Effect of teat skin disinfection on the rate of infection and interval to infection in cows exposed to high levels of Staphylococcus aureus

Richard F. Sheldrake and Roderic J. T. Hoare

Journal of Dairy Research (1981), 48, 1–5

All teats of a herd of lactating cows were contaminated after each milking with Staphylococcus aureus. Two teats of each cow were subsequently dipped in an iodine solution, while the remaining 2 teats were not dipped. For cows with a single infection the interval to infection in dipped and undipped quarters was 64 and 60 d respectively, while for the first infected quarter of cows with multiple infections the interval to infection in dipped and undipped quarters was 50 and 39 d respectively. For the second quarter infected in cows with multiple infections, the period between the first and second infection was 21 and 12 d for dipped and undipped quarters respectively. It is suggested that while teat skin disinfection was effective in reducing the rate of new infection, it had little effect on the process of infection in those quarters becoming infected. The occurrence of an infection in one quarter doubled the rate of infection in the other quarters.

The action of rennet on the stability of casein micelles has been described theoretically by combining Michaelis–Menten enzyme kinetics with von Smoluchow-ski's theory for slow coagulation. The casein micelle is considered to be stabilized by stcric interactions arising from a highly hydrated surface protein layer. The stability factor, in von Smoluchowski's theory, is then considered as a variable and determined by the concentration of unhydrolysed κ-casein.

The theoretical approach describes semi-quantitatively experimental observations on clotting time (CT) as a function of temperature, rennet concentration, and total protein concentration. The model also predicts that, under certain circumstances, the CT is inversely proportional to the enzyme concentration. Furthermore, the model agrees with other workers' experimental observations on the effect of κ-cascin concentration, the presence of a lag phase, and that the lag phase is commonly about 60% of the CT.

The mass flow-rate (FR) of milk from the bovine teat within individual pulsation cycles was measured by continuous computation of the average optical density of the milk–air mixture within a teatcup. Measurements during milking were made on both live and excised teats using a range of teatcup liners, vacuum levels and pulsation characteristics. During a typical pulsation cycle, soon after the start of milking, milk FR from live teats reached a peak within 100 ms of the start of the flow period and fell to about 60% of the maximum value between 0·5 and 1·5 s of flow. If the liner was held open for longer than 1·5 s, the FR typically declined a further 10% over the next 10–15 s. After 10–15 s little or no further change occurred in milk FR at normal milking vacuum levels. Milk flow from excised teats reached a steady value within 100 ms of the start of the flow period and showed no subsequent fall in FR.

Within individual pulsation cycles the changes in FR from live teats are affected by: (i) the degree of compressive load applied by the collapsed liner to the teat in the preceding cycle; and (ii) the duration of application of the compressive load. Increasing the degree of compressive load or the length of time of loading increases the peak milk FR of the next cycle by reducing the degree of congestion in the tissues of the teat surrounding the teat canal. Muscular effects are of secondary importance. The relatively small changes in milk FR associated with muscle contractions of the teat can be seen most clearly at low milking vacuum levels (below about 30 kPa) or with an unsupported teat (e.g. using a very distensible, thin-walled liner).

Thirty-nine British Friesian cows and 17 British Friesian heifers were used to measure the effects on milk yield, food intake and digestion, and N balance of increasing the protein concentration in the diet from 111 to 147 g crude protein/kg dry matter during the first 16 weeks of lactation. The comparison was made for animals given food either ad lib. or at a level which allowed them to gain 0·5 kg/d during the 8 weeks before parturition, and for animals given ad lib. diets of either 60:40 or 40:60 hay: concentrate ratio during early lactation. Milk yield was increased by a higher protein concentration in the diet (P < 0·01) and the increase was evident from the first week of lactation. Mean increases over the first 8 weeks of lactation were 3 kg/d for animals on the diet of high roughage content and 7 kg/d for animals on the diet of low roughage content. A higher crude protein content in the diet also increased food intake (P < 0·05) and digestion (P < 0·05), and the increases were greater for animals given the diet of low compared with high roughage content. Level of feeding before parturition did not influence food intake or milk yield and there were no differences between cows and heifers in their response to treatments. N balance over the first 10 weeks of lactation averaged –13 g/d for cows and 6 g/d for heifers (P < 0·025). and was not significantly altered by treatments. Mean N balance of all animals was estimated to increase from –37 g/d immediately after parturition to zero at 8 weeks of lactation. The results demonstrate that an increased protein concentration in the diet during early lactation is associated with increased food intake and digestibility of food, which are in turn largely responsible for higher levels of milk production.

