If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Despite its high prevalence within the bovine mammary gland, Corynebacterium bovis is considered a minor pathogen and of limited clinical significance. It has been suggested that intramammary infection with C. bovis may protect quarters against subsequent infection with other pathogens. The literature has produced much conflicting data on the subject. A possible explanation for some of the divergence of opinion on the subject is incorrect identification of isolates in previous studies.
This paper describes a novel method for differentiating C. bovis from other lipophilic Corynebacterium species based on endonuclease restriction analysis. The 16S rRNA gene sequences for all known lipophilic Corynebacterium species were obtained from published data and analyzed. It was predicted that endonuclease restriction with HindIII and SmaI could be used to differentiate C. bovis from all other known lipophilic Corynebacterium species.
The method was successfully employed to identify 741 of 762 (97.2%) lipophilic Corynebacterium species as C. bovis. Twenty one (2.8%) were identified as species other than C. bovis. Using this technique, it was demonstrated that it is not safe to assume that all lipophilic coryneform organisms isolated from bovine milk samples are C. bovis.
This method is an alternative to more traditional methods of identification in large scale studies until methods such as 16S rRNA gene sequencing become more widely available.
C. bovis is occasionally associated with disease in other species. It has recently been identified as the aetiological agent of hyperkeratotic dermatitis of athymic nude mice (
Analysis of the 16S rRNA gene sequence of the coryneform bacterium associated with hyperkeratotic dermatitis of athymic nude mice and development of a PCR-based detection assay.
Corynebacterium bovis is of interest to mastitis research workers because of the observation that quarters infected with C. bovis are significantly less likely to become infected with other more pathogenic organisms (
The susceptibility of bovine udder quarters colonized with Corynebacterium bovis to experimental infection with Staphylococcus aureus or Streptococcus agalactiae.
Experimental Staphylococcus aureus intramammary challenge in late lactation dairy cows: Quarter and cow effects determining the probability of infection.
). Conversely other authors have demonstrated the opposite affect, i.e., infection with C. bovis predisposes the gland to infection with other organisms (
Corynebacterium species can be divided into lipophilic and nonlipophilic subgroups; the growth of members of the lipophilic subgroup being enhanced by the addition of free fatty acid (e.g., 1% Tween 80) to the growth medium (
) and is currently the only lipophilic Corynebacterium species that has been isolated from milk of bovine origin; although a new lipophilic species (Corynebacterium mastitidis) has been isolated from sheep with subclinical mastitis (
Many laboratories presumptively diagnose C. bovis based on the presence of tiny off-white nonhemolytic colonies after 48 to 72 h growth on blood agar cultured at 37°C (
). Some strengthen their diagnosis based on Gram stain, morphology, catalase test, and demonstration of enhanced growth with the addition of Tween 80 to the culture media. None of these tests will, however, definitively identify C. bovis. For routine mastitis bacteriological examination, the misclassification of some isolates as C. bovis is of little consequence because of its limited clinical significance. However, for studies and trials that ascribe either pathogenicity or biological significance e.g., a “protective” effect, to the presence of C. bovis it is necessary to confirm the identity with more definitive tests.
Identification to a species level can be performed by comparing the biochemical reactions and fermentation profiles of test isolates to those of type strains (
). Two commercial identification kits based on these reactions are available (API Coryne and the Biolog system); however, they correctly identified only 88.0 and 54.0% of C. bovis isolates in a recent study (
Sequencing and phylogenetic analysis of the 16S rRNA gene is now recognized as one of the reference techniques for differentiating closely related species. It is particularly useful for Corynebacterium species because 16S rRNA gene sequence divergence rates are relatively high (
). However, gene sequencing is still relatively expensive and not within the capabilities of many laboratories, which makes it an unsuitable technique in large-scale studies.
Gene sequencing is currently used extensively as a tool for differentiating closely related species and identifying novel species. Consequently the 16S rRNA gene sequences for the type strains of all currently recognised Corynebacterium species are published and available for analysis. This paper describes a technique for differentiating C. bovis from other lipophilic Corynebacterium species based on endonuclease restriction analysis of the 16S rRNA gene sequence.
Materials and Methods
Bacterial Isolates
Aseptic quarter milk samples were collected for microbiological analysis from 505 cows at three sampling time points (end of lactation, at calving, and 7 to 14 d after calving). Samples were frozen immediately after collection and submitted to an accredited laboratory for culture and identification. Samples were cultured using standard techniques (
). Corynebacterium species were presumptively identified based on Gram stain, morphology, haemolisin reaction, and catalase test.
