Acute Phase Protein as Biomarker for Diagnosis of Sub-Clinical Mastitis in Cross-Bred Cows

Bovine mastitis is an inflammatory reaction of one or more quarters of the mammary gland to bacterial, chemical, thermal or mechanical injury and cost around 72 billion per year (Kumar et al., 2014). Despite worldwide efforts, mastitis has remained economically the most important disease in dairy cattle and despite different control programs it is still a major challenge for the dairy industry (Bradley, 2002).
       
Sub-clinical mastitis (SCM) is more challenging than clinical mastitis due to absence of visible changes in the udder or milk and causes two-third losses of the total milk production. Animals with sub-clinical mastitis should be considered as a constant risk of infection within and between herds (Persson et al., 2011).
       
During intra mammary infections (IMI), somatic cells such as neutrophils increase greatly in the mammary tissue and milk (Dohoo and Meek,1982). Somatic cell count (SCC) values are usually used in diagnostic laboratories as indicators of mastitis. Currently the most common way to diagnose SCM on dairy farms is somatic cell count (Akerstedt et al., 2007).  However, SCC is affected by various factors other than IMI, such as physiological status, seasonal variation and has also been shown to remain in the milk long after the removal of infectious causes of mastitis from milk (Pyorala, 2003).
       
Sub-clinical mastitis is frequently diagnosed by cultural examination, California mastitis and electrical conductivity and by SCC. These tests have their own limitations; therefore, it is of great importance to investigate biomarkers (acute phase protein) that could be used for early and rapid detection of sub-clinical mastitis (Kumar et al., 2014).
       
In view of the above the aim of the present study was to evaluate the levels and to compare the sensitivity of serum amyloid-A (SAA), milk amyloid-A (MAA) and somatic cell count (SCC) in different udder health conditions.

Selection of animals
 
A total of 40 apparently healthy cross-bred cows, in the mid-lactation period from Livestock Farm, College of Veterinary Science and Animal Husbandry, Mhow and from nearby villages were included in this study. The clinical cases were not included in this study.
 
Collection of milk samples
 
Ten ml of milk samples were collected from individual quarter of each animal for assessment of California Mastitis Test (CMT). The animals with CMT test negative for all four quarters milk sample was included in control group. CMT test positive quarter samples of one animal were pooled to make a single sample for somatic cell countand bacterio­logical examination within 24 hours. Further, samples were stored at -18°C until the MAA assay was performed.
 
Collection of blood samples
 
Five ml of blood samples were collected aseptically in a tube without anticoagulant from jugular vein of each cross-bred cow and allowed for clotting. The serum samples were transferred to sterilize tubes and kept frozen at -20°C until the SAA assay was performed.
 
Bacteriological examination
 
Bacteriological examination of milk sam­ples were performed as per standard methods (Malinowski and Klossowska, 2002). The milk samples collected aseptically were streaked on Mueller-Hinton agar media (HI-MEDIA) in order to obtain growth of bacteria. Plates were incubated at 37°C for 24 - 48 hours. The presence of bacteria was observed by growth of colonies.
 
California mastitis test (CMT)

California mastitis test (CMT) was performed separately for milk samples of all four quarters of each animal as per the method described by Schalm and Noorlander (1957). The reaction was graded by intensity of gel formation. Results were scored as 0 (negative or trace), + (weak positive), ++ (distinct positive) and +++ (strong positive).
 
Somatic cell count
 
The somatic cell count of milk samples was performed as described by Schalm et al., (1971). Milk (0.01 ml) was withdrawn by micropipette and spread evenly on a grease free glass slide in 1 square cm area. The smear was dried in air and stained by immersing the slide in Newman’s Lampert stain (this stain fixes the film and remove the fat from milk) for 15 seconds. Then allowed the slide to dry for 30 sec, washed the slide with water, dried the slide and examined under the oil immersion lens.
 
