In vitro synergistic interaction of colistin and other antimicrobials against intrinsic colistin-resistant Morganella morganii isolates

Abstract

Morganella morganii, a non-negligent opportunistic pathogen of the family Enterobacteriaceae, enlisted recently in the global priority pathogens by WHO for its swift propensity to acquire drug-resistant genes, engendering enhanced death rates. A combination of diverse antimicrobials could be recycled to overcome the ongoing acquisition of resistance mechanisms by M. morganii. Herein, we investigated the in vitro synergistic effect of colistin with meropenem, rifampicin, minocycline and linezolid against three intrinsic colistin-resistant M. morganii strains collected from critical departments of tertiary care hospitals. The strains were identified and tested for antimicrobial susceptibility by VITEK 2 automated system. The 16S rRNA sequencing was used to reconfirm the species identification. Minimum inhibitory concentrations (MICs) of colistin, meropenem, rifampicin, minocycline and linezolid were determined by the broth microdilution method. Synergistic interactions were studied by checkerboard and time-kill assay. The VITEK 2 identification and 16S rRNA sequencing confirmed that the strains were M. morganii. The automated antimicrobial susceptibility test revealed that all three isolates were multi-drug resistant. The checkerboard analysis demonstrated the synergy of all four combinations with FICI values ranging from 0.06 to 0.31 in all three isolates. These results suggest a potential role of meropenem as an adjuvant for treating M. morganii infections. The current work presented the first evidence of synergy between colistin and other antibiotics against M. morganii infection, which needs validation through in vitro and in vivo studies using a larger number of isolates.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13205-023-03551-w.

Morganella morganii, a non-negligent opportunistic pathogen, belongs to the Proteeae tribe of the Enterobacteriaceae family. The accumulation of acquired and intrinsic drug-resistant genes in M. morganii makes it untreatable, which leads to a high death rate of ~ 15% worldwide. The World Health Organization (WHO) articulates a global priority pathogen list where Morganella spp. was counted in the critical group (Asokan et al. 2019). Inherent resistance to colistin, glycopeptides, rifampicin, daptomycin, tetracyclines, nitrofurantoin, macrolides, lincosamides, streptogramins, fusidic acid and acquired resistance to carbapenemases complicate the treatment strategies and antibiotics selection (Zaric et al. 2021). M. morganii has been described for various invasive infections in the literature to date: pericarditis, intra-abdominal abscess, peritonitis, rhabdomyolysis, orbital abscess and neonatal sepsis, etc. (Zaric et al. 2021). In addition, it is often associated with nosocomial infections like urinary tract infection (UTI) and in a few cases with non-hospital settings as well as healthy individuals (Atmış et al. 2020). Due to the paucity of published literature and evidence-based guidelines, there is no consensus regarding treating M. morganii infections. Thus, combinational therapy is being increasingly used to boost the antibacterial properties of existing drugs against multidrug-resistant strains (Zaric et al. 2021). In the absence of alternative empirical therapy, administering gentamycin in conjunction with second and third-generation cephalosporin helped recover M. morganii infections but its imprudent misuse led to the sporadic emergence of resistance (Zaric et al. 2021). Due to the growing resistance to aminoglycosides and other first-line antibiotics, M. morganii is classified under mDTR (modified difficult-to-treat resistance) (Gajdács et al. 2020). Therefore, mentioned problems invoke screening suitable antibiotic combinations and colistin as an alternate treatment regimen made available for physicians.

Clinical and Laboratory Standards Institute (CLSIs’) endorsement on use of colistin in conjunction with one or more antibiotics rather than monotherapy (Dhandapani et al. 2021), recommendation of glycopeptides for intrinsically resistant Gram-negative bacteria in combination with colistin (Armengol et al. 2019) and the synergism exhibited by colistin with carbapenems and other antibiotics against drug-resistant Enterobacteriaceae members (Brennan-Krohn et al. 2018) supports our rationale of exploring a combinatorial therapy of colistin with other frequently used antibiotics.

Hence, our work aims to investigate the in vitro synergistic effect of colistin with meropenem, rifampicin, minocycline and linezolid against intrinsic colistin-resistant (ICR) M. morganii strains isolated from critical departments of a tertiary care hospital which should be used further treatment of the infection.

