search
for
 About Bioline  All Journals  Testimonials  Membership  News


Actinomycetes
University of Udine, Mycology Department
ISSN: 0732-0574
Vol. 7, Num. 2, 1996
Actinomycetes, Vol. 7, Part 2, 1996 pp.47-54

COMPUTER ASSISTED IDENTIFICATION OF STREPTOMYCES SPECIES WITH HIGH EXTRACELLULAR PROTEASE ACTIVITY

S. R. CHAPHALKAR and S. DEY

Division of Microbial Sciences, Agharkar Research Institute, G.G. Agarkar Road, Pune 411 004, India


Code Number: AC96008
Sizes of Files: 
    Text: 20.0K
    Graphics: Line drawings (gif) - 9.0K

Abstract.

A computer assisted probabilistic identification system was used to identify 7 species of Streptomyces, characterised by high extracellular protease activity, which were isolated from different ecosystems. The probability matrices for major and minor clusters using 25 and 28 phena respectively were provided by MICRO-IS. Sixty-three morphological, physiological and biochemical tests were used and the results quantified in probability terms. Out of the 7 isolates, 5 were identified as Streptomyces diastaticus, S.anulatus, S.albidoflavus (2 strains) and S.exofoliatus and the remaining 2 to the minor clusters as S.misakiensis and S.longisporoflavus, all with a score of 0.9900 to 0.9343.

Streptomyces is an industrially important genus with many species. As the physiological and biochemical characteristics vary between strains, it is difficult for taxonomists to identify unknown Streptomyces species (Pridham et al., 1958). In early taxonomic studies of streptomycetes and related taxa (Jones, 1949; Pridham and Gottlieb, 1948) emphasis was given to monothetic classifications, based on a limited number of subjectively weighed properties, resulting in a variety of schemes. In contrast, in numerical taxonomic procedures equal weight is given to the various features (Goodfellow and Cross, 1984). Computer assisted classification was first applied to Streptomyces by Silvestri et al. (1962) using a probabilistic identification matrix listing the numerically defined taxa. After that many new matrices were prepared for classification and identification of Streptomyces (Willcox and Lapage, 1975; Willams et al., 1983a, 1983b). In all these systems a posteriori weighting was also applied (Willcox et al., 1980). In this way numerical taxonomy gives results that are valuable for both classification and identification (Goodfellow et al., 1992; Sackin and Jones, 1992). In the present note a computer- aided rapid identification was applied to 7 industrially important Streptomyces species, having high extracellular protease activity, isolated from different soils.

Materials and Methods

Origin and cultivation of the isolates.

In a search for Streptomyces species producing different extracellular enzymes, especially proteases, strains were selectively isolated from garden soils, polluted soils and volcanic crater silts (Table 1).

For growth and protease production isolates were incubated at 30 C in GYP broth, pH 8, containing (% w/v) glucose (1), yeast extract (0.5), peptone (0.5) and CaCl2.2H2O (0.2) in a rotary shaker incubator (180 rpm). Spore suspensions were preserved at -70 C in presence of 25% (v/v) glycerol. Morphological and biochemical characteristics of the isolates were studied after 5dd incubation. The production of proteases remained constant after this length of time.

-------------------------------------------------------------- 
Table 1. Origin and protease activity of the isolates
--------------------------------------------------------------
Strain     Origin                Protease activity 
                               (units mg-1 protein)
--------------------------------------------------------------
S1         Polluted soil             2560
S2         Garden soil               2050
S3         Polluted soil             1905
S4         Volcanic crater silt       158
S5         Polluted soil             1225
S6         Garden soil               1100
S7         Volcanic crater silt       990
------------------------------------------------------------- 

Selection of the strains.

The isolates were selected on the basis of extracellular protease production (Table 1) quantified by the azocaseinase method (Ginther, 1979).

Genus determination.

The generic status of the isolates was determined by analysing cell wall (Lechevalier and Lechevalier, 1970) and fatty acid (Brian and Gardner, 1967) composition and G+C mol% content (Mandel and Marmur, 1968).

Morphological and cultural characteristics.

Spore surface and spore chain morphology was investigated by scanning electron microscopy and replica techniques on 5dd old cultures on GYP (Chaphalkar et al., 1993; Langham et al., 1989). Cultural characteristics were determined on the same medium.

Antimicrobial activity.

