Molecular epidemiology and antimicrobial susceptibility of Pseudomonas spp. and Acinetobacter spp. from clinical samples at Jimma medical center, Ethiopia

Abstract

Introduction

Pseudomonas aeruginosa (P. aeruginosa) and Acinetobacter baumannii (A. baumannii) can cause difficult-to-treat infections. We characterized molecular epidemiology of ceftazidime-resistant P. aeruginosa and carbapenem-resistant A. baumannii at a tertiary hospital in Ethiopia.

Materials and methods

Non-fermenting gram-negative bacilli (n = 80) isolated from admitted patients were subjected for species identification by MALDI-TOF. Pseudomonas species resistant to ceftazidime or meropenem, and Acinetobacter species resistant to meropenem, or imipenem were selected for whole genome sequencing. DNA extracted with EZ1 Advanced XL instrument (Qiagen, Hilden, Germany) was sequenced on Illumina (HiSeq2500) using libraries prepared by NEXTRA-kits (Illumina). Raw reads were assembled using SPAdes 3.13.0, and assembled genomes were used to query databases for resistome profile and sequence types.

Result

Among Pseudomonas species isolated, 31.7% (13/41), and 7.3% (3/41) were non-susceptible to ceftazidime, and meropenem, respectively. Carbapenem-resistance was 56.4% (22/39) among Acinetobacter species. Moreover, 92% (12/13) of Pseudomonas species non-susceptible to ceftazidime and/or meropenem, and 89.4% (17/19) of Acinetobacter species encoded multiple resistance genes for at least three classes of antimicrobials. The prevalent β - lactamase genes were bla_OXA–486 (53.8%, 7/13), bla_CTX–M–15 (23.0%, 3/13) among Pseudomonas, and bla_GES–11 (57.8%, 11/19) among Acinetobacter. The bla_OXA–51-like β - lactamase, bla_OXA–69 (63.1%, 12/19) was the most prevalent carbapenemase gene among Acinetobacter isolates. Single isolates from both P. aeruginosa, and A. baumannii were detected with the bla_NDM–1. Sequence type (ST)1 A. baumannii and ST274 P. aeruginosa were the prevalent sequence types. A cgMLST analysis of the ST1 A. baumannii isolates showed that they were closely related and belonged to the international clonal complex one (ICC1). Similarly, ST274 P. aeruginosa isolates were clonally related.

Conclusion

The prevalence of MDR isolates of Pseudomonas and Acinetobacter spp. was high. A. baumannii isolates were clonally spreading in the admission wards at the hospital. Emergence of bla_NDM–1 in the intensive care, and surgical wards of the hospital is a severe threat that requires urgent intervention.

Introduction

Pseudomonas aeruginosa (P. aeruginosa) and Acinetobacter baumannii (A. baumannii) are among the main causes of nosocomial infections (De Oliveira et al., 2020). They belong to the group of bacteria known as “ESKAPE” (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, A. baumannii, P. aeruginosa, and Enterobacter species). Moreover, these group of bacteria are difficult to treat in most cases (Bergogne-Bérézin and Towner, 1996; De Oliveira et al., 2020; Ma et al., 2020).

The prevalence of antimicrobial resistant sub-populations of these strains has rapidly increased over the last few decades (Wong et al., 2017; Horcajada et al., 2019). Carbapenemase-producing A. baumannii (CRAB) and P. aeruginosa were listed top two of the three critical priority pathogens for which new antimicrobials are urgently needed (WHO, 2017). Several Acinetobacter and Pseudomonas species were previously reported from different clinical samples from both animal, and human infections (Chusri et al., 2014; Wong et al., 2017; Agnese et al., 2018). Many of them were resistant to multiple classes of antibiotics primarily by several intrinsic resistance mechanism they encode, and secondly by acquired resistance mechanisms (Bello-López et al., 2020; Meng et al., 2020). Studies have also shown that multidrug-resistant strains of P. aeruginosa and A. baumannii were the main drivers of hospital-acquired infections (Djahmi et al., 2014; Eichenberger and Thaden, 2019; Kazmierczak et al., 2020). A recent review of global epidemiology of carbapenemase-producing Gram-negative bacteria reported that carbapenemase-producing P. aeruginosa (CRPA) were associated with high mortality and morbidity among hospitalized patients with pneumonia and bloodstream infections in the United States (Brink, 2019). Though regional variations are common, colonization by carbapenem-resistant A. baumannii increased the risk of acquisition of bloodstream infection four-fold (Munoz-Price et al., 2016; Bassetti et al., 2017). In low-income countries like Ethiopia, comprehensive microbiological data is lacking.

