Mastitis is the most economically important disease in the dairy industry, and subclinical mastitis represents the major portfolio of mastitis cases in the world (Hegde et al., Reference Hegde, Isloor, Prabhu, Shome, Rathnamma, Suryanarayana, Yatiraj, Prasad, Krishnaveni, Sundareshan, Akhila, Gomes and Hegde2013). While the diagnosis of clinical mastitis associated with demonstrable changes in the udder and secreted milk is relatively simple (Safi et al., Reference Safi, Khoshvaghti, Jafarzadeh, Bolourchi and Nowrouzian2009), many challenges are faced in the accurate diagnosis of subclinical mastitis due to its insidious nature. The cow-side diagnostic tests currently in practice targeting changes in the cellular content and pH of the milk are less proficient and incompatible with most of the modern milking systems (Eckersall et al., Reference Eckersall, Young, Nolan, Knight, McComb, Waterston, Hogarth, Scott and Fitzpatrick2006). Failure in prompt identification, isolation and treatment of animals affected with sub clinical mastitis during the early stages of the disease can lead to delay in management and control of the disease, which in turn will facilitate further spread of the pathogens within the herd. Thus, there is scope for identifying and evaluating specific biomarkers capable of detecting subclinical mastitis at a very early stage of the disease (Akerstedt et al., Reference Akerstedt, Waller and Sternesjo2007).
Assessment of changes in the levels of protein constituents of milk and serum as biomarkers for detection of subclinical mastitis has attracted considerable attention in the last decade (Eckersall et al., Reference Eckersall, Young, McComb, Hogarth, Safi, Weber, McDonald, Nolan and Fitzpatrick2001). The potential of acute phase proteins (APP) in this context has been recognized (Biggadike et al., Reference Biggadike, Ohnstad, Laven and Hillerton2002). Mastitis essentially is an inflammatory response of the body against irritants, principally instigated by bacterial, parasitic, algal, and potentially viral and/or environmental elements (Wellenberg et al., Reference Wellenberg, Van der Poel and Van Oirshot2002). Bacterial invasion of the mammary gland leads to a cascade of inflammatory events including an increase in the concentrations of acute-phase proteins in the blood and milk (Simões et al., Reference Simões, Campbell, Viora, Gibbons, Geraghty, Eckersall and Zadoks2017). This increase may precede the onset of clinical signs, making them suitable to be used as specific biomarkers for early detection of subclinical mastitis.
The most important APPs concerned with bovine mastitis are haptoglobin (Hp) and serum amyloid A (SAA) (Thomas et al., Reference Thomas, Waterston, Hastie, Parking, Haining and Eckersall2015), the concentrations of which were reported to increase several fold in the serum and milk during both clinical and subclinical cases. Levels of other APPs like α-1 acid glycoprotein (AGP) (Guha et al., Reference Guha, Guha and Gera2013) and C- reactive protein (CRP) (Thomas et al., Reference Thomas, Waterston, Hastie, Parking, Haining and Eckersall2015) also increase considerably during mammary gland inflammation. There are certain negative APPs also like albumin and transferrin whose concentrations in serum decrease as a result of inflammation. Milk components other than APP have also been investigated as potential biomarkers. Lactoferrin (Lf) in milk was found to be one such reliable biomarker for diagnosis of subclinical mastitis by Sadek et al. (Reference Sadek, Saleh and Ayoub2016).
Subclinical mastitis caused by bacteria has a multifactorial etiology. Even though Staphylococci predominates, Streptococci, Corynebacterium and Mycoplasma are also implicated in many instances (Gerardo et al., Reference Gerardo, Bernardini, Elia, Ferrari, Iob and Segato2009). Differences in pathogenesis among these bacteria could be reflected in the inflammatory response of the body towards their presence, as the essential driver of irritation can decide the kind of cell reaction that happens against it in the body (Sadek et al., Reference Sadek, Saleh and Ayoub2016).
Research on the utilization of the potential of APP as biomarkers for diagnosis of subclinical mastitis have so far involved only small study populations with one or two APPs being studied at a time. Another ambiguity has been on whether the variations in the levels of acute phase proteins during mastitis were confined to the milk or serum alone, or for both (Fathi and Farahzadi, Reference Fathi and Farahzadi2011, Tothova et al., Reference Tothova, Nagy and Kovac2014). This study sought to provide comprehensive data on potential use of Hp, SAA, AGP, CRP and Lf both in serum and milk as biomarkers for diagnosis of subclinical mastitis in a larger sample size of dairy cattle. The specificity of these markers with regard to the etiological agent was also investigated to further refine their diagnostic value.
