Cell size of various lactic acid as determined by scanning electron microscope and image analysis Athanasios Kokkinosa, Constantinos Fasseas, Elias Eliopoulos, George Kalantzopoulos

To cite this version:

Athanasios Kokkinosa, Constantinos Fasseas, Elias Eliopoulos, George Kalantzopoulos. Cell size of various as determined by scanning electron microscope and image analysis. Le Lait, INRA Editions, 1998, 78 (5), pp.491-500. ￿hal-00929612￿

HAL Id: hal-00929612 https://hal.archives-ouvertes.fr/hal-00929612 Submitted on 1 Jan 1998

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Lait (1998) 78, 491-500 491 © InralElsevier, Paris

Original article

Cell size of varions lactic acid bacteria as determined by scanning electron microscope and image analysis

Athanasios Kokkinos-", Constantinos Fasseas", Elias Eliopoulos'', George Kalantzopoulos"

a Laboratory of Dairy Technology, Department of Food Science and Technology, b Electron Microscopy Laboratory, Department of Agricultural Biology and Biotechnology,

C Genetics Laboratory, Department of Agricultural Biology and Biotechnology, Agricultural University of Athens, lera odos 75, 118 55 Athens, Greece

(Received 25 November 1996; accepted 6 April 1998)

Abstract - Cells from 48 pure lactic acid bacterial strains from the ACA-DC collection were stu- died under the scanning electron microscope (SEM) and the images were photographed and pro- cessed with an Image Analysis software. The diameters of several cocci and the lengths and widths of several lactobacilli strains were measured, then the size parameters were statistically analyzed by variance analysis. Significant differences were found between the cell sizes of the cocci strains (Streptococcus 1.07 to 1.21 um, Enterococcus 0.87 to 1.01 um, 0.75 to 0.95 um), but not between lengths and widths of the strains because too high variations were found in the last measurements. © InraJElsevier, Paris. lactic acid bacteria / bacterial cell / SEM / image analysis / cell size

Résumé - Étude granuIométrique des cellules de bactéries lactiques par microscopie élec- tronique à balayage associée à l'analyse d'images. Des cellules de 48 souches pures de bactéries lactiques de la collection ACA-DC ont été étudiées par microscopie électronique à balayage (MEB). Les photographies obtenues ont été scannées puis analysées à l'aide de l'analyseur d'images Sigmascan. Les diamètres des cocci ainsi que les longueurs et largeurs des lactobacilles ont été mesurés. L'ana- lyse des variances montre des différences caractéristiques entre les différents genres de cocci (Strep- tococcus de 1,07 à 1,21 um, Enterococcus de 0,87 à 1,01 um, Lactococcus de 0,75 à 0,95 um). En revanche, dans les conditions expérimentales utilisées, il n'a pas été possible de montrer de différences significatives entre les genres de lactobacilles. © Inra/Elsevier, Paris. bactérie lactique / cellule bactérienne / MEB / analyse d'image / taille de cellule

* Correspondence and reprints. 492 A. Kokkinos et al.

1. INTRODUCTION because of its great accuracy, repetition and fidelity. Bacterial cell size is still a useful mor- The aim of this work was to determine phological characteristic used in the classi- whether the small differences in cell sizes of fication and identification of bacteria. Mor- the lactic acid bacteria (LAB) we observed phology is easy to study and analyze, with the LM were real and stable and could particularly in eucaryotic micro-organisms be used for their identification and classifi- and the more complex procaryotes. In addi- cation. tion, morphological comparisons are valuable because structural features which The most commonly employed method depend on the expression of many genes, for measuring bacteria is by means of an are usually geneticaIly stable [16]. Thus, ocular micrometer. Measurements can also morphological sirnilarity is a good indication be made by using a camera lucid attachment of phylogenetic relatedness. Although the and drawing oculars, or by projecting the Iight microscope (LM) has always been a real image on a screen and measuring the very important tool, its resolving power of bacteria. In this work, an effort was made about 0.211m decreases its value in viewing to use image analysis techniques. Since these smaller micro-organisms and structures. techniques require well-defined images, During the last decades the electron micro- SEM visualization was considered the most scope has bec orne a powerful tool in the appropriate because it gives highly magni- examination of bacterial ceIls. The trans- fied, well-focused, high resolution images. mission electron microscope (TEM) is mainly now used by microbiologists for the study of the intracellular or cell wall struc- 2. MATERIALS AND METHODS ture [12], or for a quick examination of the bacterial cell with negative staining. The A total of 48 strains of LAB from the ACA- OC collection of the Laboratory of Dairy Tech- scanning electron microscope (SEM) is used nology of the Agricultural University of Athens, for the study of the cell surface and its pro- ail isolated from traditional Greek products, were perties [1, 5, 6, 19] or in the greater field of examined. The number of strains from each spe- food science and technology, usually for the cies and their collection code is shown on table /. structure analysis offood products [8] such as products [11,13], chee se [3] meats, raisins, etc. Others have used the SEM for 2.1. Growth conditions the observation of the microflora in sorne food products [2, 17]. The SEM together Ail the strains were in the fonn of pure frozen with the LM, according to Zenseky et al. cultures and were subcultured three times in [20], is becoming increasingly more useful growth media containing 10% skimmed milk powder supplemented with 0.2% yeast extract in imaging techniques for microbial study. (Difco). The growth temperature was 37 "C for Moreover; image analysis, a new and ail strains except those of the Lactococcus lactis developing technique, is now applicable in subsp lactis which were incubated at 30 "C, The almost ail sciences dealing with images in inoculum amount was 3% for the Lactobacillus delbrueckii subsp bulgaricus strains and 2% for great detail. The major scope of the image ail the others. The incubation time was 6 h for analysis is to define parameters such as size, the cocci strains and 8 h for the bacilli strains. number, shape, position and optical density of the objects which are recognized in an image [7]. In collaboration with image pro- 2.2. Sample preparation for SEM cessing that improves the image in viewing (e.g., zooming, sharpening, segmentation), After the third subculture, 0.5 mL were taken it is becoming a powerful tool for scientists from the coagulated media and washed with Cell size of lactic acid bacteria 493