The classical phenotyping methods commonly used in low-income countries cannot reliably define mechanism of resistance in both Pseudomonas and Acinetobacter species. Lack of sufficient standardized genotypic methods for detection and tracking of multidrug-resistant or extensively drug-resistant isolates was one of the challenges to understand epidemiology of antimicrobial resistance in sub-Saharan African countries (Eichenberger and Thaden, 2019). In most cases, global reports on antimicrobial resistance lack data from African countries. Despite discrepancies in availability of data, there is sufficient overall evidence that carbapenemase-producing Gram-negative bacilli have become a threat to global health. Rapid detection and tracking of any ongoing spread of resistant strains is necessary.

We aimed to analyze the phenotypic and molecular characteristics of Pseudomonas species and Acinetobacter species isolated from clinical samples at Jimma Medical Center (JMC), a tertiary hospital in Ethiopia.

Materials and methods

Isolation, identification, and selection of strains

As part of a large epidemiological study, a total of 1,087 clinical samples (urine, stools, wound secretions, and sputum) were collected from patients with suspected infections seeking medical care from June to October 2016 at JMC, Ethiopia. Pseudomonas species and Acinetobacter species were isolated on MacConkey agar and sheep blood agar. Species identification was performed by MALDI-TOF (Bruker Daltonik GmbH, Bremen, Germany) at Karolinska University Hospital (KUH), Clinical Microbiology laboratory, and a full panel of antimicrobial susceptibility testing was performed by using the EUCAST 2021 v11 guideline.¹

Antimicrobial susceptibility testing

All Pseudomonas species and Acinetobacter species isolated were subjected to disk-diffusion susceptibility testing. Antibiotic discs of ceftazidime, meropenem, piperacillin-tazobactam, gentamicin, amikacin, ciprofloxacin was used for Pseudomonas species. Similarly, all Acinetobacter isolates were tested by using meropenem, imipenem, gentamicin, amikacin, ciprofloxacin, trimethoprim-sulfamethoxazole. Then, isolates with reduced susceptibility to ceftazidime and/or meropenem for Pseudomonas spp., and meropenem or imipenem for Acinetobacter spp. were selected for antimicrobial susceptibility testing using the newer antimicrobials (cefiderocol, ceftazidime-avibactam, ceftolozane-tazobactam, and imipenem-relebactam for Pseudomonas spp., and cefiderocol, meropenem, imipenem, and imipenem-relebactam for Acinetobacter spp. using microbroth dilution technique), and whole genome sequencing (WGS). Patients’ clinical data like admission, presence of underlying chronic illnesses, current use of antibiotics, and other factors was collected using a structured questionnaire.

DNA extraction, whole genome sequencing and analysis of genomic data

Genomic DNA was extracted using Qiagen kits on EZ1 automated DNA extractions system. The extracted DNA was quantified using Qubit™ 3.0 (Massachusetts, United States) and library preps were performed using NEXTRA-kit (Illumina) and sequenced using HiSeq2500 (Illumina). Raw reads were assembled using SPAdes ver. 3.13.0, and the assembled draft genomes were used for querying different databases, MLST-typing 2.0, hosted at center for genomic epidemiology, and detection resistome profile by using ResFinder 4.1.^{2 ,3} Epidemiologic analysis of relatedness between the isolates, and to other international isolates was performed by the minimum spanning tree using the isolate genomes deposited at the public domain for A. baumannii at⁴, and P. aeruginosa at.⁵ The Genome sequences were deposited at the NCBI, SRA database (Bioproject number: PRJNA593604, Biosample accession: SUB11593554).

Results

Clinical and demographic characteristics of the patient

From a total of 1,087 non-repeat clinical samples collected during the study period, non-duplicate, non-fermenting Gram-negative bacilli that belong to either Pseudomonas spp. or Acinetobacter spp. were isolated from 80 patents. Most of these patients, 73.7% (59/80) were male, and 26.3% (21/80) were female. Ninety percent (72/80) of these patients were admitted to different wards in the hospital [surgical ward (n = 52), intensive care unit (n = 6), pediatric ward (n = 3), and medical ward (n = 11)]. Clinical samples collected include urine for urinary tract infections, sputum for lower respiratory tract infections, wound swab for wound infections/abscess, and stools for diarrhea. Most of these patients were admitted to the hospital for other underlying diseases and developed infection after admission, and most of them were treated with locally available antimicrobial therapy. Among Pseudomonas isolates 68.2% (28/41) were from surgical ward, followed by 17.0% (7/41) from medical ward, and 9.7% (4/41) from the intensive care unit (Table 1A). Similarly, 69.2% (27/39) of the Acinetobacter species were isolated from the surgical ward and 20.5% (8/39) from the medical ward (Table 1B).

TABLE 1A.

Socio-demographic and clinical characteristics patients and isolation Pseudomonas species.