During a routine surveillance study conducted in a tertiary care hospital, Bhubaneswar, India, we collected three multidrug-resistant (MDR) isolates of M. morganii. The strains UM573, UM869 and UM169, were isolated from the urine of 74 and 75-year-old females and blood samples of 37-year-old male patients, respectively. The two female patients had the symptoms of frequent urination and a history of recurrent UTI with 10–16 days of hospital stay, while the male patient suffered from soft tissue injury and underwent plastic surgery and stayed in the hospital for 20 days. During treatment, the female patients were administered cefixime as the first empirical treatment followed by gentamycin, while the male patient was under the regimen of fluoroquinolone and imipenem. The demographic data, including patient details, sample specimens, wards they belong to and laboratory findings, were obtained from the hospital’s data centre. Thus, this study was exempted from ethical committee approval and patient consent criteria.

Identification and antimicrobial susceptibility profiles of the isolates were done by VITEK 2 COMPACT system (BioMerieux, France). The isolates were further identified through 16S rRNA gene amplification and sequencing using universal primers (8F and 1492R) (Sahoo et al. 2014). All the 16S rRNA sequences were submitted to GenBank under the accession numbers ON533443, ON533444 and ON533445. The phylogenetic tree was constructed from the 16S rRNA gene sequence by the UPGMA method with 1000 bootstrapping and visualized using FigTree v1.4.4 (Rambaut 2018). The sensitivity profile of the isolates to colistin, rifampicin, linezolid, minocycline, and meropenem was evaluated using the broth microdilution method and interpreted using CLSI breakpoint criteria for Enterobacteriaceae (CLSI 2020). The E. coli ATCC 25,922 was used as a reference strain. The checkerboard assay was performed in 96-well flat-bottom microtiter plates (Genexy, India) containing colistin and one of the four other antibiotics (meropenem, linezolid, minocycline and rifampicin). Each antibiotic was diluted from the ¼th minimum inhibitory concentration (MIC) value up to nine subsequent two-fold dilutions along the abscissa and ordinate with an inoculum ~ 5 × 10⁵ CFU/mL. After 16 h of incubation at 37 °C, the fractional inhibitory concentration index (FICI) was calculated for each combination and interpreted as FICI ≤ 0.5, synergism; 0.5 < FICI < 4, no interaction; FICI < 4, antagonism (Odds 2003). Time kill assay was performed as described by Nguyen et al. (2021). The test combinations were added to 2 mL of inoculum (~ 10⁵ CFU/mL) to maintain the final concentration equivalent to ¼MIC. Each set of the experiment includes a growth control consisting of the bacterial strain without the test compound. The change in bacterial concentration was evaluated by plotting the log10 CFU/mL vs time (hours). Bactericidal activity is defined as a reduction of < 3log10 CFU/mL relative to the initial inoculum (Sim et al. 2014). Time-dependent data were analysed using SPSS software v21.0 to test the significance of the observations. The student’s paired t-test was applied to analyse data with a significance level set at p < 0.05, and values were interpreted as mean ± Standard Deviation (SD). All the experiments were performed as independent triplicates.

The ICR strains were subjected to antimicrobial susceptibility testing, which showed resistance to all groups of antibiotics except UM169. UM169 showed susceptibility to sulphonamides and glycylcycline class of antibiotics. The VITEK 2 profile revealed that the strains belong to the MDR (Supplementary Table S1). The strains were confirmed as M. morganii using 16S rRNA sequencing and a UPGMA-based phylogenetic tree was built by removing ambiguity in the sequences retrieved from the GenBank. The BLASTn of 16S rRNA sequence against the ribosomal RNA database revealed that all the strains showed > 99% identity at > 99% sequence coverage with M. morganii strain NBRC 3848 (Accession: NZ_BCZU00000000.1). The strains UM869, UM573, and UM169, were clustered together in a separate sub-clade having high bootstrap values > 60 with a sub-clade of reference strains of M. morganii (Fig. 1). The MICs of colistin and other antibiotics were determined where all the isolates exhibited resistance to colistin (≥ 1024 µg/mL), minocycline (≥ 8 and 16 µg/mL), rifampicin (≥ 512 µg/mL) and linezolid (512 and 1024 µg/mL) whereas, the MIC of meropenem ranged from 16 to 1024 µg/mL among the strains (Table 1). The checkerboard analysis demonstrated synergy with colistin-containing combinations, with FICI values ranging from 0.06 to 0.31 (Table 2). In the study of (Cannatelli et al. 2018), the checkerboard assay of resveratrol and colistin has an apparent dose-dependent synergistic efficacy against inherently ICR strains like S. marcescens and P. mirabilis.

Fig. 1.

Phylogenetic relationship in the form of a cladogram of 16SrRNA partial sequences of Morganella morganii strains UM169, UM573, and UM869 from this study (highlighted in blue color) along with the available partial sequence of other closely related species of the Morganella genus. The bootstrap score in percentage is shown as a node label and branches are coloured according to the respective bootstrap score. The tree is generated by ClustalW v2 using the UPGMA/Average linkage (Unweighted Pair Group Method using arithmetic Averages) method and visualized in Fig Tree v1.4.4

Table 1.

