Isolation frequency ofcarbapenem-nonsusceptible Bacteroides spp. and detection methods for cfiA-positive Bacteroides fragilis

Yukiko Takemori-Sakai^*1, †Yasunori Iwata^*2,3, Hatsumi Otani^*1, Akiko Maekawa^*1, Hiroyasu Oe^*1, Megumi Oshima^*1,2, Mika Mori^*1,4, Yoshio Sakai^*5, Toshifumi Gabata^*1

Original
Lab Med Int 2025; 4(2): 62-68

†Division of Infection Control, Department of Nephrology and Laboratory Medicine, Kanazawa University Hospital13-1 Takara-machi, Kanazawa 920-8641, Japan
E-mail: iwatay”@”staff.kanazawa-u.ac.jp
Received March 20, 2023; accepted January 13, 2025
*1 Division of Clinical Laboratory Medicine, Kanazawa University, Kanazawa, Japan
*2 Department of Nephrology and Laboratory Medicine, Kanazawa University, Kanazawa, Japan
*3 Division of Infection Control, Kanazawa University, Kanazawa, Japan
*4 Department of Cardiovascular and Internal Medicine, Kanazawa University, Kanazawa, Japan
*5 Department of Gastroenterology, Kanazawa University, Kanazawa, Japan

ABSTRACT

Aims: Bacteroides fragilis is the most common anaerobic bacteria causing infectious diseases in humans. A cfiA gene encoding metallo-β-lactamase, which degrades carbapenem, has been reported as a resistance mechanism against carbapenem. Therefore, the detection of cfiA gene is important for determining appropriate antibiotic therapy. Although cfiA gene is detected via polymerase chain reaction (PCR), simpler and easier methods are required in clinical settings. We compared the detectability of cfiA-positive B. fragilis using multiple methods. Moreover, the isolation frequency of carbapenem-nonsusceptible Bacteroides spp. was evaluated. 
Methods: Bacteroides spp. was isolated from clinical specimens at the Kanazawa University Hospital. Antimicrobial susceptibility testing was performed via broth microdilution. We explored whether matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) and modified carbapenem inactivation method (mCIM) can show cfiA gene positivity. 
Results: Around 20-50 Bacteroides spp. strains/year were isolated from June 2015 to December 2020. Among these, 0-8 strains exhibited meropenem and/or imipenem nonsusceptibility (intermediate and resistant) in each year. An increased rate of carbapenem resistance, including meropenem- and imipenem-nonsusceptible Bacteroides spp., was observed. All carbapenem-nonsusceptible Bacteroides spp. were susceptible to metronidazole. Four (66.7%) of 6 strains showed cfiA gene positivity based on PCR. 
Conclusions: MALDI-TOF MS and mCIM detected cfiA-positive strains. MALDI-TOF MS and mCIM are useful for detecting cfiA-positive B. fragilis. Because the isolation frequency of carbapenem-nonsusceptible Bacteroides spp. is increasing, determining their susceptibility is important for appropriate antimicrobial therapy.

〔Lab Med Int 2025; 4(2): 62-68〕

Key Words

Bacteroides spp., carbapenem resistance, cfiA gene

I. Introduction

Bacteroides spp. are anaerobic bacteria that form normal microbiota in the intestinal tract. They are important pathogenic bacteria for humans in diseased conditions. In particular, antimicrobial-resistant Bacteroides spp. infection reportedly shows poor prognosis and prolonged hospitalization¹⁾. Bacteroides fragilis is the most common anaerobic bacteria causing infectious disease in humans ²⁾.
Carbapenem resistance in B. fragilis was first reported in 1983 in Japan³⁾ and in 1986 in the United States⁴⁾. Although their resistance mechanism to carbapenem remains unclear, the involvement of cfiA gene encoding metallo-β-lactamase (MBL), which degradates carbapenem, has been reported⁵⁾. Considering the poor prognosis of carbapenem-resistant B. fragilis infection, the detection of cfiA gene is important. While cfiA gene is detected via polymerase chain reaction (PCR), this method is available only in some hospitals. Therefore, simpler and easier methods are required in clinical settings. In this study, the detectability of cfiA-positive B. fragilis was compared using multiple methods. Moreover, the isolation frequency of carbapenem-nonsusceptible Bacteroides spp., especially cfiA-positive strains, was evaluated.