Materials and methods
Animals
The study was carried out in two medium-sized organized dairy farms of Thrissur district, Kerala State, S. India, in 110 lactating cows. The animals were cross-breds of Holstein-Friesian and Jersey breeds with the local non-descript cattle. They were of different parity and in various stages of lactation. Routine clinical examination was done on all animals before collecting samples and all udder quarters were checked to assess their apparent health status.
Diagnosis of subclinical mastitis
California mastitis test (CMT) and measurement of electrical conductivity (EC) of milk samples from individual quarters were done by direct stripping of milk into the designated wells of the CMT paddle, and the cups of a DraminskyR multi quarter EC meter. Somatic cell count (SCC) was done by direct, manual microscopic cell counting using modified Newman's staining to reduce the reported errors due to clots and flakes encountered in automated counting (Eckersall et al., Reference Eckersall, Young, McComb, Hogarth, Safi, Weber, McDonald, Nolan and Fitzpatrick2001).
Grouping of animals
Milk was collected individually from all lactating quarters of the 110 animals and subjected to CMT, SCC and EC tests. Forty cows with CMT score below two, EC score above 300 and SCC below 200 000 cells/ml in milk from each of all the milking quarters were grouped as healthy cows (Group 1) and 40 cows with CMT score above two, EC score below 300 and SCC above 200 000 cells/ml in milk from each of all the secreting quarters were grouped as sub-clinically infected (Group 2) (Sharma et al., Reference Sharma, Pandey and Sudhan2010). Determination of APP levels in milk was done using milk from the quarter with the lowest SCC score for Group 1 animals and the highest SCC score for Group 2 animals. Milk and blood samples were collected from all the 80 selected animals. Serum was separated and stored at −20°C along with milk samples until further use.
Culture, isolation and identification of pathogens
Direct isolation of bacteria was done from milk by streaking on brain heart infusion (BHI) agar plates followed by incubation at 37°C for 24 h. Plates were examined after 24 h for growth. Isolated colonies were selected and representative samples were inoculated into BHI broth and incubated at 37° C for 24 h. The isolates were identified based on polymerase chain reaction (PCR) for the etiological agents for bovine subclinical mastitis as below. Details of primers used are given in Online Supplementary Table S5.
Pathogen specific variation of APP
Group 2 was subdivided into three subgroups, namely A with only Staphylococcal infection, B with only Streptococcal infection and C with mixed infection of both Staphylococcus and Streptococcus spp.
Measurement of acute-phase protein levels and albumin
The levels of APP, Hp and CRP were measured in both milk and serum samples. Serum amyloid A was measured in serum samples alone and AGP and Lf levels were determined from milk samples only. Solid-phase ELISA was used for the quantitative analysis of the APP levels using commercially available ELISA kits specific for each APP, developed by Life Diagnostics, USA, according to the manufacturer's instructions. A standard curve was constructed for each ELISA and concentrations of that particular APP in the samples tested were determined from the curve using quadratic equations. The values obtained were multiplied with the dilution factor to get the concentration in ng/ ml of sample. The levels of albumin in serum samples were estimated using Liquick Cor-ALBUMIN kits (Cormay Group, India). The test was based on spectrophotometry for measuring the color intensity of the complex formed between serum albumin and bromocresol green in the presence of succinate buffer (Doumas et al., Reference Doumas, Watson and Biggs1971). Milk samples were centrifuged at 3000 g for 15 min at 4° C to separate skim milk. Skim milk was centrifuged with 0.1 N HCl to separate milk serum (whey). Then the albumin level was estimated like that of the blood serum.
Statistical analysis
The concentrations of the APP obtained were not in normal distribution, so they were transformed into natural logarithm prior to statistical analysis. Independent sample t test was used for comparing the means of the healthy and infected groups. The pathogen specific variation of the APP was analyzed by one way ANOVA.
