Meningococcal Antibiotic Resistance: Molecular Characterization of Isolates from Patients with Invasive Meningococcal Disease (IMD) in Greece

For effective case management and chemoprophylaxis of Invasive Meningococcal Disease (IMD), prompt antibiotic treatment is required. N. meningitidis is usually susceptible to antibiotics, but reduced susceptibility to penicillin, ciprofloxacin, and rifampicin is increasing worldwide, jeopardizing patients’ outcome. We assessed, phenotypically and genotypically, the antimicrobial resistance patterns of 192 strains isolated from IMD cases from all over Greece during 2010–2021. Antimicrobial susceptibility to penicillin, rifampicin, and ciprofloxacin was determined using the E-test. All isolates were genotyped by Multilocus Sequence Typing (MLST). penA, rpoB, and gyrA genes were amplified by PCR and sequenced. Of the 192 isolates, 37% (72/192) were penicillin-susceptible/had increased exposure, and 11% (21/192) were penicillin-resistant. Among those, 40 penA alleles were identified; penA1, penA27, and penA3 were highly associated with susceptibility to penicillin; penA14, penA25, and penA22 related to reduced susceptibility to penicillin, while penA9, penA910, and penA295 had resistance to penicillin. Two ciprofloxacin-resistant isolates harbored the gyrA346 allele, while one rifampicin-resistant isolate harbored the rpoB5 allele. Resistance to ciprofloxacin and rifampicin remains rare. As Greece is one of the countries with high antimicrobial resistance, continued monitoring of antibiotic resistance is important to ensure timely detection of emerging resistance for treatment and prevention guidelines.


Introduction
Invasive meningococcal disease (IMD), caused by Neisseria meningitidis, is a severe, life-threatening illness posing a major worldwide health problem as an important cause of morbidity and mortality. Meningitis, or septicaemia, are the main clinical presentations of IMD. Globally, there are approximately half a million cases of IMD each year [1], with the incidence varying across geographical regions and case fatality rates ranging from 4.1% to 20% [2]. In Europe, an average incidence rate of 0.6 cases per 100,000 people was reported [3], with the highest incidence rates observed in children <1 year old, followed by a second peak amongst adolescents and young adults [3]. Worldwide, the serogroups responsible for the majority of IMD cases are A, B, C, W, X, and Y [1,4], while, in Europe, the most prevalent serogroups are B and C. Although, in recent years, an increase in IMD due to serogroup W has been reported [3]. Multilocus sequence typing (MLST) classifies meningococcal strains into different sequence types (STs), with related STs grouped into lineages termed clonal complexes (ccs) [5].
Due to the severity of the disease, suitable and prompt patient management is of great importance with regards to the administration of antibiotics, even prior to admission to the

Susceptibility to Penicillin in Relation to Clonal Complexes
Overall, the most prevalent clonal complex was cc269 (21%), followed by cc41/44 (15%) (all related to MenB), while the most prevalent clonal complexes among Pen S isolates were cc269, cc32, and cc23. In contrast, cc213 and cc865 were most prevalent among the Pen R isolates ( Figure 3). The majority of Pen R isolates belonged to MenB (95%; 20/21), while 5% (1/21) belonged to MenW. A decrease in the proportion of MenB Pen S isolates was observed over the study period, ranging from 79% in 2012 to 11% in 2018 (data not shown). No Pen R isolates were detected among MenY and MenC isolates.