II. Materials and methods

Strains and identification
Bacteroides spp. were isolated from clinical specimens at the Kanazawa University Hospital between June 2015 and December 2020. The specimen was applied and cultured on brucella agar medium (Kyokuto, Tokyo, Japan) for 2 days under anaerobic conditions at 35℃. The isolated strains were identified using a Bacteroides Bile Esculin agar (BBE Agar, Kyokuto Pharmaceutical Industrial Co., Ltd., Tokyo, Japan) and AccuDia^TM Bacteroides Agar (BA, Shimadzu Diagnostics Corporation, Tokyo, Japan) according to the each directions. B. fragilis group were identified by the change of only BBE Agar color from dark brown to black that appears as zones around the colonies. Non-B. fragilis group were identified by the growth on BA and no color change in BBE Agar. Those that grew under aerobic conditions were excluded. Eight carbapenem-nonsusceptible strains, which were stored at our hospital, were re-identified using matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) (MALDI Biotyper, Bruker, Billerica, MA, USA). The strain was applied to the MSP 96 target polished steel BC (Bruker). After drying, 1 µL of 70% formic acid (Beckman Coulter, Brea, CA, USA) was added. Then, the sample was loaded into MALDI Biotyper after 1 µL of HCCA portioned (Bruker) was added. The spectra were analyzed using the MALDI Biotyper Compass software (MBT Compass software) (Bruker) following the manufacturer’s instructions.

Antimicrobial susceptibility testing
The strain was suspended in ABCM broth (Eiken Chemical Co., Ltd., Tokyo, Japan) and adjusted with McFarland 2 standards. Then, 25 µL of sample were applied to Brucella Broth Eiken (Eiken Chemical Co., Ltd.). After 100 µL of the bacterial solution was added to Dry Plate Eiken (Eiken Chemical Co., Ltd.), the plate was incubated under anaerobic conditions at 35℃ for 48 h. After 48 h, the growth of the strain was visually observed. The minimum inhibitory concentration (MIC) and interpretive category were determined according to the Clinical and Laboratory Standards Institute M100-ED32 ⁶⁾. The additional MIC measurement of metronidazole (MNZ) was performed on 8 carbapenem-nonsusceptible strains, which were stored at our hospital.

cfiA gene detection
cfiA gene was detected via PCR. DNA extraction was prepared as follows. The strains were suspended in 200 µL of sterilized water. Then, 5 µL of proteinase K (350 U/mL, Takara Bio Inc., Shiga, Japan) was added, followed by incubation at 50℃ for 5 min. After 7 min of incubation in a boiling water bath, the lysate was centrifuged at 10,000 rpm for 5 min. Then, the supernatant was used as a template. The primers were 5’-CCCAACTCTCGGACAAAGTG-3’ and 5’-ACGATCTGCTTGGTATGCTC-3’ ⁷⁾. PCR was performed using 1 µL of template DNA in a total reaction volume of 25 µL consisting of EmeraldAmp PCR Master Mix (Takara Bio Inc., Shiga, Japan) and 25 pmol of each primer. Reactions were run in a thermal cycler (Biometra, Gottingen, Germany). The PCR program was as follows: 98℃ for 60 s; 30 cycles of 98℃ for 10 s, 56℃ for 30 s, and 72℃ for 60 s; and 72℃ for 60 s. Then, 5 µL of amplicon was analyzed using 2% agarose gel. Strains in which an amplicon was observed at 625 bp were determined to have cfiA gene.

Detection of cfiA-positive B. fragilis by MALDI-TOF MS
MALDI Biotyper was used in the same way as for bacterial identification. The MALDI Biotyper Subtyping Module with the MBT Compass software automatically identifies not only bacterial species but also possession of cfiA gene. Peaks shifts were reported from 4711 Da and 4817 Da (cfiA-negative) to 4688 Da and 4826 Da (cfiA-positive), respectively⁸⁾. Strains that matched with the spectra of cfiA-positive B. fragilis were considered “presumptive cfiA-positive”.

Detection of cfiA-positive B. fragilis using modified carbapenem inactivation method
We identified carbapenemase-producing B. fragilis using modified carbapenem inactivation method (mCIM)⁶⁾ with a few changes. Briefly, 1 µL of inoculation loop of anaerobically cultured strain was firmly suspended in 2 mL of BD BBL^TM Trypticase^TM Soy Broth (Becton, Dickinson and Company, New Jersey, USA). A meropenem (MEPM) or imipenem (IPM) disk (Eiken Chemical Co., Ltd.), each containing 10 µg of antibiotic, was added and suspended. After incubation under anaerobic conditions at 35℃ for 4 h, the disk was removed from the suspension and placed on a BD BBL^TM Muller–Hinton II agar medium (Becton, Dickinson and Company, New Jersey, USA) inoculated with a susceptible Escherichia coli indicator (ATCC25922). After incubation under aerobic conditions at 35℃ for 24 h, the diameter of growth inhibition zones was measured. MBL-producing Klebsiella pneumoniae was used as a positive control in mCIM analysis. The negative control was a culture medium only.