Results
Culture and isolation-culture of 146 quarter milk samples (QMS) from 40 cows in Group 2 produced bacterial growth in 132 from which 261 pure isolates were obtained (Table 1). Eighty nine isolates (34%) showed positive result for catalase indicating Staphylococcus organisms and remaining 172 (66%) were Streptococci. Genus and species specific PCR of 89 catalase positive samples recognized Staphylococci spp. of which 37 (14%) were S. aureus and 52 (20%) were other Staphylococci spp. Species specific PCR for 172 catalase-negative isolates revealed 98 (38%) S. agalactiae, 48 (18%) S. dysagalactiae and 26 (10%) S. uberis. Fifty-six samples yielded only a single genus of bacteria of which 12 were Staphylococcus and 44 Streptococcus. Species level characterization revealed single infections with S. aureus, S. agalactiae, S. dysagalactiae and S. uberis in seven, 17, eight and two isolates respectively. In all other samples multiple infections were identified.
Number of different species of bacteria isolated and identified in the milk samples from 40 cows in Group 2
APP concentrations and pathogen specificity
Comparison of Groups 1 (healthy) and 2 (sublinical mastitis) are given in Table 2, and the pathogen-specific APP data for Group 2 are in Table 3. In this latter case no significant differences were noticed between the three sub-groups.
Comparison between the levels of acute-phase proteins: Hp, CRP, SAA, AGP, albumin, and Lf in the serum and milk of 40 healthy cows (Group 1) and 40 cows with subclinical mastitis (Group 2)
Means with different superscripts differ significantly, n.s. is non significant (P > 0.05)
Hp- Haptoglobin, CRP- C-reactive protein, SAA- serum amyloid A, AGP-α-1 Acid glycoprotein, Ln- natural logarithm
Comparison of variation of acute-phase proteins against the bacterial isolates in the milk of 40 cows from Group 2
Group 2A – Only Staphylococcus, group 2B – Only Streptococcus, Group 2C – Both Staphylococcus and Streptococcus
N.S. non significant (P > 0.05)
Hp- Haptoglobin, CRP- C-reactive protein, SAA- serum amyloid A, AGP-α-1 Acid glycoprotein, Ln- natural logarithm
The mean of Hp levels in the milk of infected cows (Group 2) was significantly higher (P < 0.05) than that of the healthy cows (Group 1) whereas significant differences were not observed between the mean values of Hp in serum. Mean values of AGP and SAA were significantly higher (P < 0.01 and P < 0.001 respectively) in the subclinical mastitis group, as was Lf (P < 0.001) Significant differences were not detected in the mean values of either CRP or albumin in milk or serum.
Discussion
The large sample size of the research, and inclusion of all the major acute phase proteins have provided comprehensive information on the potential use of lactoferrin and acute phase proteins as potential bio-markers for diagnosis of sub-clinical mastitis in dairy cattle. Levels of lactoferrin and three acute phase proteins, namely Hp, AGP and SAA were significantly elevated in dairy cows affected with the disease, while the CRP and albumin levels were not. Interestingly, these changes were found to be independent of the specific etiology for the disease.
Etiological agents of SCM like Staphylococcus spp. and Streptococcus spp. cause formation of micro abscesses, pin point necrotic changes and fibrosis which can lead to tissue injury leading to increase in Hp levels in milk. Elevation of Hp in milk subsequent to mammary tissue injury is caused by local expression of Hp mRNA rather than influx from the serum, making it a specific marker for inflammatory changes in mammary tissue. This is borne out by the findings of this study where only the levels of milk Hp, and not the serum Hp, were found to be significantly enhanced. Our results also points to the non-specificity of serum Hp levels as a true indicator for SCM as Hp values among the healthy cows were actually relatively high, possibly due to co-morbidities. Kovac et al. (Reference Kovac, Popelkova, Tkaaikova and Ihnat2007) report Hp values in the serum of healthy cows to be below detectable levels, but such high levels of animal health are yet to be achieved by the Indian dairy industry. Safi et al. (Reference Safi, Khoshvaghti, Jafarzadeh, Bolourchi and Nowrouzian2009) find an insignificant increase in the serum Hp concentration in case of SCM, whereas Eckersall et al. (Reference Eckersall, Young, Nolan, Knight, McComb, Waterston, Hogarth, Scott and Fitzpatrick2006) describe the increase in serum Hp of cows with SCM to be significant. Only the β subunit of haptoglobulin is present in milk during subclinical matitis while the serum contains all three subunits (Upadhyaya et al., Reference Upadhyaya, Thanislass, Veerapandyan, Badami and Antony2016). Hence, the specificty of the commercial ELISA kits for Hp could be increased if they were to target only the β subunits, instead of targeting both beta and alpha chains as done now.