The trypsin-inhibitor capacity of bovine milk was seen to increase in mastitits. This capacity showed good correlation with the California Mastitis Test score, somatic cell count (Coulter Counter method) and bovine serum albumin (BSA) content (radial immunodiffusion). The elevated antitrypsin level proved to be a safer indicator of mastitis than BSA alone. Using gel-filtration chromatography, the major trypsin-inhibitor activity of mastitic milk and plasma was eluted with BSA, indicating a similar mol. wt (70000). This fraction showed electrophoretic polymorphism. A large molecular weight fraction cross-reacted immunologically with human α2-macroglobulin. The antitrypsin activity in mastitic milk was clearly different from the principal activity in colostrum, which had a mol. wt of about 13000.

Subcutaneous injections of 30 mg atropine into lactating cows induced a 20–40% decrease of free amino acid (AA) levels in arterial plasma. Minimum levels were observed after 30–50 min. The decline persisted for more than 6 h. The greatest fall in concentration was noted for tyrosine, methionine, lysine, arginine, phenylalanine and threonine. Arterial glucose levels remained unaffected. The effect of atropine on milk secretion was studied in 2 cows which were milked every hour with the aid of oxytocin. Maximal effects were observed after 3–4 h. They included reduction in concentration of casein and ±-lactalbumin (±-la) and a decline in production of milk (20%), casein (35%), ±-la (45%) and lactose (18%). Uptake by the lactating udder over a period of about 1 h after injection of atropine was studied in 2 cows. Mammary blood flow and glucose uptake remained unaffected. There was a positive correlation between arteriovenous differences of essential AA and arterial plasma concentrations. The uptake of essential AA decreased by approximately 50 %. There was no evidence that atropine has a direct inhibiting effect on the udder. It is suggested that the decrease of ±-la synthesis might induce an inhibition of lactose synthesis and milk production.

In 3 consecutive experiments, each using 20 cows, the application of Escherichia coli to teat ends after milking led to high rates of intramammary infection. These infections were not prevented by disinfection of the teats before milking, by the installation of shields in the short milk tubes of the milking cluster or by the use of an individual quarter milking cluster. Rates of infection were significantly lower when teat contamination was applied 1 h before milking compared to contamination applied immediately after milking.

These data suggest that penetration of the teat duct by the E. coli occurred in the period between contamination and milking. Seventy four percent of infections occurred in hindquarters and there were variations in the susceptibility of cows to infection.

The leucocytes in sheep colostrum, milk and involution secretion were studied using the electron microscope. The predominant cell type in colostrum was the polymorphonuclear leucocyte (PMNL) (41–84%), followed by the macrophage (8–49%), and the lymphocyte (6–11 %). Plasma cells were present in low numbers (1–2%) while no secretory epithelial cells were observed. In mid-lactation the cell components changed so that the macrophage was the predominant cell (83–86%) followed by the lymphocyte (10–17%). In early involution secretion, PMNL reappeared but declined in secretion obtained 21 d after weaning, when the macrophage was again the predominant cell. Associated with the whole cells were membranous extracellular materials and ‘sunburst’ fragments of the secretory epithelial cells. These were phagocytosed by both PMNL and macrophages, which also engulfed fat droplets from the milk. The lymphocytes were examined for plasma membrane markers and the T-cell percentage (E+ rosettes) fluctuated within wide limits, (0–80 %). However, there was no discernible trend associated with stage of lactation. The B-cell markers (C1 and Fcγ) also fluctuated widely and it was evident that contaminating monocytes reduced the accuracy of the count. However, it appeared that more B-cells were present in milk than in blood.

The acid-soluble ribonucleotides of cow's, goat's and sheep's milks were measured by enzymic and ion-exchange chromatographic procedures at different stages of lactation. Colostra and milk from ruminant species contained orotic acid and 13 well-identified nucleotides: AMP, CMP, GMP, UMP, UDP, GDP, UDP-glucose, UDP-galactose, UDP-N-acetyl-glucosamine, UDP-N-acetyl-galactosamine, UDP-glucuronate, GDP-mannose and GDP-fucose. Cow's, goat's and sheep's colostrum contained significant amounts of nucleotides, which increased in amount from the moment of parturition, to reach a maximum 24–48 h later. The nucleotide concentration decreased thereafter with advancing lactation. Cow's milk contained substantial amounts of orotic acid, which increased during lactation, whereas in goat's and sheep's milk no increase took place. The cytidine and adenosine derivatives in ruminants' milk did not change significantly during lactation. The UDP-glucose/UDP-galactose ratio was measured in ruminants' milk at different stages of lactation.

The effect of subclinical mastitis on the citric acid concentration of cow's milk was studied by comparing milk from the mastitic quarter with that from a healthy quarter of the same udder. Forty-two pairs of fore-milk from the morning milking of 17 cows were examined and the citric acid content was measured by an automated carboxylic acid analyzer. The citric acid concentration decreased in subclinical mastitic milk and the extent of decrease was proportional to the degree of mastitis as expressed by the quarter difference value in concentration of Na+Cl in mm/1, i.e. the increase in the Na and Cl concentrations in the abnormal milk over that found in normal milk from the same udder. The correlation coefficient between the quarter difference value of Na+Cl and the value for citric acid was –0·77, (d.f. = 46, P <0·001) and a linear regression equation was obtained. The result showed clearly that subclinical mastitis is one of the causes of variation in the citric acid content of cow's milk. The quantitative change in citric acid concentration in subclinical mastitic milk is discussed with reference to the admixture theory for the formation of abnormal milk.