Table I. Strains studied and their collection codes. Tableau I. Souches étudiées avec leur code dans la collection.

Strain name ACA-DC collection codes

Streptococcus thermophilus 2,3,4,5,6,7,8,9, 10, 11, 12, 15 Lactococcus lactis subsp lactis 47,48,49,50,51,52,53,54,55,56,57,58 Enterococcusfaecalis 138,239,240,252 Enterococcusfaecium 165, 174,241,225 Enterococcus durans 242,223,226,228 Lactobacillus delbrueckii subsp. bulgaricus 84,85,86,87,88,90, 100, 101, 102, 103, 104, 105

0.5 mL of phosphate buffer (0.1 mol-L:", pH 3. RESULTS AND DISCUSSION 7.2). After three washes, each was followed by a centrifugation (2 000 rpm, 10 min), the cells were Tables II, IV, VI, VII and VIII show the resuspended in 1 mL of the buffer. From that data and statistics obtained with the image media a small amount (approximately 200 ilL) analysis software. Each table is for one was taken and mounted directly on aluminium genus. Before any discussion we divided SEM stubs without prior fixation, air dried at 40 "C and sputter coated with gold. Preliminary the means of each genus into classes, as experiments showed that fixation, dehydration shown on tables III, V and IX for better com- and critical point drying offered no advantages to parison. The class intervals in each of these the preservation of the cells. The samples were tables were calculated using the model of then observed and photographed with a Cam- Sturges where the class interval (C) is given bridge Stereoscan S-150. Because the method by the type C = Rlq where R = Max value - relies on the accuracy of measurements, care was Min value and q = 1 + 3.32 10glO(n) where taken to always use the microscope at the same n is the number of measurements. viewing angle. Beginning the interpretation of the results for each genus separately, we can see on 2.3. Scanner settings and image analysis tables II and III that all the Streptococcus thermophilus strains (figure la) have dia- To scan the received pictures we used the GT meters apparently bigger than 1.00 um with 9000 EPSON scanner model under the following the most common class interval between settings: l.07 and l.21!lm. Eight out of the 12 strains gray levels: 64; sharpness: very sharp; bright- are found in this class interval. According ta ness: normal; resolution x, y: 663 dpi; graphie [4] cells of the genus Streptococcus are file type: bitmaps for Windows (.bmp). smaller than 2 um in diameter while the For the image analysis of the micrographs we Streptococcus thermophilus cells have a dia- used the SigmaScan image software (Jandel meter between 0.7 to 0.9 um, although these Scientific, San Rafael, CA, USA). The analysis figures differ from our observations regard- was limited in measuring the diameters of the ing cellular sizes. The Lactococcus lactis cocci celIs and the widths and lengths of the subsp lactis strains (figure lb), as shown on bacilli cells. After calibrating the program accor- tables IV and V, are smaller than the Strep- ding to the scale bar on the micrographs, celIs tococcus thermophilus ones with a diameter were magnified for better definition of the cell edges; the cells were then measured by moving of approximately 0.75 to 0.95 um. Accord- the pointer from one cell edge to the other. The ing to Mundt [15], the Streptococcus lactis program was keeping the measurements in (known now as Lactococcus lactis) cells are memory and calculating sorne statistical values. 0.5 to 1.0 um in diameter which is similar to 494 A. Kokkinos et al.

