Molecular Characterization of Invasive Neisseria meningitidis Strains Isolated in Chile during 2010–2011

Background With the upcoming licensure of Outer Membrane Protein-based vaccines against meningococcal disease, data on disease incidence and molecular characteristic of circulating N. meningitidis strains in Latin American countries is needed. Chile is, to date, one of the few countries in the region that has performed this type of work in a comprehensive collection of disease-associated strains from two consecutive years, 2010–2011. Methods A total of 119 N. meningitidis strains isolated from patients with invasive disease in Chile in 2010–2011 were characterized by the National Reference Laboratory. Serogroup determination, MLST and porA typing were performed. Results Serogroup B was predominant in both study years, but W135 experienced a noticeable increase in 2011 compared to 2010. ST-11 complex, ST-41/44 complex ST-32 complex were the most prevalent among the isolates, and were strongly associated with serogroups W135 (ST-11 Complex) and B (ST-41/44 and ST-32 complexes). Likewise, the major porA types detected were strongly associated with these three clonal complexes: P1.5,2 was found exclusively among W135:ST-11 isolates, whereas P1.7, 2–3 was only detected in C:ST-11. ST-41/44 isolates mainly had P1.10-8, and ST-32 complex were associated with a P1.18-8 porA. Conclusions Our data show disease-associated N. meningitidis circulating in Chile are similar to those found in other parts of the world. The increase on W135:ST-11 isolates observed in 2011 foretold the unusual epidemiological situation experienced in the country in 2012, and MLST data show that this strain is indistinguishable from the one linked to the global Hajj 2000-related outbreak that occurred in 2001. Finally, this work demonstrates the importance of maintaining a strong national surveillance program integrating clinical, epidemiological and laboratory data and incorporating gold standard diagnostic and characterization techniques that allow the data to be compared all over the world.


Introduction
Neisseria meningitidis, the meningococcus, is an exclusively human Gram negative pathogen that is carried asymptomatically by 8-20% of the population and can cause invasive disease in the form of meningitis and septicemia [1]. Invasive meningococcal disease (IMD) generally develops rapidly and has an associated mortality approaching 10%. Approximately 20% of survivors suffer from neurological and disabling sequelae in spite of prompt antibiotic therapy [2,3]. The disease is a major concern in public heath worldwide and can occur as sporadic cases, outbreaks, and large epidemics [4]. Disease epidemiology varies among continents and countries, and therefore maintaining a well-established and coordinated clinical and laboratory-based surveillance system is crucial for identifying disease trends and timely identification of outbreaks, as well as evaluating the impact of preventive measures, including vaccination strategies [5].
Phenotypic characterization of N. meningitidis includes serogroup determination and is usually complemented by antigen typing, which is based on the variable regions of outer membrane proteins (OMP) and determines the serotype (porB) and serosubtype (porA).
The development of molecular biology techniques has resulted on classic monoclonal antibodies-based serotyping being replaced by DNA sequencing. Also, notwithstanding the importance of microbiological and phenotypic characterization, MLST has become the ''gold standard'' for the characterization of Neisseria meningitidis. Both MLST and sequencing of OMP, namely porA and fetA as recommended by the European Meningococcal Disease Society (EMGM) [6], have been incorporated as part of the routine work by national reference laboratories of many European countries and in the United States, making data comparable among countries, facilitating the identification of global spread of hyperinvasive strains and allowing investigations into the complex structure of bacterial populations [7,8,9].
However, these types of data are very scarce in Latin American countries, making it difficult to understand the disease epidemiology and how it relates globally. More so, the implementation of preventive measures against meningococcal disease, and specially vaccine development, is becoming almost exclusively dependent on molecular characteristics of strain, so it is now more important than ever to collect sufficient information before considering the possibility of including one of the existing or upcoming vaccines into the National Vaccination Scheme [10,11,12,13,14,15,16].
Disease associated serogroups vary widely across continents. Serogroup A is associated mainly with large scale epidemic outbreaks in the Africa Meningitis Belt [5]. Serogroup B is the leading cause of disease in Australia and New Zealand, as well as in Europe. In the United States, serogroups B, C and more recently Y account for the majority of invasive disease [5]. Data on serogroup distribution in Latin America and Caribbean countries is scarce. However, those obtained through the SIREVA II network indicates that although serogroups B and C are the main cause of IMD in the region, serogroup Y and specially W135 have increased in recent years [17,18,19].
In Chile, IMD is a disease of compulsory notification and all clinical cases must be reported to the epidemiology department of Ministry of Health. N. meningitidis isolates from reported cases are sent to the reference laboratory at the Instituto de Salud Pública (ISP) in the capital, Santiago [17]. Disease trends have experienced a progressive decline in the country since 2001, with incidence rates rarely surpassing 0.5/100,000 and serogroup B being the first cause of disease, followed by C and W135 [20,21].
Three quadrivalent polysaccharide conjugate vaccines that use different protein carriers and cover serogroups A, C, W135 and Y are currently available and have been proved to be an effective measure to control IMD caused by these serogroups [22]. However, with the exception of strain-specific vaccines developed as a response to long-standing outbreaks in Cuba, Norway and New Zealand, no effective vaccine is available against serogroup B disease. There are currently several vaccines, some of them in very advanced stages of development, that target different OMPs. These vaccines have been designed using those OMP variants that are more commonly found among hyperinvasive meningococci and have shown promising results against all meningococcal strains, including serogroup B [22]. As these vaccines are being developed on the basis of genetic data, the implementation of molecular characterization into routine laboratory surveillance is of paramount importance for decision-making bodies for an evidence-based decision on the implementation of preventive measures against the disease.
This study presents a comprehensive analysis of IMD causing strains in Chile during a two-year period, from 2010 to 2011, combining traditional microbiological characterization with molecular typing of strains, and will provide an insight into the characteristics of circulating meningococcal strains in the country, as well as allow data to be compared to that from other parts of the world.

