Intramammary infections (IMI) causing inflammation of the udder (mastitis) can lead to serious economic losses to the dairy industry because of the reduced milk production (Leitner et al., Reference Leitner, Merin and Silanikove2011; Gelasakis et al., Reference Gelasakis, Angelidis, Giannakou, Filioussis, Kalamaki and Arsenos2016), treatment costs and animal welfare issues (Rainard et al., Reference Rainard, Foucras, Fitzgerald, Watts, Koop and Middleton2018), as well as physicochemical changes that negatively impact the quality of milk and dairy products (Forsbäck et al., Reference Forsbäck, Lindmark-Månsson, Andrén and Svennersten-Sjaunja2010). Changes in goat milk composition caused by subclinical IMI have been reported, such as the reduction of lactose (Leitner et al., Reference Leitner, Merin and Silanikove2011; Silanikove et al., Reference Silanikove, Merin and Gabriel2014), and increased serum protein levels (Leitner et al., Reference Leitner, Merin, Silanikove, Ezra, Chaffer, Gollop, Winkler, Glickman and Saran2004b), as well as alterations in somatic cell count (SCC) and total bacteria count (TBC) (Koop et al., Reference Koop, van Werven, Schuiling and Nielen2010).
Staphylococcus other than S. aureus species, also referred as non-aureus staphylococci (NAS) have been reported as leading subclinical IMI agents in the caprine species (Gosselin et al., Reference Gosselin, Lovstad, Dufour, Adkins and Middleton2018), and have been associated with negative changes in goat milk yield, quality, and nutritional value (Silanikove et al., 2014). However, such impact is controversial, since in bovine species, NAS have been reported to cause only moderate and transitory inflammation leading to limited changes in milk (Tomazi et al., Reference Tomazi, Gonçalves, Barreiro, Arcari and dos Santos2015).
The accurate identification of staphylococci by means of conventional biochemical methods has been a bottleneck to evaluate the real impact of those NAS species on milk quality traits in goats. In the last decade, matrix assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS) has been increasingly used for a rapid and accurate species-level diagnostic tool both in human and veterinary medicine, including the identification of mastitis-causing agents (Gosselin et al., Reference Gosselin, Lovstad, Dufour, Adkins and Middleton2018). By means of MALDI-TOF, the aim of this study was to investigate the effects of pathogens associated with subclinical intramammary infections on yield, composition and quality indicators (total bacteria and somatic cells counts) of goat milk.
Material and methods
Study design and samplings
In this longitudinal study, half udder milk samples were collected at early (50 d), intermediate (100 d), and late stages of lactation (150 d). Based on sampling guidelines (NMC, 2004), a total of 132 milk samples were collected from ¾ Alpine (n = 12) and Anglo Nubian (n = 10) multiparous dairy goats in a commercial herd in Paraiba state, Brazil. Goats were milked twice a day manually and milk yield determined weekly. Milk production per animal was calculated considering both morning and afternoon yields.
Somatic cell count (SCC) and physicochemical composition
Milk samples (50 ml) from each udder half were collected in sterilized plastic bottles containing bronopol, and immediately shipped to a national reference lab (PROGENE, UFRPE, Brazil) for milk compositional analysis and somatic cell count by means of Fourier Transform Infrared Spectroscopy (FTIR, Bentley Instruments, USA) and flow cytometry (Somacount 300, Bentley), respectively. Milk samples were also evaluated by titratable acidity and cryoscopy (MK 540 FLEX, Brazil).
Total bacterial count (TBC), microbial isolation and identification
TBC was determined by means of aerobic mesophilic bacteria count using pour plate technique. Microbial isolation was performed by means of conventional microbiological procedures. All isolates were further identified in triplicates by means of Matrix Assisted Laser Desorption Ionization Time of Flight Mass Spectrometry (MALDI-TOF MS) (Bruker Daltonics, Bremen, Germany).
Data analyses
Based on a longitudinal design, the response variables were considered independent among the animals since the repeated measurements were nested within the individual animal. We considered mixed infections whenever two or more different bacterial species were detected in a same udder half. Estimated reductions of daily milk yield were determined for each of the following categories: no infection, S. aureus infection, coagulase negative Staphylococcus infection, and mixed infection.
The putative effects of the intramammary infections on milk production, physicochemical composition, SCC, and TBC were evaluated using two linear mixed models using the PROC MIXED from Statistical Analysis System (SAS) version 9.2 package. Model 1 was applied to determine the effects of the subclinical intramammary infection on milk production (kg/day), SCC, TBC, and physicochemical composition. The values of the dependent variables SCC and TBC were logarithmically transformed (Log). The animal was used as random effect and intramammary infection and lactation period were used as fixed effects, according to the following mathematical model:
$$Y_{{\rm ijl}}\,{\rm}= \,{\rm I}{\rm M}_{\rm i}\,{\rm}+ \,{\rm P}{\rm L}_{\rm j}\,{\rm}+ \,A_{{\rm l\ +\ }}e_{{\rm ijl}}$$where: Y ijl is the dependent variable; μ is a general constant inherent to each observation; IMi is the fixed effect of the intramammary infection; PLj is the fixed effect of lactation period; A l is the random effect of the animal; e ijl is the error associated to each observation.