r<)OOl"'-"

3~8~8~~88333 000000000000

1 1 1 1

~~"

og~~::!:l:-~ ~::: 0 000000000000

<) "0 o u C "§ (/J Cell size of lactic acid bacteria 495

~~~::!;~~gn~~~~8 00 0 0 000

1"- trl ..0 trl .,f' trl r<"ltrl trltrlN 00 "":ÔV) V) trl trl 0\ "<1" 00 "<1" r-: "<1" 88gg~30·3333~3 00000 000000

~~~388888838 "

«)"<1" trll"-trl N\O NI"- 000000001"-0\0000000000 00000000000 .g o U .~l: CIl 496 A. Kokkinos et al.

0\ 0\ C'l \0 \0 r- r- 0\ 0000

\0 Ir) Ir) '

ggg8 g~8~ 0000 0000

C'l C'l C'l C'l C'l C'l C'l 0000 0000 0000 0000

00. 8 ~ o 00

C'l00C'l'

Q) Q) "0 o "8 u u ::: ::: 0g "g ~ ~ Cel! size of lactic acid bacteria 497

a. Streptococcus thermophilus ACA-OC 9 (x4000)

c. Enterococcusfaecalis ACA-OC 252 (x4000)

b. Lactococcus /actis subsp /actis ACA-OC 53 (x4000)

d. Enterococcusfaecium ACA-OC 174 (x4000) e. Enterococcus durans ACA-OC 226 (x4000)

Figure la-e. Pictures of the cocci genera studied as received from SEM. Figure la-e. Photographies des genres de cocci étudiés obtenues par MES. 498 A. Kokkinos et al.

S what we found. The strain ACA-DC 58 is ;:l"',,"", S g S Ir) 00 00 bigger in diameter than the other Lactococ- 0><; ~ ::l -: ~ ~ ~ ro ;> '""' eus lactis subsp lactis strains (1.17 um), a :::s fact that causes sorne queries. In tables VI-IX we can see that most of the Entera- S cocci strains (figure 1c-e) - a total of 9 - ;:l ~"""' S ;:l S have diameters ranging between 0.87 and 'S (;j :::t Ir) 0\ 00 '0 .- > '-" r- r- r- r- 1.01 um while the other three have more or :::s 0 0 0 0 less the same diameters. The Streptococcus faecalis cells (known now as Enterococcus â faecalis), as described by Mundt [14], are 6 0.5 to 1.0 um in diameter which is more or >< '0 00 Ir) 00 ~ ~ less similar to what we found. "H- ..0 r- N "" ""N A look at the statistics of the cocci genera ::0 can persuade us that the results are quite ""-0o ...... ;:l acceptable. The standard deviation, which is o ~ ::::.D 0 <.; an indication of great importance for the :::: S S ;:l '" 0 0 'T accuracy of the results, is low or around i:: 5= N N 0 "" ~ o "" "" 10% of the mean value. This percentage might be decreased by increasing the num- ::l 0"""' \.)'"' \.) (,)S ber of measurements; however, 10% is quite (;)

Table IX. Classification into groups of the Enterococci strains according to the mean of their œil diameter. Tableau IX. Classification des souches d'Enterococci en groupes suivant leur diamètre cellulaire moyen.

Groups (in um) Strains

0.8 -0.87 138,225 0.87 -0.94 239,240,165,241,242,223,226 0.94 - LOI 174,228 LOI - 1.08 1.08 - U5 252

Table X. ANOV A table. Tableau X. Résultats de l' ANOV A.

Source Sum of squares d.f. Mean square F-ratio

Between groups SSA =0.598 2 MSA=0.299 38.33 Within groups SSE= 0.260 33 MSE=0.0078 Total SST= 0.858 35

Figure 2. Picture of a lac- tobacilli strain as received from SEM. Figure 2. Photographie d'une souche de lactoba- cille obtenue par MEB. distinguish and therefore measurement was and the lengths of each strain despite the impossible. For those with a straight cell fact that ail the culture conditions were the and clearly visible, their length varied consi- same for ail strains. This should be consi- derably between 2.8 um and 5.35 um. As dered for further research. far as we know the Lactobacillus delbruec- By comparing the data of the three cocci kii cells are 0.5 to 0.8 um in width and 2 to genera, it becomes evident that there is a 9 um in length [9], which is quite similar significant difference in the size of their to our results. Moreover, it was not possible cells. The Streptococcus thermophilus to find any relationship between the widths strains seem to be the biggest, with the Ente- 500 A. Kokkinos et al.