The Distribution of penA Alleles
Forty (40) penA alleles were identified (Table S1). penA1, penA27, and penA3 were highly associated with a susceptibility to penicillin (95%, 94%, and 71%, respectively). Alleles penA14, penA25, and penA22 were highly associated with a reduced susceptibility to penicillin (85%, 78%, and 46%, respectively) (MIC values 0.094-0.25 mg/L). In contrast, penA9, penA910, and penA295 were highly associated with resistance to penicillin (100% for penA9 and penA910 and 75% for penA295) (Figure 4). A reduction in the penA27 allele was observed from 2010 to 2013, following a further reduction from 2014 to 2021, while alleles penA1 and penA3 were relatively stable throughout the study period. Similarly, there was not a significant change among the alleles highly associated with a reduced susceptibility to penicillin (penA14, penA22, and penA25) during the study period. In contrast, the alleles highly associated with penicillin resistance (penA9, penA910, and penA295) were detected from 2015 onwards ( Figure 5). A reduction in the penA27 allele was observed from 2010 to 2013, following a further reduction from 2014 to 2021, while alleles penA1 and penA3 were relatively stable throughout the study period. Similarly, there was not a significant change among the alleles highly associated with a reduced susceptibility to penicillin (penA14, penA22, and penA25) during the study period. In contrast, the alleles highly associated with penicillin resistance (penA9, penA910, and penA295) were detected from 2015 onwards ( Figure 5). Two new penA alleles were identified (Table S2) Alleles penA3 and penA27 were found in MenB isolates and were highly associated Two new penA alleles were identified (Table S2)

Discussion
The present study describes, for the first time in Greece, the genotypic detection of antibiotic resistance to penicillin, ciprofloxacin, and rifampicin among N. meningitidis isolated from IMD cases. In a similar previous Greek study, during the years 1989 to 1991 [24], no resistance to ciprofloxacin and rifampicin was reported. However, nearly half the isolates (48.3%) were reported with reduced susceptibility to penicillin. This percentage was quite high, mainly due to different penicillin breakpoints implemented at that time. In the present study, 52% of isolates were penicillin-susceptible, and standard exposure (Pen S ) was in line with studies from Italy (55%) [7] and lower than in the UK (63%) [8]. Furthermore, 37% exhibited penicillin susceptibility, increased exposure (Pen I ) in agreement with studies from UK (34%) [8], and lower than Italy (45%) [7] and Australia (59%) [13]. The present study showed that 11% of isolates were penicillin-resistant, in agreement with data from Australia (13%) [13] and higher than those from the UK (3%) [8], the USA (<0.1%) [9], and Italy (0.7%) [7].
During the 12-year study period, the percentage of Pen I isolates has increased over time, which is consistent with global trends, as multiple countries have reported an increased proportion of Pen I isolates since 2000 [7,10,11,27]. Although there is no evidence to suggest that isolates with reduced susceptibility are associated with treatment failure, the increased Pen I isolates could still pose a concern for individuals with hereditary or acquired complement deficiencies and persons being treated with complement inhibitors, who sometimes receive long-term penicillin prophylaxis [4,28].
In total, 40 penA alleles were identified; alleles penA9, penA910, and penA295 were highly associated with penicillin resistance in line with other studies, for which penA9 was also identified in Pen R isolates in studies from the UK [8], Belgium [10], and Japan [12], while the penA295 allele was also detected among the Pen R isolates in a study from the UK [8]. The penA910 allele was identified among the Greek Pen R isolates; to our knowledge, this allele has not been reported in Pen R isolates from other studies. Alleles penA14, penA25, and penA22 were found to be highly associated with reduced susceptibility to penicillin. Our findings come to an agreement with studies from Italy [7] and the UK [8], where the penA14 has also been reported in Pen I isolates. The identification of Pen I isolates harboring the penA25 allele is in agreement with studies from the UK [8] and Italy [7], although we detected it in a higher frequency. The penA22 allele was identified mostly in Pen I isolates, which is in agreement with the UK study [8]. However this allele has also been identified in Pen S isolates, in agreement with a study from Ireland [27]. Alleles penA1, penA27, and penA3 were found to be highly associated with susceptibility to penicillin. The penA1 allele has been detected in Pen S isolates in studies from the UK [8], Ireland [27], and Brazil [29], although in a higher proportion compared to our study. In contrast, the penA27 allele has also been identified in Pen S isolates in studies from the UK [8] and Ireland [27], but in a lower proportion compared to our findings. penA3 was mostly associated with Pen S isolates, which is in line with studies from the UK [8], Ireland [27], and Brazil [29].
