Nocardia, Rhodococcus & Others

TAXONOMY Back to top

The word “actinomycete” is derived from two Greek roots (actino- and -mycete) meaning

“ray” (and hence also “rod”) and “fungus,” respectively. The anaerobic organisms now in the

genus Actinomyces and the aerobic organisms grouped together as the “aerobic

actinomycetes” were previously presumed to be related to one another on the basis of

shared features of organismal and colonial morphology. The aerobic actinomycetes, a group

for which no agreed-upon operational definition currently exists, are now known to be an

evolutionarily heterogeneous assemblage of genera. At some stage they all form grampositive

rods, and most of the more commonly isolated species exhibit at least rudimentary

branching under certain growth conditions; all grow better under aerobic than anaerobic

conditions, a feature distinguishing them from most organisms in the

genus Actinomyces. Figure 1 shows the current classification of genera included in this

chapter according to J. P. Euzeby (List of Prokaryotic Names with Standing in Nomenclature

[LPSN;www.bacterio.cict.fr]).

The organisms containing mycolic acids in their cell walls (included in the genera Dietzia,

Gordonia, Nocardia, Rhodococcus, Segniliparus, Tsukamurella, and Williamsia) are rather

closely related on the basis of molecular genetic studies (36, 191); these mycolic-acidcontaining

genera appear phylogenetically more closely related to the

genera Corynebacterium and Mycobacterium (both of which are sometimes considered

aerobic actinomycetes) than to the other non-mycolic-acid-containing genera usually also

included with the aerobic actinomycetes. According to a recent classification scheme, these 7

genera of actinomycetes, along with the genera Corynebacterium and Mycobacterium, are

classified together in the suborder Corynebacterineae (LPSN). Note in Fig. 1 that the

genera Corynebacterium, Williamsia, and possibly Dietzia are the only genera in the

suborder Corynebacterineae that are not at least weakly acid-fast.

The number of recognized pathogenic species of aerobic actinomycetes has been rising

rapidly. Only genera containing species of clinical significance are dealt with in this chapter.

Unfortunately, the increasingly fine discrimination of species has made it increasingly difficult

to delineate important species-specific differences in geographic distribution, pathogenic and

other biological mechanisms, disease associations, and antimicrobial susceptibility patterns.

This enormous proliferation of distinct species for which association with human disease has

been claimed presents several problems for the clinical microbiologist. First, phenotypic

testing has been rendered virtually useless for accurate discrimination among species.

Particularly for the aerobic actinomycetes, the number of phenotypic tests available in the

clinical laboratory (and in most research laboratories) is far too small for accurate

differentiation among so many species, and often, information on the percent positivity for a

specific reaction in a given species is not known. Furthermore, precisely the same

biochemical testing format has generally not been employed with isolates of all the species,

making the usefulness of multistudy comparisons uncertain. Second, some types of testing,

such as analysis of cell wall constituents and of whole-cell sugars, are available in only a few

research settings and are rarely of use for species-level identification. Some general

information regarding these features is provided herein, but the references should be

consulted for performance procedures and additional details. Third, only gene sequencing is

currently adequately discriminatory, reproducible, and sufficiently available to be useful for

precise species identifications in clinical laboratories. Fourth, because of the growing

problems with accurate species determinations, the clinical literature is rife with erroneous

identifications; nowhere is this problem more apparent than with organisms in the

genus Nocardia. Additionally, as some species have been described on the basis of only a

single isolate, and for others only a handful of reports exist, little if any meaningful clinical

information can be associated with many species names. Particularly when molecular

methods are not available, precise identification of isolates may be impossible—and is also

frequently not of immediate clinical utility.

There are a few terminological issues that, while they pertain to all bacteria, seem to cause

confusion particularly frequently with regard to the aerobic actinomycetes. First, for every

bacterial species, there is only one “type strain,” the strain on which the original description

of the species was based. Other strains contained in culture collections that are thought to

belong to the same species as a given type strain could be referred to as “reference strains,”

but they are not type strains. Second, the term “sensu stricto” means “in the strict sense.”

The term, when used with a species name, should be restricted to mean organisms

belonging to that particular species, as determined by the best available methods. So, to

refer to isolates that are “Nocardia asteroides sensu stricto” should mean isolates that by

gene sequencing are identical to the type strain of N. asteroides (ATCC 19247T).

DESCRIPTION OF THE GENERA Back to top

Basic information about specific microscopic and colonial morphologies of the following

genera is presented inTable 1; additional information is included in the text below for some

genera. See Table 2 for a comparison of the chemotaxonomic characteristics and lysozyme

resistance that distinguish these genera. In the listings that follow, enumeration of species

within a genus is taken from LPSN.

Amycolata and Amycolatopsis

The genera Amycolata and Amycolatopsis were proposed by Lechevalier et al. for Nocardialike

organisms that were gram positive and modified acid-fast negative but lacked mycolic

acids (128) (Fig. 2F and 3B). Among the species transferred to the

genus Amycolatopsis was Nocardia orientalis, two isolates of which were reported from

clinical specimens (80). A study based on 16S rRNA gene sequences recommended

combining the generaAmycolata and Pseudonocardia into an emended

genus, Pseudonocardia (214). Currently, there are 4 validly named species in the

genus Amycolata and approximately 44 validly named species or subspecies in the

genus Amycolatopsis.

Dermatophilus

The two species currently making up the genus Dermatophilus are probably not closely

related to most of the other organisms considered in this chapter. Aerial mycelium may be

produced by colonies if the organism is grown in an atmosphere of increased CO2; the

organisms are facultatively anaerobic (77). The microscopic morphology is unusual and

striking; this branching organism develops both longitudinal and transverse septa (Fig. 3D).

The resulting chains of coccoid cells may occur in as many as eight parallel rows (232). The

coccoid cells may develop into motile zoospores under favorable environmental conditions.

Dietzia

The species Dietzia maris was removed from the genus Rhodococcus because of

chemotaxonomic and 16S rRNA gene sequence differences from Rhodococcus species (174).

Seven additional species have since been included in the genus. These organisms apparently

rarely, if ever, branch, but like organisms in the genusRhodococcus, they may show coccal

and rod forms.

Gordonia

The biology of the genus Gordonia has been recently reviewed (5). On Gram

stain, Gordonia species appear coryneform and could easily be mistaken for a component of

the normal oral biota in sputum specimens (Fig. 3C). The genus “Gordona” (now Gordonia),

originally described by Tsukamura, was revived in 1988 by Stackebrandt et al. (192) to

contain species with mycolic acids of approximately 46 to 66 carbon atoms in length (73)

and a predominant menaquinone of nine isoprene units. Gordonia and Rhodococcus were

also found to be separable on the basis of their 16S rRNA gene sequences. There are

approximately 29 validly named species in the genus (Fig. 2I and 3C).

Nocardia

The most commonly isolated aerobic actinomycete human pathogens belong to the

genus Nocardia. In direct Gram smears, organisms generally appear as very long, obviously

branching, thin, and finely beaded gram-positive rods (Fig. 3E and K). Unlike the individual

cells composing a streptococcal chain, the beads generally do not abut one another (Fig.

3H and O). Particularly when prepared from cultures, smears may show streptococcus-like

chains or small branching filaments, probably as a result of the fragile nocardial mycelium

breaking during smear preparation (157, 158) (Fig. 3E, F, G, K, and L).

Colonial morphology varies from species to species and frequently varies from isolate to

isolate within a species (Fig. 2D, E, G, and L). Colony color may best be seen on the reverse

when colonies are grown on translucent media such as Sabouraud agar, as color may

become obscured on the surface by the powdery aerial hyphae typically produced.

A brief account of the nomenclatural history of the designation “Nocardia asteroides” is

necessary to clarify a confusing and widely unappreciated taxonomic issue. In 1888, Edmond

Nocard obtained an isolate of an organism thought to be the causative agent of bovine farcy.

