Clinical optochin resistant Streptococcus pneumoniae and Streptococcus pseudopneumoniae strains in Tunisia

Introduction: Streptococcus pneumoniae can be responsible for severe human infections. Optochin resistance has been a potential cause of misidentification of pneumococcus and other members of the mitis group. Hence, rapid and easy optochin resistant (Optr) S. pneumoniae identification is essential. Methodology: Atypical pneumococci were characterized using optochin susceptibility, bile solubility based on spectrophotometric reading, serotyping, pulsed field gel electrophoresis (PFGE), 16S rRNA sequencing and PCR-based assays targeting pneumococcal genes lytA, ply, pspA, cpsA, Spn9802 and Spn9828. Results: Optical density values for the bile solubility test suggest the identification of four Optr S. pneumoniae and one Streptococcus pseudopneumoniae. All Optr pneumococci harbored cpsA, lytA, ply, Spn9802, Spn9828 and pspA genes. Only ply, spn9802 and Spn9828 genes were detected in S. pseudopneumoniae. The 16S rRNA sequencing differentiates between these two species. Optr S. pneumoniae strains belonged to different genotypes and serotypes (14, 19A, 3 and 9V). Three Optr S. pneumoniae isolates were typed as pspA family 2, while one belonged to pspA family 1. Sequencing of the atpA and atpC gene of the Optr variants revealed three mutations in the ATPase a-subunit (L99I, M23V and V52I) and one mutation in ATPase c-subunit (V48I). Conclusions: Our data indicate that bile OD-values provides an accurate, fast and easy method to discriminate between Optr S. pneumoniae and other Streptococcus mitis group. Moreover molecular techniques, confirming the bile test, can be used in order to prevent these atypical pneumococci and alert clinical microbiologists of the presence of these strains in the community.


Introduction
Streptococcus pneumoniae is one of the most important human pathogen worldwide and a leading cause of invasive and non-invasive infections in children and elderly. Differentiation between S. pneumoniae and the other alpha-hemolytic streptococci is conventionally based on biochemical and phenotypic characteristics, most commonly optochin susceptibility and bile solubility. However, since 1987, the emergence of optochin-resistance has been described from diverse geographic areas [1][2][3][4][5][6][7][8]. In Tunisia, Raddaoui et al. reported three Opt r S. pneumoniae strains isolated from immunocompromised patients between 2005 and 2011 [7].
Most of clinical microbiology laboratories used the identification procedure based on optochin susceptibility testing. Therefore, optochin resistance has been a potential cause of misidentification of pneumococcus due to the complexity to distinguish Opt r S. pneumoniae from S. pseudopneumoniae or other closely related streptococci [9][10][11][12][13][14][15][16][17][18]. The low incidence and potential pathogenic role of S. pseudopneumoniae has been demonstrated in a murine model as well as in humans [11,13]. This bacterium may be associated with chronic obstructive pulmonary disease [12,18]. Thus, the differentiation of S. pneumoniae from S. pseudopneumoniae remains important. Despite having 99% 16S rRNA gene identity with S. pneumoniae, S. pseudopneumoniae is phenotypically distinct [9]. S. pseudopneumoniae is optochin resistant in the presence of 5% CO 2 and optochin susceptible in ambient atmosphere, bile insoluble and lacks the pneumococcal capsule [9].
As biochemical tests are often insufficient to distinguish between these species several molecular methods have been proposed. Among these methods 16S rRNA sequencing, PCR targeting the capsular polysaccharide biosynthesis gene A (cpsA) or specific pneumococcal virulence factors such as autolysin A (lytA), pneumolysin (ply), pneumococcal surface protein A (pspA), DNA fragment of unknown function (Spn9828 and Spn9802), and competence-stimulating peptide ComC sequence (comC) were described [9,10,13,14,17]. In addition, Multilocus sequence typing (MLST) has already been established as important diagnostic approach, but this method is difficult to carry out for large-scale and routine clinical diagnosis.
The aim of this study is to use an accurate phenotypic and molecular testing for identification of optochin resistant S. pneumoniae and S. pseudopneumoniae isolates.

Clinical isolates
Five atypical Opt r pneumococci isolates were collected between 2014 and 2016 at Habib Bourguiba University Hospital Sfax-Tunisia. Of these strains, four isolates (Sp22, Sp125, Sp358, Sp345) were bile soluble and one strain (P1058) showed a difficulty to interpret the bile test based on visual evaluation. All strains were from non-invasive sites; sputum and ear exudates ( Table 1). To ensure the phenotypic methods, two susceptible optochin S. pneumoniae, one Streptococcus mitis and one Streptococcus oralis strains were included in our study.

