Research in VeterinaryScience 1991, 51, 299-305
Examination of milk samples from experimentally infected heifers by an N-acetyl-fl,D-glucosaminidase test and an antigen-capture ELISA H. J. BALL, D. GREER, D. FINLAY, D. P. MACKIE, Veterinary Research Laboratories, Stoney Road, Stormont, Belfast BT4 3SD
Milk samples collected from normal and experimentally infected quarters of five heifers throughout their first lactation were examined by bacterial culture, milk cell count (MCC), N-acetyl-~-D-glucosaminidase (NAGase)test and a monoclonal antibody based antigen-capture ELISA. The results were analysed according to the presence (mastitis positive) or absence (mastiffs negative) of bacteria on culture, and the numbers of false negative and false positive results of the other three tests defined in relation to this. Similar numbers of false negative results were observed with the MCC (20), NAGase (18) and ELISA (13). False positive results due to the physiological factors present in early lactation were evident in the MCC, prominent in the NAGase test and absent from the EHSA. The major difference in false positive results associated with experimental infection between the three tests was the more rapid return to negative values of the ELISA following resolution of infection, compared with MCC and NAGase.
limited information on such things as efficiency of milking machinery, lactation period, antibiotic treatment and other factors which non-specifically influence MCC was provided with the samples. In addition, badly collected samples with high bacterial contamination limited inclusion of the bacterial results in the analysis. Thus only a comparison of the NAGase test and ELISAwith the MCC was possible. An M¢¢ of 400,000 cells ml-~ was taken as the cut-off point to distinguish between mastitic and nonmastitic milk. The false negative (over 400,000 cells ml -~ M¢¢ with low test level) and false positive results (under 400,000 cells m l - ~with high test levels) could not be evaluated further. The aim of the present study was systematically to evaluate the NAGase test and the ELISA for the detection of subclinical mastitis in a controlled experiment. Milk samples were collected regularly from five heifers throughout their first lactation, during which experimental infection was induced. All samples were examined by bacterial culture as well as by MCC, NAGasetest and ELISA.
A L T H O U G H bacterial culture together with the estimation of milk cell counts (MCC) is the standard method for detecting subclinical bovine mastitis, levels of enzymes in milk have also been investigated Materials and methods as possible indicators (review, Kitchen 1981). The most successful application of these has been the Experimental samples detection of N-acetyl-fl-D-glucosaminidase (NAGase), The five heifers calved at the laboratory. They were using microtitre methods with a fluorometric sub- milked with quarter milking machines and their milk strate (Mattila and Sandholm 1985, Ball and Greer yield recorded. Samples for laboratory testing were 1991). In addition, an ELISA using monoclonal collected at least every other day during the complete antibody to a bovine 23.5 kD antigen was recently lactation. Each teat was cleaned and disinfected with applied to subclinical mastitis detection (Ball et al methylated spirits soaked cotton wool and the fore 1991). The antigen detected by the ELISA was only milk discarded before the samples were collected in characterised by its molecular mass, but appeared to sterile bottles. be associated with the inflammatory response Heifers 1, 2, 3 and 4 were infected by intraconnected with infectious mastitis. mammary infusion in one quarter with cultures of Problems were encountered with the interpretation Streptococcus agalactiae, a mixture of Streptococcus of results of both NAGasetest and ELISAcarried out on uberis and Staphylococcus aureus or S aureus alone, herd surveys in previous studies (Ball and Greer 1991, at various times. The experimental protocol is sumBall et al 1991). The problems arose because only marised in Table 1. All bacterial strains used were 299
300
H. J. Ball, D. Greer, D. Finlay, D. P. Mackie
TABLE 1 : Intramammary infections and treatment of infections in the experimental animals
Cow
Quarter*
1
2
3
4
5
* t ++ §
Intramammary inoculation (lactation day)
Duration of infection? (between lactation days)
Resolution of infection by (number of treatments)
39-50 65-68 89-115
Antibiotic treatment ( x 1 ) Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (11 ) S uberis + S aureus (64) Sterile control (11 ), S aureus (88)
RF RH LF LH
Control, uninoculated S agalactiae (11 $, 24++, 37) S uberis + S aureus (64) Sterile control (11, 24, 37), S aureus (88)
38-64 65-68 101-129
Antibiotic treatment ( x 1 ) Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (11 ++, 225, 37) S uberis + S aureus (64) Sterile control (11, 22, 37), S aureus (86)
41-45 65-66 88-113