This organism was given the name Nocardia farcinica by Trevisan in 1889, and thereupon, it

became the type strain for both the genus and species. Gordon and Mihm (79) found, using

their battery of phenotypic tests, that N. farcinica could not be distinguished from isolates to

which the name Nocardia asteroides had been applied. Because of some uncertainty relating

to the isolate obtained by Nocard and what was presumed to be the conspecificity of the

organisms then known as N. farcinica and N. asteroides, an appeal was made to the Judicial

Commission to have the type species of the genus changed to N. asteroides, with strain

ATCC 19247T selected as the type strain of the species. The appeal was accepted, but the

isolate of N. farcinica was retained as the type strain for that particular species, as not all

were convinced that the two species were truly identical.

Nocardiopsis

The genus Nocardiopsis was originally described by Meyer to accommodate an organism that

at the time was called Actinomadura dassonvillei but differed chemotaxonomically and in

certain colonial morphologic features from other organisms in the

genus Actinomadura (147). The cell wall of this organism does not contain madurose. The

substrate mycelium fragments into coccal forms, and the aerial hyphae fragment into

variable-sized spores (147). Currently, there are approximately 41 validly named species or

subspecies in the genus.

Pseudonocardia

Pseudonocardia are characterized by the microscopic morphology of their aerial hyphae that

are segmented as a result of elongation by budding. Aerial and substrate hyphae are often

zigzag shaped (91). Currently, there are approximately 31 validly published species names in

the genus.

Rhodococcus

The microscopic morphology of rhodococci can range from coccoid to bacillary depending on

species and specimen type and on the stage of growth of the organism (52). The organisms

exhibit a rod-coccus growth cycle. Rod forms are best visualized by growing isolates in a

liquid medium, and under such conditions some branching of individual cells may be found.

Generally, however, the organisms appear as gram- positive, beaded to solidly staining

coccobacilli (Fig. 3H). The modified acid-fast stain must be performed and interpreted with

particular care when dealing with isolates that may belong to this genus, as only a tiny

fraction of the cells may retain the stain (Fig. 3I). Modified acid-fast smears prepared from

isolates growing on trypticase soy agar with 5% sheep blood or chocolate agar may appear

to be acid-fast negative (60).Rhodococcus species can easily be dismissed as “diphtheroids”

because of their Gram stain morphology (Fig. 3H) and their frequent failure to develop

obvious pigmentation during the first few days of growth. At 37°C, colonies may be only

about 1 mm in diameter after 24 h of incubation. An aerial mycelium is generally not

macroscopically visible but may occasionally be seen microscopically (72). Rhodococcus has

been the subject of several recent reviews (13, 74, 83). Some data suggest that it may be

justifiable to separate the genus into several additional genera (83). There are approximately

34 validly named species in the genus.

Saccharomonospora

The approximately eight validly described species of the genus Saccharomonospora are

characterized by the presence of single spores tightly packed on aerial hyphae. The

organisms are commonly found in soil, lake sediment, peat, compost, and manure. They are

moderately thermophilic, with optimum growth occurring at 35 to 50°C (138).

Saccharopolyspora

Species in the genus Saccharopolyspora were initially thought to be related

to Nocardia and Streptomyces by biochemical characteristics but are phylogenetically distinct

from those genera. The organism was originally isolated from sugar cane and has a

microscopic morphology similar to species in the genus Nocardiopsis. It is so named because

of the presence of beadlike chains of sheath-enclosed spores formed by segmentation of the

aerial hyphae. The organism grows in temperatures between 25 and 50°C, with optimum

growth at 37 to 40°C (119). There are approximately 16 validly described species in the

genus.

Segniliparus

The presence of unique high-pressure liquid chromatography mycolic acid patterns in 4

clinical isolates thought to belong to the genus Mycobacterium initiated an investigation into

the characteristics of this recently described genus. By 16S rRNA gene sequence, the genus

is most closely related to members of the genusRhodococcus, and cell wall chemistry places

it in the suborder Corynebacterineae with Rhodococcus and other related genera

(31). Segniliparus is the only aerobic actinomycete (besides Mycobacterium) that is strongly

acid-fast (Fig. 2M and 3M and N).

Streptomyces

The genus- and species-level taxonomy of this huge group, which contains approximately

597 validly named species or subspecies, remains problematic (3, 141). Many of these

species have been patented because of the commercially useful products they synthesize (3)

(Fig. 2C).

Thermoactinomycetes

The approximately eight species of the genus Thermoactinomycetes are characterized by

their production of abundant endospores that are resistant to heat and easily become

airborne. Organisms are found in soil, moldy and decaying plant materials, and composts. By

16S rRNA gene sequence and G+C content, members of this genus are more closely related

to the genus Bacillus than they are to the other genera considered aerobic actinomycetes,

but they are considered with this group because of their similar morphologic features.

Optimum growth temperatures vary by species, but most species grow between 35 and 58°C

(120).

Tsukamurella

The genus Tsukamurella was created to accommodate organisms with a specific cell wall

chemistry that separated it from other aerobic actinomycetes (41). The type species of the

genus, T. paurometabola(corrected from T. paurometabolum), was previously known

as Corynebacterium paurometabolum. There are approximately 11 validly named species in

the genus. The taxonomy of the genus remains confusing; see below for additional details

(Fig. 2J and K and 3J).

Williamsia

The genus Williamsia was recently established to include environmental organisms that

resembled those in genera belonging to the family Nocardiaceae but which had an unusual

cell morphology. When examined by electron microscopy, cells of Williamsia show hairy

structures distributed over the whole surface of the cell. These structures are not visible in

negative-stained preparations (105). The genus contains approximately five species (Fig.

2N and O and 3O).

EPIDEMIOLOGY AND TRANSMISSION Back to top

While the aerobic actinomycetes are widely distributed in the environment, the extent to

which particular species are geographically restricted is not well known. Their primary

ecological niche is probably the decomposition of plant material (141).

The majority of infections caused by aerobic actinomycetes stem from environmental

sources, and even most nosocomial infections appear attributable to an environmental

source such as dust from construction work (18). There has not yet been a documented case

of direct patient-to-patient spread of an aerobic actinomycete infection without the

intermediation of another human agent or of environmental contamination.

Most of the reports of outbreaks and pseudo-outbreaks of infection have been attributed

to Nocardia species. For a brief discussion of these outbreaks, see the Manual of

Clinical Microbiology, 9th ed. (48)

A seven-patient outbreak of Gordonia bronchialis sternal wound infections was traced to a

nurse from whose hands and other body sites the organism was isolated; it was also isolated

from two of the nurse’s dogs (176).

Two pseudo-outbreaks of infection with N. asteroides associated with the use of the BACTEC

460 TB system and inadequate needle sterilization have been reported (132, 161).

An outbreak of pseudoinfection attributed to T. paurometabola, involving specimens from 10

different patients and attributed to a common source somewhere in the laboratory, has also

been reported (8).

CLINICAL SIGNIFICANCE Back to top

While the number of recognized species of aerobic actinomycetes is rapidly increasing,

assessment of the clinical significance of many species is becoming increasingly difficult. In

some cases, a new species is described on the basis of a single isolate, and while careful

attention is paid to its molecular features, little or no information may be provided to

document that the isolate was actually a cause of disease in the patient from whom it was

isolated. On the other hand, reports continue to be published regarding organisms that,

while unquestionably the cause of disease, have been misidentified because of failure to use

molecular methods; in other cases, even when molecular methods have been used, the

accuracy of the reported identifications may be in doubt because of the provision of

inadequate detail regarding the molecular technique used and the methods of interpretation

used. Furthermore, many identifications reported in the older literature would be considered

incorrect by currently accepted species criteria; perhaps the most glaring example is that

of Nocardia asteroides, which to the best of the authors’ knowledge, has never been

documented to be a human pathogen using currently accepted taxonomy.

Unfortunately, all these factors hinder the accurate association of clinically useful information

with a given species. There is perhaps a growing tendency to assemble species into “groups”

on the basis of certain features such as antimicrobial susceptibility, as was originally done by

Wallace et al. for what were considered antibiogram types of N. asteroides but which are now

all divided into different species. While currently “species” are delineated and identified

primarily, if not entirely, on the basis of molecular features, possibly a separate but more

relevant clinically based taxonomy may emerge, based not only on molecular features but on

such aspects as pathogenicity, antimicrobial susceptibility, and type of disease produced—all

of which presumably do have a basis in the fundamental molecular features of the organism.