The majority of MenB cc269 and cc32 isolates were Pen S and harbored the penA27 and penA3 alleles, respectively. The majority of MenB cc162 Pen I isolates were highly associated with the penA25 allele, in contrast to a previous study where the penA14 allele has been reported in Pen I MenB cc162 isolates [7]. Half of the MenB cc213 isolates harbored the penA295 allele and were either Pen I or Pen R , in line with a previous study [7]. The majority of the MenY isolates, belonging to cc23 [7,[30][31][32][33], harbored the penA22 allele and were either Pen S or Pen I , in contrast to previous studies [7], where the penA20 allele was identified among the MenY cc23 Pen I isolates. Most of the isolates belonging to the hypervirulent clone cc11, mainly associated with MenC [7,30,31], were Pen I , harboring mainly the penA248 allele, in agreement with a previous study [7]. Although the invasive MenW cc11 lineage has been detected in our study, the majority of isolates were Pen S , highly associated with the penA1 allele, while only one isolate was Pen R , harboring the penA9 allele that has been reported in Pen R strains circulating in several countries [12,34].
Although ciprofloxacin resistance is rare worldwide [7][8][9]12,29,30], a recent study from China has shown that the average ciprofloxacin susceptibility rate across all serogroups was 24.9% [35], and, since 2004, all ciprofloxacin-resistant isolates of the various clonal complexes contained a T91I mutation in the gyrA gene, with more genetic diversity for gyrA compared with susceptible strains [36]. This causes an ongoing concern regarding the increasing prevalence of ciprofloxacin-resistant N. meningitidis globally. In the present study, only two isolates were ciprofloxacin-resistant (MIC = 0.25 mg/L), harboring the gyrA346 allele with a T91I mutation, belonging to ST-3129 (unassigned cc) and closely related to an isolate from China, indicating that the ciprofloxacin-resistant isolates were imported. Further, 99% of the isolates were sensitive to ciprofloxacin; the gyrA4 allele was the most prevalent and in agreement with other studies [31].
Almost all isolates were sensitive to rifampicin (99.5%; 191/192). The most frequent allele was rpoB4. However, one strain was found resistant and derived from a sporadic case with no evidence of being a close contact of an index case, harboring the rpoB5 allele. However, no mutations in the rpoB gene associated with high rifampicin resistance was found, in contrast to a previous study in Brazil where the allele rpoB14, which is associated with high rifampicin resistance, was detected in one isolate [29]. Nonetheless, rifampicin resistance seems to be stable in Greece, as it remains low and in agreement with the previous study [24]. Further analysis is needed to identify the possible mechanisms responsible for the resistance to rifampicin of this particular strain.
Some limitations in our study include that our results reflect only culture confirmed cases due to early antibiotic treatment prior to sampling (40% vs. 60% non-culture confirmed). Furthermore, due to the COVID-19 pandemic, the number of cases and isolates was low during the years 2020-2021 due to the sharp decrease in IMD cases in Greece [37] and worldwide [38]. Nonetheless, this study provides the first insight into the molecular characterization of invasive N. meningitidis isolates in Greece as data collected represent the whole country.
From a public health perspective, as Greece has a high percentage of antimicrobial resistance in both the community and hospital settings, monitoring the antimicrobial susceptibility of N. meningitidis is of utmost importance. Sustained genomic surveillance of AMR among meningococci is essential to monitor for the emergence and evolution of resistant clones carrying specific resistance genes that could jeopardize the effective national IMD chemoprophylaxis and treatment strategies.

