Raman Spectroscopic Characterization of ... - ACS Publications

Feb 7, 2016 - ABSTRACT: The intracellular lifestyle of L. pneumophila within protozoa is considered to be a fundamental process that supports its surv...
4 downloads 0 Views 7MB Size
Letter pubs.acs.org/ac

Raman Spectroscopic Characterization of Packaged L. pneumophila Strains Expelled by T. thermophila Dragana Kusić,† Anuradha Ramoji,‡,§ Ute Neugebauer,‡,§,∥ Petra Rösch,*,† and Jürgen Popp†,‡,§,∥ †

Institut für Physikalische Chemie and Abbe Center of Photonics, Friedrich-Schiller-Universität Jena, Helmholtzweg 4, D-07743 Jena, Germany ‡ Leibniz Institute of Photonic Technology (IPHT), Albert-Einstein-Straße 9, D-07745 Jena, Germany § Center for Sepsis Control and Care (CSCC), Jena University Hospital, Erlanger Allee 101, D-07747 Jena, Germany ∥ InfectoGnostics Forschungscampus Jena e.V., Zentrum für Angewandte Forschung, Philosophenweg 7, D-07743 Jena, Germany S Supporting Information *

ABSTRACT: The intracellular lifestyle of L. pneumophila within protozoa is considered to be a fundamental process that supports its survival in nature. However, after ingesting the cells of L. pneumophila, some protozoa expel them as compressed live cells in the form of small rounded pellets. The pellets of tightly packaged viable but not culturable forms (VBNCFs) as well as highly infectious mature intracellular forms (MIFs) of L. pneumophila are considered as infectious particles most likely capable to cause human infection. Since L. pneumophila cells are hardly culturable from these pellets, detection methods for packaged live L. pneumophila forms remaining in water should be cultivation free. Hence, we demonstrate the potential of Raman microspectroscopy to directly sort pellets containing L. pneumophila cells, expelled by T. thermophila, and to characterize them on the basis of their Raman spectra.

A

and U8W), we therefore attempted to confirm whether pellets containing tightly packaged cells of those strains could be detected and differentiated from the morphologically similar expelled food vacuoles without intact bacteria via Raman microspectroscopy. To obtain the Raman spectra of released pellets containing undigested food material without intact bacteria, two samples were prepared. The first control sample includes incubation of T. thermophila in filter-sterilized tap water for 24 h at 30 °C. The second sample involves the incubation of both immunolabeled L. pneumophila cells with fluorescent antibodies and T. thermophila in sterile-filtered tap water for 24 h at 30 °C allowing pellets with most likely degraded bacteria (pellets without fluorescence)8 to be easily sorted by using the fluorescence mode of the Raman microscope (BPE) (Supplemental Experimental Section) and the Raman spectra to be collected. All the experiments were performed in triplicate. Furthermore, the potential of Raman microspectroscopy to distinguish between pellets containing diverse L. pneumophila strains was assessed. Besides Raman microspectroscopy, confocal laser scanning microscopy (CLSM) (Supplemental Experimental Section) was used to visualize T. thermophila in the control sample as well as the pellet

thin Gram negative facultative intracellular bacterium, Legionella pneumophilia, is the major cause of Legionnaires’ disease. In water, L. pneumophila grows within protozoan hosts. However, some protozoa such as Tetrahymena spp. efficiently ingest L. pneumophila into intracellular food vacuoles afterward expelling live cells packaged into free spherical pellets in the surrounding environment.1 It has been found that T. thermophila supports rapid in vivo differentiation of the infectious stationary phase forms (SPFs) of L. pneumophila into highly infectious mature intracellular forms (MIFs) without bacterial cell division.2 Such pellets containing tightly packaged MIFs of L. pneumophila could infect both amoeba and macrophages and could therefore be responsible for the propagation of severe Legionnaires’ disease.1,3 In addition, it has been shown that viable but not culturable forms (VBNCFs) of L. pneumophila, packed by T. thermophila, remain infectious.4 To detect pellets containing VBNCFs of L. pneumophila as well as pellets in which L. pneumophila cells rapidly decrease in culturability,5 detection methods without cultivation have to be considered. For instance, targeting morphological markers with OmpS-specific antibodies and secondary fluorescent antibodies were used to detect L. pneumophila within pellets by fluorescence microscopy in concentrated water samples.6 However, it has been established that Raman microspectroscopy of single Legionella cells permits a direct identification of pathogens without both cultivation and additional labeling of the samples.7 By using the cultures of T. thermophila with three different L. pneumophila strains (Philadelphia-1, Los Angeles-1, © 2016 American Chemical Society

Received: December 11, 2015 Accepted: February 7, 2016 Published: February 7, 2016 2533

DOI: 10.1021/acs.analchem.5b04699 Anal. Chem. 2016, 88, 2533−2537

Letter

Analytical Chemistry formation after feeding T. thermophila with L. pneumophila cells. Figure 1 shows the bright field image of a control sample of T. thermophila as well as released food vacuoles most likely with

Figure 1. Bright field image of T. thermophila cells as well as food vacuoles (white arrows) released by protozoa after 24 h of incubation in sterile-filtered tap water (control sample).