Tables 3 and 4 list species of aerobic actinomycetes other than those in the

genus Nocardia and those in the genus Nocardia, respectively, that are considered relatively

frequent human pathogens or, in a few cases, are clearly established pathogens but probably

have been previously unrecognized as separate species. Tables 5and 6 list species of aerobic

actinomycetes other than those in the genus Nocardia and those in the genusNocardia,

respectively, that have only rarely (often only once) been reported as human isolates, whose

actual causative role in human disease does not seem well established or for which there are

no recent reports.Tables 5 and 6 should by no means be considered exhaustive. Whenever a

laboratory isolates an aerobic actinomycete for which sequencing indicates that the organism

belongs to a species that has been rarely isolated, the circumstances surrounding the

isolation of the organism should be carefully evaluated to assess its potential clinical

significance. A search of the current literature may be conducted to determine what is

already established regarding that species, including its known pathogenicity and antibiotic

susceptibilities. For such searches, LPSN and PubMed (www.pubmed.gov) may be good

starting points; current nomenclature and taxonomic standing should be followed as stated

in LPSN.

Actinomadura

A useful clue to the identification of organisms in the genus Actinomadura is the nature of

the lesion from which an isolate originates. Most commonly, this organism causes a

mycetoma, a chronic, invasive, slowly progressive infection usually occurring in the foot

(Madura foot) and nearby anatomic structures and nearly always resulting from traumatic

implantation of the organism. Draining sinuses are typically present in a mycetoma;

macroscopically visible grains (organism aggregates or microcolonies) may be visible in the

discharge from the lesions. There are not adequate data available to allow using grain color

for species or even genus level discrimination of the etiologic agent. The infection occurs

most frequently in tropical regions, where people are more likely to walk barefoot. The word

“mycetoma” is used only to describe the clinical nature of the infection, not the etiologic

agent. Actinomycotic mycetomas are caused by aerobic actinomycetes; eumycotic

mycetomas are caused by true fungi. Organisms in this genus have very rarely been

implicated in other types of infection. Actinomadura madurae and A. pelletieri have been the

two species most frequently reported as pathogens. In many reports mentioning the species

causing mycetomas, identification has been made solely on the basis of the histological

appearance of the grains, on the basis of a small number of phenotypic tests, or in ways not

specified in any detail (34, 113, 133, 134). As with virtually all the aerobic actinomycetes,

molecular methods are the only accurate procedures to use for definitive species level

identification of organisms in this genus.

Amycolata and Amycolatopsis

There are no recent reports documenting human infection caused by species in either the

genus Amycolata orAmycolatopsis. Three species in the genus Amycolatopsis and several

other aerobic actinomycete species may be causative agents of equine placental infection

and abortion (118).

Dermatophilus

D. congolensis causes dermatitis in a wide variety of animals worldwide, including cattle,

horses, goats, and sheep, but has only rarely been noted as a cause of human infection

(198). In humans, the organism has been reported to cause a variety of cutaneous

manifestations, including scaling and exudative lesions, pustules, pitted keratolysis (69), and

hairy leukoplakia of the tongue (30). Filamentous and coccoid forms of the organism may be

visualized directly in tissue specimens. Optimal therapy has not been defined, but infections

may be self-limiting. Dermatophilus chelonae has been reported to cause disease in several

species of animals (137, 217).

Dietzia

Dietzia maris is the species in this genus most frequently associated with human infection

(Table 5). In one reported case, the organism was isolated from blood and from an

intravascular catheter. Biochemical and chemotaxonomic studies of the isolate were

performed (14), but the isolate was noted to be modified Kinyoun stain negative, although

from the description of the species, D. maris would be expected to be modified acid-fast

positive. The other reported case involved infection of a hip prosthesis; a variety of

identification techniques were employed, including 16S rRNA sequencing. The closest match

was with the D. maris type strain with a sequence similarity of only 98% (166). The patient

responded to treatment with teicoplanin. A case of aortic dissection associated with this

organism has also been reported (175).

Gordonia

There have been relatively few reports of infections attributed to species in the

genus Gordonia. However, an unknown number of Gordonia infections may be missed, either

because the isolate is considered an insignificant coryneform gram-positive rod or the isolate

is misidentified as belonging to another genus such as Nocardia or Rhodococcus (17). In 5

patients with catheter-related Gordonia species infection, the organism was correctly

identified to the genus level in only one case until 16S rRNA gene sequencing was employed

(17). In three of these five patients, Gordonia terrae was the infecting organism, one was G.

bronchialis, and one was G. otitidis. Recently, there has been a review published on medical

device-associated Gordoniainfections in connection with a report of infection of an orthopedic

device caused by G. araii; most of the infections reviewed were attributed to G. terrae (99).

One of the very few clusters of infection attributable to any aerobic actinomycete involved

seven patients who developed sternal wound infections following coronary artery bypass

surgery (176). The organism involved, Gordonia bronchialis (then called Rhodococcus

bronchialis), was identified by biochemical testing and cell wall mycolic acid analysis only.

Immunocompromise and/or the presence of foreign bodies appear to have been contributing

factors in many of the infections caused byGordonia species.

Nocardia

Nocardia infections generally result either from trauma-related introduction of the organism

or, particularly in immunocompromised patients, from inhalation and the resulting

establishment of a pulmonary focus. The brain is one of the most common secondary sites of

infection (158). An initial advance in the clinically useful categorization of pathogenic

nocardial isolates was provided by Wallace and his coworkers (211). They divided organisms

phenotypically resembling N. asteroides into six different drug pattern types and one

additional miscellaneous group. With more recent work, numerous different species have

been described within this set of organisms, which came to be known as the Nocardia

asteroides complex. These include N.abscessus (drug pattern I), N. cyriacigeorgica (drug

pattern VI), N. farcinica (drug pattern V), N. nova (drug pattern III), N. wallacei (drug

pattern IV), and isolates of drug pattern II. Other new Nocardia species are continually being

described, and undoubtedly, by current species definition criteria, many more will be

described in the future.

Given the current state of the literature, it is possible to make only the most tentative

statements regarding geographic distribution and disease correlates of different species.

While accurate species assignment allows one to make some predictions regarding likely

antimicrobial susceptibility patterns, such a species assignment today requires the use of

molecular methods (38). Nonetheless, susceptibility testing of all clinically significant isolates

is recommended, whether or not molecular techniques have been used for their

identification.

Nocardia abscessus

Nocardia abscessus was formally named in 2000 by Yassin et al. (228). ATCC strain 23824,

the reference strain of N. asteroides drug pattern type I (195), was one of the isolates found

to be N. abscessus on the basis of 16S rRNA sequencing and DNA-DNA hybridization. In their

description in 1988 of the antimicrobial susceptibility patterns of 78 clinical isolates

of Nocardia from various sources, Wallace et al. noted that 20% of the isolates had this drug

pattern type (211). Several of the strains in the report naming the species were from

abscesses; a subsequent report from Japan reported several isolates from pulmonary

sources and one from a brain abscess (103). Two isolates have been reported from

Germany, one from pericardial fluid (218) and the other from a posttraumatic wound (94).

More recent cases have included a report of disseminated infection in an AIDS patient (56).

Nocardia asteroides and the “N. asteroides Complex”

There are innumerable reports of human infection attributed to N. asteroides (48); it has

probably been considered the most commonly isolated human pathogenic Nocardia species.

However, molecular analyses of the pathogenic isolates attributed to this species that have

been conducted thus far have indicated that all belong to some other named or as-yetunnamed

species. It is currently believed that N. asteroides sensu stricto is rarely, if ever,

pathogenic.

The term “N. asteroides complex” has been used for organisms phenotypically resembling

the N. asteroidestype strain. However, several distinct species have now been described

within that complex, and precisely what the complex is intended to designate is usually

unclear. Given the current ability to identify Nocardiaisolates molecularly, the phrase

“N. asteroides complex” should be avoided. The term “complex” is best restricted to groups

of species that are related on a molecular basis and have similar phenotypic features. Such

groups would include the “N. nova complex” and the “N. transvalensis complex.” Whenever

the term “complex” is used, it is best to state initially precisely which species are intended to

be included in the complex.