The activity of the isolates was tested by the overlay method (Waksman and Curtis, 1916) against Bacillus subtilis, Micrococcus luteus, Streptomyces murinus, Candida albicans and Saccharomyces cerevisiae. Antifungal activity was checked against Aspergillus niger according to Mller (1958) on potato dextrose agar after 5dd incubation at 30 C.

Antibiotic resistance.

The isolates were checked, on GYP medium and after 5dd incubation at 30 C, for their ability to grow in the presence of neomycin (50æg ml^-1), rifampicin (50æg ml^-1), oleandomycin (100æg ml^-1) and penicillin (10IU ml^-1).

Degradation tests.

Degradation of adenine, tyrosine, xanthine, casein, gelatine, starch, elastin and arbutin was detected by measuring after 5dd incubation at 30 C the transparent zone around the colony on GYP agar.

Growth tests.

Tolerance to different pH values (from 6.0 to 9.5) was tested by growing the strains at 30 C for 5 days on GYP agar. Growth at different temperatures (25-45 C) was determined, after 5dd incubation, on GYP agar at pH 8.0.

Utilisation of different compounds as sole sources of carbon and nitrogen.

The isolates were tested for utilisation of different carbon sources (1%, w/v) in the basal medium at 30 C for 5dd. Growth was compared with negative (without glucose) and positive (with glucose) controls.

Nitrogen sources (0.1%) were added to the basal medium containing 1% (w/v) glucose. Utilisation was determined by comparison of the growth in the positive (0.5% yeast extract as nitrogen source) and the negative (no nitrogen source) control.

Growth inhibition.

Inhibition of growth by sodium azide (0.01%, w/v), NaCl (7%, w/v) and phenol (0.1%, w/v) was tested on the basal medium, containing 1% glucose and 0.5% yeast extract, after 5dd incubation at 30 C. Growth was compared with that on the medium without an inhibitor, absence of growth or very poor growth was recorded as negative.

Data coding.

All characters studied were noted as 1 (positive) or 0 (negative) (Rogosa et al., 1986). Characters showing poor separation value were deleted from the matrix. The most diagnostic characters were selected as described by Williams et al. (1983b). Initial selection of a reduced number of characters was carried out by deleting those with no or poor separation values. A total of 63 characters were selected and assessed by MICRO-IS (Microbial Information System, Potyrata and Krichevsky, 1993). Out of these 63 characters, 24 were common to both the major and minor cluster matrices, while 25 were exclusive to the major matrix and 14 to the minor one.

Computer assisted identification.

Data were processed by IBM PC/AT 386 and the software base for computing identification scores was MICRO-IS.

Since with MICRO-IS theoretical evaluation of single member clusters is not possible, TAXAN (release 2.0, Information Resources Group, Maryland Biotechnology Institute, USA) was used to determine the similarity levels among the isolated Streptomyces and the single member clusters using the simple matching coefficient. TAXAN has 32 OTUs consisting of 25 single member clusters with the addition of 7 new isolates and 22 characters which are the same as those of minor clusters.

Results and Discussion

Strain characteristics.

All strains were aerobic, formed substrate and aerial mycelia and chains of spores on the latter. Isolates S2 and S5 produced a small amount of brown pigment after 15 days on GYP.

LL-diaminopimelic acid and glycine (cell wall chemotype I) were present. All isolates contained only saturated fatty acids, with iso and anteiso components, varying from 45.2 to 79% (Table 2). G + C mol% of the isolates varied from 68 to 72%. On the basis of these data, all seven isolates were attributed to the genus Streptomyces.

--------------------------------------------------------------

Table 2. Genus characteristics of the isolates. All strains
contain LL-diaminopimelic acid and glycine in their cell walls
(chemotype I)
--------------------------------------------------------------
                                  Strain
                   -------------------------------------------
Characteristics     S1    S2    S3    S4    S5    S6    S7
--------------------------------------------------------------
Iso- and anteiso- 
fatty acids (%)    45.2   60    79    56    60    56    65

G+C Mol %           72    68    70    69    72    71    70
-------------------------------------------------------------

Identification.

The major cluster matrix consisted of 25 phena and the minor cluster one of 28 phena. Identification scores and cluster identification are shown in Table 3. Isolate S4 (identified as S.albidoflavus) shows a Willcox probability (Willcox and Lapage, 1975) of 0.9900 and the identification score for strains S2, S3, S5 and S6 was 0.9800. Isolate S1, identified to S.diastaticus, (0.9642) was characterised by atypical tests (activity against Bacillus sp. and growth in the presence of tellurite), but the score was substantially better than that against the next alternatives: S.chromofuscus (0.0238), S.cyaneus (0.0099) and S.rochei (0.0019). All other strains showed only one atypical character (e.g., xylitol and meso- inositol utilisation, pectin hydrolysis). Isolate S7 showed only one atypical result (0.9343) and was identified as S.longisporoflavus.

