Heterogeneity Within Corynebacterium minutissimum

CLINICAL MICROBIOLOGY AND INFECTIOUS DISEASE Original Article

Heterogeneity Within Corynebacterium minutissimum Strains Is Explained by Misidentified Corynebacterium amycolatum Strains ANNELIES S. ZINKERNAGEL, ALEXANDER VON GRAEVENITZ, MD, AND GUIDO FUNKE, MD

against O/l29, produced large amounts of propionic acid, and mycolic acids were not detected. In combination with quantitative DNA-DNA hybridizations, it was demonstrated that strains of the second cluster belonged, in fact, to C amycolatum. Furthermore, it was observed that a few C minutissimum strains may also ferment mannitol. These data indicate that the clinical microbiologist must be careful not to misidentify C amycolatum strains as C minutissimum. (Key words: Corynebacterium minutissimum; Corynebacterium amycolatum; Identification; Numerical taxonomy; Infection) Am J Clin Pathol 1996; 106:378383.

It is generally believed that the identification of coryneform bacteria (ie, asporogenous, irregular gram-positive rods) is one of the most difficult tasks for the clinical microbiologist. The reason for this is two-fold: (1) the group of coryneform bacteria is extremely heterogeneous;1 and (2) there are very few identification systems for coryneform bacteria, and their databases cover only a certain proportion of this large group of bacteria.2"4 Corynebacterium minutissimum is considered to be one of the more frequently encountered true Corynebacterium species in clinical specimens. Originally, it was thought that C minutissimum caused erythrasma, but later studies questioned this disease association and revealed that erythrasma is probably a polymicrobial process. '

Since 1990, the Department of Medical Microbiology at the University of Zurich (DMMZ) has isolated or received many clinically significant strains tentatively identified as C minutissimum. As a cursory review of these strains indicated that they were rather heterogeneous we decided to characterize them further. Heterogeneity within C minutisssimum strains was also reported by Hollis and Weaver who divided C minutissimum isolates into two groups corresponding to strains NCTC 10288 (ATCC 23348) (sucrose fermenting) and ATCC 23346 (sucrose nonfermenting).5 By applying phenotypical, chemotaxonomical and molecular methods, we could demonstrate that a large proportion of strains identified as C minutissimum actually correspond to C amycolatum. MATERIALS AND METHODS

From the Department ofMedical Microbiology. University ofZurich, Zurich. Switzerland.

Strains

Guido Funke was supported by a grant from the Sassella-Stiftung, Zurich, Switzerland. Manuscript received December 27, 1995; revision accepted March 30, 1996. Address reprint requests to Dr. Funke: Department of Medical Microbiology. University of Zurich, Gloriastrasse 32, CH-8028 Zurich, Switzerland.

The clinical strains used in this study were cultured from clinical materials using standard procedures or were referred to the DMMZ for identification between 1990 and 1995. Some strains were also received from

378

Downloaded from http://ajcp.oxfordjournals.org/ by guest on March 20, 2016

Korty-eight clinical strains that were tentatively identified as Corynebacterium minutissimum on the basis of standard biochemical reactions (Hollis-Weaver tables) as well as by the use of the API (RAPID) Coryne system were examined further. Two different groups of strains were observed. Thefirstgroup (including the type strain of C minutissimum) contained 27 strains showing creamy colonies. These strains grew homogeneously in 6.5% NaCl broth, exhibited DNase activity, were susceptible to the vibriocidal compound O/l 29, produced succinic acid, and contained mycolic acids. The second group comprised 21 strains with dry colonies. They grew in clumps at the surface of 6.5% NaCl broth, DNase activity was not detected, they were resistant

ZINKERNAGEL ET AL.

379

C minutissimum versus C amycolatum other reference centers, namely, the LCDC Special Bacteriology Reference Laboratory, Ottawa, Canada, and the Culture Collection of the University of Goteborg, Sweden. For comparative investigations strains ATCC 23348 (the type strain of C minutissimum), ATCC 23346 (C minutissimum), and NCFB 2768 (the type strain of C amycolatum) were used.