Nocardia brasiliensis

N. brasiliensis appears to be the most common cause of actinomycotic mycetoma (see

“Actinomadura” above) in the Western Hemisphere, especially in Mexico (33, 34). The

organism probably occurs worldwide; there are reports of infection from Australia (68), West

Bengal (134), and Europe (135) and many from North America (189). A variety of cutaneous

manifestations in addition to mycetoma have been reported, including cellulites, abscesses,

and lymphocutaneous infection. Nearly all cases are a result of trauma, including that caused

by thorns (135), cat scratch (21), and insect bite (159). Most of the trauma-related

infections have occurred in immunocompetent individuals. Disseminated infection, usually

originating from a pulmonary focus, has also been reported (189); such infections are more

likely to occur in immunocompromised patients (115). Some cases, such as one of a brain

abscess resulting from dissemination from a pulmonary focus (154), do occur in patients who

appear to be immunocompetent. However, most of the cases of invasive disease, as well as

some of the cases of cutaneous infection, attributed to N. brasiliensis (prior to 1995) almost

certainly have been caused by N. pseudobrasiliensis (see below).

Nocardia brevicatena

In 1982, Goodfellow and Pirouz examined 108 phenotypic characteristics of

sporoactinomycetes that contained meso-diaminopimelic acid in their cell walls. Results

showed that one species they examined,Micropolyspora brevicatena, shared many

characteristics with Nocardia isolates included in the study. The authors recommended that

this organism be transferred to the genus Nocardia as Nocardia brevicatena (76). Brown et

al. suggested in 1997 that Nocardia isolates sharing both an unusual drug susceptibility

pattern and one or another of three different RFLP patterns of an amplified portion of the 65-

kDa heat shock protein (HSP) gene should be considered to form the Nocardia

brevicatena complex (B. A. Brown, R. W. Wilson, V. A. Steingrube, Z. Blacklock, and R. J.

Wallace, Jr., Abstr. 97th Gen. Meet. Am. Soc. Microbiol. 1997, abstr. C-65, p. 131, 1997).

These organisms included 19 clinical isolates from the United States, 10 clinical isolates from

Australia, and three ATCC reference strains of N. brevicatena. There have been no

subsequent reports of clinical isolates or taxonomic studies of these organisms. Given the

different restriction fragment length polymorphism (RFLP) patterns involved, several

different species may be included in this complex, one of which may be Nocardia

paucivorans (see below).

Nocardia cyriacigeorgica

N. cyriacigeorgica (spelling corrected from the original N. cyriacigeorgici) was described on

the basis of an isolate obtained from the sputum of a patient with chronic bronchitis (229).

Subsequent isolates were obtained from brain abscesses in an immunocompromised patient

(66) and from a patient with pneumonia following a near-drowning incident, from whom N.

farcinica and several other pathogens were also isolated (203). Isolates of this species were

reported to constitute 13 of 96 (14%) clinical isolates of Nocardia from Thailand (169) and

13 of 86 (15%) such isolates from Belgium (215). The 16S rRNA sequence of the type strain

of this species (1,400 bp) was found to be identical to that of the reference strain of “N.

asteroides drug pattern type VI” (ATCC 14759) (179, 195). DNA-DNA hybridization studies

subsequently established that this drug pattern type VI reference strain and N.

cyriacigeorgica belong to the same species (49). In the original work describing the

different N. asteroides drug pattern types, type VI was the most commonly isolated strain

(35%) (211). N. cyriacigeorgica is probably the most frequent human nocardial pathogen, at

least in areas where actinomycotic mycetomas are relatively rare. While the species N.

cyriacigeorgica has been designated an emerging pathogen (183), it is actually a relatively

common and long-recognized pathogen that has recently acquired a valid name. In fact,

many isolates previously reported as “N. asteroides” almost certainly belong to this species.

Note that the designation “N. asteroides drug pattern type VI” is not a valid species name.

Nocardia farcinica

The organisms initially described by Wallace et al. as belonging to the group “N.

asteroides drug pattern type V” (211) were subsequently found to belong to N.

farcinica (212). Isolates of this species are perhaps the most resistant of all Nocardia isolates

(Table 7). N. farcinica isolates that showed in vitro resistance to trimethoprimsulfamethoxazole

have been reported to respond to meropenem alone (92) and to the

combination of linezolid and minocycline (130). This species may have a particular

propensity for causing disseminated disease; Wallace et al. reported that among 30 patients

for which disease extent was known, 57% had disseminated disease and one-third had

central nervous system (CNS) involvement (212). Most patients infected by this species,

especially those with disseminated disease, have some type of immunocompromise (212),

but cutaneous and other infections have been reported to occur in the apparently

immunocompetent as well (182). The lung is a common site of involvement, affecting 43%

of patients according to one review (196). As verified by molecular analysis, N. farcinica has

been isolated from brain abscesses (122, 148, 190), blood (35, 122), cases of keratitis (57),

and an infected cochlear implant (123). N. farcinica has also been isolated from the

bronchoalveolar wash fluid of a cystic fibrosis patient (163). Isolates of this species were

reported to make up 34 of 96 (35%) clinical isolates of Nocardia from Thailand (169) and 38

of 86 (44%) such isolates from Belgium (215).

Nocardia nova

N. nova was described by Tsukamura in 1982 and was distinguished from N. asteroides by

several phenotypic tests (199) and confirmed to be a separate and distinct species by DNADNA

hybridization (224). Wallace et al. found that 18% of 78 clinical isolates fell into their

type III drug susceptibility pattern, characterized by susceptibility to ampicillin and

erythromycin and resistance to carbenicillin (211). In a subsequent study of 223 clinical

isolates, employing both biochemical and susceptibility testing procedures, 17% of the

isolates, as well as the type strain of N. nova, had similar characteristics, including the type

III drug pattern; these isolates were all then considered members of N. nova (210). Of the

patients for whom clinical information was available, 35% were thought to have

disseminated disease; organisms were obtained from many sites, including blood, lung, CNS,

skin and soft tissue, joints, and cornea. An upper extremity sporotrichoid form of nocardiosis

attributed to N. nova in a human immunodeficiency virus (HIV)-positive patient who

sustained a thumb injury while working in a field has been reported (96). Hamad et al.

reported the isolation of N. novaidentified by 16S rRNA gene sequencing from a case of

spondylodiscitis and psoas abscess (86).

Recent investigations have revealed that the organisms identified as N. nova by phenotypic

testing, including antibiogram, as well as by the RFLP patterns obtainable from the hsp65

gene, actually may belong to other species in addition to N. nova, including N. africana, N.

kruczakiae, and N. veterana (43). These species can be distinguished from one another and

from N. nova sensu stricto only by gene sequencing. Phylogenetic analysis using sequence

data of the HSP and secA1 genes consistently place these species in the same clade as N.

nova sensu strict, indicating the close relationship among these species (P. S. Conville and F.

G. Witebsky, unpublished data). Phylogenetic analysis of the 16S rRNA, HSP,

and secA1 genes also place Nocardia aobensisand N. elegans in the N. nova complex clade;

no phenotypic characteristics of these two species have been examined to determine their

similarity to other species in the complex. All species included in the N. novacomplex

(including N. aobensis and N. elegans) have been isolated from humans or implicated in

human disease. Isolates identified only by phenotypic testing as belonging to one or another

of these species are probably best reported as members of the “N. nova complex.”

Nocardia otitidiscaviarum

The initial isolate of the species N. otitidiscaviarum, described by Snijders in 1924, was

obtained from the infected middle ear of a guinea pig; for a time this species was known

as N. caviae (78). This species is a relatively infrequent cause of human infection. Clark et al.

reviewed 28 cases of cutaneous infection, including several of mycetoma, attributed to this

species; many of those for which information was available were considered trauma related

(37). There are a few reports of infection at other sites, including brain abscess (90),

pyothorax (231), catheter-related infection (129), and disseminated infection

(162, 164, 186). This species is relatively reliably identified on the basis of its decomposition

reactions; it decomposes xanthine and hypoxanthine but not casein or tyrosine. In a recent

molecular study of numerous clinical isolates, several different sequences were obtained

from nine different strains that had been phenotypically identified as N.

otitidiscaviarum (160). These results were interpreted as suggesting the presence of several

different species within this group, but some of the results also suggested that individual

isolates might contain two or more differing copies of the 16S rRNA gene.