Minimum Inhibitory Concentration of colistin (CL), minocycline (MN), meropenem (MR), rifampicin (RF) and linezolid (LN) against three Morganella morganii isolates

Microdilution assay (μg/mL)
Isolates	CL	MN	MR	LN	RF
UM869	≥ 1024	≥ 8	1024	1024	≥ 512
UM169	≥ 1024	16	32	512	≥ 512
UM573	≥ 1024	16	≥ 16	512	512

Table 2.

Fractional inhibitory concentration index (FICI) of colistin combined with each antibiotic against the three Morganella morganii isolates

Isolates	CL + MN				CL + MR				CL + LN				CL + RF
	CL (μg/mL)	MN (μg/mL)	FICI	Effect	CL (μg/mL)	MR (μg/mL)	FICI	Effect	CL (μg/mL)	LN (μg/mL)	FICI	Effect	CL (μg/mL)	RF (μg/mL)	FICI	Effect
UM869	128	0.5	0.19	Synergy	16	256	0.25	Synergy	2	32	0.06	Synergy	4	128	0.13	Synergy
UM169	256	1	0.31	Synergy	64	4	0.13	Synergy	256	32	0.31	Synergy	64	128	0.31	Synergy
UM573	256	1	0.31	Synergy	128	2	0.13	Synergy	128	32	0.19	Synergy	64	64	0.19	Synergy

Time-kill (TK) assays were conducted to investigate if colistin in combination with other antibiotics had a bactericidal effect. The time-dependent killing curve suggested a reduction in the colony-forming unit (CFU) from 8 to 12 h in most of the drugs combined with colistin while tested in all isolates. The colistin-meropenem combination showed rapid bactericidal activity against UM573 and UM169 after 4 h of exposure. This was clearly evident from a ≥ 3log10 decrease in viable cell counts of UM573 (from 9.05 ± 0.02 to 8 ± 0.0163 log10 CFU/mL) and UM169 (from 8.9 ± 0.01 to 7.78 ± 0.024 log10 CFU/mL) (Fig. 2A, B and Table S2). Before adding meropenem, all strains were resistant to colistin, with a MIC of 1024 µg/mL. However, colistin MIC values decreased after adding meropenem ranging from 2 to 256 µg/mL (fold change 4–8). Colistin-meropenem had a substantial bactericidal action when compared to colistin-linezolid and colistin-rifampicin (p < 0.05), determined at ¼MIC and hence could be the best eight-hourly drug regimen combination for M. morganii infection. This is concordant with a previous report on a time-kill study on Enterobacteriaceae that showed a bactericidal effect on bacterial growth when colistin-meropenem was tested (Brennan-Krohn et al. 2018). Similarly, the colistin-minocycline combination showed sustained synergism and bactericidal activity in UM573 and UM169 (Fig. 2A and B). However, in UM869, the combination did not result in the complete killing of bacteria at 24 h (Fig. 2C). For colistin-rifampicin, the combination failed to achieve complete eradication of bacteria in UM869 and UM169 at 24 h, whereas for UM573, no viable bacterial counts were observed at 24 h. The current work presented the first evidence of synergy between colistin and other antibiotic compounds against M. morganii infection. The most promising results point to combining colistin and meropenem as a new strategy for treating MDR M. morganii, with therapeutic potential to be proved in the future through in vitro and in vivo studies accommodating a larger number of MDR M. Morganii isolates against a range of most commonly prescribed antibiotics. Synergy testing on colistin with other antimicrobial combinations can be carried out to help clinicians to choose optimal antibiotic combinations against Gram-negative microorganisms.

Fig. 2.

Time-dependant killing analysis of colistin (CL) in combinations with minocycline (MN), meropenem (MR), Rifampin (RF) and Linezolid (LN) against A UM573, B UM169 and C UM869 at ¼MIC. Bacterium growth without any antibiotics (Untreated) was taken as control (blue)

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization: DUB, SD and ES; Formal analysis: DUB, KR, MG and SS; Investigation: DUB, KR; Methodology: ES and SD; Project administration: ES; Resources: BR and MR; Supervision: ES and RKS; Writing—original draft: DUB, SD, KR and ES; Writing—review and editing: SD, MG, KR, RKS and ES. All authors critically revised the document and agreed on its content.

Data Availability

All data have been included in the manuscript.

Declarations

Conflict of interest

The authors declare that they have no conflict of interest in the publication.