Identification of Haemophilus influenzae and Haemophilus ...

3 downloads 0 Views 323KB Size Report
Sep 10, 2013 - DOI : 10.1007/s10096-013-1958-x. Cite this ... of Haemophilus (growth requirement for X and V factor), and multilocus sequence typing (MLST).
Eur J Clin Microbiol Infect Dis (2014) 33:279–284 DOI 10.1007/s10096-013-1958-x

ARTICLE

Identification of Haemophilus influenzae and Haemophilus haemolyticus by matrix-assisted laser desorption ionization-time of flight mass spectrometry J. P. Bruin & M. Kostrzewa & A. van der Ende & P. Badoux & R. Jansen & S. A. Boers & B. M. W. Diederen

Received: 28 May 2013 / Accepted: 12 August 2013 / Published online: 10 September 2013 # Springer-Verlag Berlin Heidelberg 2013

Abstract Generally accepted laboratory methods that have been used for decades do not reliably distinguish between H. influenzae and H. haemolyticus isolates. H. haemolyticus strains are often incorrectly identified as nontypeable Haemophilus influenzae (NTHi). To distinguish H. influenzae from H. haemolyticus we have created a new database on the matrixassisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) bio-typer 2 and compared the results with routine determination of Haemophilus (growth requirement for X and V factor), and multilocus sequence typing (MLST). In total we have tested 277 isolates, 244H. influenzae and 33H. haemolyticus. Using MLST as the gold standard, the agreement of MALDI-TOF MS was 99.6 %. MALDI-TOF MS allows reliable and rapid discrimination between H. influenzae and H. haemolyticus.

Introduction Respiratory tract infections associated with H. influenzae have a high morbidity and mortality in developed and nonindustrialized nations [1]. The majority of these respiratory infections are caused by non-encapsulated H. influenza J. P. Bruin (*) : P. Badoux : R. Jansen : S. A. Boers : B. M. W. Diederen The Regional Laboratory of Public Health, Haarlem, The Netherlands e-mail: [email protected] M. Kostrzewa Bruker Daltonics GmbH, Bremen, Germany A. van der Ende Academic Medical Center, Department of Medical Microbiology and The Netherlands Reference Laboratory for Bacterial meningitis, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands

(NTHi) [2]. NTHi rarely causes invasive disease in healthy hosts, but is a significant cause of localized respiratory tract infections, such as acute otitis media, sinusitis, bronchitis, and conjunctivitis [3]. NTHi is also the most commonly seen bacterium in acute exacerbations of chronic obstructive pulmonary disease (COPD) [4, 5]. The majority of invasive H. influenzae infections in children, such as bacteremia, meningitis, pneumonia, epiglottitis and septic arthritis, are caused by the encapsulated H. influenzae type b [3]. Since the effect of vaccinations on the bacterial flora became apparent, several studies have mapped the bacterial flora of the vaccinated population [6, 7]. During these studies it became clear that traditional culturing and serotyping methods that were designed for (invasive) H. influenzae isolates are not satisfactory for identification of Haemophilus isolates from the pharyngeal flora [8, 9]. The phylogenetically closely related H. haemolyticus is also found in the pharynx of healthy adults and is considered a respiratory tract commensal, and has rarely demonstrated pathogenicity [3, 10, 11]. The mainstay of the classification of bacteria in clinical microbiology laboratories is colony morphology, microscopic examination, differential growth on selective media and various biochemical tests. Direct identification of microorganisms using molecular diagnostic methods (16S ribosomal RNA sequencing or real-time PCR) can be used as a alternative method [12, 13]. The identification of H. influenzae in clinical laboratories is based on the growth requirement for hemin (X-factor) and nicotinamide adenine dinucleotide (V-factor). H. haemolyticus can be phenotypically differentiated from NTHi in the laboratory by its ability to produce a clear hemolytic zone on horse blood agar. However, beta-hemolysis has been shown to be a poor indicator for distinguishing H. influenzae and H. haemolyticus [12], possibly caused by the loss of the hemolytic activity of H. haemolyticus on subculture [3]. In recent years, matrix-assisted laser desorption ionization–time of flight mass spectrometry (MALDI-TOF MS) has increasingly been applied in clinical

280

Eur J Clin Microbiol Infect Dis (2014) 33:279–284

microbial diagnostics for routine species identification of bacterial and fungal pathogens [14, 15]. Multilocus sequence typing (MLST) provides a precise method for the characterization of the bacterial isolates by sequencing internal fragments of the housekeeping genes [13]. Therefore, MLST can be used as a gold standard for the differentiation of H. haemolyticus from H. influenzae. The aim of this study was to determine whether MALDITOF MS can be used as a routine method for fast and reliable discrimination between H. influenzae and H. haemolyticus clinical isolates at the species level in our laboratories.