Characterization of Coryneform Isolates as Lipophilic or Nonlipophilic
Coryneform isolates were subcultured onto brain heart agar (Merck, Darmstadt, Germany) with and without the addition of 1% Tween 80, vol/vol, to differentiate lipophilic and nonlipophilic isolates (
Genomic diversity and phylogenetic relationships among lipid requiring diphtheroids from humans and characterization of Corynebacterium macginleyi sp nov.
). Plates were compared after aerobic incubation at 37°C for 48 and 72 h.
Lipophilic species were considered those that demonstrated no or minimal growth on agar without Tween 80 and increased or luxuriant growth on agar containing Tween 80.
Compilation of a 16S rRNA Gene Sequence Database
A 16S rRNA gene sequence database was created from published data for all lipophilic Corynebacterium species. Sequence data was imported from GenBank, held by the National Center for Biotechnology Information (www.ncbi.nlm.nih.gov), into sequence manipulation software (Omiga 2.0, Genetics Computer Group, Inc., Madison, WI). At least one 16S rRNA gene sequence was available for the 11 currently validated lipophilic Corynebacterium species [(C. accolens (
Taxonomic study of Corynebacterium group Anf-1 strains. Proposal of Corynebacterium afermentans sp nov containing the subspecies C. afermentans subsp afermentans subsp nov and C. afermentans subsp lipophilum subsp nov.
Genomic diversity and phylogenetic relationships among lipid requiring diphtheroids from humans and characterization of Corynebacterium macginleyi sp nov.
Sequences downloaded into the database were aligned using the sequence manipulation software. Forward and reverse primers were designed to be universal for all available 16S rRNA gene sequences from lipophilic Corynebacterium species, located outside the hypervariable regions and positioned close to either end of the 16S rRNA gene. They had a G + C content of 50 to 55%, similar melting temperatures (approximately 60°C) and at least one G or C at the 3′ end. Homopolymeric regions, sequence repeats, and self-complimentary areas were avoided. Forward primer: 5′ – GCG AAC GGG TGA GTA ACA CG – 3′; Reverse primer: 5′ – TCT GCG ATT ACT AGC GAC TCC G – 3′. All 16S rRNA gene sequences held within the database were shortened at the 5′ and 3′ ends, so that sequences began with the forward primer and ended with the reverse primer.
Prediction of Restriction Endonuclease Digest Patterns
Using the 16S rRNA sequence database, restriction endonuclease digest patterns (number and size of resultant fragments) were predicted for all lipophilic Corynebacterium species using the following common enzymes: AluI, BamHI, BgII, BgIII, EcoRI, EcoRV, HaeIII, HindIII, HinfI, KpnI, MluI, MspI, NcoI, NotI, NruI, PstI, PvuII, RsaI, SalI, SfiI, SmaI, SphI, TaqI, XbaI, and XhoI.
Based on the number and size of fragments, it was predicted that endonuclease restriction with SmaI and HindIII could be used to differentiate C. bovis from all other lipophilic species (Table 1).
Table 1Predicted number of restriction sites and size of fragments produced, in base pairs, after endonuclease restriction of lipophillic Corynebacterium species 16S rRNA gene sequences with HindIII and SmaI.
The sequence data from 49 C. bovis isolates was available for analysis, 3 representative isolates are presented here, data for the other 46 isolates is not shown.
The sequence data from 49 C. bovis isolates was available for analysis, 3 representative isolates are presented here, data for the other 46 isolates is not shown.
The sequence data from 49 C. bovis isolates was available for analysis, 3 representative isolates are presented here, data for the other 46 isolates is not shown.
Predicted size of fragments (in base pairs) after endonuclease restriction.
886
889
299
45
58
888
58
299
58
888
1132
1132
1131
888
1131
SmaI
1
1
1
0
0
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
0
488
488
487
494
493
485
472
483
746
746
747
747
747
737
732
740
1A = C. bovis (AJ222965), B = C. bovis (X84444), C = C. bovis (AF311389), D = C. accolens (X80500), E = C. afermentans (X81874), F = C. afermentans (X82054), G = C. afermentans (X82055), H = C. apendicis (AJ314919), I = C. F-1 (X81904), J = C. F-1 (X81905), K = C. G-2 (X80498), L = C. jeikeium (X84250), M = C. jeikeium (X82062), N = C. jeikeium (U87815), O = C. jeikeium (U87816), P = C. jeikeium (U87823), Q = C. lipophiloflavum (Y09045), R = C. macginleyi (X80499), S = C. mastitidis (Y09806), T = C. urealyticum (X81913), and U = C. urealyticum (X844439).