Categorization of milk samples according to udder health conditions (n=10)
 
• Group I (G I): Control group; Bacteriologically negative milk samples, California mastitis test (CMT) negative.
• Group II (G II): Bacteriologically negative milk samples with SCC ≤ 200,000 cells/ml milk (CMT positive).
• Group III (G III): Bacteriologically negative milk samples with SCC≥200,000 ≤500,000 cells/ml milk (CMT positive).
• Group IV (G IV): Bacteriologically positive milk samples with SCC ≥200,000 ≤500,000 cells/ml milk (CMT positive).
 
Determination of serum amyloid A
 
The concentration of serum amyloid A was determined in ten serum samples (n=10) selected from each of the above mentioned groups by solid phase sandwich Enzyme Linked Immuno Sorbent Assay (ELISA) kit procured from Tridelta Development Ltd., Ireland. Prior to testing the serum samples were diluted as 1:500. The results were expressed in µg/ml.
 
Determination of milk amyloid A
 
The concentration of milk amyloid A was determined in ten milk samples (n=10) selected from each of the above mentioned groups by solid phase sandwich Enzyme Linked Immuno Sorbent Assay (ELISA) kit procured from Tridelta Development Ltd., Ireland. Prior to testing the milk samples were diluted as 1:50. The results were expressed in µg/ml.
 
Receiver operating characteristic (ROC) analysis for sensitivity and specificity
 
Receiver operating characteristic (ROC) analysis was used to evaluate and compare the sensitivity, specificity and accuracy of somatic cell count and amyloid A in serum and milk for the detection of sub-clinical mastitis based on bacterial culture results.
 
Ethical approval
 
The present study was conducted with the approval of Institutional Animal Ethical Committee, College of Veterinary Science and Animal Husbandry, Mhow, Madhya Pradesh.

California mastitis test (CMT)
 
On the basis of score card of California mastitis test of 57 milk samples from apparently healthy udder of cross-bred cows, the results were interpreted as negative (-), trace, weak positive(+), distinct positive (++) and strong positive (+++). 12 (21.05%) samples were found negative (-), 06 (10.52%) samples were found trace, 16 (28.07%) samples were found weak positive (+), 12 (21.05%) samples were found distinct positive (++) and 11 (19.29%) samples were found strong positive (+++).
       
Ondiek et al., (2013) examined milk samples from 41 lactating cows, 34.1 per cent samples were found CMT positive for sub-clinical mastitis.Ayanoet_al(2013) examined a total of 546 milking cows, out of which 224 (41.02 per cent) were positive for sub-clinical mastitis on basis of California mastitis test (CMT).These findings are partially in accordance with the findings of the present study.
 
Somatic cell count (SCC)
 
The mean values of somatic cell count in the milk samples of group-I, group-II, group-III and group-IV were 1.24±0.13, 1.45±0.12, 3.77±0.32 and 3.98±0.33 (10⁵/ml), respectively. A highly significant (P<0.01) differences in the mean valuesof somatic cell count (SCC) were observed between groups except group I and group II and group III and group IV.
       
The present findings are in agreement with the findings of Safi et al., (2009). They reported the mean concentration of SCC was higher in culture positive cows than culture negative cows. Chahar (2001) examined milk of 78 cows and reported the average mean ± SE value of somatic cell counts in sub-clinical mastitic milk of cows were 1.900 ± 0.53 million cells/ml (ranged between 0.5412 ± 8.0479 million cells/ml) and the mean ± SE value of SCC in normal milk of cows were 0.4621 ± 0.09 million cells/ml. These findings are also in accordance with the present findings.
       
On the basis of number of somatic cell count (10⁵ cells/ml), 07 (12.28%) samples were found SCC in between 0.70-1.0, 14 (24.56%) samples were found SCC in between 1.0-2.0, 09 (15.78%) samples were found SCC in between 2.0-3.0, 09 (15.78%) samples were found SCC in between 3.0-4.0 and 18 (31.58%) samples were found SCC in between 4.0-5.0 (105 cells/ml).
       