Material and methods Bacterial isolates For building the new Hi/Hh database, for H. haemolyticus an in house reference isolate #2030823 and ATCC 3339, and for H. influenzae ATCC 49766 was added to the manufacturer’s spectra protein (MSP) database. The reference isolate #2030823 was isolated from the nasopharynx of a healthy person. The characteristics of this isolate were required X and V factor, non-hemolytic, 7 F3 epitope of P6 not present;

100

99

98

97

96

Concatenated sequences (Boers)

Key

Genus

Species

080714015339

.Haemophilus

influenzae

081110015407

.Haemophilus

influenzae

081208015287

.Haemophilus

influenzae

081008015372

.Haemophilus

influenzae

090120015366

.Haemophilus

influenzae

090112015508

.Haemophilus

influenzae

090213015458

.Haemophilus

influenzae

080701015293

.Haemophilus

influenzae

090220015378

.Haemophilus

influenzae

090206015495

.Haemophilus

influenzae

090213015196

.Haemophilus

influenzae

090122015424_ii

.Haemophilus

influenzae

ATCC_49766

.Haemophilus

influenzae

080722015265

.Haemophilus

influenzae

090105015424_i

.Haemophilus

influenzae

081002015397

.Haemophilus

influenzae

080826015423

.Haemophilus

haemolyticus

080828015334

.Haemophilus

haemolyticus

080828015228

.Haemophilus

haemolyticus

080911015417

.Haemophilus

haemolyticus

ATCC_33390

.Haemophilus

haemolyticus

080903015256

.Haemophilus

haemolyticus

090220015386

.Haemophilus

haemolyticus

2030823

.Haemophilus

haemolyticus

080725015295_i

.Haemophilus

haemolyticus

080918015294

.Haemophilus

haemolyticus

090212015373

.Haemophilus

haemolyticus

080918015295_ii

.Haemophilus

haemolyticus

090202015446

.Haemophilus

haemolyticus

080811015442

.Haemophilus

haemolyticus

090216015489

.Haemophilus

haemolyticus

080828015388

.Haemophilus

haemolyticus

080828015389

.Haemophilus

haemolyticus

090302015428

.Haemophilus

haemolyticus

080808015649

.Haemophilus

haemolyticus

080711015264

.Haemophilus

haemolyticus

Fig. 1 Dendrogram multilocus sequence analysis control set of 33 Haemophilus isolates including the reference strains

Eur J Clin Microbiol Infect Dis (2014) 33:279–284

281

Table 1 Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS) Hi/Hh and high-throughput multilocus sequence typing (HI-MLST) results of 244 clinical isolates Material

n=

Hi (%)

Hh (%)

Hi STs

Sputa Broncho-alveolar lavage Middle ear fluid Conjunctiva Nasopharynx Genital tract Blood CSF Lymph node

74 18 4 2 1 2 129 13 1

65 (88) 14 (78) 4 2 1 2 128 13 1

9 (12) 4 (22) 0 0 0 0 1 0 0

53 12 4 2 1 2 77 9 1

Hi H. influenzae, Hh H. haemolyticus, Hi STs number of H. influenzae sequence types

performing genotyping; adk, atp G, mdh, pgi rec A, infB, fuc K: result H. haemolyticus . For testing the new Hi/Hh database 33 isolates, identified by the growth requirement for hemin (X-factor) and nicotinamide adenine dinucleotide (V-factor) as H. influenza were used. The 33 isolates were selected, for biodiversity, based on of the results of slide agglutination serotyping (SAST). This control set of 33 isolates was also tested for calibration accuracy, signal-to-noise ratio, and homogeneity of spectra and checked using FlexAnalysis at the Bruker Daltonics laboratory. For validation of this new Hi/Hh database we added 244 Haemophilus isolates. Of these, 101 (74 cultured from sputa, 18 from broncho-alveolar lavage, 4 from middle ear fluid, 2 from conjunctiva, 1 from the nasopharynx, and 2 from the genital tract) were randomly picked from routinely obtained isolates and 143 (129 cultured from blood, 13 from CSF and 1 from lymph node tissue) were obtained from the Netherlands Reference Laboratory for Bacterial Meningitis (NRLBM, Academic Medical Center, University of Amsterdam) and obtained from multiple laboratories in the Netherlands. The 143 invasive isolates were received in 2011 by the NRLBM, based on the growth requirement for hemin (X-factor) and nicotinamide adenine dinucleotide (V-factor), determined as Haemophilus influenzae isolates.