Figure 2. CLSM images of immuno-labeled L. pneumophila cells (green) with fluorescent antibodies (a); accumulation of food vacuoles within T. thermophila containing fluorescently labeled bacteria (green) 3 h after incubation in sterile-filtered tap water (b), where the white arrow in panel b points to a pellet being expelled; expelled pellets containing fluorescently labeled bacteria (green) depicted after 24 h of incubation in sterile-filtered tap water (c) and staining of membrane damaged immuno-labeled bacteria (red) with Propidium Iodide (PI) after 24 h of incubation in sterile-filtered tap water (d). The white arrow in panel d points to a membrane damaged single bacterial cell not being ingested by protozoa, while L. pneumophila within the pellets (green) maintained the membrane integrity.

undigested foodstuff (white arrows). For this figure, T. thermophila was incubated in sterile-filtered tap water without L. pneumophila for 24 h at 30 °C. However, when incubated with L. pneumophila, T. thermophila rapidly accumulates vacuoles laden with L. pneumophila and expels free spherical pellets containing live cells.1,5,8 Figure 2 shows immuno-labeled Philadelphia-1 cells with fluorescent antibodies (a) as well as the accumulation of packaged bacteria into vacuoles within T. thermophila 3 h after incubation (b). Twenty-four h after incubation, pellets containing the immuno-labeled Philadelphia1 cells with fluorescent antibodies were expelled by T. thermophila (c), wherein the existence of viable cells within pellets was confirmed via staining with Propidium Iodide (PI) (d) as well as with Live/Dead staining shown in Figure S1, which is consistent with previous studies.1,5,8 In addition, Figure 3 shows bright field images (a, c) as well as fluorescence images (b, d) of immuno-labeled L. pneumophila (a, b) as well as the accumulation of vacuoles within T. thermophila (c, d). For this figure, packaged L. pneumophila within T. thermophila were captured 3 h after incubation in sterile-filtered tap water via the fluorescence mode of the BPE. Furthermore, expelled pellets loaded with L. pneumophila are depicted in Figure 4 24 h after incubation, where diverse morphologies of the pellets can be observed, which is in accordance with previous studies.1,2 Morphologically distinct expelled pellets containing immunolabeled Philadelphia-1 cells are shown as bright field images (Figure 4). Due to the dense packaging of L. pneumophila1 by T. thermophila, the single bacterial cells cannot be observed within tightly packaged pellets via the fluorescent mode of the BPE (b, d and f) as observed within the pellets shown in panels h, j, l, and n (Figure 4). Spherical-shaped pellets with smooth membrane containing tightly wrapped bacteria (a) were also observed in the moment when the pellets were expelled (Figure S2). Therefore, we can say that they might correspond to the late-stage vacuoles with smooth membrane and tightly packed bacteria identified by Faulkner et al.2 In contrast, fluorescence image n in Figure 4 shows a pellet that actually contains several intact bacteria as well as a fluorescence-free interspace most

Figure 3. Bright field images (a, c) as well as fluorescence images (b, d) of both L. pneumophila Philadelphia-1 cells (a, b) and T. thermophila containing Philadelphia-1 cells within the food vacuoles (c, d). The L. pneumophila Philadelphia-1 cells were immuno-labeled with fluorescent antibodies before feeding the protozoa and visualized 3 h after incubation in sterile-filtered tap water. 2534

DOI: 10.1021/acs.analchem.5b04699 Anal. Chem. 2016, 88, 2533−2537

Letter

Analytical Chemistry

Figure 5. Bright field image (a) and fluorescence image (b) as well as the obtained Raman spectrum (c) of the interspace without bacteria within the pellet at a position without fluorescence. The white point, whose diameter corresponds to 1 μm, indicates the measuring spot within the pellet. The tentative assignment of the Raman bands is given in Table S1.

pellet due to the degradation of bacterial cells within the food vacuole. Nevertheless, virulent L. pneumophila able to resist the intravacuolar digestion were selectively and tightly packaged into spherical pellets by Tetrahymena.1 We compared the mean Raman spectra of single cells (a) and tightly packaged pellets with a smooth membrane (b) containing L. pneumophila Philadelphia-1 (Figure 6), Los Angeles-1, and U8W cells