Natural elimination Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (26) S uberis + S aureus (46) Sterile control (26), S aureus (82)
RF RH LF LH
Control, uninoculated Control, uninoculated Control, uninoculated Natural infection, S aureus (97)
28-39 47-50, 53-78§ 84-195
97-195
Natural elimination Antibiotic treatment ( x 3) Antibiotic treatment ( x 3)
Dry c o w therapy only
RF Right fore, RH right hind, LF left fore, LH left hind From first to last isolation Infection not established S uberis only after first treatment
freshly isolated from mastitis cases. Some glands eliminated the infection naturally, while others were treated with various intramammary antibiotics. Heifer 5 was not experimentally infected as a control, but naturally developed a S aureus infection in one gland, which was left untreated until the end of lactation. Bacteriology Approximately 20 #1 of milk from all samples were streaked onto blood agar plates which were incubated in air at 37°C for 24 hours. Colonies that developed were identified by standard methods (International Dairy Federation 1971). Milk cell count Milk samples (10 ml) were fixed by overnight incubation at 30°C with 0.2 per cent formalin and the cell level measured electronically using a Coulter counter. Cell levels above 400,000 cells ml-~ were regarded as MCC positive. N-acetyl-13-D-glucosaminidase
test
Duplicate samples of milk (20 tzl) were tested in
microtitre trays (Fluoroplate; Flow Laboratories) by the protocol described by Ball and Greer (1991). Briefly, after cooling to + 4°C, 40/~l aliquots per well of substrate (2.25 mM 4-methyl umbelliferyl-Nacetyl-~-D-glucosamide [Sigma M2133] in 0.25 M citrate buffer, pH 4.6) were added to the milk and the plate reincubated at + 4 ° C for 30 minutes. The reaction was stopped by the addition of 200/~l 0.2 M glycine/sodium hydroxide buffer, pH 10.7, to each well, and read on a fluorometric microplate reader (Titertek Fluoroskan II, Flow Laboratories) at an excitation wavelength of 335 nm and emission wavelength of 480 nm, using an arbitrary scale. A mastitic milk control was included on each plate and used to standardise all readings. The control was stored in small volumes at - 20°C. A sample thawed for each test was used in a series of doubling dilutions to produce a graph plot for each plate. Each plot together with the test plate values was adjusted to standard values in a computer program containing a standard curve. This latter had been calculated by taking the average of 10 graph plots produced from the control sample. This procedure enabled the direct comparison of samples tested at different times. A fluorometric value of 220 was used to distinguish between positive and negative results.
301
Mastitis detection by NAGase and ELISA ELISA
7.2, containing 0.05 per cent Tween 20. A two hour incubation of test samples was followed by a one hour incubation with biotinylated mAb, a one hour incubation with streptavidin-peroxidase (Sigma) and finally 10 minutes with the substrate which consisted of 0-4 mg o-phenylenediamine dihydrochloride (Sigma) m l - l with 0.4 /A m1-1 (v/v) 30 per cent hydrogen peroxide in 0.1 M citrate phosphate buffer, pH 5.0. The substrate reaction was stopped by the addition of 50/zl 2.5 M sulphuric acid per well. The
Duplicate volumes of milk (100/zl) were tested by an antigen-capture format using monoclonal antibody (mAb) 1B3 as described by Ball et al (1991). Briefly, the wells of microtitre plates were coated with mAb in 0" 05 M carbonate buffer, pH 9" 5, overnight at + 4°C. This and all other reagents were used in 100 tzl volumes, all further incubations were at 37°C, and after each incubation stage the wells were washed with five changes of 0- 01 M phosphate buffered saline, pH
2000
~
[
it'~' ,
°00_
V 7
it
•
/fill
tl-"
ii7
=
Examination of milk samples from experimentally infected heifers by an N-acetyl-fl,D-glucosaminidase test and an antigen-capture ELISA H. J. BALL, D. GREER, D. FINLAY, D. P. MACKIE, Veterinary Research Laboratories, Stoney Road, Stormont, Belfast BT4 3SD
Milk samples collected from normal and experimentally infected quarters of five heifers throughout their first lactation were examined by bacterial culture, milk cell count (MCC), N-acetyl-~-D-glucosaminidase (NAGase)test and a monoclonal antibody based antigen-capture ELISA. The results were analysed according to the presence (mastitis positive) or absence (mastiffs negative) of bacteria on culture, and the numbers of false negative and false positive results of the other three tests defined in relation to this. Similar numbers of false negative results were observed with the MCC (20), NAGase (18) and ELISA (13). False positive results due to the physiological factors present in early lactation were evident in the MCC, prominent in the NAGase test and absent from the EHSA. The major difference in false positive results associated with experimental infection between the three tests was the more rapid return to negative values of the ELISA following resolution of infection, compared with MCC and NAGase.