III. Results

Patient characteristics
The study period is from June 2015 to December 2020. Carbapenem-nonsusceptible Bacteroides spp. were detected in 17 patients. The mean age of patients was 65.0 years, and 10 patients (59.0%) were female. Their underlying diseases were malignant tumors (11 patients, 64.7%), gastrointestinal disorders (6 patients, 35.3%), and hepatobiliary diseases (4 patients, 23.5%). The strains were detected from pus (9 patients, 52.9%), abdominal cavity and subdiaphragmatic drains (4 patients, 23.5%), bile (3 patients, 17.6%), and blood (1 patient, 5.9%). Fifteen specimens (88.2%) were of abdominal origin. Fourteen patients were treated, and 11 patients (78.5%) received antimicrobials. Especially, 7 (63.6%) of 11 patients were treated with noneffective carbapenems (Fig. 1).

Isolation frequency of carbapenem-nonsusceptible Bacteroides spp.
The characters of isolated 233 Bacteroides spp. were described in Fig. 2. Two hundred strains were B. fragilis group, and 33 strains were non-B. fragilis group. Around 20-50 strains/year were isolated in Kanazawa University Hospital. Among these, 0-8 strains exhibited MEPM and/or IMP nonsusceptibility (intermediate and resistant) in each year. Although no MEPM-nonsusceptible Bacteroides spp. was detected in 2015, its isolation rate peaked at 16% in 2019 and decreased to 7% in 2020 (Fig. 2a). In contrast, the frequency of IPM-nonsusceptible Bacteroides spp. increased from 0% in 2015 to 8% in 2020 (Fig. 2b). An increased rate of carbapenem resistance, including MEPM- and IPM-nonsusceptible Bacteroides spp., was observed during the period.

Antimicrobial susceptibility of carbapenem-nonsusceptible Bacteroides spp.
Nineteen strains of carbapenem-nonsusceptible Bacteroides spp. were isolated between 2015 and 2020. Seventeen strains were B. fragilis group and 2 were non-B. fragilis group. Two patients were positive for both B. fragilis group and non-B. fragilis group. The antimicrobial susceptibility of these strains is shown in Table 1. Eight of 19 strains were available for susceptibility tests for MNZ. All of the 8 strains were B. fragilis group and showed susceptibility to MNZ.

Detection of cfiA gene from carbapenem-nonsusceptible B. fragilis
Six of 8 carbapenem-nonsusceptible strains that were stored at our hospital were re-identified as B. fragilis via MALDI-TOF MS. The other 2 strains were re-identified as B. thetaiotaomicron and B. uniformis, respectively. We investigated whether these 6 B. fragilis strains possessed cfiA gene by performing PCR. Four strains (66.7%) were cfiA-positive. Among the preserved strains, a cfiA-positive strain was identified in 2016 and 2020 in each and 2 strains were identified in 2019. Then, MALDI-TOF MS was performed to identify strains with cfiA gene (Table 2) and was compared to mCIM.

MALDI-TOF MS detected strains with cfiA gene
Of 6 B. fragilis isolates, 4 strains were detected as presumptive cfiA-positive based on MALDI-TOF MS. The other 2 strains were detected as presumptive cfiA-negative (Fig. 3). The results of cfiA-positive B. fragilis based on MALDI-TOF MS were consistent with those of cfiA gene based on PCR.

mCIM revealed the susceptibility of strains with/without cfiA gene
Neither MEPM nor IPM disks inhibited the growth of E. coli in mCIM analysis of 4 cfiA-positive B. fragilis. In the analysis of the other 2 cfiA-negative B. fragilis, E. coli formed growth inhibition zones around MEPM disks, with a diameter larger than 20 mm. IPM disks also inhibited the growth of E. coli in the analysis of 2 cfiA-negative B. fragilis, with growth inhibition zones larger than 18 mm (Fig. 4).

Figure 1   Treatment of patients infected with carbapenem-nonsusceptible Bacteroides spp. Here, 63.6% of patients were treated with noneffective carbapenems.

Figure 2   Changes in carbapenem sensitivity from 2015 to 2020 Sensitivity to meropenem (a) and imipenem (b).  An increased rate of carbapenem-nonsusceptible Bacteroides spp. was observed in meropenem and imipenem. dot: susceptible, oblique line: intermediate, black: resistant, S: susceptible, I: intermediate, R: resistant.

