the media and under the conditions recommended for testing other bacteria (see Chapter 12 for more information regarding validated testing methods), this does not necessarily mean that interpretable and reliable results will be produced. Chapter 12 should be reviewed for preferable strategies that can be used to provide susceptibility information when validated testing methods do not exist for a clinically important bacterial isolate.

Prevention

Many of these organisms are ubiquitous in nature, and many are part of the normal human flora commonly encountered without deleterious effects on healthy human hosts. Currently, there are no recommended vaccination or prophylaxis protocols for prevention and treatment of diseases caused by these organisms.