SAA is a group of three proteins produced by the liver as a response to acute inflammatory changes after induction by pro-inflammatory cytokines like IL-1, IL-6 and TNF alpha. Significant increase in SAA levels among affected cows in our study supplement the findings of Eckersall et al. (Reference Eckersall, Young, McComb, Hogarth, Safi, Weber, McDonald, Nolan and Fitzpatrick2001) and Razak et al. (Reference Razak, Hussain, Dar and Mir2015), but Gerardo et al. (Reference Gerardo, Bernardini, Elia, Ferrari, Iob and Segato2009) report the SAA levels to be of little use for discriminating between subclinical and clinical mastitis. It could be that the larger size and homogeneity of our samples nullified the influence of external factors like hidden inflammatory diseases suspected in their study. We obtained a low correlation between SCC and SAA values (data not shown), which contradicts the theory of ingress of SAA into milk during mastitis (O'Mahony et al., Reference O'Mahony, Healy, Harte, Walshe, Torgerson and Doherty2006).
AGP is produced mainly in liver and secreted. Extra hepatic production of AGP does occur (Ceciliani et al., Reference Ceciliani, Pocacqua, Lecchi, Fortin, Rebucci, Avallone, Bronzo, Cheli and Sartorelli2007) and there is evidence of influx of serum AGP into organs for maintenance of homeostasis by reducing the tissue damage associated with the inflammatory process (Murata et al., Reference Murata, Shimada and Yoshioka2004). The significant elevation of AGP detected among Group 2 animals in our study points to the potential of AGP being a sensitive indicator of chronic inflammatory changes associated with SCM in cattle.
C-reactive protein (CRP) is a pentamer, which has the capacity to bind with non-self antigens, thus playing an important role in the recognition and protection against infections and clearance of damaged tissue (Mold et al., Reference Mold, Rodic-Polic and Du Clos2002). Only a non-significant numerical increase in CRP levels in Group 2 animals was detected in the present study. According to Panicker et al. (Reference Panicker, Gopalakrishnan and George2014) the CRP levels increase only by very moderate levels during infections in cows, although Thomas (Reference Thomas2015) consider CRP in milk to be a potential marker for mastitis detection while their second paper (Thomas, Reference Thomas2015) reports that M-SAA3 and CRP do not correlate with Hp or SCC or between themselves. Significant increase of CRP in blood in human mastitis (Fetherston et al., Reference Fetherston, Wells and Hartmann2006) and canine clinical mastitis (Vasiu et al., Reference Vasiu, Dabrowski, Martinez-Subiela, Ceron, Wdowiak, Pop, Brudasca, Pastor and Tvarijonaviciute2017) have also been reported. All these studies were conducted with sample sizes that were confined to a small location and were much smaller than ours, and could have been influenced by inflammatory responses against co-morbidities in one or a few animals in the study population. Another interpretation could be that the synthesis of different acute phase proteins is not uniform, and cows in different stages of inflammatory response may be having different levels of specific proteins. Zalewska et al. (Reference Zalewska, Kawecka-Grochocka, Słoniewska, Marczak, Jarmuz, Zweirzchowski and Bagnicka2020) did not find any significant differences between the SAA and Hp levels in the cistern lining epithelium of cows affected with persistent infection of coagulase negative staphylococci and healthy cows, leading them to conclude that during chronic infections, the APP production is downgraded to some extent to ensure only minimal damage to the host cells.
Even though numerous APP have been identified, the rise in their serum levels is neither common nor uniform during inflammatory processes due to different diseases. The seat of infection during subclinical mastitis is limited to the mammary gland alone, favouring increase in production of APP that also have extra hepatic sources, like the mammary gland for AGP, Hp and Lf. The levels of only Hp and AGP in milk and SAA in serum were found to increase significantly during SCM while milk and serum concentrations of CRP and albumin remained insignificant. This is in agreement with the results of a study by Thomas et al. (Reference Thomas, Waterston, Hastie, Parking, Haining and Eckersall2015) when the samples were collected without prejudice to the clinical condition of the cows. Thomas et al. (Reference Thomas, Geraghty, Simões, Mshelbwala, Haining and Eckersall2018), using an experimental infection had found the profile of the three APP (CRP, Hp and SAA-3), to closely mirror each other during clinical mastitis. Dissimilarities in the onset and magnitude of fold increase of each APP in infected milk were observed in that study, indicating a possible influence of the stage of infection on the APP profile. Interestingly, the variations in levels were found to be insufficient to differentiate between clinical and subclinical mastitis. This points to the necessity of identifying the specific APP as biomarker for each disease conditions.