limited information on such things as efficiency of milking machinery, lactation period, antibiotic treatment and other factors which non-specifically influence MCC was provided with the samples. In addition, badly collected samples with high bacterial contamination limited inclusion of the bacterial results in the analysis. Thus only a comparison of the NAGase test and ELISAwith the MCC was possible. An M¢¢ of 400,000 cells ml-~ was taken as the cut-off point to distinguish between mastitic and nonmastitic milk. The false negative (over 400,000 cells ml -~ M¢¢ with low test level) and false positive results (under 400,000 cells m l - ~with high test levels) could not be evaluated further. The aim of the present study was systematically to evaluate the NAGase test and the ELISA for the detection of subclinical mastitis in a controlled experiment. Milk samples were collected regularly from five heifers throughout their first lactation, during which experimental infection was induced. All samples were examined by bacterial culture as well as by MCC, NAGasetest and ELISA.
A L T H O U G H bacterial culture together with the estimation of milk cell counts (MCC) is the standard method for detecting subclinical bovine mastitis, levels of enzymes in milk have also been investigated Materials and methods as possible indicators (review, Kitchen 1981). The most successful application of these has been the Experimental samples detection of N-acetyl-fl-D-glucosaminidase (NAGase), The five heifers calved at the laboratory. They were using microtitre methods with a fluorometric sub- milked with quarter milking machines and their milk strate (Mattila and Sandholm 1985, Ball and Greer yield recorded. Samples for laboratory testing were 1991). In addition, an ELISA using monoclonal collected at least every other day during the complete antibody to a bovine 23.5 kD antigen was recently lactation. Each teat was cleaned and disinfected with applied to subclinical mastitis detection (Ball et al methylated spirits soaked cotton wool and the fore 1991). The antigen detected by the ELISA was only milk discarded before the samples were collected in characterised by its molecular mass, but appeared to sterile bottles. be associated with the inflammatory response Heifers 1, 2, 3 and 4 were infected by intraconnected with infectious mastitis. mammary infusion in one quarter with cultures of Problems were encountered with the interpretation Streptococcus agalactiae, a mixture of Streptococcus of results of both NAGasetest and ELISAcarried out on uberis and Staphylococcus aureus or S aureus alone, herd surveys in previous studies (Ball and Greer 1991, at various times. The experimental protocol is sumBall et al 1991). The problems arose because only marised in Table 1. All bacterial strains used were 299
300
H. J. Ball, D. Greer, D. Finlay, D. P. Mackie
TABLE 1 : Intramammary infections and treatment of infections in the experimental animals
Cow
Quarter*
1
2
3
4
5
* t ++ §
Intramammary inoculation (lactation day)
Duration of infection? (between lactation days)
Resolution of infection by (number of treatments)
39-50 65-68 89-115
Antibiotic treatment ( x 1 ) Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (11 ) S uberis + S aureus (64) Sterile control (11 ), S aureus (88)
RF RH LF LH
Control, uninoculated S agalactiae (11 $, 24++, 37) S uberis + S aureus (64) Sterile control (11, 24, 37), S aureus (88)
38-64 65-68 101-129
Antibiotic treatment ( x 1 ) Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (11 ++, 225, 37) S uberis + S aureus (64) Sterile control (11, 22, 37), S aureus (86)
41-45 65-66 88-113
Natural elimination Antibiotic treatment ( x 1 ) Natural elimination
RF RH LF LH