Patient ID	Age	Sex	Inpatient/ outpatient	Current diagnosis	*Current antibiotic	Specimen	Underlying disease	Pseudomonas species
I020	28	M	Inpatient	Surgical site infection	CRO, MET	Wound swab	Surgical incision	P. aeruginosa
I032	28	M	Inpatient	Urinary tract infection	CRO, MET	Urine	Trauma	P. aeruginosa
I038	30	M	Inpatient	Urinary tract infection	CRO	Urine	Severe head injury	P. putida
I043	17	M	Inpatient	Urinary tract infection	CRO	Urine	Aspiration pneumonia	P. aeruginosa
M019	70	M	Inpatient	COPD	CRO, VAN	Sputum	Cor pulmonale	P. aeruginosa
M030	60	M	Inpatient	Community-acquired pneumonia	CRO	Sputum	**COPD	P. aeruginosa
M074	50	M	Inpatient	COPD	VAN, CIP	Sputum	Asthma	P. aeruginosa
M119	40	M	Inpatient	Pneumonia	CRO	Sputum	Post TB fibrosis	P. aeruginosa
M304	40	M	Inpatient	Community-acquired pneumonia	No	Sputum	Post TB fibrosis	P. aeruginosa
M334	35	F	Inpatient	Severe community-acquired pneumonia	CRO	Sputum	No	P. aeruginosa
M521	60	F	Outpatient	Community-acquired pneumonia	CRO	Sputum	**T2DM	P. aeruginosa
P014	14	M	Inpatient	Wound infection	AUG	Wound swab	No	P. aeruginosa
P109	4	F	Inpatient	Diarrhea	AMOX, GENT	Stool	SAM, pneumonia	P. aeruginosa
S007	32	M	Inpatient	Necrotizing fasciitis	No	Wound swab	No	P. aeruginosa
S010	5	M	Inpatient	Wound infection	CAF, CLO	Wound swab	Trauma	P. aeruginosa
S011	7	M	Inpatient	Wound infection	AMP, CAF	Wound swab	Trauma	P. aeruginosa
S017	18	M	Inpatient	Necrotizing fasciitis	CRO, MET	Wound swab	No	P. aeruginosa
S019	60	M	Inpatient	Foot ulcer	CRO, MET	Wound swab	Diabetes mellitus	P. aeruginosa
S020	30	M	Inpatient	Wound infection	CRO, MET	Wound swab	Trauma	P. aeruginosa
S036	17	F	Inpatient	Wound infection	No	Wound swab	Unstable pelvis fracture	P. aeruginosa
S047	10	M	Inpatient	Wound infection	AMP	Wound swab	Chronic osteomyelitis	P. aeruginosa
S048	15	M	Inpatient	Wound infection	AMP, CAF	Wound swab	Fracture of femoral shaft	P. aeruginosa
S077	3	M	Inpatient	Wound infection	AMP, GENT	Wound swab	Colostomy, imperforation	P. aeruginosa
S116	50	F	Inpatient	Wound infection	AMOX, MET	Wound swab	Uterine cancer	P. aeruginosa
S129	77	M	Inpatient	Wound infection	CRO, VAN	Wound swab	Amputation of leg	P. aeruginosa
S133	45	M	Inpatient	Wound infection	CRO, MET	Wound swab	Neck injury trauma	P. aeruginosa
S114	25	F	Inpatient	Acute kidney infection	CRO, MET	Urine	Rib fracture	P. aeruginosa
S155	21	F	Inpatient	Wound infection	AMP, CAF	Wound swab	Infected skin graft	P. aeruginosa
S174	60	M	Inpatient	Wound infection	CAF, AMP	Wound swab	Infected fracture site	P. aeruginosa
S192	75	M	Inpatient	Urinary tract infection	CRO	Urine	**BOO	P. putida
S195	21	F	Inpatient	Wound infection	CRO, MET	Wound swab	Skin graft infection	P. aeruginosa
S198	57	M	Outpatient	Wound infection	CLO, CAF	Wound swab	Left femoral fracture	P. aeruginosa
S209	18	F	Inpatient	Wound infection	CRO	Wound swab	Burn wound	P. aeruginosa
S248	20	F	Inpatient	Contaminated wound		Wound swab	No	P. aeruginosa
S288	50	M	Inpatient	Pneumonia	No	Sputum	Abdominal mass	P. aeruginosa
S319	40	M	Inpatient	Urinary tract infection	No	Urine	BPH	P. aeruginosa
S325	37	M	Inpatient	Wound infection	CAF, AMP	Wound swab	Compound distal fracture	P. aeruginosa
S328	60	M	Inpatient	Pneumonia	CRO, MET	Sputum	3rd degree burn	P. aeruginosa
S332	30	M	Inpatient	Wound infection	CRO, MET	Wound swab	2nd degree burn	P. fulva
S356	30	M	Inpatient	Wound infection	CRO, MET	Wound swab	Infected palate of right knee	P. aeruginosa
S371	64	M	Inpatient	Surgical site infection	CLO, CAF	Wound swab	Laparotomy	P. aeruginosa

*Current antibiotics: CRO, ceftriaxone; Met, metronidazole; VAN, vancomycin; CLO, cloxacillin; CAF, chloramphenicol; CIP, ciprofloxacin; AMP, ampicillin; GENT, gentamicin.