TABLE

CLINICAL STRAINS INCLUDED IN THE PRESENT STUDY

Organism (no. of isolates) C minutissimum

Source (no. of isolates)

(27)

Blood cultures (10) Ear infection (4) Urinary tract infection (2) 1 ntravascular catheter (1) CAPD catheter (1) Deep wound (1) Abscess (I) Perianal fistula (I) Arterial aneurysm (I) Unknown (5)

Biochemical Profiles

Antimicrobial Susceptibility

Patterns

Minimal inhibitory concentrations (MICs) were determined for eight antimicrobial agents that exhibit activity against coryneform bacteria. We followed the methods for performance and interpretation of MICs published by the National Committee for Clinical Laboratory Standards. 78 Susceptibility to the vibriocidal compound 0/129 was measured with the disk diffusion method by placing a 150 ng disk (Oxoid, Basingstoke, UK) on Mueller-Hinton agar (supplemented with 5% sheep blood), which was inoculated by streaking a McFarland 0.5 suspension onto it. Inhibition zones were measured after 24 hours of incubation at 35 °C.

C mimttissimum-Uke (21) (C amycolatum)

Wound infection (5) Ear infection (4) Intravascular catheter (2) Blood culture (1) Joint infection (1) Chronic osteitis (I) Abscess (1) Infected hematoma (I) Chronic fistula (abdominal wall) (I) Urinary tract infection (1) Unknown (3)

CAPD = continuous ambulatory peritoneal dialysis.

matography. 6 The guanine+cytosine content of the bacterial DNA was determined as outlined before.10 DNA-DNA

Hybridizations

Cultures were grown for 24 hours in 500 mL brain heart infusion broth (Becton Dickinson). Cells were harvested, suspended in 20 mL 0.2 M sucrose 0.05 M Tris pH 8.0 (Merck, Darmstadt, Germany) and incubated in a shaker for 2 hours at 37 °C in the presence of 2.5 mg lysozyme (Sigma, St. Louis, MO) per mL. Then, 20 mg sodium dodecyl sulfate (Sigma) and 100 Mg proteinase K (Sigma) were added per mL suspension followed by further incubation for 2 hours at 55 °C. Extracted and purified DNA was labelled with 3H-labelled nucleotides (Amersham, Little Chalfont, UK) by using a Megaprime kit (Amersham). DNA-DNA hybridizations were carried out at 60 °C for 16 hours in 0.42 M NaCl using the SI nuclease-trichloroacetic method." 1 2 RESULTS

Chemotaxonomical

Investigations

Analysis of the cellular fatty acids (CFAs) of the strains was performed with the Sherlock system (Microbial ID, Newark, DE) as reported previously.9 Mycolic acids were detected from whole-cell hydrolysates by thin-layer chro-

The origins of the 48 clinical isolates included in this study that were tentatively identified as C minutissimum are given in Table 1. All strains were nonlipophilic and showed the following basic biochemical reactions: nitrate reduction negative, urea and esculin hydrolysis neg-

Vol. 106-No. 3


Testing of biochemical reactions was done by application of the commercial API Coryne and API ZYM systems (bioMerieux, La Balme les Grottes, France) and has been outlined in detail before.6 Cells used for inoculation of the commercial systems were previously grown on Columbia agar plates (supplemented with 5% sheep blood) for 24 hours. All tests were carried out at 37 °C. The API 50CH system (bioMerieux) for fermentation of carbohydrates was used according to the instructions provided by the manufacturer. Growth in 6.5% NaCl was determined by inoculating about 108 cells in 5 mL Trypticase broth (Becton Dickinson, Cockeysville, MD) +6.5% NaCl and observing growth after incubation of 48 hours. DNase activity was tested with DNase test agar with methyl green (Difco, Detroit, MI) by incubation of about 108 cells for up to 10 days. Hippurate hydrolysis was tested in tubes containing 0.2 mL 1% sodium hippurate solution by incubating one bacterial colony for 2 hours and then adding one drop of a 3% aceton-ninhydrine solution.