Table 2 Results of mass spectrometry identification and multilocus sequence analysis for the 277 tested haemophilus isolates

MSI mass spectrometry identification, MSA multilocus sequence analysis

Mass spectrometry identification Identification of the isolates was performed using MALDITOF MS (Bruker Daltonics, Bremen, Germany) with integrated Flex control software (Biotyper 2.0), according to the manufacturers' instructions. A part of a colony of each isolate, taken directly from the agar plate after 18–24 h incubation, was deposited on a single spot of the target steel plate (Bruker Daltonics) and allowed to dry at room temperature. After drying, 1 μl of matrix solution (saturated solution of acyano-4-hydoxycinnamic acid in 50 % acetonitrile) was added to the sample, which was then crystallized by airdrying at room temperature for 5 min. Each run with MALDI-TOF MS included a control sample (E. coli) to make sure that the spectrometer was set properly. Raw spectra of the isolates were analyzed using MALDI-biotyper 2.0 software (Bruker Daltonics) using the default settings: laser frequency 60 Hz, spectrum size 20, 624 pts; delay 9,813 pts; ion source I 20 kV; ion source II 16 kV; lens voltage 7 kV, and mass range 2,000–20,137 kDa. Results with score values ≥2.000 were considered correct for species-level identification and a score of 1.700 to 1.999 for genus-level identification. Creating new database entries For making the new Hi/Hh database entries the reference strains were added to the manufacturer’s spectra protein (MSP) database. To create the new database entries we followed the procedure described by the manufacturer. In short, freshly grown isolates were suspended in 300 μl of distilled water. Thereafter, 900 μl ethanol was added. The subsequent sample preparation was carried out in accordance with the standard ethanol/formic acid extraction protocol. After centrifugation the ethanol pellet was dried very carefully (i.e., drying up to 60 min) to maximize the formic acid treatment in the next step. A suited volume of formic acid (70 %) was added to suspend, solubilize, and to disrupt the cells. The same volume of acetonitrile was added and after mixing and centrifugation 1 μl of the supernatant was transferred to the MALDI target position. Each single reference strain was spotted on eight positions onto the target and measured three times per spot.

H. influenzae

Test isolates Clinical isolates Invasive isolates Total

H. haemolyticus

Agreement

n=

MSI n(%)

MSA n (%)

MSI n (%)

MSA n (%)

33 101 143 277

14 (42.4) 88 (87.1) 142 (99.3) 244 (88.1)

14 (42.4) 87 (86.1) 142 (99.3) 243 (87.7)

19 (57.6) 13 (12.9) 1 (0.7) 33 (11.9)

19 (57.6) 14 (13.9) 1 (0.7) 34 (12.3)

(%) 100.0 99.0 100.0 99.6

282

Multilocus sequence typing Partial DNA sequences of the adk, atpG, frdB, fucK, mdh, pgi, recA, and infB genes were generated using the highthroughput multilocus sequence typing (HiMLST) strategy as described by Boers et al. [16]. The oligonucleotides used for PCR amplification reported in the standardized MLST schemes by Meats et al. and McCrea et al. were equipped with universal tails for the HiMLST protocol [12, 13]. An unweighted pair group method with arithmetic mean (UPGMA) cluster analysis of the pair-wise distances derived from the concatenated sequences was achieved using Bionumerics v6.6 (Applied Math) software.

Results Haemophilus species identification by multilocus sequence typing All of the 280 Haemophilus isolates available, including three control strains, were subjected to MLST. Cluster analysis based on the concatenated sequences of the sequence fragments of the eight genes showed two distinct clusters (Fig. 1). Cluster I consisted of 127 distinct genotypes derived from 244 isolates and included the genotype derived from H. influenzae ATCC 49766, and was therefore indicated as the H. influenzae cluster. Cluster II consisted of 35 distinct genotypes derived Fig. 2 Result of principle component analysis. Principle component analysis of 33 Haemophilus strains confirmed the sepatation of H. influenzae and H. heminolyticus