Nocardia paucivorans

The species N. paucivorans was described by Yassin et al. in 2000 on the basis of a

respiratory isolate from a patient with chronic lung disease (227). By 16S rRNA gene

sequence, this organism showed 99.6% sequence similarity to N. brevicatena; results of

DNA-DNA hybridization indicated that they were distinct species. N. paucivorans and N.

brevicatena, along with isolates belonging to the unnamed group N. asteroides drug pattern

II (211), are sometimes considered to belong to the “N. paucivorans/N.

brevicatena complex” based on similar phenotypic characteristics and antibiograms (28).

N. paucivorans has been recovered from cerebrospinal fluid in a case of cerebral nocardiosis

in an immunocompromised individual (58), an intracerebral abscess in an immunocompetent

patient (112), and a mitral valve (24). Two of 86 (2.3%) clinical isolates from Belgium were

identified as belonging to this species (215). Gray at al. reported that in a retrospective

study of Nocardia isolates recovered from Australian patients over a 20-year period, 32

isolates were identified as N. paucivorans by 16S rRNA gene sequence. No indication of the

identification criteria used for species assignment was presented. These isolates were

recovered from a variety of sources including skin, lung, blood, brain, pleural fluid, and

lymph node (81).

Nocardia pseudobrasiliensis

In 1995, Wallace et al. (209) reported on a subset of organisms that had been identified

as N. brasiliensis that were sometimes isolated from cutaneous infections but were

predominantly associated with noncutaneous invasive disease. This subset of isolates

appeared to belong to another taxon, formally named N. pseudobrasiliensis the following

year (180). Most infections have occurred in immunocompromised patients; cases have been

reported from North and South America, Japan, and Australia (25, 101, 209).

Unlike true N. brasiliensis isolates, most N. pseudobrasiliensis isolates hydrolyze adenine, are

susceptible to ciprofloxacin and clarithromycin, and are resistant to minocycline. The two

species also differ in their mycolic acid patterns and 16S rRNA, HSP, and secA1 gene

sequences. Most cases of infectious disease of a nonmycetomatous nature previously

attributed to N. brasiliensis have probably been caused by this species (see “Nocardia

brasiliensis” above).

Nocardia transvalensis and the N. transvalensis Complex

N. transvalensis was originally described in 1927 on the basis of an isolate from an African

patient with a mycetoma (168). In 1997, Wilson et al. published a study of 58 clinical

isolates from the United States and Australia that showed resistance to amikacin (220).

Using biochemical and molecular methods, these authors were able to separate the isolates

into four distinct groups. One group comprised 53% of the isolates studied, were all

recovered from patients in the United States, and were similar to isolates classified as N.

asteroidesdrug pattern type IV as defined by Wallace et al. (211). Isolates belonging to this

group were later officially designated Nocardia wallacei (42) (see below). Another group

(17% of isolates) included organisms similar to the type strain of N. transvalensis. Two

additional groups, new taxon 1 (14% of isolates) and new taxon 2 (16% of isolates), were

also defined. Australian isolates belonged to the N. transvalensis sensu stricto group and to

new taxon 1. Isolates belonging to new taxon 1 have been given the species designation N.

blacklockiae (42). Because of their similar phenotypic and molecular characteristics, Wilson

proposed that these three species and the unnamed new taxon 2 be considered to belong to

the “N. transvalensis complex.”

There have been numerous reports in the literature describing the isolation

of N. transvalensis from clinical specimens; many of these reports base the identification of

the isolate on a small number of phenotypic tests. It is unclear if these reports represent

accurate species assignments, given the similarity of related species as described above.

McNeil et al. reported on 16 patients from whom isolates attributed to this species had been

obtained (142). The isolates from 10 of the patients were considered clinically significant;

the other isolates were considered colonizers or of uncertain significance. A variety of sites of

infection were involved, some of the infections were disseminated, and several patients were

known to be immunocompromised. There are a few other case reports of molecularly

characterized disseminated infection attributed to the N. transvalensiscomplex, including a

brain abscess in a cancer patient (230), brain and cutaneous infection in a heart transplant

patient (131), and pulmonary infection and subcutaneous abscess in a patient with

histiocytosis X (2). Three clinical isolates assigned to this species complex have been

reported from Japan (102).

Nocardia veterana

N. veterana was first described on the basis of an isolate obtained from the bronchial lavage

fluid of a patient with pulmonary lesions, but that isolate was thought not to be clinically

significant (84). Subsequently, N. veterana has been implicated as the causative agent in

cases of mycetomas (108), three patients with pulmonary disease (170), ascitic fluid

infection in an HIV patient (70), and bloodstream infection in a cancer patient (4). In a

report of three additional clinical isolates, two of which were shown to have been causative

agents of pulmonary disease, it was noted that the isolates had an antimicrobial

susceptibility pattern essentially identical to the patterns of N. africana and N. nova (44),

demonstrating that species identification could not be definitively established by

susceptibility testing alone. It was also noted that the isolates of this species that were

studied showed 16S rRNA gene similarities to the type strains of N. africana and N. nova of

99.0 and 97.7%, respectively. N. veterana is one of the species included in the “N.

nova complex” (see“Nocardia nova” above). HSP, secA1, and 16S rRNA gene sequencing

place N. veterana in a clade with other members of the complex.

N. wallacei

Officially described in 2008, N. wallacei is the most commonly isolated member of the N.

transvalensis complex (42). 16S rRNA, HSP, and secA1 gene sequencing showed ≥99.8%

sequence similarity among five clinical isolates obtained from sputum samples; one of these

was also recovered from pleural fluid. This species was initially designated as N.

asteroides drug pattern IV by Wallace et al. and comprised 5% of 78 isolates identified as “N.

asteroides” by phenotypic methods but unique in their resistance to amikacin (211).

Nocardiopsis

Nearly all of the very few infections attributed to organisms in the genus Nocardiopsis have

been attributed toN. dassonvillei. In a letter to the editor regarding a case of

actinomycetoma attributed to this species, it was stated that N. dassonvillei was “regularly

encountered” at the Centers for Disease Control and Prevention, with 21 isolates having been

identified from 1981 through 1986; no clinical details were provided (1). In the case of a

blood isolate, a variety of methods, including 16S rRNA gene sequencing, was used to

identify the isolate; the same report provided references to a number of other cases of

infection, including mycetomas and cutaneous infections, attributed to this species (11).

Rhodococcus

The most commonly isolated pathogen in the genus Rhodococcus is Rhodococcus equi. While

other species have occasionally been reported to cause disease, in most but not all of these

reports, gene sequencing was not used, and it is impossible to obtain reliable species-level

identifications without this technique (13, 74). There is some evidence that R. equi itself may

be a complex of several different species (139). Infections caused by R. equi have been the

subject of several reviews (6, 52, 110, 197). The organism has long been known as a

significant pulmonary pathogen in horses (hence the species name) (171). The initial report

of human disease caused by the organism (reported as Corynebacterium equi) involved a

patient on high-dose steroids with a pulmonary abscess; he had worked briefly in stockyards

contaminated with animal feces shortly before the onset of his illness (71). Herbivore

manure provides an ideal growth medium for the organism, and inhalation of the organism is

presumed to be the major mode of infection in horses (171) and is probably also the

principal mechanism for human infections. Direct inoculation and oral ingestion are other

possible routes of infection (216). Most patients who develop R. equi infection are

immunocompromised, and at least until recently, approximately two-thirds of infected

patients were also HIV infected (216). Essentially any body site can be involved, but the lung

is a site of involvement in approximately 80% of immunocompromised patients and at least

40% of immunocompetent patients. Bacteremia has been reported in >80% of

immunocompromised patients and approximately 30% of immunocompetent patients (216).