-------------------------------------------------------------- 
Table 3. A summary of the identification scores of the
isolates
--------------------------------------------------------------
                              Identification scores
        Cluster          -------------------------------------
Isolate Identi-  Cluster Willcox Taxonomic Standard Characters
        fication         probab- distance  error    against
                          ility
--------------------------------------------------------------
S1  S.diastaticus  Major  0.9642   0.130    -0.699      2
S2  S.anulatus     Major  0.9800   0.182    -2.812      1
S3  S.misakiensis  Minor  0.9800   0.362     1.205      1
S4  S.albidoflavus Major  0.9900   0.207    -1.905      1
S5  S.exfoliatus   Major  0.9800   0.327    -2.473      1
S6  S.albidoflavus Major  0.9800   0.220     0.481      1 
S7  S.longisporo-  Minor  0.9343   0.385     0.148      1
      flavus
--------------------------------------------------------------

    Figure 1. Simple matching coefficient of the isolates together with the single member clusters

Although all 7 isolates were identified to a definite cluster group, comparison with single member clusters was also carried out (Fig. 1). None of the isolated Streptomyces species showed a 0.9000 similarity level with any of the 25 single member clusters. The greatest similarity was found between S4 and S6, which belonged to the same major cluster. The lowest similarity (0.4450) was found between S.prunicolor and S.bikiniensis.

Results obtained showed that all the isolates, characterised by high protease activity, were members of the genus Streptomyces (Locci, 1989) and could be identified as shown in Table 3. Streptomycetes are well known antibiotic and enzyme producers, however their taxonomy is a serious problem and a workable, polythetic system using a number of characteristics is still required for the specific identification. Numerical classification has provided a suitable database for the construction of an identification matrix using 63 diagnostic characteristics for the 7 strains investigated.

All the isolates gave best identification scores of more than 0.9000 but less than 0.9900, with the exception of strain S4. The criteria adopted for identification were: (a) Willcox probability greater than 0.9000, with low scores for taxonomic distance and standard error, (b) first group scores substantially better than those against the next best two alternatives, (c) characters against none or few.

Identification scores followed the general pattern as those of Williams et al. (1983b). Most isolates were placed into clusters and, since these did not show any significant overlapping, it seemed appropriate to regard them as species groups. The most diagnostic characters included in the matrices were selected as objectively as possible

The number of discriminating characters chosen for the construction of the matrix appears rather large, however they may reflect variation within the cluster.

In MICRO-IS, as in MATIDEN programs (Sneath, 1979a), the most frequently used criterion is Willcox probability with a score of 0.9999 (Schindler et al., 1979). In other studies 0.9000 has been accepted as an indication of positive identification (Sneath, 1979b). The Willcox probability score can be supplemented by taxonomic distance and its standard error to increase the accuracy of identification (Williams et al., 1985).

The matrix presented here provides a workable system for the identification of unknown Streptomyces species and should be useful also in dealing with a wider range of isolates.

References

Brian, B.L. & E.W.Gardner (1967). Preparation of bacterial fatty acid methyl esters for rapid characterization by gas chromatography. Appl. Microbiol., 15: 1499- 1500

Chaphalkar, S., R.Dongre, D.Joshi & S.Dey (1993). A rapid method of microbial sample preparation for scanning electron microscopy. Biotech. & Histochem., 68: 166- 168

Ginther, C.L. (1979). Sporulation and the production of serine protease and cephamycin by S.lactamadurens. Antimicrob. Ag. Chemother., 15: 522-526

Goodfellow, M. & T.Cross (1984). Classification. In: M.Goodfellow, M.Mordarski & S.T.Williams (eds.) The Biology of Actinomycetes. Academic Press, London, pp. 7-194

Goodfellow, M., E.V.Fergusson & J.J.Sanglier (1992). Numerical classification and identification of Streptomyces species- a review. Gene, 115: 225-233

Jones, K.L. (1949). Fresh isolates of actinomycetes in which the presence of sporogenous aerial mycelia is a fluctuating characteristic. J. Bact., 57: 141-145