**Underlying diseases: COPD, congestive obstructive pulmonary disease; T2DM, type-2 diabetes mellitus; BOO, Bladder outlet obstruction; BPH, Benign prostatic hyperplasia.

TABLE 1B.

Socio-demographic and clinical characteristics patients and isolation Acinetobacter species.

Patient ID	Age	Sex	Inpatient/ outpatient	Current diagnosis	*Current antibiotic	Specimen	Underlying disease	Bacterial species
I027	45	F	Inpatient	Aspiration pneumonia	CRO, MET	Sputum	Stroke/hemiparalysis	A. baylyi
I030	28	F	Inpatient	Skin infection	No	Wound swab	No	A. baumannii
M029	22	M	Inpatient	Urinary tract infection	CRO	Urine	Retroviral infection	A. baumannii
M057	60	F	Inpatient	COPD	No	Sputum	No	A. baumannii
M135	70	F	Outpatient	Community-acquired pneumonia	CRO	Sputum	No	A. junii
M212	25	M	Inpatient	Pneumonia	CRO, VAN	Sputum	Disseminated tuberculosis	A. calcoaceticus
M217	35	M	Inpatient	Community-acquired pneumonia	No	Sputum	**COPD	A. schindleri
M328	50	F	Outpatient	Severe community-acquired pneumonia	No	Sputum	**T2DM	A. baumannii
M344	17	M	Outpatient	Pneumonia	No	Sputum	Electrical burn	A. junii
M431	50	F	Inpatient	Pneumonia	CRO	Sputum	Rt&Lt femoral fracture	A. baumannii
P015	30	M	Inpatient	Wound infection	CLO, CAF	Wound swab	Compound fracture of right tibia	A. baumannii
P038	35	M	Outpatient	Wound infection	AMP, CAF	Wound swab	Infected fracture site	A. baumannii
S073	15	M	Inpatient	Wound infection	CRO	Wound swab	Chronic osteomyelitis	A. baumannii
S082	45	M	Inpatient	Wound infection	CIP	Wound swab	Femoral fracture	A. baumannii
S126	40	M	Inpatient	Wound infection	CRO, MET	Wound swab	Surgical site infection	A. baumannii
S130	30	M	Inpatient	Wound infection	CRO, MET	Wound swab	3rd deg. burn	A. baumannii
S147	30	F	Inpatient	Wound infection	CRO	Wound swab	Colostomy	A. baumannii
S161	50	M	Inpatient	Wound infection	AMP, CAF	Wound swab	No	A. parvus
S165	8	M	Inpatient	Wound infection	CAF, CLO	Wound swab	No	A. baumannii
S167	40	M	Inpatient	Wound infection	CRO, MET	Wound swab	Scalp abscess	A. baumannii
S170	27	M	Inpatient	Wound infection	CRO, MET	Wound swab	No	A. baumannii
S171	32	M	Inpatient	Wound infection	CRO, MET	Wound swab	Retroviral infection	A. baumannii
S176	25	F	Inpatient	Necrotic wound infection	CRO, MET	Wound swab	No	A. baumannii
S209	50	M	Outpatient	Pneumonia	CRO, MET	Sputum	No	A. baumannii
S210	25	M	Inpatient	Wound infection	CRO	Wound swab	No	A. baumannii
S212	40	M	Inpatient	Wound infection	CRO, MET	Wound swab	Compound fracture left leg	A. baumannii
S217	48	M	Inpatient	Pneumonia	CRO, MET	Sputum	colostomy/post-operation	A. baumannii
S219	39	M	Inpatient	Wound infection	No	Wound swab	Tibia-fibular fracture	A. baumannii
S226	24	M	Inpatient	Urinary tract infection	No	Urine	Urethral stricture	A. haemolyticus
S247	40	M	Outpatient	Wound infection	AMP, CAF	Wound swab	Infected incision site	A. baumannii
S260	18	F	Inpatient	Wound infection	CRO	Wound swab	Burn wound	A. baumannii
S267	30	F	Inpatient	Wound infection	CRO, MET	Wound swab	Surgical site infection	A. baumannii
S270	26	M	Inpatient	Wound infection	CRO, MET	Wound swab	Chronic osteomyelitis	A. baumannii
S275	48	M	Inpatient	Wound infection	No	Wound swab	Compound fracture of femur	A. baumannii
S294	70	M	Inpatient	Urinary tract infection	CRO	Urine	Bladder outlet obstruction	A. baumannii
S296	10	M	Inpatient	Wound infection	AMP, CAF	Wound swab	Left femoral fracture	A. baumannii
S315	2	M	Inpatient	Diarrhea	AMP, GENT, CLO	Stool	No	A. lwoffii
S327	29	M	Inpatient	Severe community-acquired pneumonia	CRO, MET	Sputum	No	A. baumannii
S347	18	F	Inpatient	Urinary tract infection	AMP, CAF	Urine	Cervical Spine fracture	A. baumannii

*Current antibiotics: CRO, ceftriaxone; Met, metronidazole; VAN, vancomycin; CLO, cloxacillin; CAF, chloramphenicol; CIP, ciprofloxacin; AMP, ampicillin; GENT, gentamicin.