Pulmonary cavitation is a frequent finding. Malacoplakia, an aggregation of histiocytes

containing concentrically layered basophilic structures known as Michaelis-Gutmann bodies,

has been noted as a histopathologic finding in several studies (82, 197). Two virulenceassociated

antigens (VapA and VapB) encoded by plasmids have been identified. Isolates

producing the VapA antigen are the predominant, possibly the sole, cause of disease in

horses, but isolates producing the VapA antigen, the VapB antigen, or neither antigen have

been isolated from human cases of infection (156). A combination of antimicrobials is

generally used for the treatment of infections. Agents used include aminoglycosides,

erythromycin, imipenem, quinolones, rifampin, and vancomycin. Linezolid may also be

efficacious (216); in an in vitro study of 102 R. equi isolates obtained from humans and

animals, the linezolid MIC ranged from 0.5 to 2.0 μg/ml (23). Some rifampin-resistant

strains with rpoB gene mutations have been reported (7). Without careful assessment, R.

equi isolates can easily be mistaken for coryneform bacteria and dismissed as insignificant

(Fig. 2H). A recent report describes two patients who died of overwhelming R. equi infection;

for both patients, isolates from multiple positive blood cultures were initially misidentified as

a Corynebacterium species (201).

Segniliparus

Although the two species that make up this genus (S. rotundus and S. rugosus) were

originally recovered from clinical material, no information was provided concerning the

significance of these isolates except that they were from “nonsterile human sources” (31).

Published reports of the recovery of S. rugosus seem to indicate the predilection of this

organism for patients with cystic fibrosis. A report of the recovery of S. rugosus from the

bronchoalveolar lavage fluid or induced sputum of three cystic fibrosis patients emphasized

the microscopic and colonial similarities of this species with those of rapidly growing

mycobacteria (Fig. 2M and 3M and N). All 3 isolates were initially identified phenotypically as

either Mycobacterium abscessus or Mycobacterium chelonae; sequence analysis of the first

500 bases of the16S rRNA gene of two of the isolates (from each of two siblings) showed

100% sequence similarity to the type strain of S. rugosus (32). Drug susceptibility testing on

these isolates was problematic because of insufficient growth in cation- adjusted Mueller-

Hinton broth; Middlebrook 7H9 broth was necessary to achieve adequate growth. Using

breakpoints recommended forNocardia and for rapidly growing mycobacteria (38), results

indicated that the S. rugosus isolates were susceptible to imipenem, rifabutin,

sulfamethoxazole, and trimethoprim-sulfamethoxazole and resistant or intermediate to

amikacin, amoxicillin-clavulanate, ceftriaxone, ciprofloxacin, clarithromycin, linezolid,

minocycline, and tobramycin. Using bacterial breakpoints, the isolates were susceptible to

moxifloxacin. In a subsequent report from Australia, S. rugosus was isolated from the

mycobacterial culture of sputum from a cystic fibrosis patient; the sputum had undergone a

standard mycobacterial decontamination process. PCR of a region of the 16S rRNA gene used

to identify mycobacteria was negative, but sequence analysis of a 1,250-bp region of the

16S rRNA gene showed 100% similarity to the type strain of S. rugosus (87).

Streptomyces

The most common type of infection attributed to species in the genus Streptomyces is

mycetoma (see“Actinomadura” above), and the most commonly mentioned etiologic agent

is Streptomyces somaliensis, although in many reports the identification procedures

employed could not have ensured that other species were not involved. There are a few

reports implicating other species in the genus as occasional pathogens. In case reports of

bacteremias attributed to Streptomyces bikiniensis (150) and Streptomyces

thermovulgaris (59), molecular methods were used in identifying the isolates, and such

methods were described in detail in a case report of a mycetoma attributed to Streptomyces

albus (136). While the majority of isolates from nonmycetomatous lesions probably

represent either contamination or colonization, these organisms are capable of occasionally

causing disease other than mycetoma. Recently, six cases of invasive Streptomycesinfections

have been reported, along with a literature review of such infections (107). Because of the

huge number of validly described species of Streptomyces and the lack of information about

clinical significance of many of these species, identification to the genus level is probably

sufficient in most cases. In a study of the susceptibility of 92 Streptomyces species from

clinical specimens, 100% of those tested were found to be susceptible to amikacin and

linezolid, 77% to minocycline, 67% to imipenem, and 51% to clarithromycin and amoxicillinclavulanate

(178).

Tsukamurella

Tsukamurella infections have been most commonly reported in connection with a foreign

body, such as an intravenous catheter (22), but have also been reported even in apparently

immunologically normal patients in the absence of any foreign body (188). The literature on

catheter-related infections caused by Tsukamurellaspecies has recently been reviewed in

connection with two additional reports of such infection (22); several other of the relatively

few infections reportedly caused by Tsukamurella species are also briefly mentioned in that

review. Rhodococcus aurantiacus ATCC 25938 (the initial type strain of the species), which

has had a convoluted taxonomic history, has been placed in the

species T. paurometabola (41). In the older literature, there are a few reports of infection

attributed to R. aurantiacus, such as pulmonary infection, meningitis, peritonitis, and

subcutaneous abscesses (200). Organisms isolated from clinical material have generally been

found to resemble strain ATCC 25938. An outbreak of pseudoinfection (8) was also attributed

to T.paurometabola.

Williamsia

The first clinical isolate of the genus Williamsia was recovered from a protected bronchial

brush sample of a patient with bilateral alveolar infiltrates following aortic valve replacement.

Gram stain of the sample showed numerous gram-positive rods, and culture of the fluid grew

>1,000 CFU/ml of an organism identified by 16S rRNA gene sequencing as Williamsia

muralis (99.9% sequence similarity). Using breakpoints established forNocardia spp.,

susceptibility testing of the W. muralis isolate indicated that the isolate was susceptible to

amoxicillin-clavulanate, cefotaxime, imipenem, ciprofloxacin, tobramycin, gentamicin, and

trimethoprim-sulfamethoxazole (55). The recovery of W. muralis was also reported from a

case of endophthalmitis. The isolate showed only 98.4% 16S rRNA gene similarity, but 100%

DNA-DNA hybridization, with the type strain ofW. muralis (153). Two isolates

of W. deligens have been reported from human blood, but no clinical information was

provided (226).

Other Genera

Some other genera of aerobic actinomycetes do have roles in human diseases but are

unlikely to be encountered in the clinical laboratory. Organisms in three genera of

thermophilic aerobic actinomycetes,Saccharomonospora, Saccharopolyspora,

and Thermoactinomyces, have been implicated in the etiology of hypersensitivity

pneumonitis. Spores of these organisms may be encountered when performing air sampling;

molecular methods are helpful for species-level identification when needed (89, 222).

COLLECTION, TRANSPORT, AND STORAGE Back to top

In temperate climates, the respiratory tract is the most frequent portal of entry for the

aerobic actinomycetes and therefore the primary site of nocardial infections in the

immunocompromised host. Sputum is the most easily obtained pulmonary specimen, and

examination of several fresh early-morning samples collected on separate days (213) may

maximize the chances of organism recovery; isolation of an aerobic actinomycete from

multiple samples may help to establish the clinical significance of the isolated organism

(73, 117).Nocardia species may, however, be difficult to recover from sputum even in

documented cases of pulmonary infection (158), either because of low numbers of organisms

present in the sample or because contaminating bacteria in the sample may overgrow the

more slowly growing aerobic actinomycetes. More invasive procedures, such as

bronchoalveolar lavage or fine needle or open lung biopsy, may be required to obtain a

definitive diagnosis (158). These more invasive procedures may be necessary for diagnosis

of as many as 44% of primary pulmonary infections (68); macrophage-rich samples may be

necessary to maximize recovery of organisms such as rhodococci, which tend to localize

within these cells (52).

Exudates from abscesses or mycetomas should be delivered to the laboratory in a sterile

container for macroscopic and microscopic examination for characteristic granules and for

smear and culture. In the case of disseminated cutaneous lesions or small lesions secondary

to trauma, a skin biopsy can be useful. The use of swabs is not recommended, as fibers can

make smear interpretation difficult (213). To optimize isolation ofDermatophilus, scabs or

crusty lesions should be removed and soaked in sterile distilled water, and the fluid should

be inoculated onto blood agar plates (232).