Eur J Clin Microbiol Infect Dis (2014) 33:279–284

from 35 isolates and included the genotype derived from H. haemolyticus ATCC 33390, and is therefore indicated as the H. haemolyticus cluster. In addition, all isolates in the cluster II were PCR-negative for the fuculokinase gene (fuc K), supporting the fact that these isolates are H. haemolyticus [13, 17]. Results of mass spectrometry identification Determination of the control set of 33 Haemophilus isolates by using the Hh/Hi database, 14 of the 33 isolates (42.4 %) were identified as H. influenzae and 19 (57.6 %) as H. haemolyticus . Using this control set of isolates the same results were obtained at the Bruker Daltonics laboratory. By using the Hi/Hh database of the 101 tested clinical isolates 13 (12.9 %) were determined to be H. haemolyticus and 88 isolates (87.1 %) as H. influenzae. Of the 143 tested invasive isolates with the Hi/Hh database 1 isolate was determined to be H. haemolyticus and the other 142 isolates H. influenzae. Overall, of the tested Haemophilus isolates (n =277), 244 (88.1 %) were identified as H. influenzae and 33 (11.9 %) as H. haemolyticus (Tables 1 and 2). For the 277 isolates tested we found 99.6 % agreement (Table 2) between multilocus sequence analysis (MSA) and mass spectrometry identification (MSI). One H. haemolyticus isolate (MSI) was determined by MSA as H. influenzae, even after repeated MSI testing (score ≥2.0). The PCR for this isolate was found to be negative for the genes fuculokinase (fucK) and adenylatekinase

Eur J Clin Microbiol Infect Dis (2014) 33:279–284

(adk). Other characteristics were growth factors V and X both required, non-hemolytic, satellite growth with S. aureus.

Discussion Today the laboratory determination of Haemophilus is predominantly based on the growth requirement for hemin (X-factor) and nicotinamide adenine dinucleotide (V-factor). Based on these requirements no differentiation can be made between H. influenzae and H. haemolyticus. Difficulties were also reported in the serotyping of H. influenzae isolates by slide agglutination serotyping (SAST) [8, 18]. Accurate identification of NTHi with the conventional microbiology methods has become increasingly difficult, but molecular diagnostic techniques are able to solve this problem. Different molecular diagnostic techniques for differentiating H. influenzae from H. haemolyticus are available and have been evaluated [12, 13, 19, 20]. Meats et al. have described a MLST scheme for the precise characterization of encapsulated and non-capsulated isolates of H. influenzae. Genotyping with MLST has a great advantage over other genotyping methods: isolates characterized in different laboratories can be readily compared and the allelic profiles of isolates can be held in a single database [13]. However, none of the genotyping methods can be easily incorporated as bench side tools in clinical microbiology laboratories for the identification of H. haemolyticus. The technique MALDI-TOF MS allows a highly reliable rapid identification of most of the pathogenic bacterial species available in pure culture [14, 15]. For the identification of the two Haemophilus species, the peak list of the unknown spectrum is matched against the manufacturer’s spectra protein (MSP) in the database. In a MSP dendrogram (Fig. 1), the isolates of the two Haemophilus species clustered separately. A cross-wise pattern matching created a score matrix that was used for the dendrogram. Principal component analysis (PCA) carried out using ClinProTools 3.0 also confirmed the separation of both species (Fig. 2). Of the 143 tested invasive isolates with HiMLST, 1 isolate was identified as H. haemolyticus and 142 isolates as H. influenzae. The finding of a H. haemolyticus isolate from a patient with bacteremia is in line with the recent publications reporting that H. haemolyticus has rarely caused invasive disease [10, 11].

Conclusion Matrix-assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF MS) is a reliable and rapid technique for differentiating Haemophilus influenzae from Haemophilus haemolyticus. MALDI-TOF MS has a comparable performance compared with molecular techniques and allows

283

correct identification within a few minutes at a low per-sample cost. This method can be recommended for use in appropriately equipped laboratories. Accurate identification of pathogenic H. influenzae is important and may contribute to a reduction in unnecessary antibiotics used in the treatment of misidentified H. influenzae. Funding This study was funded by The Regional Laboratory of public Health, Haarlem. Conflict of interest M. Kostrzewa are employed at the mass spectrometry company Bruker Daltonics GmbH, Bremen, Germany and therefore has potential conflicts of interest. All other authors have declared that no competing interests exist.