Control, uninoculated S agalactiae (26) S uberis + S aureus (46) Sterile control (26), S aureus (82)
RF RH LF LH
Control, uninoculated Control, uninoculated Control, uninoculated Natural infection, S aureus (97)
28-39 47-50, 53-78§ 84-195
97-195
Natural elimination Antibiotic treatment ( x 3) Antibiotic treatment ( x 3)
Dry c o w therapy only
RF Right fore, RH right hind, LF left fore, LH left hind From first to last isolation Infection not established S uberis only after first treatment
freshly isolated from mastitis cases. Some glands eliminated the infection naturally, while others were treated with various intramammary antibiotics. Heifer 5 was not experimentally infected as a control, but naturally developed a S aureus infection in one gland, which was left untreated until the end of lactation. Bacteriology Approximately 20 #1 of milk from all samples were streaked onto blood agar plates which were incubated in air at 37°C for 24 hours. Colonies that developed were identified by standard methods (International Dairy Federation 1971). Milk cell count Milk samples (10 ml) were fixed by overnight incubation at 30°C with 0.2 per cent formalin and the cell level measured electronically using a Coulter counter. Cell levels above 400,000 cells ml-~ were regarded as MCC positive. N-acetyl-13-D-glucosaminidase
test
Duplicate samples of milk (20 tzl) were tested in
microtitre trays (Fluoroplate; Flow Laboratories) by the protocol described by Ball and Greer (1991). Briefly, after cooling to + 4°C, 40/~l aliquots per well of substrate (2.25 mM 4-methyl umbelliferyl-Nacetyl-~-D-glucosamide [Sigma M2133] in 0.25 M citrate buffer, pH 4.6) were added to the milk and the plate reincubated at + 4 ° C for 30 minutes. The reaction was stopped by the addition of 200/~l 0.2 M glycine/sodium hydroxide buffer, pH 10.7, to each well, and read on a fluorometric microplate reader (Titertek Fluoroskan II, Flow Laboratories) at an excitation wavelength of 335 nm and emission wavelength of 480 nm, using an arbitrary scale. A mastitic milk control was included on each plate and used to standardise all readings. The control was stored in small volumes at - 20°C. A sample thawed for each test was used in a series of doubling dilutions to produce a graph plot for each plate. Each plot together with the test plate values was adjusted to standard values in a computer program containing a standard curve. This latter had been calculated by taking the average of 10 graph plots produced from the control sample. This procedure enabled the direct comparison of samples tested at different times. A fluorometric value of 220 was used to distinguish between positive and negative results.
301
Mastitis detection by NAGase and ELISA ELISA
7.2, containing 0.05 per cent Tween 20. A two hour incubation of test samples was followed by a one hour incubation with biotinylated mAb, a one hour incubation with streptavidin-peroxidase (Sigma) and finally 10 minutes with the substrate which consisted of 0-4 mg o-phenylenediamine dihydrochloride (Sigma) m l - l with 0.4 /A m1-1 (v/v) 30 per cent hydrogen peroxide in 0.1 M citrate phosphate buffer, pH 5.0. The substrate reaction was stopped by the addition of 50/zl 2.5 M sulphuric acid per well. The
Duplicate volumes of milk (100/zl) were tested by an antigen-capture format using monoclonal antibody (mAb) 1B3 as described by Ball et al (1991). Briefly, the wells of microtitre plates were coated with mAb in 0" 05 M carbonate buffer, pH 9" 5, overnight at + 4°C. This and all other reagents were used in 100 tzl volumes, all further incubations were at 37°C, and after each incubation stage the wells were washed with five changes of 0- 01 M phosphate buffered saline, pH
2000
~
[
it'~' ,
°00_
V 7
it
•
/fill
tl-"
ii7
=