**Underlying diseases: COPD, congestive obstructive pulmonary disease; T2DM, type-2 diabetes mellitus.

Species diversity and antimicrobial susceptibility

Species diversity

From a total of 41 isolates of Pseudomonas spp., 92.6% (38/41) were P. aeruginosa, and 7.3% (3/41) were another Pseudomonas spp. Among 39 isolates of Acinetobacter spp. (n = 39), 79.5% (31/39) were A. baumannii, and 20.5% (8/39) were other Acinetobacter spp. [A. junii which account for 5.1% (2/39), and A. baylyi, A. lwoffii, A. calcoaceticus, A. haemolyticus, A. parvus, and A. schindleri each account for 2.5% (1/39)].

Antimicrobial susceptibility pattern

Among isolates of Pseudomonas spp., 31.7%, (13/41) were non-susceptible to ceftazidime, and 7.3% (3/41) to meropenem. Among isolates of Acinetobacter spp., 56.4% (22/39) were carbapenem-resistant. Most of the ceftazidime-resistant Pseudomonas spp. were also resistant to piperacillin-tazobactam 84.6% (11/13), ciprofloxacin 30.7% (4/13), ceftazidime-avibactam 46.1% (6/13) and ceftolozane-tazobactam 53.8% (7/13). However, most of these isolates were susceptible to cefiderocol 84.7% (11/13), imipenem-relebactam 92.3% (12/13), and all Pseudomonas isolates were susceptible to amikacin (Table 2A). Most of the Acinetobacter isolates were resistant to meropenem, imipenem, and imipenem-relebactam 54.5%, (12/22). However, a lower rate of resistance was detected to ciprofloxacin 18.1% (4/22), gentamicin 27.2% (6/22), amikacin 18.1% (4/22), and cefiderocol 9.1% (2/22) (Table 2B).

TABLE 2A.

Antimicrobial susceptibility pattern of ceftazidime resistant Pseudomonas spp. isolated at JMC, Ethiopia.

Bacterial species		Type of antimicrobials used for antimicrobial susceptibility testing
		Ceftazidime	Meropenem	Imipenem	Piperacillin-tazobactam	Ciprofloxacin	Amikacin	Cefiderocol	Ceftazidime-avibactam	Ceftolozane-tazobactam	Imipenem-relebactam
Pseudomonas aeruginosa (n = 10)	S (%)	0 (0.0%)	7 (70)	8 (80)	0 (0)	7 (70)	10 (100)	9 (90)	5 (50)	6 (60)	9 (90)
	R (%)	10 (100)	3 (30)	2 (20)	10 (100)	3 (30)	0 (0)	1 (10)	5 (50)	4 (40)	1 (10)
Other Pseudomonas species (n = 3)	S (%)	1 (32.4)	1 (32.4)	3 (100)	1 (32.4)	2 (66.6)	3 (100)	2 (66.6)	2 (66.6)	0 (0)	3 (100)
	R (%)	2 (66.6)	2 (66.6)	0 (0)	2 (66.6)	1 (32.4)	0 (0)	1 (32.4)	1 (32.4)	3 (100)	0 (0)

S, susceptible; R, resistant, other Pseudomonas species [Pseudomonas putida (n = 2), Pseudomonas fulva (n = 1)].

TABLE 2B.

Antimicrobial susceptibility pattern of carbapenem resistant Acinetobacter species isolated at JMC, Ethiopia.

Bacterial species		Type of antimicrobials used for antimicrobial susceptibility testing
		Meropenem	Imipenem	Imipenem-relebactam	Cefiderocol	Ciprofloxacin	Trimethoprim/sulfamethoxazole	Gentamicin	Amikacin
Acinetobacter baumannii (n = 22)	S (%)	5 (27.8)	9 (50)	9 (50)	17 (94.4)	14 (77.8)	12 (66.7)	12 (66.7)	15 (83.3)
	R (%)	13 (72.2)	9 (50)	9 (50)	1 (5.6)	4 (22.2)	6 (33.3)	6 (33.3)	3 (16.7)
Other Acinetobacter spp.	S (%)	1 (25)	1 (25)	1 (25)	4 (100)	1 (25)	0 (0)	4 (100)	4 (100)
	R (%)	3 (75)	3 (75)	3 (75)	0 (0)	3 (75)	4 (100)	0 (0)	(100)

S, susceptible; R, resistant; Other Acinetobacter spp. (n = 4). (A. calcoaceticus, A. baylyi, A. junii, and A. lwoffii).