Similar finding was reported by Skrzypek et al., (2004). They revealed the somatic cell count of milk was in between 50,000 and 100,000 cells/ml in healthy udder and 200,000 cells/ml in animals affected with sub-clinical mastitis.
 
Bacteriological examination
 
On the basis of bacteriological examination of milk, 27 (47.36%) cross-bred cows were found bacteriologically positive and 30 (52.63%) cross-bred cows were found bacteriologically negative.
       
Farag et al., (2017) observed that single bacterial isolates of Escherichia coli and Staphylococcus aureus were constituted 23 (19.2%) and 17 (14.2%) of 120 mastitic milk samples respectively, Meanwhile the mixed infection was 35 (29.16%) isolates. These findings are partially accordancewith the findings of the present study.
       
Somatic cell count (SCC) revealed that, of the examined 57 apparently healthy milk samples from cross-bred cows, 21 (36.84%) samples had SCC≤200,000 cells/ml and 36 (63.15%) samples had SCC ≥200,000 ≤500,000 cells/ml. Out of 21 samples with  SCC ≤200,000 cells/ml, 20 (95.23%) samples were found negative for bacteriology and 01 (4.76%) sample was reported positive for bacteriology. Of the 36 milk samples with SCC≥200,000 ≤500,000 cells/ml, 10 (27.77%) samples were reported negative for bacteriology and 26 (72.22%) samples were found positive for bacteriology.
 
The similar result was reported by Safi et al., (2009). Out of 173 cows, milk samples were collected from 692 quarters. Of these, 536 (77.46%) milk samples from 134 cows had negative CMT and a SCC<200,000cells/ml.
 
Preparation of calibration curve
 
A calibration curve was prepared for determination of serum and milk amyloid A using six standard calibrators for bovine with the concentrations of 0.3, 0.15, 0.075, 0.0375, 0.0188 and 0 µg/ml.

Serum amyloid A (SAA)
 
The mean values of serum amyloid A in group-I, group-II, group-III and group-IV were 10.80±2.11, 23.10±4.42, 56.60±7.46 and 88.00±0.66 µg/ml, respectively (Table 1). Highly significant (P<0.01) difference in mean values of SAA was reported between all the groups whereas no significant difference was observed between group I and group II.
 

       
Szczubial et al., (2008) reported the mean SAA concentration in milk from healthy cows was (11.67±7.40) µg/mL and was significantly lower (P<0.01) compared to that in milk from cows with the different forms of mastitis. Safi et al., (2009) observed the mean concentration of serum amyloid A was higher in culture positive sample as compared to negative samples.  Horadagoda et al., (1999) determined the SAA level in apparently healthy cows was <8.8 mg/l. These findings were in agreement with the findings of the present study.
 
Milk amyloid A (MAA)
 
The mean values of milk amyloid A in group-I, group-II, group-III and group-IV were 2.30±0.60, 9.20±1.61, 12.90±0.85 and 13.19±0.91 µg/ml, respectively (Table 1). Highly significant (P<0.01) difference in mean values of MAA was reported between all the groups but no significant difference was observed between group III and group IV.
       
Hussein et al., (2018) observed that there was a prompt increase in MAA concentration in Group III (group of milk samples had SCC £200,000 cells/ml and bacteriologically positive) than Group I (group of milk samples with SCC ≤500,000 cells/ml and bacteriologically negative). Safi et al., (2009) observed the concentration of milk amyloid A was higher in positive bacterial culture group than negative bacterial culture group. These observations are in agreement with the report of the present findings. Higher concentration of SAA activity in milk from cows with subclinical mastitis compared to that in healthy cows indicate that this protein is good marker of inflammatory processes in the udder, even mild ones (Szczubial et al., 2008).
 
Association among acute phase proteins (APPs) and somatic cell count (SCC)
 
To explore the interdependencies among serum amyloid A (SAA), milk amyloid A (MAA) and somatic cell count, Pearson’s correlation technique was applied to the observed values. Results of the correlation matrix revealed significantly (P<0.01) strong positive correlations of SCC with acute phase proteins, serum amyloid A (r = 0.718**, P<0.01) and milk amyloid A (r = 0.548**, P<0.01) concentration in serum and milk, respectively and the correlation serum amyloid A with milk amyloid A was (r = 0.485**, P<0.01) (Table 2).
 