Bacterial Isolates
A total of 119 meningococcal strains isolated in 2010 (55 isolates) and 2011 (64 isolates) from an equal number of patients with IMD were included in this investigation. Isolates were received by the ISP as part of the routine disease surveillance and came from all over the country (Figure 1). Strain information was available from the laboratory report. Isolates were confirmed by standard microbiology techniques, and serogroup was determined by slide agglutination using commercial antiserum against meningococcal capsular polysaccharides (DIFCO, Beckton Dickinson).

DNA Extraction
Isolates were retrieved from storage at 270uC and grown on chocolate agar plates for 16 hours at 37uC in a 5% CO 2 atmosphere. About 5 colonies were suspended in 100 ml of molecular grade water and boiled for 10 minutes. Particulate matter was removed by sedimentation at 12,0006g, and the supernatant was used in subsequent analyses.

Multilocus Sequence Typing (MLST) and porA Typing
MLST and porA typing were performed as previously described using the primers listed in the PubMLST website (http://pubmlst. org/neisseria/) [23,24]. Sequencing reactions were performed using the BigDye terminator cycle sequencing kits (Applied Biosystems), and run in 3130 DNA sequencer (Genetic analyzer 3130, Applied Biosystems). Designation of Sequence Types (ST) and Clonal Complexes (CC) were retrieved by comparison on the PubMLST website. The strains for which the allelic profile query did not find an ST in the MLST database were submitted to the database curator for validation and assignment of ST numbers. PorA variable regions 1 and 2 (VR1 and VR2) were identified by querying the porA VR sequence database located at http:// neisseria.org/nm/typing/pora/.

Clinical Isolates
A total of 119 N. meningitidis strains were collected from Health Services located throughout Chile between 2010 and 2011 ( Figure 1). Out of 119 isolates, 71 (59.6%) came from the Metropolitan region where the capital, Santiago, is located and houses over 40% of the country's residents. Fifteen isolates (12.6%) came from region V, which corresponds to Valparaiso, a popular vacation and tourist resort. The rest of the isolates came from the rest of the country, and numbers ranged from six to one. Isolates were obtained mainly from blood (64, 53,7%) and CSF (52, 43,6%). The remaining isolates (2,5%) came from joint fluid (2 isolates, 1,6%) and amniotic fluid (1 isolate, 0,8%). Only one isolate per patient was processed. A total of 48 (40.3%) strains were isolated from patients presenting meningitis (defined as the presence of meningitis symptoms with or without sepsis), and 18 (15%) came from patients with septicemia. Diagnostic information was not recorded in the laboratory report for 53 isolates (44,5%). Patients were aged between one month and 90 years, and 66 strains (55.4%) were isolated from patients under five years. Isolates from patients aged 20 and older represented 27,7% of the total.

Serogroup Distribution by Age-groups
Overall serogroup distribution and by age-group in 2010 and 2011 is shown in Table 1. Serogroup B was the most frequently detected, representing 63.6% and 51.6% in 2010 and 2011 respectively. The prevalence of serogroup C remained unchanged during the two-year period, with an average of 11.5%. A significant change on the prevalence of serogroup W135 was noticed, increasing from 7.8% in 2010 to 33.9% in 2011. This increase was specially noticed among those aged under the age of 5, and adults .20 years. In 2011 serogroup W135 became the most prevalent serogroup among isolates obtained from patients aged ,1 year, representing 62.5% of all isolates from that agegroup. On the contrary, serogroup B isolates obtained from patients ,5 and .20 years of age showed the opposite trend.