Resistome profile of Pseudomonas and Acinetobacter species

Resistome profiling showed that both Pseudomonas and Acinetobacter spp. encoded multiple β - lactamase genes. The bla_OXA–486, bla_OXA–50 and bla_CTX–M–15 genes were most common among Pseudomonas isolates, and the bla_OXA–51-like bla_OXA–69, bla_OXA–66, bla_OXA–91, bla_OXA–180 and bla_GES–11 were the most common among Acinetobacter isolates (Table 3A).

TABLE 3A.

Sequence types and resistance genes observed among carbapenem-resistant Acinetobacter species isolated at Jimma Medical Center, Ethiopia (n = 19).

Isolate ID	STs	Number and type of antimicrobial resistance genes
		Carbapenemase	ESBL and other β -lactamases	Aminoglycosides	Trimethoprim	Sulfonamides	Tetracycline	Macrolides
I027AB*	ST2	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
I030AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
M217AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA1,	sul1
M328AB	164	bla OXA–91	blaCARB–5, blaCARB–49				tet(39)
P015AL	NA		blaCTX–M–15, blaOXA–1,	aac(6′)-Ib-cr			tet(A)	mdf(A), mph(A), mph(D)
S126AB	ST2090	bla OXA–259/180
S130AB	ST85	blaNDM–1, blaOXA–94		ant(2″)-Ia, aph(3′)-VI		sul2		mph(E) msr(E)
S161AB	ST2	bla OXA–66	blaTEM–1D,	aac(3)-Ia, aph(6)-Id		sul1
S167AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib-cr, aac(6′)-Ib3	dfrA7	sul1
S170AB	ST164	bla OXA–91	blaCARB–49, blaCARB–16				tet(39)
S171AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S176AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S209AB	ST2	bla OXA–66		aac(3)-Ia, aadA1, aph(6)-Id		sul1sul2	tet(B)
S212AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr,	dfrA7	sul1
S270AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S275AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S296AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S315AB	ST1	bla OXA–69	bla GES–11	aac(6′)-Ib3, aac(6′)-Ib-cr	dfrA7	sul1
S327AB	ST1	blaOXA–69

Moreover, two isolates, P. aeruginosa (n = 1) and A. baumannii (n = 1) carrying the bla_NDM–1 carbapenemase gene were detected. Furthermore, the resistome profile shows that resistance genes to the antimicrobial classes of aminoglycosides, fluoroquinolones, and trimethoprim were prevalent among these two isolates (Tables 3A,B).

TABLE 3B.

Sequence types and resistance genes observed among ceftazidime-resistant Pseudomonas spp. isolated at Jimma Medical Center, Ethiopia (n = 13).

Strain ID	STs	Number and type of antimicrobial resistance genes
		Carbapenemase	β -lactamase genes	Aminoglycosides	Fluoroquinolones	Sulfonamides	Phenicols	Fosfomycin
I020PA*	11		blaOXA–486, blaPAO	aph (3′)-IIb			catB7	fosA
I032PA	2948	bla NDM–1	blaOXA–10, blaOXA–50	aadA1, ant(2″)-Ia, aph(3′)-IIb	crpP	sul1	catB7, cmlA1	fosA
I038PP*	NA		blaCTX–M–15, blaOXA–1	aac(6′)-Ib-cr,
M019PA	1228		blavEB–1, blaOXA–486 blaOXA–10	aadA1, aadA2 ant(2″)-Ia, aph(3′)-IIb	crpP qnrD1	sul1	catB7 cmlA1	fosA
M119PA	274		blaOXA–486, blaPAO,				catB7	fosA
M304PA	274		blaOXA–486, blaPAO				catB7	fosA
S116PA	500		blaOXA–486, blaPAO		crpP		catB7	fosA
S114PP*	NA		blaCTX–M–15, blaOXA–10 blaOXA–1	aac(3)-Iia, aac(3)-Ib aac(6′)-Ib, aac(6′)-Ib-cr, ant(3″)-Ia, aph(3″)-Ib	oqxB qnrVC1	sul1
S155PA	274		blaOXA–486, blaPAO	aph(3′)-IIb			catB7	fosA
S248PA	840		blaOXA–486, blaPAO	aph(3′)-IIb	crpP		catB7	fosA
S332PF**	NA		blaPER–1,	aph(6)-Id	qnrVC6	sul1
S356PA	646		blaCTX–M–15, blaOXA–494 blaOXA–50, blaTEM–1B	ant(2″)-Ia, aph(3′)-IIb aph(3′)-Ia, ph(6)-Id		sul1	catB7, floR	fosA
S047PA	244		bla OXA–50	aph(3′)-IIb			catB7	fosA

*A. baumannii; PA, Pseudomonas aeruginosa; PP**, Pseudomonas putida; PF**, Pseudomonas fulva.