Figure 4. Representative bright field images (left) as well as fluorescence images (right) of morphologically diverse single pellets expelled by T. thermophila containing immuno-labeled L. pneumophila Philadelphia-1 cells 24 h after incubation. Figure 6. Normalized mean Raman spectra of single cells (a) and pellets with smooth (b) and wrinkled (c) membranes of tightly packaged Philadelphia-1 cells as well as their spectral differences a − b and b − c. Scale bar: 5 μm.

likely occupied with vesicular and membranous material.1 In fact, the Raman spectrum (c) of the interspace without fluorescence of the corresponding pellet is shown in Figure 5, where the most prominent Raman bands could be assigned to a variety of biological molecules such as proteins, polysaccharides, and lipids of membranous material (Table S1). Such pellets containing outer membrane fragments of digested legionellae cells and few intact bacteria were described by Berk et al. and defined as membranous pellets.1 Ingestion of already dead L. pneumophila cells, whose presence was detected via PI staining and depicted in Figure 2d, by T. thermophila may result in the accumulation of outer membrane fragments in the

(Figure S3-A,-B) by calculating the spectral differences (a − b). Substantially higher nucleic acid (1578, 1480, 1331, 1098, 778, 725, and 690 cm−1) as well as protein signals (1669, 1605, 1440, 1238, 1035, 1004, and 856 cm−1) were detected in the Raman spectra of the single cells of L. pneumophila strains in regard to the corresponding bacteria that were packaged into pellets, which might suggest a reduced metabolic activity of those cells within pellets. Faulkner et al. observed that the 2535

DOI: 10.1021/acs.analchem.5b04699 Anal. Chem. 2016, 88, 2533−2537

Letter

Analytical Chemistry

from the pellets holding the cells of L. pneumophila strains via a support vector machine (SVM) analysis. The obtained 10-fold cross-validation results of the four classes, which include Philadelphia-1, Los Angeles-1, and U8W cells closely packed by T. thermophila into spherical pellets as well as pellets without bacteria (pellets*) are shown in Table 1. Overall, an accuracy of

differentiation of L. pneumophila SPFs into MIFs while in transit through Tetrahymena happens in less than 1 h without bacterial replication.2 Therefore, a reduced metabolic activity in L. pneumophila within pellets compared to the free single cells might be due to an absence of bacterial replication within pellets. As a consequence, the reduced expression of cell division proteins can be expected.9 On the other side, negative features in the difference spectra (a − b) demonstrate a higher polysaccharide (2940 and 2980 cm−1) as well as lipid (2890 and 2850 cm−1) contribution in the Raman spectra of pellets containing tightly packaged bacteria, which might indicate the presence of membranous material within pellets. In fact, the occurrence of the layers of membranes as well as the rapid accumulation of lipids within pellets containing tightly packaged bacteria was previously reported by Hojo et al.5 Furthermore, the spectral differences (b − c) among the mean Raman spectra of tightly packaged pellets with smooth membranes (b) and those with wrinkled membranes in appearance (c) differ for all tested strains and show comparable features, which can be observed in Figures 6 and S3. The spectral differences (b − c) demonstrate a higher contribution of both proteins (1669, 1605, 1440, 1238, 1035, 1004, and 856 cm−1) and nucleic acids (1578, 1480, 1331, 1098, 778, 725, and 690 cm−1) as well as a diminished accumulation of both lipids (2890 and 2850 cm−1) and polysaccharides (2940 and 2980 cm−1) in the Raman spectra of wrinkled pellets which might suggest a reduced amount of membranous material and the presence of more metabolically active L. pneumophila in the wrinkled pellets compared to those with a smooth membrane. Specifically, we observed similarities between Raman spectra either by measuring vacuoles in the control sample or by tracking the fluorescence free pellets by means of the fluorescence mode of the Raman microscope. Figure 7 depicts

Table 1. Confusion Matrix for the Differentiation between the Spherical Pellets Containing Three L. pneumophila Strains and Vacuoles without Bacteria Expelled by T. thermophila Philadelphia-1 Los Angeles-1 U8W pellets* sensitivity/% specificity/% accuracy/%