For the second model, the SCC values were categorized into low SCC (<2.5 × 105 cells/ml), mean SCC (>2.5 × 105 and < 1.3 × 106 cells/ml) and high SCC (>1.3 × 106 cells/ml) counts considering the 25th (2.5 × 105 cells/ml) and 75th (1.3 × 106 cells/ml) percentiles. Those categories were used as a fixed effect to evaluate the effect of SCC on the changes in physicochemical composition of milk, according to the following mathematical model:
$$Y_{{\rm ijkl}}\,{\rm}= \,{\rm SC}{\rm C}_{\rm i}\,{\rm}+ \,{\rm P}{\rm L}_{\rm j}\,{\rm}+ \,A_{{\rm l\ +\ }}e_{{\rm ijl}}$$where: Y ijl is the dependent variable; μ is a general constant inherent to each observation; SCCi is the effect of SCC; PLk is the fixed effect of the lactation period; A l is the random effect of the animal; e ijl is the error associated to each observation.
Differences in the occurrence of IMI between different lactation stages and breed were analyzed using a binomial negative distribution. The Kruskal-Wallis test was also performed to evaluate differences in SCC among the three lactation stages for healthy mammary udder halves. Data were analyzed using a significance level of 5% with Statistical Analysis System (SAS) version 9.2 package.
Results
The reduction in milk yield in animals affected with subclinical mastitis caused by S. aureus (−0.2 Kg/animal/day), NAS (−0.1 kg/animal/day), or both S. aureus and NAS (−0.4 Kg/animal/day) were not statistically different from the milk yield of goats with no subclinical mastitis (1.8 kg/day), as observed in online Supplementary Table S1.
The pathogens cultured from the infected udder halves were identified as S. aureus (50%), non-aureus species (NAS) (46.1%), Micrococcus luteus (3.8%), and Klebsiella pneumoniae (1.9%). In the present study, only staphylococci were included in the statistical models because of the low frequency of infections attributed to other organisms. The most common NAS detected in the present study were S. caprae (9.6%), S. epidermidis (7.7%), and S. simulans (5.8%). The IMI effects (A) and the effects of etiological agent (B) on the physicochemical composition, SCC and TBC of milk are shown in Table 1.
Physicochemical composition, and logarithmically-transformed somatic cell counts (SCC) and total bacteria count (TBC) in milk produced by ¾ Alpine (n = 12) and Anglo Nubian (n = 10) goats according to the occurrence of subclinical intramammary infection (A) caused by Staphylococcus aureus (S. aureus) and non-aureus staphylococci (NAS) (B)1
1 Linear mixed model = Animal as random effect; period of lactation and intramammary infection as fixed effects.
2 Standard error of the mean (sem).
Means with different letters in the rows differ statistically by the Tukey-Kramer test (P < 0.05).
In terms of quality indicators, SCC and TBC were higher (P < 0.05) in subclinical mastitis samples (A). Considering the effects of pathogen-specific agents on the quality traits of milk (B), SCC and TBC were higher in IMI associated with S. aureus compared to NAS (P < 0.05; Table 2). In terms of composition, IMI decreased both fat contents and total solids (P < 0.05). No other significant changes in physicochemical composition were attributed to IMI (A) or specific pathogens (B).
Physicochemical composition, and logarithmically-transformed total bacteria count (TBC) in milk produced by ¾ Alpine (n = 12) and Anglo Nubian (n = 10) goats according to somatic cell count levels (low, medium, and high)1
1 Linear mixed model = Animal as random effect; lactation period and SCC categories as fixed effects.
2 Low: < 2.5 × 105; Medium: >2.5 × 105 and <1.3 × 106; High: > 1.3 × 10 6 cells/ml.
3 Standard error of the mean (sem).
Means with different letters in the rows differ statistically by the Tukey−Kramer test (P < 0.05).
In our linear mixed model, it was observed that the increase of SCC in milk was associated with increased protein (P < 0.001), fat (P < 0.05) and total solids (P < 0.05), as well as with a decreased TBC (P < 0.001; Table 2). IMI occurrence was lower in the late lactation stage when compared to early and intermediate stages, whereas there was no difference in IMI between breeds (Supplementary Table S2). In non-infected udder halves, SCC counts were lower (P < 0.05) in the early lactation period as compared with the other stages. Mean values were 1.1 × 106, 2.2 × 106, and 2.7 × 106 cells/ml at 50, 100 and 150 d, respectively.