380

CLINICAL MICROBIOLOGY AND INFECTIOUS DISEASE Original Article TABLE 2. API (RAPID) CORYNE PATTERNS OF THE STRAINS STUDIED*

Organism C mimtlissimum

Numerical API Code (no. of strains)

Similarity Index

T Value

2100124(2)

C minutissimum Cjeikeium

81.5 17.6

0.89 0.76

2100125(4) 2100324(2)

C minutissimum C minutissimum Cjeikeium

97.1 55.4 42.3

0.92 0.97 0.93

6100124(2)


81.8 17.7

0.64 0.51

6100125(1)

C minutissimum C xerosis C minutissimum C xerosis C minutissimum Cjeikeium

74.1 24.7 74.1 24.7 56.5 43.2

0.67 0.59 0.17 0.09 0.72 0.68


55.4 42.3

0.97 0.93

6100135(1) 6100324(1) C minutissimum-\\ke (C amvcolatum)

2100324(8)

• Thirteen C. minutissimum and eight C. minutissimum-like strains were tested.

ative, fermentative production of acid from glucose and maltose but not from xylose, and variable acid production from sucrose. This pattern of biochemical reactions was compatible with an identification of the 48 strains as C minutissimum? Two strains were also able to produce acid from mannitol. The 48 strains were divided in two separate groups according to their macroscopic morphology: the first group included 27 strains with whitish, creamy colonies of 1 to 2 mm in diameter (after 24 hours) and the second group 21 strains with white-grayish, dry and rough colonies of 1 to 2 mm in diameter (after 24 hours). The type strain of C minutissimum belonged to the first group and, therefore, these strains were designated C minutissimum, whereas strains of the second group were referred to as C minutissimum-like. Corynebacterium minutissimum as well as C minutissimum-like strains came from various sources, with blood cultures predominating for C minutissimum strains. Many strains of both groups were isolated from sites anatomically closely related to the skin. None of the strains examined came from a respiratory specimen. Gram stains of strains belonging to either the first and the second group showed typical diphtheroids (ie, club-shaped gram-positive rods compatible with the genus Corynebacterium). Thirteen strains of the C minutissimum and 8 strains of the C minutissimum-like group were also tested with the API Coryne system. All strains were identified as C minutissimum by the system when the first rank of identification was considered (Table 2). However, because of

the marked morphological differences between C minutissimum and C minutissimum-like strains additional biochemical tests were applied to possibly further differentiate the two groups. Activities of 19 different enzymes were tested with the API Zym system for all 48 strains but no difference was noted between the two groups except that most C minutissimum strains showed leucine arylamidase activity, whereas this enzyme was detected in two of the C minutissimum-like strains only (Table 3). All 48 strains exhibited activity of acid and alkaline phosphatase, esterase (C4), and esterase lipase (C8). However, DNase activity was detected in all C minutissimum strains between 3 to 10 days (mean 6 days), whereas no activity was found in C minutissimum-like strains. Growth of C minutissimum in 6.5% NaCl broth resulted in an increased turbidity, whereas C minutissimum-like strains grew in clumps at the surface of the broth. Hydrolysis of hippurate was variable in both groups. Eleven strains of C minutissimum and 6 of C minutissimum-like strains were tested for their capability to ferment any of the 49 carbohydrates in the API 50CH system. Only the fermentations of gluconate, Nglucosamine, and D-tagatose differentiated between the two groups of bacteria (Table 3). Furthermore, we determined the end products of glucose metabolism and found that C minutissimum strains mainly produced succinic acid, whereas C minutissimum-like strains produced large amounts of propionic acid from glucose. Palmitic and oleic acids were the predominant CFAs of C minutissimum strains, whereas in C minutissimum-like

A.J.C.P.- September 1996


Identification

Z I N K E R N A G E L ET AL.

381

C minutissimum versus C amycolatum TABLE 3. CHARACTERISTICS DIFFERENTIATING C MINUTISSIMUM FROM C MINUTISSIMUM-UKE (C AMYCOLATUM) STRAINS C minutissimum (n = 27)

Characteristic DNase Growth in 6.5% NaCl Leucine aminopeptidase Fermentation off Gluconate N-acetyl-Glucosamine D-Tagatose Production of Propionic acid Succinic acid Cellular fatty acids (percentage of total CFA)| CI 6:0 C18:la>c/'.v9 CI 8:0 Presence of mycolic acids

Uniform turbidity