       
The results of the present findings are in accordance with Singh et al., (2015). They observed highly significant positive correlation of SCC and MAA (r=0.810**, P<0.01) concentration in milk.
 
Receiver operating characteristic (ROC) analysis for sensitivity and specificity of somatic cell count
 
In ROC analysis, the sensitivity (Se), specificity (Sp), positive predictive value, negative predictive value and accuracy of somatic cell count were (100, 26.66, 31.25, 100 and 45), (100, 66, 50, 100 and 75) and (90, 73.33, 52.94, 95.65 and 77.50) at various proposed cut-off levels of 1.5 (10⁵ cells/ml), 2.0 (10⁵ cells/ml) and 2.5 (10⁵ cells/ml), respectively. The area under curve (AUC) was 0.95 for somatic cell count (Table 3).
 

       
O’ Mahoney et al., (2006) reported the Se and Sp of 80% and 81%, respectively, for detecting mastitis in quarters with a SCC>150,000 cells/ml. Safi et al., (2009) observed that the Se and Sp of 89.6% and 72%, respectively and area under curve was 0.948 at cut-off level of >130000cells/ml. Jaeger et al., (2017) determined the Se and Sp of  90.3% and 71.8%, respectively where cut-off level of SCC was 1,50,000 (cells/ml). Mcdermott et al., (1982) reported Se of 92% and Sp of 53% at a threshold of 100,000 cells/ml and using a higher threshold of 200,000 cells/ml, they reported Se of 89% and Sp of 75% for subclinical mastitis. These findings were partially in accordance with the findings of the present study.
 
Receiver operating characteristic (ROC) analysis for sensitivity and specificity of serum amyloid A
 
In ROC analysis, the sensitivity (Se), specificity (Sp), positive predictive value, negative predictive value and accuracy of serum amyloid A were (100, 83.33, 66.66, 100 and 87.50), (100, 93.33, 83.33, 100 and 95) and (60, 96.66, 85.71, 87.87, 87.50) at various proposed cut-off levels of 50 (µg/ml), 74 (µg/ml) and 87 (µg/ml), respectively. The area under curve (AUC) was 1.0 for serum amyloid A (Table 4).
 

       
Safi et al., (2009) reported the Se and Sp of SAA was 90% and 72.1% at the cut-off level >159.1 (mg/l). Our findings were partially similar with that of Eckersall et al., (2001) who reported high Sp (100%) and a reasonable Se (93%) with measurement of SAA concentrations in serum.
 
Receiver operating characteristic (ROC) analysis for sensitivity and specificity of milk amyloid A
 
In ROC analysis, the sensitivity (Se), specificity (Sp), positive predictive value, negative predictive value and accuracy of milk amyloid A were (90, 50, 37.50, 93.75 and 60), (90, 56.66, 40.90, 94.44 and 65) and (80, 60, 40, 90 and 65) at various proposed cut-off levels of 7 (µ/ml), 10 (µ/ml) and 13 (µ/ml), respectively. The area under curve (AUC) was 0.85 for milk amyloid A (Table 5).
 

       
The similar results were reported by Safi et al., (2009). They investigated that the concentration of milk amyloid Awas the most accurate of the 5 tests, with a Se of 90.6% and Sp of 98.3%. The present findings are partially in accordance with Singh et al., (2015). They revealed high Se (100%) and high Sp (100%) of MAA at cut off value of 0.264 μg/ml^-1 and area under curve is 1.0 which indicates very good test. Haghkhah et al., (2010) also reported highest Se and Sp (100%) using MAA as a diagnostic parameter compared to other acute phase proteins.

 The authors are thankful to the Hon’ble Vice Chancellor, Nanaji Deshmukh Veterinary Science University, Jabalpur, Madhya Pradesh for providing research facilities in form of infrastructures and consumables.