Molecular epidemiology using cgMLST

Epidemiologic typing using the seven gene multi-locus sequence typing demonstrated that ST274 (23.0%, 3/13) among P. aeruginosa and ST1 (MLST-Pasteur) A. baumannii (63.1%, 12/19) were the most prevalent sequence types (Tables 3A,B). A cgMLST analysis showed that the ST1 A. baumannii isolates were highly similar with no allelic differences between them. Most of the Acinetobacter isolates belonged to the international clonal complexes, ICC1 (includes ST1) (n = 12) and ICC2 (n = 2) (Table 3A).

Discussion

In this study, nearly all isolates of both Pseudomonas and Acinetobacter spp. were from patients admitted to the hospital for more than 72 h. Nosocomial acquisition of MDR isolates of Pseudomonas spp. and Acinetobacter spp. is worrisome. The situation is complicated by limited availability of antimicrobial agents, lack of prescription guidelines, and insufficient standard routine microbiology laboratory services to support antibiotic selection. In such cases, the safety of patients admitted to the hospital can be severely compromised.

The prevalence of ESBL-producing strains among P. aeruginosa, and carbapenem non-susceptible isolates among A. baumannii strains was high in the present study. Previous phenotypic studies from Ethiopia also show that MDR strains of Pseudomonas and Acinetobacter were prevalent (Motbainor et al., 2020; Birhane Fiseha et al., 2021). However, most of these studies did not describe the genotypes and resistome profile of the isolates, and mechanism of resistance is difficult to compare between studies. So far, to our knowledge, there is no report of a genotypic study on the prevalence of ESBL-producing P. aeruginosa in Ethiopia. Generally, there are limited studies conducted on carbapenemase-producing P. aeruginosa in Africa. The few reports available are from northern Africa and mainly from Egypt (Osei Sekyere and Reta, 2020; Rizk et al., 2021). In northern Africa, the prevalence ranges from 0 to 96% (Gaballah et al., 2018). A finding from Uganda (7.4%) was comparable to the present study (Aruhomukama et al., 2019). But, one study from South Africa (51.0%) reported a higher prevalence of ceftazidime resistant P. aeruginosa compared to the present study (Hosu et al., 2021). The phenotypic studies from Ethiopia, and other African countries showed higher prevalence as compared to the present study.

In Ethiopia, a high prevalence of carbapenem-resistant Acinetobacter spp. was reported from one previous phenotypic study (Ayenew et al., 2021), which is comparable to the prevalence of carbapenem resistance herein. On the other hand, a systematic review and meta analysis on carbapenemase producing P. aeruginosa and A. baumannii in Africa showed that the lowest prevalence of carbapenemase-producing A. baumannii was 4.7% (n = 21), and the highest prevalence was 100% (n = 7) (Kindu et al., 2020). Studies conducted on P. aeruginosa and A. baumannii strains from Africa were limited to small sample sizes and were mainly phenotypic studies, which makes the comparison with genotypic studies difficult (Kindu et al., 2020; Osei Sekyere and Reta, 2020; Ayenew et al., 2021; Mekonnen et al., 2021). However, most available studies including a study for hospital environment (Solomon et al., 2017) reported higher prevalence of the MDR P. aeruginosa and A. baumannii in Ethiopia, calling for the application of genotypic methods to studies on mechanisms of resistance and spread.

The multiple genetic variants of antibiotic resistance observed among both Pseudomonas and Acinetobacter spp. pose a huge challenge on the limited therapeutic options available to low-income countries. Among Acinetobacter spp., the presence of the bla_GES–11 (ESBL-genotype and weak carbapenemase) and the OXA-51-like (bla_OXA–66 and bla_OXA–69) carbapenemases encoding genes is a serious threat. The OXA-51-like intrinsic carbapenemase encoding ST1 A. baumannii Isolates were reported from India (Rose et al., 2021), but isolates from the present study encoded additionally the weakly carbapenem hydrolyzing bla_GES–11 gene. The bla_GES–11 ESBL-genotypes, and the bla_OXA–51-like carbapenemases have not been previously reported from Ethiopia, and the present study is to our knowledge the first report of bla_OXA–51 like and bla_GES–11 from Ethiopia. Generally from Africa, only one study from Tunisia has documented clinical isolates of A. baumannii encoding the bla_GES–11 (Chihi et al., 2016). The emergence of bla_NDM–1 encoding isolates of A. baumannii at surgical ward, and P. aeruginosa at intensive care unit can severely compromise the safety of vulnerable patients admitted to the hospital. Moreover, detection of two isolates encoding the bla_NDM–1 which showed resistance to the newer antimicrobials, A. baumannii for (cefiderocol and imipenem-relebactam), and P. aeruginosa for (cefiderocol, ceftazidime-avibactam, ceftolozane-tazobactam) compromises the already limited treatment alternatives for vulnerable groups of patients at this hospital.