In immunocompromised patients, the aerobic actinomycetes, especially Nocardia, can

disseminate to almost any organ. A biopsy sample or aspirate, when obtainable, may be the

best specimen to evaluate for the presence of such organisms. Normally sterile body fluids

should be collected in a sterile container and sent immediately to the

laboratory. Nocardia isolates are infrequently recovered from cerebrospinal fluid, even if

numerous brain lesions are present, because the organisms may be confined to the brain

abscess itself. Blood should be inoculated directly into blood culture media; many

commercially available blood culture systems as well as lysis centrifugation methodologies

have been shown to support the growth of Nocardia species. Various aerobic actinomycetes

have been implicated as the cause of catheter-related bacteremia; potentially infected

catheter tips should be transported to the laboratory in a sterile container and cultured by

appropriate methods.

In all cases where infection with an aerobic actinomycete is suspected, it is of utmost

importance that the laboratory be notified of the suspected diagnosis. This will ensure that

samples will receive appropriate handling, that the correct direct smears will be prepared,

and that the sample will be inoculated onto the appropriate media and incubated for an

extended period of time (a minimum of 2 weeks, preferably for up to 3 weeks) at the

appropriate temperature.

All samples should be transported promptly to the laboratory following the specimen

collection and handling procedures outlined in chapter 16.

DIRECT EXAMINATION Back to top

Microscopy

Careful microscopic examination of clinical specimens suspected of containing aerobic

actinomycetes is extremely important. The observation of organisms characteristic

of Nocardia and other aerobic actinomycetes should alert laboratory personnel to inoculate

appropriate media and to extend incubation at the appropriate temperature. In addition, the

detection of organisms directly in the patient specimen may assist in the interpretation of

culture results; smears of such specimens may show more diagnostic morphologies of

organisms than smears from colonial growth.

Smears can be made directly from sputum, drainages, and aspirates; however, liquid

specimens such as bronchoalveolar lavage or normally sterile body fluids that are not

excessively cellular should be concentrated before smear preparation and medium

inoculation. For example, a 5-ml aliquot of the sample can be centrifuged for 10 minutes at

2,800 ×g and the pellet can be used for smears and medium inoculation. Alternatively,

smears can be prepared using a cytocentrifuge. For tissue samples, smears can be made

from ground material and from touch preps. Two important stains that should be used in the

clinical laboratory for direct samples are the Gram stain and the modified acid-fast stain.

Histopathologic examination of fixed tissue by use of special stains (including Fite stain and

Grocott-Gomori methenamine silver stain) may also reveal the presence of organisms

belonging to these genera (221).

Gram stain morphologies of the aerobic actinomycetes vary by genus; organisms appear as

gram-positive rods ranging in shape from coccoid to bacillary. Filamentous and branching

forms may be present, depending on the species involved and the stage of growth in infected

tissues (Fig. 3). See Table 1 for a summary of the microscopic morphologies of the aerobic

actinomycetes.

The modified acid-fast stain used on direct specimens may more accurately reflect the true

partially acid-fast nature of the organisms than do modified acid-fast stains prepared from

colonial growth. The modified acid-fast stain uses a weaker decolorizer (1% H2SO4) than

does the mycobacterial stain (3% HCl). See chapters 17and 28 for details on reagent

preparation and staining methodology. Because of the difficulty of standardizing this

technique, it is imperative that positive and negative controls be run simultaneously with

patient smears. The control slides can be made from growing suspensions

of Streptomyces species (negative control) andNocardia species (positive control). Smears

should be evaluated by experienced laboratory personnel, and the quality of the stain itself

should be evaluated before results are reported. Gordonia, Nocardia,

Rhodococcus, andTsukamurella (and possibly Dietzia) are known to be partially acid-fast with

this stain; Segniliparus is strongly acid-fast.

Careful attention should be paid to the cellular material present in the sample (Fig.

3K and L). Nocardia is frequently seen in association with polymorphonuclear leukocytes

(28). Phagocytized gram-positive or acid-fast organisms can sometimes be seen within

macrophages and mononuclear cells; in the modified acid-fast smear these may appear as

“beaded” cells with strongly acid-fast granules within non-acid-fast or weakly acid-fast rods

(52, 157).

In cases of suspected actinomycetoma, aspirated material should be examined grossly for

the presence of granules by spreading the sample in a sterile petri dish. Granules found

should be washed and crushed. Smears should be prepared from the crushed material, and

appropriate media should be inoculated. Granules are most often seen in infections caused

by N. brasiliensis or Actinomyces species but can also be seen in infections caused by other

species of Nocardia (28).

Nucleic Acid Detection

Several authors have reported the use of molecular methodologies for the detection

of Nocardia directly from clinical specimens. Targets include the 65-kDa HSP gene (172), the

16S rRNA gene (53, 208), and the secA1gene (G. Fahle, personal communication). These

techniques have the potential to provide rapid diagnosis of nocardiosis from samples

exhibiting bacterial morphologies suggestive of this genus. Further assessment of these

techniques will be needed to assess their clinical utility.

R. equi chromosomal DNA and vapA plasmid DNA (a gene that codes for, and mediates

expression of, a virulence-associated protein on the cell wall surface) have been directly

detected from tracheal wash fluid and other specimens of potentially infected foals. PCR with

amplicon detection by gel electrophoresis (185) and real-time PCR (88) have been shown to

be highly specific and more sensitive than culture or serologic testing for the diagnosis of R.

equi infection.

ISOLATION PROCEDURES Back to top

Blood agar, chocolate agar, brain heart infusion agar, Sabouraud dextrose agar, and

Lowenstein-Jensen medium support the growth of most aerobic

actinomycetes; Dermatophilus congolensis may not grow on Sabouraud dextrose agar or

Lowenstein-Jensen medium (232). Buffered charcoal yeast extract agar (BCYE) is particularly

useful for the recovery of Nocardia species. Specimens from sterile sites or concentrated

sterile body fluids can be inoculated directly onto these media. Specimens from respiratory

sites, skin, and other potentially contaminated sites, such as mycetomas, should additionally

be inoculated onto selective media, such as modified Thayer-Martin agar (152) and selective

BCYE (containing polymyxin B, anisomycin, and either vancomycin or cefamandole) (206).

Sabouraud agar with added chloramphenicol may not be useful, as it may also suppress the

growth of some Nocardia species (85). A specialized medium for the recovery of R.

equi using a Mueller-Hinton agar-based medium with added ceftazidime and novobiocin has

been described (207).

Cultures for aerobic actinomycetes should be handled as fungal cultures, thus ensuring that

the cultures will be incubated and regularly examined over an extended period. Two BCYE

plates (for samples from sterile sites) or selective BCYE plates (for respiratory or other

potentially contaminated sites) should be inoculated. One BCYE or selective BCYE plate

should be incubated at 30°C and the other at 35°C, both in ambient air. It should be noted,

however, that Streptomyces species may show best growth at 25°C and that Dermatophilus

congolensis grows better and shows enhanced production of aerial hyphae in increased

CO2 (232). For all cultures, plates should be held for a minimum of 2 weeks, preferably for

up to 3 weeks, and should be sealed to prevent dehydration.

A low-pH decontamination procedure has been successfully used for pretreatment of heavily

contaminated specimens suspected of harboring Nocardia species. The sample is diluted 1:10

in 0.2 M HCl–0.2 M KCl at pH 2.2, mixed, and allowed to stand for 3 to 5 minutes, after

which it is inoculated onto selective and nonselective media (20, 206). Murray et al. reported

a drop in viability of Nocardia species after 30-min exposures to N-acetyl-L-cysteine (NALC),

NaOH-NALC, or Zephiran-trisodium phosphate (151). The key to improved recovery

of Nocardia from NaOH-NALC-treated specimens may be a shorter exposure to the

decontaminating reagents (15 min), as is in fact also recommended for recovery of

mycobacteria (111).

Aerobic actinomycetes have been recovered from blood using a variety of commercially

available blood culture systems, including conventional 2-bottle systems, biphasic bottles,

systems using radiometric and nonradiometric detection, and lysis-centrifugation systems.