rococci strains being smaIler, whereas those [5] Herald P.T., Zottola E.A., Scanning electron of the Lactococcus lactis subsp lactis being microscopie examination of Yersinia enteroco- lytica attached to stainless steel at selected tem- the smallest. This was also verified by peratures and pH values, 1. Gen. Microbiol. 137 applying the ANOY A (analysis of variance) (1988) 549-559. statistical method to the mean values of the [6] Hood S.K., Zottola E.A, Electron microscopy sizes of the different strains of each different study of the adherence properties of Lactoba- cillus acidophillus, J. Food Sei. 52 (1987) genus. Our aim was to find any statistically 791-792,805. significant differences of the mean values [7] Joyce-Loebl Ltd., Image Analysis: Principles of the cell diameter of the three cocci genera and Practice. Joyce-Loebl Ltd., Gateshead, at the 5% level of significance. The features England, 1989. of the analysis of variance are shown on [8] Kalàb M., Allan-Wojtzas P., Shea Miller S., Microscopy and other imaging techniques in table X. Since Fz, 33; 0.05 = 3.3 and food structure analysis. Trends Food Sei. Tech- F = 38.33 > 3.3 = Fz 33' 0.05' we can safely nol. 6 (1995) 177-217. say that there is a sigilificant difference bet- [9] Kandler O., Weiss N., List of the species of the ween the diameter of the three cocci genera. genus Lactobacillus, in: Bergey's Manual of This also supports the impression we had Systematic Bacteriology Vol. 2, The Williams and Wilkins Company, Baltimore, MD, USA, when examining these genera with the light 1986. microscope or analyzing the results with the [10] Knaysi G., Cytology ofbacteria, Botan. Rev. 4 image analysis. (1938) 83. [II] LagoueyeteN., LabIée J., Lagaude A, Tarodo de The application of this work in nume- la Fuente B., Temperature affects microstruc- rous different strains from different genera, ture of renneted milk gel, J. Food Sei. 59 (1994) could give us the ability to create a data- 956-959. base containing their cell sizes, which would [12] Lortal S., Rousseau M., Boyaval P., Van Hiei- be a very useful data for the numerical taxo- jenoort J., Cell wall and autolytic system of Lac- tobacillus helveticus ATCC 12046, J. Gen. nomy of bacteria. Consequently, the cell Microbiol. 137 (1991) 549-559. size in addition to the classic methods can [13] McManus W.R., McMahon D.J., Oberg C.1., provide us with a more precise identification High resolution scanning electron microscopy of and classification of an unknown strain, and milk products: a new sample preparation pro- cedure. Food Struct. 12 (1993) 475-482. that is of great importance. [14] Mundt 0.1., Enterococci. in: Bergey's Manual of Systematic Bacteriology Vol. 2, The Williams and Wilkins Company, Baltimore, MD, USA, ACKNOWLEDGMENTS 1986a. [15] Mundt 0.1., Lactic acid Streptococci. in: Ber- We thank E. Manolopoulou in charge of the gey's Manual of Systematic Bacteriology Vol. ACA-DC Collection and 1. Psarokostopoulos for 2, The Williams and Wilkins Company, Balti- technical assistance with the SEM. more, MD, USA, 1986b. [16] Prescot L., Harley J., Klein D., Microbial taxo- nomy, in: Microbiology, 2nd Ed, William C. REFERENCES Brown Communications, Inc, Dubuque, Iowa, USA, 1993. [1] Bottazzi V., Bianchi P., l Microorganismi Lat- [17] Rousseau M., Study of the surface flora oftra- tiero-Caseari al Microscopio Elettronico a Scan- ditional Camembert by scanning elec- sione, Edi-Ermes, Milan, Italy, 1984. tron microscopy, Milchwissenschaft 39 (1984) [2] Bottazzi V., Battistoti B., Bianchi F., Microco- 129-135. lonies of thermophilic lactic acid bacteria in [18] Salle J., Fundamental Principles of Bacterio- Grana cheese, Microbiol. Aliment. Nutr. 1 logy, 3rd Ed., McGraw-Hill, New York, USA, (1983) 285-291. 1948. [3] Bryant A., Ustunol Z., Steffe J., Texture of [19] Zani G., Severi A, Cellular ultrastructure and Cheddar chee se as influenced by fat reduction, morphology in Bifidobacterium bifidum, Micro- J. Food Sei. 60 (1995) 1216-1219, 1236. biologica 5 (1982) 255-267. [4] Hardie J., Genus Streptococcus, in: Bergey's [20] Zenseky Y., Tokunaga J., Tawara J., Atlas of Manual of Systematic Bacteriology Vol. Z, The Scanning Electron Microscopy in Microbiology, Williams and Wilkins Company, Baltimore, Georg Thieme Publishers, Stuttgart, Igaku Shoin MD, USA, 1986. Ltd., Tokyo, 1976.