Table 1 Antimicrobial susceptibility of carbapenem-nonsusceptible Bacteroides spp.

S/A: sulbactam-ampicillin, T/P: tazobactam-piperacillin, MEPM: meropenem, IPM: imipenem, CLDM: clindamycin, MNZ: metronidazole, S: susceptible, I: intermediate, R: resistant.

Table 2 Detection results of cfiA-positive B. fragilis via MALDI-TOF MS and mCIM

PCR: polymerase chain reaction, MALDI-TOF MS: matrix-assisted laser desorption ionization–time of flight mass spectrometry, mCIM: modified carbapenem inactivation method, MEPM: meropenem, IPM: imipenem.

Figure 3   MALDI-TOF MS peaks of cfiA-negative and cfiA-positive B. fragilis.
White arrows indicate peaks of cfiA-positive B. fragilis. Black arrows indicate peaks of cfiA-negative B. fragilis. 

Figure 4 Results of modified carbapenem inactivation method
B. fragilis strains 1, 2, 3, 4 were cfiA-positive, whereas strains 5 and 6 were cfiA-negative. NC is the result of a bacteria-free test. PC is the result of metallo-β-lactamase-producing Klebsiella pneumoniae.
MEPM: meropenem, IPM: imipenem, NC: negative control, PC: positive control.

IV. Discussion

The isolation frequency of carbapenem-nonsusceptible Bacteroides spp. was investigated, and methods for detecting cfiA-positive B. fragilis were examined. This study revealed that the maximum isolation frequency of MEPM- and IPM-nonsusceptible Bacteroides spp. was 16% and 8%, respectively. The isolation rate of carbapenem-nonsusceptible Bacteroides spp. increased over the past 6 years. Recent reports indicate that carbapenem-nonsusceptible Bacteroides spp. have been isolated in some countries with various frequencies^9)-11). The increased frequency of carbapenem-resistant Bacteroides spp. was also reported in some studies^9),12),13). The isolation frequency increased similar to that reported in other studies in Japan⁹⁾. Therefore, monitoring the isolation frequency of carbapenem-nonsusceptible Bacteroides spp. is important in hospitals. Moreover, the detection of carbapenem-nonsusceptible Bacteroides spp. is important for determining appropriate antibiotic therapy. We should have investigated the isolation frequency of carbapenem-nonsusceptible B. fragilis. Although MALDI-TOF MS analysis identify strains to B. fragilis, the equipment was not available during the study period in our facility. Therefore, we investigated the isolation frequency of carbapenem-nonsusceptible strains in the genus Bacteroides.

The isolation rate of MEPM-nonsusceptible Bacteroides spp. peaked at 16% in 2019 and decreased to 7% in 2020. Although the reason for the reduced MEPM-nonsusceptible Bacteroides spp. frequency is not clear, small sample number may affect the change of frequency in this case.

In our study, while 4 strains (66.7%) of carbapenem-nonsusceptible B. fragilis were cfiA-positive, the other 2 strains did not possess the gene. Other than cfiA gene, the mechanisms of carbapenem nonsusceptibility remain unclear¹⁴⁾. One study showed that functional changes in efflux pumps affect carbapenem sensitivity¹⁵⁾. Further studies are needed to clarify the resistance mechanisms, especially in 2 carbapenem-nonsusceptible B. fragilis without cfiA gene.

To date, there is no routine method to detect cfiA-positive B. fragilis. In this study, MALDI-TOF MS and mCIM were useful in detecting cfiA-positive B. fragilis. A recent study reported that the sensitivity and specificity of MALDI-TOF MS in detecting cfiA-positive B. fragilis were 100.0% and 99.7%, respectively¹⁶⁾. Although MALDI-TOF MS is a simple and quick method, it can be performed only in the facilities with the required equipment. In facilities where MALDI-TOF MS is unavailable, mCIM may be useful. Because mCIM does not require specialized equipment or reagents, it is widely used and can be performed at any facility. Therefore, MALDI-TOF MS and mCIM might provide earlier and more accurate results in determining cfiA-positive strains. This will help determine appropriate antimicrobial use for treating carbapenem-nonsusceptible B. fragilis infection.

In summary, even if MALDI-TOF MS or PCR is not available, mCIM is useful for detecting cfiA-positive B. fragilis. Because the isolation frequency of carbapenem-nonsusceptible Bacteroides spp. is increasing, determining its susceptibility is important for appropriate antimicrobial therapy.