Source of Specimens
A total of 486 IMD cases were confirmed for the period 2010-2021 (average incidence 0.4 per 100,000 population). All samples, cerebrospinal fluid (CSF), and/or blood (depending on the patient's clinical presentation), as well as isolates, were sent to the National Meningitis Reference Laboratory (NMRL) from hospitals throughout the country for further identification. However, due to early administration of antibiotic treatment prior to sampling, 40% (192/486) of the cases were cultured and confirmed, while 60% (294/486) were confirmed solely by PCR, for which no culture isolates were available. As the present study aims susceptibility testing, we focused on the 192 culture confirmed cases. Patients' age ranged from <12 months to 88 years old (median age 18 years), while 90 patients were males, and 102 were females.

Identification
Meningococcal isolates were cultured on Chocolate Columbia Agar (OXOID Ltd., Basingstoke, UK) and incubated at 37 • C and 5-10% CO 2 for 24 h, with a further N. meningitidis confirmation by the application of a multiplex PCR, as previously described [39]. Serogroups were determined by a slide agglutination test (Remel Europe Ltd., Dartford, Kent, UK) according to the manufacturer's instructions.

Antibiotic Susceptibility Testing
E-test was deployed for determining the Minimum Inhibitory Concentration (MIC) for penicillin, rifampicin, and ciprofloxacin by the use of MIC test strip methods (LIOFILCHEM S.r.l, Teramo, Italy) on Mueller-Hinton agar supplemented with 5% sheep blood (OXOID Ltd., Basingstoke, UK) and incubated at 37 • C according to the manufacturer's instructions. The values were interpreted according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST; v13.0; 1 January 2023).
Isolates of penicillin MIC value ≤0.064 mg/L were categorised as 'susceptible, standard dosing regimen (S)' (where there is a high likelihood of therapeutic success using a standard dosing regimen of the agent) and isolates of penicillin MIC value >0.25 mg/L were categorised as 'resistant (R)' (where there is a high likelihood of therapeutic failure even when there is increased exposure). Isolates exhibiting intermediate penicillin MIC values 0.094-0.25 mg/L were categorised as 'susceptible, increased exposure (I)' (where there is a high likelihood of therapeutic success because exposure to the agent is increased by adjusting the dosing regimen or by its concentration at the site of infection) [40].
Isolates of ciprofloxacin MIC value ≤0.016 mg/L and rifampicin MIC value ≤0.25 mg/L were categorized as susceptible.

Molecular Identification of penA, gyrA and rpoB Genes
A 402 bp fragment (penA) of the NEIS1753 (PBP2) gene was characterised by PCR by the use of penA1F and penA1R primers, as previously described by Taha et al. [15]. A 525 bp fragment of the Quinolone Resistance-Determining Region (QRDR) of gyrA gene was amplified by PCR, by the use of gyrA1F and gyrA1R primers, previously described by Hong et al. [42], while a 660 bp fragment of the rpoB gene, was characterized by PCR by the use of RpoB1F and RpoB1R primers, as previously described by Taha et al. [43] (Table S3). Positive and negative controls were included in each respective PCR assay. In particular, the N. meningitidis reference strains 4954, 4955, and 4956 were obtained from the UK NEQAS (in the frame of IBDLabNet External Quality Assurance scheme), which were used as positive controls, while PCR-grade water (Invitrogen TM , Life Technologies Corporation, Waltham, MA, USA) was used as a negative control.

PCR Product Purification and Sequencing
PCR products were purified according to the PCR-clean-up protocol, NucleoSpin ® Gel, and a PCR clean-up kit (Macherey-Nagel, Düren, Germany) in a 30 µL final elution volume. Further, in order to test the purification yield, 5 µL of the purified product was stained with 1 µL GelRed loading buffer (6× Gel loading dye, Biotium, Fremont, CA, USA). Purified products were subjected to electrophoresis in 2.0% (w/v) agarose gel (Nippon Genetics, Tokyo, Japan) and visualized under ultraviolet fluorescence light. The purified products were sequenced by Sanger on an ABI 3730xl DNA analyzer (Applied Biosystems™, ThermoFisher Scientific, Waltham, MA, USA) using the BigDye™ Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems™, ThermoFisher Scientific, Waltham, MA, USA).

Conclusions
In conclusion, we demonstrated that, during 12 years of IMD surveillance in Greece, resistance to ciprofloxacin and rifampicin remained rare. However, decreasing susceptibility was observed for penicillin. Phenotypic antimicrobial resistance surveys of isolates collected from Greek IMD cases, along with genetic investigations into the mechanisms of resistance, are important for ensuring that current antibiotic treatment and prophylaxis recommendations remain relevant.
Funding: This research received no external funding.
Institutional Review Board Statement: Ethical review and approval were waived for this study as depersonalized data from notifiable disease registry and laboratory findings were used.
Informed Consent Statement: Patient consent was waived due to use of depersonalized data from a notifiable disease registry and laboratory findings.
Data Availability Statement: Data upon request.