Phenotypical methods for bacterial identification
The identification of the isolates was carried out using conventional tests, including observation of Gram stain morphology, colony characteristics on blood agar plates, optochin susceptibility testing and bile solubility testing.
Optochin susceptibility was performed by disk diffusion, using commercially available optochin disks (5µg, BioRad, Hercules,USA) applied onto Mueller Hinton agar supplemented with 5% defibrinated horse blood and incubated simultaneously at 37 °C overnight in 5% CO 2 atmosphere and in ambient atmosphere. Isolates were considered resistant to optochin if they displayed inhibition zones of less than 14 mm in 5% CO 2 .
Firstly bile-solubility test was performed in tube and interpreted by the naked eye. Colonies were transferred in 2 mL of saline solution and divided in two equal volumes. Deoxycholate was added to one tube and the two tubes were incubated at 37 °C for 30 minutes. The test was positive when clearing or loss of turbidity of the suspension has occurred in the deoxycholate tube.
Secondly, bile solubility testing was performed with the a densitometer as described previously [16]. Briefly, this technique is based on the measurement of the differences of absorbance in the test tube containing 10% sodium deoxycholate versus the control tube, after incubation for 10 minutes at 37 °C by spectrophotometer. The species with an optical density value (OD-value) of 2.1 or above were identified as S. pneumoniae, while finding an isolate with an OD-value in the range of 0.9 to 2.1 suggested the identification of S. pseudopneumoniae [16].

Molecular method for bacterial identification
Chromosomal DNA was extracted from the overnight cultures of the isolates using InstaGene kit (Biorad, Hercules, USA) according to the manufacturer's instructions.
Detection of cpsA, lytA, ply, pspA, Spn9828 and Spn9802 genes, was achieved by PCR [14,19]. Amplification of approximately 1000bp of the 16S rRNA gene was carried out with the primer pair 536F (5'-CAGCAGCCGCGGTAATAC-3') and rp2 (5'-ACGGCTACCTTGTTACGACTT-3') [20]. Then, the PCR product was sequenced. The identification of species was determined by comparing the obtained sequence with that of existing sequences in the GenBank database using the BLAST program available at the National Center of Biotechnology Information Web site [21].

Pulsed-field gel electrophoresis
Pulsed-field gel electrophoresis (PFGE) of macrorestriction DNA fragments was performed as described by Lefevre et al. [23]. After macrorestriction using SmaI 30U (Promega), the fragments were separated in a GenePath system (Bio-Rad, Hercules, USA) according to the manufacturer's recommendations.

pspA gene family typing
The pspA genes were detected by PCR using primers previously described. R6 and TIGR4 isolates were used as control strains for pspA family 1 and family 2, respectively [19].

ATPase
PCR amplification was performed for atpA and atpC genes encoding H + -ATPase a-and c-subunit, respectively.

Ethical Approval
This study was performed with approval from the Local Medical Commitee of Habib Bourguiba Hospital, Sfax, Tunisia.

Bacterial identification
Four atypical S. pneumoniae strains, resistant to optochin and bile soluble, were recovered between 2014 and 2016. Finding an OD-value in the range of 2.6 to 2.8 suggested the identification of S. pneumoniae. In addition to these strains, we recovered one α-hemolytic streptococcus resistant strain to optochin (P1058) with a bile test not clear on a visual interpretation. The ODvalue of 1.9 implicate that could be S. pseudopneumoniae. For the control strains, the ODvalues were 2.6 and 2.9 for optochin susceptible S. pneumoniae strains. None of the two other control strains were lysed by bile salts. The OD values were 1.6 and 0 for S. mitis and S. oralis, respectively.

Molecular typing
The pneumococcal isolates have different serotypes 14, 19A, 3 and 9V/A. The 19A strain was pspA family 1 and the three remaining strains were pspA family 2. As serotyping results, analysis of PFGE profiles indicated that the isolates were not genetically related to each other ( Figure 2).

Resistance to optochin
Concerning the resistance to optochin, DNA sequencing demonstrated that two S. pneumoniae serotypes 19A and 9V/A had an ATPase-A mutation of Leu99Ile (L99I). It is noticeable that in the ATPase asubunit we identified also two mutations; one of the helix1 (M23V) and a second mutation of Val52Ile in S. pneumoniae serotype 14. For ATPase c-subunit, only isolate with serotype 3 showed a mutation of Val48Ile (V48I).