Declarations

Funding: This work was supported by the Funding Program for Female Researchers, Kanazawa University.

Conflicts of interest: The authors have declared that no conflict of interest exists.

Ethics approval: All experiments were performed in accordance with approved guidelines of Kanazawa University. This study was conducted with the approval of the ethics committee of Kanazawa University (approval number: 3596).

Authorship contributions: YTS and YI designed the study and interpreted the data. YTS performed experiment and wrote the draft of the manuscript together with YI. All authors read the manuscript and commented.

References 

1. Hartmeyer GN, Sóki J, Nagy E, et al. Multidrug-resistant Bacteroides fragilis group on the rise in Europe? J Med Microbiol. 2012; 61(Pt 12): 1784-8.PubMed
2. Jasemi S, Emaneini M, Ahmadinejad Z, et al. Antibiotic resistance pattern of Bacteroides fragilis isolated from clinical and colorectal specimens. Ann Clin Microbiol Antimicrob. 2021; 20(1): 27-34. PubMed
3. Yotsuji A, Minami S, Inoue M, et al. Properties of Novel 3-Lactamase Produced by Bacteroides fragilis. Antimicrob Agents Chemother. 1983; 24(6): 925-9.
4. Cuchural GJJr, Malamy MH, Tally FP. Beta-Lactamase-Mediated Imipenem Resistance in Bacteroides fragilis. Antimicrob Agents Chemother. 1986; 30(5): 645-8.PubMed
5. Thompson JS, Malamy MH. Sequencing the Gene for an Imipenem-Cefoxitin-Hydrolyzing Enzyme (CfiA) from Bacteroides fragilis TAL2480 Reveals Strong Similarity between CfiA and Bacillus cereus, B-Lactamase II. J Bacteriol. 1990; 172(5): 2584-93.PubMed
6. CLSI M100 ED32:2022. http://em100.edaptivedocs.net/
7. Yamazoe K, Kato N, Kato H, et al. Distribution of the cfiA Gene among Bacteroides fragilis Strains in Japan and Relatedness of cfiA to Imipenem Resistance. Antimicrob Agents Chemother. 1999; 43(11): 2808-10.PubMed
8. Nagy E, Becker S, Sóki J, et al. Differentiation of division I (cfiA-negative) and division II (cfiA-positive) Bacteroides fragilis strains by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. J Med Microbiol. 2011; 60(Pt 11): 1584-90. PubMed
9. Ueda T, Takesue Y, Matsumoto T, et al. Change in antimicrobial susceptibility of pathogens isolated from surgical site infections over the past decade in Japanese nation-wide surveillance study. J Infect Chemother. 2021; 27(7): 931-9.PubMed
10. Wang Y, Han Y, Shen H, et al. Higher Prevalence of Multi-Antimicrobial Resistant Bacteroides spp. Strains Isolated at a Tertiary Teaching Hospital in China. Infect Drug Resist. 2020; 13: 1537-46. PubMed
11. König E, Ziegler HP, Tribus J, et al. Surveillance of Antimicrobial Susceptibility of Anaerobe Clinical Isolates in Southeast Austria: Bacteroides fragilis Group Is on the Fast Track to Resistance. Antibiotic. 2021; 10(5): 479-85. PubMed
12. Copsey-Mawer S, Hughes H, Scotford S, et al. UK Bacteroides species surveillance survey: Change in antimicrobial resistance over 16 years (2000-2016). Anaerobe. 2021; 72: 102447.PubMed
13. Liu CY, Huang YT, Liao CH, et al. Increasing Trends in Antimicrobial Resistance among Clinically Important Anaerobes and Bacteroides fragilis Isolates Causing Nosocomial Infections: Emerging Resistance to Carbapenems. Antimicrob Agents Chemother. 2008; 52(9): 3161-8.PubMed
14. Wexler HM. Bacteroides: the Good, the Bad, and the Nitty-Gritty. Clin Microbiol Rev. 2007; 20(4): 593-621.PubMed
15. Pumbwe L, Chang A, Smith RL, et al. Clinical significance of overexpression of multiple RND-family efflux pumps in Bacteroides fragilis isolates. J Antimicrob Chemother. 2006; 58(3): 543-8. PubMed
16. Kawamoto Y, Kosai K, Ota K, et al. Rapid detection and surveillance of cfiA-positive Bacteroides fragilis using matrix-assisted laser desorption ionization time-of-flight mass spectrometry. Anaerobe. 2021; 72: 102448.PubMed