References 1. Foxwell AR, Kyd JM, Cripps AW (1998) Nontypable Haemophylus influenzae: pathogenesis and prevention. Microbiol Mol Biol Rev 62(2):294–308 2. Murphy TF, Faden H, Bakeletz LO, Kyd JM, Forsgren A, Campos J, Virji M, Pelton SI (2009) Nontypable Haemophilus influenzae as a pathogen in children. Pediatr Infect Dis J 28(1):43–48 3. Mukundan D, Ecevit Z, Patel M, Marrs CF, Gilsdorf JR (2007) Pharyngeal Colonization Dynamics of Haemophilus influenzae and Haemophilus haemolyticus in healthy adult carriers. J Clin Microbiol 45(10):3207–3217 4. Murphy TF, Brauer AL, Schiffmacher AT, Sethi S (2004) Persistent colonization by Haemophilus influenzae in chronic obstructive pulmonary disease. Am J Resp Crit Care Med 170:266–272 5. Sethi S, Murphy TF (2001) Bacterial infection in chonic obstructive pulmonary disease in 2000: a state-of-the-art review. Clin Microbioal Rev 14(2):336–363 6. Kirkham LS, Wiertsema SP, Mowe EN, Bowman JM, Riley TV, Richmond PC (2010) Nasopharyngeal carriage of Haemophilus haemolyticus in otitis-prone and healthy children. J Clin Microbiol 48(7):2557–2559 7. Millar EV, O’Brien KL, Watt JP, Lingappa J, Palipamu R, Rosenstein N, Hu D, Reid R, Santosham M (2005) Epidemiology of invasive Haemophilus influenzae type A disease among Navajo and White Mountain Apache children, 1988–2003. Clin infect Dis 40:823–830 8. LaClaire LL, Tondella ML, Beall DS, Noble CA, Raghunathan PL, Rosenstein NE, Popovic T (2003) Identification of Haemophilus influenzae serotypes by standard slide agglutination serotyping and PCR-based capsule typing. J Clin Microbiol 41(1):393–396 9. Falla TJ, Crook DWM, Brophy LN, Maskell D, Kroll JS, Moxon ER (1994) PCR for capsular typing of Haemophilus influenzae. J Clin Microbiol 32(10):2382–2386 10. Jordan K, Conley AB, Antonov IV et al (2011) Genome sequences for five strains of the emerging pathogen Haemophilus haemolyticus. J Bacteriol 50(4):5879–5880 11. Morton DJ, Hempel RJ, Whitby TW, Seale W, Stull TL (2012) An invasive Haemophilus haemolyticus isolate. J Clin Microbiol 50(4): 1502–1503 12. McCrea KW, Xie J, Lacross N, Patel M, Mukundan D, Murphy TF, Marrs CF, Gilsdorf JR (2008) Relationships of nontypeable Haemophilus influenzae strains to haemolytic and nonhemolytic Haemophilus haemolyticus strains. J Clin Microbiol 46(2):406–416 13. Meats E, Feil EJ, Stringer S, Cody AJ, Goldstein R, Kroll JS, Popovic T, Spratt BG (2003) Characterization of encapsulated and

284

14.

15.

16.

17.

Eur J Clin Microbiol Infect Dis (2014) 33:279–284 noncapsulated Haemophilus influenzae and determination of phylogenetic relationships by multilocus sequence typing. J Clin Microbiol 41(4):1623–1636 Bizzini A, Greub G (2010) Matrix-assisted laser desorption ionization time-of-flight mass spectrometry, a revolution in clinical microbial identification. Clin Microbiol Infect 16(11):1614–1619 Emonet S, Shah HN, Cherkaoui A, Schrenzel J (2010) Application and use of various spectrometry methods in clinical microbiology. Clin Microbiol Infect 16(11):1604–1613 Boers SA, van der Reijden WA, Jansen R (2012) High-throughput multilocus sequence typing: bringing molecular typing to the next level. PLoS One 7(7):e39630 Binks MJ, Temple B, Kirkham L, Wiertsema SP, Dunne EM, Richmond PC, Marsh RL, Leach AJ, Smith-Vaughan HC (2012)

Molecular surveillance of true nontypable Haemophilus influenzae: an evaluation of PCR screening assays. PLoS One 7(3):e34083 18. Bokermann S, Zanella RC, Lemos AP, de Andrade AL, Brandileone MC (2003) Evaluation of methodology for serotyping invasive and nasopharyngeal isolates of Haemophilus influenzae in the ongoing surveillance in Brazil. J Clin Microbiol 41(12):5546–5550 19. Murphy TF, Brauer AL, Sethi S, Kilian M, Cai X, Lesse AJ (2007) Haemophilus haemolyticus a human respiratory tract commensal to be distinguished from Haemophilus influenzae. J Infect Dis 195(1):81–89 20. Theodore MJ, Anderson RD, Wang X, Katz LS, Vuong JT, Bell ME, Juni BA, Lowther SA, Lynfield R, MacNeil JR, Mayer LW (2012) Evaluation of new biomarker genes for differentiating Haemophilus influenzae from Haemophilus haemolyticus. J Clin Microbiol 50(4): 1422–1424