Discussion
Although CoNS have been reported as leading agents causing subclinical IMI in goats (Gosselin et al., Reference Gosselin, Lovstad, Dufour, Adkins and Middleton2018), S. aureus was the most frequent pathogen found in our study, corroborating a previous report (Merz et al., Reference Merz, Stephan and Johler2016). It must be noted that S. aureus is highly contagious and associated with persistent and hard-to-treat infections in both cows and small ruminants (Rainard et al., Reference Rainard, Foucras, Fitzgerald, Watts, Koop and Middleton2018). Some strains can also produce thermostable enterotoxins capable of causing food poisoning in humans. Therefore, goat milk contamination by S. aureus is a matter of public health concern.
Our findings suggest no differences in terms of total milk yield between goats with and without IMI, corroborating previous studies that reported no effects of subclinical IMI on the total goat milk yield (Koop et al., Reference Koop, van Werven, Schuiling and Nielen2010). On the other hand, some other studies reported significant lower milk yield in infected goats (Leitner et al., Reference Leitner, Merin and Silanikove2011; Koop et al., Reference Koop, De Vliegher, De Visscher, Supré, Haesebrouck, Nielen and van Werven2012).
The increased SCC values observed in milk samples from udder halves affected by subclinical IMI corroborated previous studies (Koop et al., Reference Koop, van Werven, Schuiling and Nielen2010). The inflammatory process triggers leukocyte migration from the blood to the mammary gland (Leitner et al., Reference Leitner, Merin and Silanikove2011) that causes increased SCC. In dairy cows, there is an inverse correlation between somatic cell counts and milk quality, which makes SCC a valuable and widely used indicator of subclinical mastitis. On the other hand, the relevance of SCC in dairy goats is very controversial (Merz et al., Reference Merz, Stephan and Johler2016), as the correlation between increased SCC and subclinical IMI is not as clear as it is in the bovine species. In this study, we were able to correlate increased SCC levels with subclinical IMI using a SCC threshold of 1.3 × 106 cells/ml, which is similar the threshold used in some countries for goat milk regulatory purposes.
Higher SCC and TBC were expected in the samples positive for S. aureus compared to NAS, since S. aureus is known as a major mastitis pathogen, frequently reported to harbor several pathogenicity factors and greater potential to damage the mammary glandular tissue leading, therefore, to increased SCC. Despite predominance of NAS in IMI in goats, the influence of those microorganisms on SCC is not as evident as compared to S. aureus (Koop et al., Reference Koop, De Vliegher, De Visscher, Supré, Haesebrouck, Nielen and van Werven2012). Although there were no differences in protein and lactose contents between udder halves with and without IMI, the linear mixed model indicated that the increase in SCC was associated with the increase of proteins, fat and total solids contents, as well as with a decrease in lactose contents (Table 2). Changes in milk composition are expected during mastitis as there might be increased permeability of the blood-milk barrier conferred by leakiness of the tight junctions of the alveolar epithelial cells (Stelwagen and Singh, Reference Stelwagen and Singh2014). Higher protein contents (Åkerstedt et al., Reference Åkerstedt, Wredle, Lam and Johansson2012), and reduced lactose in milk (Silanikove et al., 2014) have been previously associated with increased SCC. In our study, we observed an increase in protein and a decrease in lactose along the lactation stages in non-infected animals, corroborating previous reports (Leitner et al., Reference Leitner, Merin, Silanikove, Ezra, Chaffer, Gollop, Winkler, Glickman and Saran2004b, Reference Leitner, Merin and Silanikove2011).
The increase in SCC along the lactation was expected and might be attributed to the desquamation of alveolar epithelium that is more marked at the later stages of lactation, as reported by Leitner et al. (Reference Leitner, Merin, Krifucks, Blum, Rivas and Silanikove2012) in Saanen, Alpine and crossbred goats without IMI. The reduction of fat and total solids observed in milk from animals with IMI could be attributed to the inflammation process. Our results corroborate Sánchez et al. (Reference Sánchez, Fernández, Contreras, Luengo and Rubert2002), who reported a lower fat concentration in milk of Murcian goats with S. caprae intramammary infection. On the contrary, previous studies found no effect of subclinical IMI on fat contents in goat milk (Leitner et al., Reference Leitner, Merin and Silanikove2004a). Such differences are probably associated with the different capacity of pathogens to trigger inflammatory reactions in the caprine udder and to the severity of the inflammation at a given lactation point, which reinforces the importance of knowing the effects of specific pathogens in longitudinal study designs.
In conclusion, whilst staphylococcal subclinical intramammary infections are associated with a significant reductions in fat and total solids in goat milk, Staphylococcus aureus species lead to higher somatic cell and total bacteria counts compared to other species. These indicators cannot be overlooked in goat milk quality assessment considering the high importance of S. aureus species as a mastitis-causing pathogen and its role in public health.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022029921000418
Acknowledgements
This study was financed by Coordenação de Aperfeiçoamento de Pessoal de Nível Superior – Brasil (CAPES) – Finance code 001 and Conselho de Desenvolvimento Científico e Tecnológico (CNPq; Brasília-DF, Brazil).