Discussion
The emergence of optochin-resistant S. pneumoniae has been reported in recent years [1][2][3][4][5][6][7][8]. In the National Center of Bone Marrow Transplantation of Tunis-Tunisia, among 59 pneumococci recovered during 2005 to 2011 only three Opt r strains have been reported (5%) Figure 1. Alignment of 16S rRNA sequence region of P1058 and SP22 with nucleotide sequence of S. pneumoniae (CP000918.1), S. pseudopneumoniae (EU156785) and S. mitis (AY518677.1) by Clustal V. Identical residues are indicated with an asterisk; Box represents region with mutation site. [7]. In our study, we recovered four Opt r S. pneumoniae strains between 2014 and 2016. The isolation frequency of Opt r pneumococci varied in different countries; it is 2.1% in Portugal, 0.5% in United States and Argentina, 0.68% in Japan and 0.8% in Brazil [4,5,24,25]. The fifth Opt r strain included in our study showed a difficulty to interpret the bile test based on visual evaluation. The challenging identification of this bacterium is due to the subjective human evaluation of this technique. Recently, Slotved et al. presented cut-off OD-values that can be used for differentiating S. pneumoniae and S. pseudopneumoniae from other mitis group streptococci [16]. Also, Ercibengoa et al. recommended the use of bile-solubility test as a technique for S. pneumoniae identification, because of the emergence of optochin-resistant pneumococci and optochin susceptible S. mitis [26]. This method helped us to differentiate between S. pneumoniae and S. pseudopneumoniae. Molecular biology techniques have been developed and used to distinguish atypical S. pneumoniae from S. pseudopneumoniae or other closely related streptococci. In our study, 16S rRNA gene sequencing allowed us to distinguish S. pneumoniae from S. pseudopneumoniae by looking at the difference in 2 distinct positions that are described as sites for reliable Streptococcus species identification [10]. Indeed, El Aila et al. designed highly specific primers targeting a region of the 16S rRNA gene including these two sites that let to distinguish S. pneumoniae from other viridans group streptococci [10].
As reported previously, in our finding ply gene and Spn9802-and Spn9828-specific primers, known to be specific to S. pneumoniae, could not differentiate S. pneumoniae and S. pseudopneumoniae [14,27]. Only cpsA, lytA and pspA genes were useful for the discrimination of both species. Therefore, phenotypic and molecular methods based on PCR can easily be used in diagnostic procedures to distinguish S. pneumoniae from S. pseudopneumoniae and from two other Streptococcus species.
Our data indicate that Opt r S. pneumoniae strains possess different serotypes and genotypes (four PFGE profiles). Of note, Pinto et al. showed that optochin resistance is not due to clonal dissemination [6]. Pneumococcal strains were non invasive and belonged to a variety of capsular types (14, 19A, 3 and 9V/A). These serotypes were commonly found in our region, South of Tunisia [28]. It has been reported that serotype distribution among Opt r S. pneumoniae strains is mostly dependent on the epidemiological profile of each area [5][6][7]. In Tunis, North of Tunisia, three Opt r strains had different serotypes including 6C, 19F and 23F corresponding to the commonly types described in this region [7]. It was noticeable that distribution of pspA gene families among Opt r S. pneumoniae strains show pspA families 1 and 2 wich are the common profiles identified in S. pneumoniae isolates [6,19].
The emergence of unusual phenotype Opt r S. pneumoniae is still limited. The mechanism leading to optochin resistance is essentially ascribed to ATPase csubunit and rarely to ATPase a-subunit [7,[29][30][31]. Majority of studies reported that the point mutations in amino acid residues 48, 49 or 50 of the H + -ATPase csubunit confer optochin resistance. In our study among four Opt r S. pneumoniae, three isolates presented mutations located in the gene coding for the a-subunit of H + -ATPase. Two strains presented changes only in codon 99 (L99I) and one strain displayed two changes: one in codon 23 (M23V) and the second in codon 52 (V52I) in α helices of the a-subunit. However the mutations in 23 (M23V), 52 (V52I) and 99 (L99I) of asubunit of the H + -ATPase found in this study had never been observed before. Only one strain showed a M: PFGE Lambda ladder; 1-4: Opt r Streptococcus pneumoniae (SP22, SP125, SP358, SP345); 5: S. pseudopneumoniae (P1058).

Figure 2.
Pulsed field gel electrophoresis (PFGE) of SmaIdigested chromosomal DNA from optochin resistant isolates. mutation previously described in codon 48 of the ATPase C. Recently Raddaoui et al. showed that for three Opt r strains recovered, only the sequencing of atpA and atpC genes showed modification in atpC gene [7].

Conclusions
The occurrence of Opt r S. pneumoniae variants has hampered the phenotypic identification of pneumococci. Accurate identification of Opt r S. pneumoniae and differentiation of this emerging pathogen from S. pseudopneumoniae is important to ensure the correct diagnosis and treatment of patients. Despite the low number of isolates, this is the first time that we have isolated Opt r S. pneumoniae and S. pseudopneumoniae in the South of Tunisia. As for optochin susceptible S. pneumoniae, the bile solubility test based on spectrophotometric give a correct identification for Opt r S. pneumoniae and can differentiate it from other mitis group streptococci. In our study the Opt r S. pneumoniae strains belonged to serotypes 14, 19A, 3 and 9V/A, commonly found in our region. Constant vigilance must be applied to monitor the emergence of the Opt r strains of S. pneumoniae and S. pseudopneumoniae.