Using various blood culture systems, aerobic actinomycetes have been recovered after 3 to

19 days of incubation, which in some cases included the incubation times of terminal

subcultures (150, 205). Studies employing newer blood culture systems have resulted in

recommendations that extended incubation is no longer necessary (9); however, such

studies have not established that such shorter incubation would generally be adequate for

the isolation of aerobic actinomycetes. If the possibility of bacteremia with a member of one

of these genera is anticipated, it would be advisable to perform fungal blood cultures (for the

expanded incubation period, at least 3 weeks) or to perform terminal subcultures if the

incubation period of routine blood cultures cannot be extended.

Cultures from all sources should be examined daily for the first week of incubation and then

weekly thereafter, preferably using a dissecting microscope, which will allow detection of tiny

colonies. Such microscopic examination is particularly important for specimens that contain

contaminating flora, as an aerobic actinomycete can be quickly overgrown by more rapidly

growing organisms. Care should also be exercised when Lowenstein-Jensen or Middlebrook

media are examined for mycobacteria or when BCYE plates are examined

for Legionella species, as Nocardia and other aerobic actinomycetes can grow on these

media.

IDENTIFICATION Back to top

Microscopic Morphology

Evaluation of Gram stain morphology of a suspected aerobic actinomycete should be the

initial step in organism identification, as microscopic morphology can vary among the genera

(Table 1 and Fig. 3). A sufficient number of fields should be reviewed to allow determination

of the most prevalent morphology and to detect the sometimes-rare branching forms. Care

should be taken not to confuse perpendicular aligning of the organisms with true branching.

Smears made from colonial growth of filamentous isolates may fragment and appear as

bacillary or coccoid forms (28).

A properly prepared and carefully interpreted modified acid-fast smear can assist in the

preliminary identification of the organism (Fig. 3F, G, I, J, and L). Quality control slides

should be stained along with stains of colonies from cultures. With the modified acid-fast

stain, the background should be blue; slides that have a pink background may be

inadequately decolorized and should be repeated. The smear should be scanned for areas

where individual cells can be seen or areas where single layers of cells allow clear

differentiation of cell borders. The acid-fast reaction of tightly packed clumps of organisms

may not represent the true partially acid-fast nature of the cells; be wary of large clumps of

cells which all appear to be acid-fast positive. Acid-fast cells will be clearly red; cells that

stain purple or light pink may or may not be truly acid-fast. A stain that shows an

unambiguous acid-fast positive reaction may frequently show only a few clearly red cells,

with a majority of blue cells. Frequently, only the beads appear acid-fast positive. If modified

acid-fast stain results are ambiguous, transfer of the organism to a lipid-rich medium, such

as Lowenstein-Jensen medium or Middlebrook 7H11 agar, and repeat staining may give a

more clear-cut stain result. Acid-fastness may become more evident as colonies age; the

acid-fast reaction has been reported to be most reliable when performed from colonies after

1 to 4 weeks of growth (202). Occasionally, coccoid forms of Streptomyces may appear

partially acid-fast; hyphae, however, are acid-fast negative. Because of the difficulties of

interpretation of the modified acid-fast smear, results of this stain should be considered

preliminary and must be used only in conjunction with results from other tests.

Slide Cultures

Slide cultures may be used to evaluate the microscopic morphology of actively growing

cultures and are probably the best way to evaluate the morphology of some genera, such

as Amycolata, Amycolatopsis, andNocardiopsis. Small blocks of a minimal agar (such as tap

water agar) are inoculated on the side with the organism of interest, and a coverslip is

placed on the top of the block. The slide culture is incubated in a humid environment at 25°C

for 2 to 3 weeks and examined regularly for the characteristic features of the various genera.

Vegetative mycelia (also called substrate hyphae) that grow beneath the surface of the agar

and aerial hyphae show various morphologies and degrees of branching based on the

organism being tested (26,93). Differences among the genera of aerobic actinomycetes in

the growth characteristics seen on slide culture may be subtle; experience in recognizing

these morphologic differences is required for correct interpretation of these tests.

Colonial Morphology

The colonial appearance of members of the aerobic actinomycetes is extremely variable

among genera and even between isolates of the same species (Fig. 2). Environmental factors

such as growth medium, incubation temperature, air circulation, presence of CO2, and age of

culture can affect the size and consistency of the colonies and the production of aerial

hyphae (15). Some species produce a diffusible pigment that can vary from strain to strain.

See Table 1 for information on specific colonial morphologies.

Aerial Hyphae

Aerial hyphae project away from the surface of the colony into the air but may not be

apparent until 7 to 14 days for some species that form such

hyphae. Nocardia and Streptomyces usually produce abundant aerial hyphae that give the

colonies their characteristic powdery or velvety appearance (Fig. 2C, E, and L); rare strains

of Nocardia produce sparse or no aerial hyphae (Fig. 2D and G). Of the genera that are

partially acid-fast, only Nocardia species regularly produce aerial hyphae. The presence of

spores and their relative number and arrangement on the aerial hyphae can also give some

clue to genus identification (93).

Genus Assignment

To determine the genus of an unknown isolate, observation of microscopic and colonial

morphology is especially important (Table 1 Fig. 2 and 3). In assessing the characteristics of

colonial growth, the presence or absence of aerial hyphae should be

determined. Nocardia and Streptomyces generally produce aerial hyphae, with other less

commonly isolated genera also showing this morphologic trait (Table 1). Positive results of a

carefully prepared and interpreted modified acid-fast smear combined with the presence of

aerial hyphae help to distinguish Nocardia from the other genera.

The lysozyme test may also assist in initial genus assignment. Lysozyme catalyzes the

hydrolysis of certain polysaccharides in the cell wall, resulting in a weakening of the cell wall

(15). Nocardia and Tsukamurella (both modified acid-fast-positive organisms) are resistant

to lysozyme and show good growth in lysozyme broth; of these, only Nocardia species show

aerial hyphae. Gordonia shows variable growth in lysozyme broth but does not show aerial

hyphae (15). Berd recommends the use of a very small inoculum into glycerol broth with

0.005% lysozyme (see chapter 17), and a second similarly inoculated glycerol broth without

lysozyme as a control. After 4 weeks of incubation, results are interpreted by comparing

organism growth in the broths with and without lysozyme. All Nocardia isolates in one study

(Brown et al., Abstr. 97th Gen. Meet. Am. Soc. Microbiol., abstr. C-65, p. 131, 1997) were

resistant to lysozyme, except for 10 lysozyme-susceptible isolates that belonged to

the N. brevicatena complex.

Laurent et al. describe a PCR-based method that allows differentiation of members of the

genus Nocardia from other genera of aerobic actinomycetes (127). It is not known if this

method allows discrimination of newly described Nocardia species.

Species Assignment

Limitations

Given the increasing number of species of aerobic actinomycetes, the use of phenotypic or

biochemical tests to obtain exact species or even genus-level identification has become

impossible. In addition, phenotypic attributes may be unstable based on environmental and

procedural variation. Among the Nocardia, most species are nonreactive in most

commercially available biochemicals, precluding the definitive identification of these isolates.

For laboratories without molecular capabilities, assignment to the genus level should be

attempted. When a precise identification is required for a significant patient isolate,

molecular testing is strongly recommended.

Some new species have been described based on a single isolate; such species present

problems for laboratories attempting to make identification decisions phenotypically, as

characteristics described may not reflect the typical reactions of the species that would be

determined if more isolates were analyzed (98).

Biochemicals

Because of the increasing number of described species and the low discrimination power and

small number of commercially available phenotypic tests, biochemical testing is not

recommended for definitive identification of these organisms. For laboratories without

molecular capabilities, the use of antibiotic susceptibility patterns and basic biochemical

results may provide preliminary identifications of frequently isolated Nocardia species or

complexes obtained from clinical specimens. Most (but not all) of the isolates of a given

species show the results listed in Table 8. When biochemical tests are performed, it is

extremely important to include appropriate positive and negative controls to ensure that

tests are inoculated, incubated, and interpreted correctly. See “Evaluation, Interpretation,

and Reporting of Results” below for recommendations on reporting preliminary identifications

obtained using these methods.