Discussion
This study represents, to the author's knowledge, the first comprehensive investigation combining microbiological and molecular data on a collection of meningococcal strains from Chile for a two-year period. These isolates are fully representative of the epidemiology of the country in 2010-11, as meningococcal disease is of compulsory notification and the country has a very efficient epidemiological surveillance system that is well coordinated with the national reference laboratory in the ISP [17].
Our data show that three clonal complexes have predominated in Chile over the two years of the study: ST-11 complex, ST-41/ 44 complex and ST-32 complex, a distribution is similar to that described in Europe [25]. Serogroup B prevailed over the rest, with a total of 67 isolates mainly found in the ST-32 complex and ST-41/44, with 27 and 29 serogroup B isolates respectively, which is also consistent with what has been observed in Europe. However, whereas ST-11 complex isolates reported from Europe have been mainly associated with serogroup C [25], the vast majority of ST-11 complex strains isolated in Chile were serogroup W135, and only 6 C:ST-11 complex strains were detected in the study period, of which four were isolated in 2010. Moreover, a marked increase in the number of cases of IMD caused by the ST-11 complex serogroup W135 was detected in 2011 compared to the previous year, especially among patients under the age of 5. The W135:P1.5,2:ST-11 meningococcal strains found in Chile are indistinguishable from the one that was  [19,28,29]. In both countries the strain has been identified as the W135:P1.5,2:ST-11. Moreover, this Hajj 2000-related strain has been detected in other countries throughout the region (Ana Belén Ibarz-Pavón, personal communication). Molecular data from other Latin American countries experiencing disease caused by W135 strains would be of great value to understand the epidemiology and anticipate the occurrence of disease outbreaks. It is now more important than ever that Chile maintains a strong epidemiological and laboratorybased surveillance of IMD incorporating molecular typing, as a nationwide vaccination campaign with tetravalent conjugate vaccines targeting children under the age of five has been launched as a control measure for the recent increase in the number of cases caused by W135 [27,30]. Molecular typing will allow the early detection of capsule switching and, more specifically, the early detection of the emergence of hyperinvasive strains. An additional 11 serogroup B strains belonged to other CCs or were STs that were reported in this investigation for the first time, which further evidences the importance of generating molecular data, as association between CC and porA and fetA variants has been described [16], and it would be reasonable to assume that there might be variants circulating in Chile that have not yet been identified. Even though our data on porA does not suggest that this is the case, the number of strains processed is still small and limited to a short timespan compared to the amount of data that has been generated from other parts of the world. A strain belonging to the newly identified ST-9914 presented an IS1301 inserted in the porA VR1 region. This insertion element has been found in various regions across the N.meningitidis genome [31], including those related to capsule expression and transport [32,33], as well as in the porA gene [34,35], resulting on the non-expression of this protein. Both diversity of porA and inactivation of the protein expression could have important implications for the decision on a vaccination strategy with new upcoming universal vaccines that include this protein among their targets.
A total of 14 serogroup C strains were characterized in this study, eight from 2010 and six from 2011. Two of these strains were isolated from two siblings and identified as C:ST-183:ST-23 complex/Cluster A3, a complex that is usually associated with serogroup Y [36]. To date, only two ST-183 strains are reported to the PubMLST database, and both belong to serogroup Y. Moreover, the porA variant in one of the database strains is different to that found in our strains.
A total of 19 STs were identified in our dataset for the first time,  and presented the same porA variant as the hyperinvasive strain linked to the Hajj 2000 outbreak. These findings highlight the fact that meningococcal populations are highly dynamic and, consequently new potentially hyperinvasive strains are constantly appearing, advocating for the need of continuous monitoring to foresee their spread.
In conclusion, data from meningococcal isolates obtained from disease cases in Chile in 2010-2011 show that even though IMD rates in the country appears to be lower than those reported from Europe and the US prior to the implementation of conjugate vaccines in national immunization schemes, disease-causing strains are similar to those found circulating among other industrialized countries [25]. This work also shows that despite serogroup B being the most common among disease isolates, a marked increase in the number of cases caused by the Hajj 2000-related W135:ST-11 strain became apparent in 2011, and highlights the importance of maintaining a molecular characterization-based epidemiological surveillance in the country. These data demonstrate the necessity of implementing such characterization techniques across Latin America, as the early detection of the Hajj 2000-related strain could allow decision-making bodies to consider the implementation of prevention measures before or in the early stages of a potential outbreak. The detection of new STs among our strain collection further advocates for the generation of molecular data and particularly, characterization of OMPs included in the forecoming vaccine formulations. Finally, our work demonstrates the importance of maintaining and strengthening National Surveillance Programs integrating clinical, laboratory and epidemiological information, and updating guidelines on all these three components so data can be traced and compared to that generated all over the world.