2 The sequence data from 49 C. bovis isolates was available for analysis, 3 representative isolates are presented here, data for the other 46 isolates is not shown.
3 Predicted number of restriction sites with each enzyme.
4 Predicted size of fragments (in base pairs) after endonuclease restriction.
. Briefly cellular mass was emulsified in double distilled water and DNA was released from cells by placing the solution in a boiling water bath for 10 min. The resultant solution was allowed to cool and centrifuged briefly to pellet cellular debris.
Polymerase Chain Reaction
Two and a half microliters of boiled product to provide DNA template was amplified by PCR using Taq DNA polymerase in a PCR premix solution (Taq PCR Mastermix, Qiagen, Crawley, West Sussex, UK). The premix solution provided a final concentration of 1.5 mM magnesium chloride and 200 μM of each dNTP. Amplification of the 16S rRNA gene sequence was performed using 50 ng of each of the primer pair described previously, using a Techne Genius FGEN02TD thermocycler (Cambridge, Cambridgeshire, UK) with a heated lid using the following temperature regime: 94°C for 2 min followed by 35 cycles of 94°C for 1 min, 56°C for 1 min, 72°C for 2.5 min, and concluding with 72°C for 10 min.
PCR Product Purification
The resulting PCR products were purified using a commercially available filter kit (Microcon PCR Centrifugal Filter Devices, Millipore Corporation, Bedford, MA), according to the manufacture's instructions. PCR product was recovered in 20 μL of ultrapure water and frozen at −20°C until required.
Gel Electrophoresis
Purified PCR products were visualized by gel electrophoresis in 1% (wt/vol) agarose gels (Agarose, Transgenomic, Crewe, Cheshire, UK). Electrophoresis was carried out in Tris-acetate buffer at between 4 and 6 V/cm. Gels were stained with 5 μg of ethidium bromide (Sigma, Poole, Dorset, UK), visualized by ultraviolet transillumination and photographed using a gel documentation system (UVP, GDS 7500, Cambridge, Cambridgeshire, UK). Gel images were captured and manipulated using GRAB IT 2.59 (Synoptics Ltd., Cambridge, Cambridgeshire, UK).
Restriction Endonuclease Cleavage
Purified PCR products were cleaved using HindIII and SmaI restriction endonuclease enzymes. Digests were performed in the buffer supplied by the manufacturer [HindIII (Buffer SB) and Sma I (Buffer SA), Sigma, Poole, Dorset, UK], using 2.5 units of enzyme (HindIII and SmaI restriction endonucleases, Sigma). The concentration of DNA in the PCR product was estimated by comparison to a known standard marker in the DNA molecular weight ladder (GeneRuler 1-kb DNA Ladder, MBI Fermentas, Vilnius, Lithuania); approximately 500 ng of DNA was used in each restriction reaction. Reactions were incubated for at least 6 h at 37°C (HindIII) and 25°C (SmaI). Cleavage products were separated and visualized in 1% agarose gels as previously described. The size of cleavage products was estimated by comparison to a DNA molecular weight ladder (GeneRuler 1-kb DNA Ladder, MBI Fermentas).
Reference Strains
Six C. bovis reference strains were obtained from a culture collection held by the Veterinary Laboratories Agency, England, and processed in the manner outlined above.
Results
Coryneform isolates (n = 939) were identified in the milk samples collected. Of these, 6 were not available for further analysis. Of the 933 isolates examined 762 (81.7%) were lipophilic, and 171 (18.3%) were nonlipophilic.
A PCR product of the predicted size (approximately 1250 bp) was generated by all 762 lipophilic isolates. From these, 741 (97.2%) restricted in the manner predicted for C. bovis, i.e., HindIII and SmaI each cut the PCR product once producing fragments of approximately 350 and 890 bp (representative examples are shown in Figure 1) and 490 and 750 bp (representative examples are shown in Figure 2), respectively. Twenty one (2.8%) lipophilic isolates did not restrict in the pattern predicted for C. bovis.
Figure 1Agarose gel electrophoresis image of the 16S rRNA PCR product of Corynebacterium bovis after endonuclease restriction with HindIII. HindIII digestion of the 16S rRNA PCR product of C. bovis produced 2 digest products (MW ∼ 350 and 890 bp), visualized on a 1% agarose gel. Lane 1: Negative control (HindIII digest reaction with no DNA), Lane 2: Isolate 47 (-ve control, no HindIII), Lane 3: Isolate 47 after HindIII digestion, Lane 4: Isolate 462 (-ve control, no HindIII), Lane 5: Isolate 462 after HindIII digestion, Lane 6: Isolate 3933 (-ve control, no HindIII), and Lane 7: Isolate 3933 after HindIII digestion.