Different bacteria elicit their pathogenicity using different virulence factors that induce variable cytokines, thereby eliciting diverse inflammatory signs and damage to udder tissue in mastitis cases (Pyörälä and Syväjärvi, Reference Pyörälä and Syväjärvi1987; Schukken et al., Reference Schukken, Wilson, Welcome, Garrison-Tinofsky and Gonzales2003; Thomas et al., Reference Thomas, Geraghty, Simões, Mshelbwala, Haining and Eckersall2018; Dalanezi et al., Reference Dalanezi, Schmidt, Joaquim, Guimaraes, Guerra, Lopes, Cerri, Chadwick and Langoni2020). C-reactive protein concentration was found to vary significantly with the causative pathogen of mastitis indicating variable levels of stimulation of its secretion by the virulence factors of different pathogens (Thomas et al., Reference Thomas, Geraghty, Simões, Mshelbwala, Haining and Eckersall2018). The micro-abscess formation due to coagulase activity of staphylococci assisting their invasion and persistence in the alveoli, and high anti-oxidant activity by streptococci are the major pathogenic mechanisms of SCM. In our study, the majority (32/40) of animals had mixed infection with both staphylococci and streptococci whereas only eight had individual infections. There was increased concentration of APP in milk and serum in all these three subgroups but there was no significant difference between the subgroups. Pyorala et al. (Reference Pyorala, Hovinen, Simojoki, Fitzpatrick, Eckersall and Orro2011) found the concentrations of APP to be low in mastitis caused by CNS and streptococci while a rise was noticed for S.aureus group. This is not borne out by our findings and the probable reason could be that the secretion of acute phase proteins in all persistent infections may be downgraded in the body so as to protect permanent damage to the host cells (Zalewska et al., Reference Zalewska, Kawecka-Grochocka, Słoniewska, Marczak, Jarmuz, Zweirzchowski and Bagnicka2020). Another reason could be the similarity in the confinement of pathogenesis to udder tissue by both of these pathogens.
Lactoferrin is an iron binding protein which functions as an innate defense factor in mastitis by preventing infection of mammary glands. The high levels of lactoferrin present in milk during the immediate post-calving period decreases to base levels as lactation progresses (Shimazaki and Kawai, Reference Shimazaki and Kawai2017). Excess secretion of lactoferrin from the mammary epithelial cells during acute clinical mastitis has been reported (Huang et al., Reference Huang, Morimoto, Hosoda, Yoshimura and Isobe2012). The concentrations of Lf in milk quickly increase in cows with subclinical and clinical mastitis (Galfi et al., Reference Galfi, Radinović, Boboš, Pajić, Savić and Milanov2016). The levels of lactoferrin in milk from Group 2 animals in our study were significantly higher, and we were able to establish a positive correlation between lactoferrin levels and Hp, SAA, as well as AGP levels (data not shown), indicating potential for lactoferrin to be used as a biomarker for detection of subclinical mastitis at very early stages of subclinical mastitis.
Batavani et al. (Reference Batavani, Asri and Naebzadeh2007) have reported increased milk albumin and Singh et al. (Reference Singh, Bhardwaj, Azad and Beigh2014) reported decreased serum albumin in case of SCM in dairy cows, but in the present study neither milk nor serum albumin was affected by health status, which supports Razak et al. (Reference Razak, Hussain, Dar and Mir2015).
We conclude that serum levels of lactoferrin and three acute phase proteins, namely Hp, AGP and SAA are significantly increased during subclinical mastitis in dairy cattle, which can potentially provide a useful tool for early diagnosis of the disease.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S002202992100056X.
Acknowledgements
The authors would like to acknowledge with gratitude the facilities and financial assistance provided by the Kerala Veterinary and Animal Sciences University for carrying out the research.