Langham, C.D., S.T.Williams, P.H.A.Sneath & A.M.Mortimer (1989). New probability matrices for identification of Streptomyces. J.gen.Microbiol., 135:121- 133

Lechevalier, M.P. & H.A.Lechevalier (1970). A critical evaluation of the genera of aerobic actinomycetes. In: H.Prauser (ed.) The Actinomycetales. VEB G.Fischer Verlag, Jena, pp. 311-316

Locci, R. (1989). Streptomyces and related genera. In: S.T.Williams, M.E.Sharpe & J.G.Holt (eds.) Bergey's Manual of Systematic Bacteriology. Williams & Wilkins, Baltimore. Vol.4, pp. 2451-2452

Mandel, M. & J.Marmur (1968). Use of CsCl density gradient analysis for determining G+C content of DNA. In: K.A.Grossman & K.Noldare (eds.) Methods of Enzymology. Academic Press, New York, Vol. XII (Part B), pp. 195-206

Muller, K.O. (1958). Studies on phytoalexins. Aust. J. Biol. Sci., 11: 275-330

Potyrata, D.A. & M.I.Krichevsky (1993). Microbial database management and analysis system. Binary, Computing in Microbiology, 4: 31-37

Pridham, T.G. & D.Gottlieb (1948). The utilization of carbon compounds by some Actinomycetes as an aid for species determination. J. Bact., 56:107-114

Pridham, T.G., C.W.Hesseltine & R.G.Benedict (1958). A guide for the classification of Streptomyces according to the selected group. Appl. Microbiol., 6: 52-79

Rogosa, M., M.I.Krichevsky & R.R.Colwell (1986). Coding Microbiological Data For Computers. Springer Verlag, New York

Sackin, M.J. & D.Jones (1992). Computer assisted classification. In: M.Goodfellow & A.G.O'Donnell (eds.) New Bacterial Systematics. Academic Press, London, pp. 281- 313

Schindler, J., J.Duben & O.Lysenko (1979). Computer aided numerical identification of gram negative fermentative rods on a desktop computer. J. Appl. Bacteriol., 47: 45- 51

Silvestri, L., M.Turri, L.R.Hill & E.Gillardi (1962). A quantitative approach to the systematics of Actinomycetes based on overall similarity. In: G.C..Ainsworth & P.H.A.Sneath (eds.) Microbial Classification. Symp. Soc. Gen. Microbiol., 12: 333-360

Sneath, P.H.A. (1979a). BASIC program for identification of an unknown with presence/absence data against identification matrix of percent positive characters. Comp. & Geosc., 5: 195-213

Sneath, P.H.A. (1979b). BASIC program for character separation indices from an identification matrix of percent positive character. Comp. & Geosc., 5: 349-357

Waksman, S.A. & R.E.Curtis (1916). The actinomycetes of the soil. Soil Sci., 1: 99-134

Willcox, W.R. & S.P.Lapage (1975). Methods used in a program for computer aided identification of bacteria. In: M.Pankhurst (ed.) Biological Identification with Computers. Academic Press, pp. 103-119

Willcox, W.R., S.P.Lapage & B.Homes (1980). A review of numerical methods in bacterial identification. Ant. van Leeuwenhoek, 46: 233-299

Williams, S.T., M.Goodfellow, G.Alderson, E.M.H.Wellington, P.H.A.Sneath & M.Sackin (1983a). Numerical classification of Streptomyces and related genera. J. gen. Microbiol., 129: 1743-1813

Williams, S.T., M.Goodfellow, G.Alderson, E.M.H.Wellington, P.H.A.Sneath & M.Sackin (1983b). A probability matrix for identification of streptomycetes. J. gen. Microbiol., 129: 1815-1830

Williams, S.T., J.C.Vickers & M.Goodfellow (1985). Application of new theoretical concepts to the identification of streptomycetes. In: M.Goodfellow, D.Jones & F.G.Priest (eds.) Computer Assisted Bacterial Systematics. Academic Press, Orlando, pp. 289-306

Copyright 1996 C.E.T.A., The International Centre for Theoretical and Applied Ecology, Gorizia


The following images related to this document are available:

Line drawing images

[ac96008a.gif]
Home Faq Resources Email Bioline
© Bioline International, 1989 - 2024, Site last up-dated on 01-Sep-2022.
Site created and maintained by the Reference Center on Environmental Information, CRIA, Brazil
System hosted by the Google Cloud Platform, GCP, Brazil