Figure 2Agarose gel electrophoresis image of the 16S rRNA PCR product of Corynebacterium bovis after endonuclease restriction with SmaI. SmaI digestion of the 16S rRNA PCR product of C. bovis produced 2 digest products (MW ∼ 490 and 750 bp), visualized on a 1% agarose gel. Lane 1: Negative control (SmaI digest reaction with no DNA), Lane 2: Isolate 47 (-ve control, no SmaI), Lane 3: Isolate 47 after SmaI digestion, Lane 4: Isolate 462 (-ve control, no Sma I), Lane 5: Isolate 462 after SmaI digestion, Lane 6: Isolate 3933 (-ve control, no SmaI), and Lane 7: Isolate 3933 after SmaI digestion.
Twenty one isolates did not restrict in the pattern predicted for C. bovis. Of these, 18 were not cut by HindIII but were cut in an identical fashion to C. bovis isolates with SmaI (Representative examples are given in Figures 3 and 4). Two were not cut by either enzyme and the remaining isolate was not cut by SmaI but was cut in an identical fashion to C. bovis isolates with HindIII.
Figure 3Agarose gel electrophoresis image of the 16S rRNA PCR product of isolates 4376, 4396, and 4465 after endonuclease restriction with HindIII. HindIII digestion of the 16S rRNA PCR product of isolates 4376, 4396, and 4465 produced no digest products, visualized on a 1% agarose gel. Lane 1: Negative control (HindIII digest reaction, no DNA), Lane 2: Positive control (Corynebacterium bovis, no HindIII), Lane 3: Positive control (C. bovis after HindIII digestion), Lane 4: Isolate 4376 (-ve control, no HindIII), Lane 5: Isolate 4376 after HindIII digestion, Lane 6: Isolate 4396 (-ve control, no HindIII), Lane 7: Isolate 4396 after HindIII digestion, Lane 8: Isolate 4465 (-ve control, no HindIII), and Lane 9: Isolate 4465 after HindIII digestion.
Figure 4Agarose gel electrophoresis image of the 16S rRNA PCR product of isolates 4376, 4396, and 4465 after endonuclease restriction with SmaI. SmaI digestion of the 16S rRNA PCR product of isolates 4376, 4396, and 4465 produced two digest products (MW ∼ 490 and 750 bp), visualized on a 1% agarose gel. Lane 1: negative control (SmaI digest reaction, no DNA), Lane 2: Isolate 4376 (-ve control, no SmaI), Lane 3: Isolate 4376 after SmaI digestion, Lane 4: Isolate 4396 (-ve control, no SmaI), Lane 5: Isolate 4396 after SmaI digestion, Lane 6: Isolate 4465 (-ve control, no SmaI), and Lane 7: Isolate 4465 after SmaI digestion.
) have been used to investigate the Corynebacterium genus by previous research workers. In both cases, the authors made no attempt to predict the outcome of restriction with the enzymes they used; they simply relied on different species restricting with different patterns. To the authors knowledge this study is the first in which members of the Corynebacterium genus have been differentiated from other species using enzymes for which the number and size of resultant DNA fragments was predicted prior to restriction. This has only become possible because of the increasingly widespread use of 16S rRNA gene sequencing as a method of identifying new species. Consequently, the sequences for all currently recognized lipophilic members of the Corynebacterium genus are freely available for analysis. Predicting the number and size of DNA fragments makes restriction analysis a much more powerful diagnostic tool because it enables enzymes to be selected for specific purposes, e.g., differentiating one species from all others within a group. It also becomes possible to create decision trees, which allow all members of a group or genus to be identified by the progressive application of different enzymes. This technique may be especially useful if species can be divided into small but distinct groups by the application of simple and reliable tests (e.g., demonstration of lipophilic nature in this study).
16S rRNA gene sequencing has recently been successfully employed to identify and differentiate C. bovis from other coryneforms present in milk (
) and is probably now the method of choice for definitively identifying Corynebacterium species. However, the technique is still relatively expensive and not within the capabilities of most diagnostic laboratories. Commercially available identifications kits are a cheaper alternative but correctly identified only 88.0 and 54.0% of C. bovis isolates in a recent study (
). Compared to sequencing, the endonuclease restriction analysis method outlined here is relatively cheap and easy to perform. Until 16S rRNA sequencing becomes more generally accessible, the method described here is a viable alternative to more traditional methods of identification (e.g., biochemical reactions and fermentation profiles) for the differentiation of C. bovis from other lipophilic Corynebacterium species. The method is particularly appropriate for large-scale field studies investigating the prevalence and significance of C. bovis. In these situations an accurate diagnosis is essential so that any protective or pathogenic affects demonstrated can be correctly accredited to C. bovis.