Sequence typing showed that both P. aeruginosa and A. baumannii strains were polyclonal. But, since a large proportion of A. baumannii isolates were ST1, the spread of A. baumannii isolates at this hospital might also be to some extent clonal. A previous study conducted at the same hospital had identified three strains of A. baumannii encoding bla_NDM–1, and all of them belonged to the ST597. Furthermore, cgMLST analysis of A. baumannii with other international isolates in pubmlst showed that isolates in the current study were distinct from isolates from other African countries (Figure 1).⁶ However, these isolates were clustered with isolates from other countries like United States, and Brazil. Though isolates from this study were polyclonal, the most prevalent isolates were those that belong to the international cluster, CC1 and CC2.

FIGURE 1.

(A) Epidemiologic relatedness of A. baumannii isolates to the isolates collected from different geographical regions and deposited in the public database (pubmlst https://pubmlst.org/organisms/acinetobacter-baumannii~-- accessed on 26-09-2021), the isolates from this study (bright green) were clustered with the global ICC1 clone marked with red circle. The numbers in the circle show the sequence type of the isolates, the size of the circle is proportional to the number of isolates that belongs to that sequence type, and the color shows the country of origin for the isolates. The minimum spanning tree was constructed for isolates from different country of origin having ≥ 10 isolates recorded in the database. (B) Minimum spanning tree of 29 A. baumannii isolates from Africa, the tree was constructed based on core genome multi-locus sequence typing. There were no previous isolates from Ethiopia in PubMLST, all isolates labeled Ethiopia in the tree are from this study.

Although A. baumannii is a well-known major cause of nosocomial infections, knowledge of its genomic epidemiology and availability of reliable data regarding the genetic basis of antibiotic resistance is limited in low-income countries. Similarly, P. aeruginosa isolates were found to be polyclonal, and different from a collection of isolates other African counties found in pubmlst (see text footnote 5) (Figure 2).

FIGURE 2.

Grape tree generated using cgMLST from genomic collection of Pseudomonas aeruginosa isolates from Africa (n = 22), Ethiopia, Nigeria, and South Africa, pubmlst database for microorganisms (https://pubmlst.org/bigsdb?db=pubmlst_paeruginosa_isolates&page=query&genomes=1) accessed on 29-11-2021. The numbers in circles/nodes indicate sequence types of the isolates, size of the nodes is proportional to the number of isolates in a particular sequence type and the numbers on the branches are the number of allelic differences between the two neighboring nodes. From 10 isolates sequenced in this study, eight were included in the tree, two of the sequences failed quality required to upload to the pubmlst database.

The current study may serve as a baseline regarding local spread of international clones and alert clinicians and other health workers, researchers, and public health policy makers to the problem. Implementation of strict infection prevention and control strategies, and antimicrobial stewardship programs are highly desirable in the admission wards where the international clones are spreading. Furthermore, despite limitation of resources, the added value of next generation sequencing is in understanding the dynamics and mechanisms of spread of MDR bacterial clones.

Conclusion

The prevalence of MDR isolates is high among both in clinical isolates of Pseudomonas species and Acinetobacter species at Jimma medical center. Emergence of the bla_NDM–1 in clinical isolates of P. aeruginosa and A. baumannii strains is worrisome. However, the susceptibility of P. aeruginosa strains to amikacin, cefiderocol, imipenem-relebactam and ceftolozane-tazobactam, and A. baumannii strains to amikacin and cefiderocol is important to consider as alternative options to carbapenems. The use of next generation sequencing is important to understand the mechanism of resistance and spread of resistant clones such as ICC1, and ICC2 A. baumannii strains detected at this hospital.

Data availability statement

The genome sequences were deposited at the NCBI, SRA database (PRJNA593604, Biosample accession: SUB11593554).

Ethics statement

The study obtained ethical approval from Addis Ababa University Institutional Review Board (AAU-IRB), Armauer Hansen Research Institute – ALERT Hospital Institutional Review Board (AHRI-ALERT-IRB), and Ethiopian National Ethics Review Committee (NERC). Patients were informed about the study and given written consent to participate in the study.

Author contributions

TS contributed to design, data acquisition, data analysis, and drafting of the manuscript. DA and YW contributed to data acquisition and write-up. AA contributed to the design, data acquisition, and revision of the manuscript. CG contributed to the overall design, data acquisition, supervision, drafting, and writing of the manuscript. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the Swedish International Development Agency (SIDA).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.