Philadelphia-1

Los Angeles-1

U8W

pellets*

280 4 1 − 98.2 95.5 95.7

11 180 3 − 92.8 98.4

8 4 101 − 89.4 99.3

− − − 124 100 100

95.7% is achieved, where the Raman spectra of pellets found in the control samples and pellets in the absence of fluorescence from the samples, in which T. thermophila was fed with immuno-labeled cells (pellets*), were correctly predicted with a sensitivity of 100% and completely separated from the Raman spectra of spherical pellets with the cells of L. pneumophila strains. From 593 Raman spectra of pellets containing cells of different L. pneumophila strains, 31 Raman spectra were misclassified. Although single planktonic cells of Los Angeles1 and U8W exhibit highly similar Raman spectra,7a the Raman spectra of spherical pellets containing the cells of those strains were differentiated among each other with relatively high sensitivities of 92.8% and 89.4%, respectively. The Raman spectra of pellets holding the Philadelphia-1 cells were predicted with a sensitivity of 98.2%. In summary, we demonstrate a very effective and simple way to differentiate spherical pellets that contain highly concentrated live L. pneumophila cells from the pellets without bacteria by using Raman microspectroscopy. In addition, the single Raman spectra of pellets containing tightly packaged cells of three closely related L. pneumophila strains are for the first time discriminated between each other via an SVM analysis. Therefore, Raman microspectroscopy could be used to directly monitor potentially infectious pellets of L. pneumophila cells expelled by Tetrahymena species without further cultivation.

Figure 7. Bright field image of released vacuoles from the control sample (a) and a normalized Raman spectrum of the marked pellet (a). Bright field and fluorescence images of morphologically similar pellets (b, c), one without bacteria (b and b1) and its normalized Raman spectrum (b), and another one containing fluorescently labeled L. pneumophila (c and c1). The tentative assignment of the Raman bands is given in Table S1.



ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.5b04699. Experimental section and additional figures (PDF)

the bright field image of vacuoles released by T. thermophila in the control sample (a) and a corresponding normalized single Raman spectrum (a) as well as the bright field images of morphologically similar pellets (b, c), in addition to one without bacteria (b1) and its normalized single Raman spectrum (b) and another one containing immuno-labeled L. pneumophila (c1). Both the pellets without fluorescence and the released vacuoles obtained from the control samples show comparable Raman spectra and could be clearly differentiated



AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Author Contributions

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. 2536

DOI: 10.1021/acs.analchem.5b04699 Anal. Chem. 2016, 88, 2533−2537

Letter

Analytical Chemistry Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The funding of the research projects RiMaTH [02WRS1276E], InfectoGnostic (13GW0096F), and the EU project “HemoSpec” [CN 611682] via the Integrated Research and Treatment Center “Center for Sepsis Control and Care’” (CSCC) [01EO1502] from the Federal Ministry of Education and Research, Germany (BMBF), as well as the financial support from the German Research Foundation (DFG) in the Collaborative Research Center ChemBioSys [SFB 1127] are highly acknowledged. This project was realized within the InfectoGnostics Research Campus Jena. The authors also thank Bernd Kampe for the help with the program GnuR and Dr. Stephan Stöckel for the critical reading of the manuscript.



REFERENCES

(1) Berk, S. G.; Faulkner, G.; Garduno, E.; Joy, M. C.; Ortiz-Jimenez, M. A.; Garduno, R. A. Appl. Environ. Microbiol. 2008, 74, 2187−2199. (2) Faulkner, G.; Berk, S. G.; Garduño, E.; Ortiz-Jiménez, M. A.; Garduño, R. A. J. Bacteriol. 2008, 190, 7728−7738. (3) Rowbotham, T. J. J. Clin. Pathol. 1980, 33, 1179−1183. (4) Al-Bana, B. H.; Haddad, M. T.; Garduño, R. A. Environ. Microbiol. 2014, 16, 382−395. (5) Hojo, F.; Sato, D.; Matsuo, J.; Miyake, M.; Nakamura, S.; Kunichika, M.; Hayashi, Y.; Yoshida, M.; Takahashi, K.; Takemura, H.; et al. Appl. .Environ. Microbiol. 2012, 78, 5247−5257. (6) Robertson, P.; Abdelhady, H.; Garduño, R. A. Front. Microbiol. 2014, 5, 670. (7) (a) Kusić, D.; Kampe, B.; Ramoji, A.; Neugebauer, U.; Rösch, P.; Popp, J. Anal. Bioanal. Chem. 2015, 407, 6803−6813. (b) Kusić, D.; Kampe, B.; Rösch, P.; Popp, J. Water Res. 2014, 48, 179−189. (8) Smith-Somerville, H. E.; Huryn, V. B.; Walker, C.; Winters, A. L. Appl. Environ. Microbiol. 1991, 57, 2742−2749. (9) Li, L.; Mendis, N.; Trigui, H.; Faucher, S. P. BMC Genomics 2015, 16, 637.

2537

DOI: 10.1021/acs.analchem.5b04699 Anal. Chem. 2016, 88, 2533−2537