Endonuclease restriction analysis of the 16S rRNA gene sequence identified 97.2% of lipophilic coryneform isolates present in milk samples as C. bovis. The other 2.8% were identified as species other than C. bovis. To the authors’ knowledge, this is the first time that lipophilic coryneform species other than C. bovis have been identified in bovine milk, although a novel lipophilic Corynebacterium species (C. mastitidis) has recently been identified in sheep with subclinical mastitis (Fernandez-Garayzabal et al., 1977). None of the isolates examined as part of the study described here restricted with the pattern predicted for C. mastitidis.
Ongoing work has demonstrated that 18 of the 21 isolates (not cut by HindIII but cut identically to C. bovis isolates with SmaI) that were identified as species other than C. bovis may well represent a novel lipophilic Corynebacterium species. These isolates will be described in a future publication. Further restriction analysis of the remaining 3 isolates with PstI, RsaI, and PvuII tentatively identified one as “C. genitalium”, a Corynebacterium species not currently validated. The remaining 2 isolates did not restrict in a pattern predicted for any currently recognised lipophilic Corynebacterium species and remained untyped at the end of the study.
Corynebacterium bovis could be differentiated from all other currently recognized lipophilic species using the restriction endonucleases SmaI and HindIII. Of the 49 C. bovis 16S rRNA gene sequences suitable for analysis, 47 had identical predicted restriction patterns after digestion with these 2 enzymes, only 2 did not. One sequence (AF311397) contained an extra SmaI restriction site and another (AF311419) contained an extra HindIII restriction site. No lipophilic coryneform isolate analyzed during the course of this study restricted in the pattern predicted for either of these sequences. If the extra restriction sites identified in these 2 sequences are real, it would appear that the prevalence of strains carrying these sequence variations are low, or they may represent strains more prevalent in other geographical locations. In fact, it appears that most of the C. bovis 16S rRNA sequences currently published are from isolates originally identified in North America.
Conclusions
This paper describes a novel method for differentiating C. bovis from other lipophilic Corynebacterium species based on endonuclease restriction analysis of the 16S rRNA gene sequence with HindIII and SmaI. Results of the analysis suggest that it is not safe to assume that all lipophilic coryneforms isolated from milk are C. bovis. Three percent of isolates were identified as species other than C. bovis by this method. For routine analysis of bovine milk samples this error is of little consequence and therefore a presumptive diagnosis of C. bovis based on phenotypic and growth characteristics is adequate. However, for epidemiological studies, investigations of C. bovis clinical mastitis outbreaks and studies relating to the properties, characteristics, and pathogenic interactions of C. bovis, further tests such as biochemical reactions, fermentation profiles, 16S rRNA gene sequencing or the endonuclease restriction technique described here are necessary to confirm species identity.
Acknowledgments
The work described in this paper was completed by J. N. Huxley while studying for a Ph.D. funded by Cross Vetpharm Group, Dublin; A. J. Bradley was supported by a fellowship funded by the Wellcome Trust. The authors would like to acknowledge and thank them for their support.
References
Bernard K.A.
Munro C.
Wiebe D.
Ongsansoy E.
Characteristics of rare or recently described Corynebacterium species recovered from human clinical material in Canada.
The susceptibility of bovine udder quarters colonized with Corynebacterium bovis to experimental infection with Staphylococcus aureus or Streptococcus agalactiae.
Analysis of the 16S rRNA gene sequence of the coryneform bacterium associated with hyperkeratotic dermatitis of athymic nude mice and development of a PCR-based detection assay.
Taxonomic study of Corynebacterium group Anf-1 strains. Proposal of Corynebacterium afermentans sp nov containing the subspecies C. afermentans subsp afermentans subsp nov and C. afermentans subsp lipophilum subsp nov.
Genomic diversity and phylogenetic relationships among lipid requiring diphtheroids from humans and characterization of Corynebacterium macginleyi sp nov.
Experimental Staphylococcus aureus intramammary challenge in late lactation dairy cows: Quarter and cow effects determining the probability of infection.
73139-2&id=gr1.jpg){kind=link}
73139-2&id=gr2.jpg){kind=link}
73139-2&id=gr3.jpg){kind=link}
73139-2&id=gr4.jpg){kind=link}