Samples of milk from silo tanks were stored at refrigeration temperatures and examined at intervals for pyruvate content and total colony count. Significant relationships between pyruvate, μg/ml, and log colony forming units (cfu)/ml were established for samples stored at 2°C for 2, 3 and 4 d (3, 4 and 5 d of age respectively) but at 5 and 7°C the relationships after 2-d storage (3 d of age) were not significant. From an overall relationship it was predicted that 4 μg pyruvate/ml would correspond to a bacterial content of 3·7 x 106 cfu/ml with 95% confidence limits of 6·6 x 10 cfu/ml. 2·1 x 107 cfu/ml.

A rapid method for the direct enumeration of somatic cells in fresh and formalin-preserved raw milk is described. Heat treatment at 80°C for 10 mn fixed somatic cells sufficiently to prevent lysis and subsequent dilution allowed the equivalent of at least 0.5 ml milk to be filtered through a 1.0 μm pore size Nuclepore membrane filter. The somatic cells were concentrated on the membrane and, after staining with acridine orange, fluoresced green, yellow or orange under an epifluorescent microscope. Three different types of cell could be distinguished on the basis of nuclear structure and cell size. These were monocytes, polymorphonuclear leucocytes and the larger epithelial and/or secretory cells. For both fresh and preserved milks the count of somatic cells on the membrane correlated well (r ≽ 0·93) with the Coulter count. Differences between counts obtained by different operators were not significant. The technique is rapid, taking about 20min, and is suitable for milks containing between 2 × 104 and 1 × 107 somatic cells/ml.

The ability of total milk proteose peptone and individual proteose peptone components to inhibit lipolysis in milk was examined. Proteose peptone (1·5 mg/ml) when added to milk inhibited lipolysis, and component 3 was a more effective inhibitor than component 5 or component 8-fast. Inhibition by proteose peptone also occurred when the milk was treated with between 2 and 20% v/v bovine blood serum or with 5 μg/ml heparin. In vitro activity of milk lipoprotein lipase and its stability were not affected by proteose peptone. It was concluded that proteose peptone does not interact with lipase or its activator and that the mechanism of inhibition involves the substrate.

The prevalence of intramammary infection in 4 commercial goat herds was studied in conjunction with electronic somatic cell count and volume analysis, determined using a Coulter Counter and volume analyser.

Neither streptococci nor mycoplasma were isolated from any half and the prevalence of intramammary infection with Staphylococcus aureus ranged from 0 to 3% between herds. For coagulase-negative staphylococci the range for infected halves was 36–71%. There was no significant difference between the mean total microscopic somatic cell count for halves infected with coagulase-negative staphylococci and those free from infection. A similar trend was observed for electronic somatic cell counts although the mean electronic cell count was greater than the mean total microscopic count on the 2 occasions that they were compared. The correlation coefficients between the 2 cell counting methods were 0·86 and 0·94. Between herds there were significant differences in mean electronic somatic cell count, with herd means ranging from 438×103 to 1684×103 cells/ml. In 2 of the 4 herds studied, milk samples from halves infected with coagulase-negative staphylococci had a significantly higher prevalence of cell volume distributions with a modal cell volume between 65 µ3 and 100 µ3. This was attributed to a higher proportion of polymorphonuclear neutrophils.

Use of electronic somatic cell count and cell volume analysis were considered of little value in predicting infection caused by coagulase-negative staphylococci as there was a high proportion of false negative and false positive predictions.

The free glucose concentration in the aqueous phase of samples of goat, sheep, cow, rat and rabbit milk was about 0·1–0·3 mn, while that in human milk was about 2 mM. During starvation the glucose concentration of goat milk fell considerably (by about 80 % in 2 d) in parallel with the decreased rate of lactose production. With rats fed ad lib., glucose concentration in the milk was greater at 12.00 h than at 18.00 h, when lactose synthesis has been shown to decrease. 3-O-Methyl-D-glucose injected into the goat mammary gland via the teat canal specifically entered the blood. These findings support the idea that glucose equilibrates across the apical membrane of mammary secretory cells, so that milk glucose concentrations reflect intracellular glucose concentrations.

Changes in the ethanol (EtOH) stability/pH profile of skim-milk induced by forewarming have been investigated as a function of temperature and duration of forewarming and of milk pH prior to heating. Forewarming increased the EtOH stability of a milk in the mid-pH range, shifting the profile laterally to more acid pH values, with little effect being observed on the maximum and minimum EtOH stabilities. No significant changes in forewarming effects were obtained on adding the sulphydryl-blocking agent, N-ethylmaleimide, indicating the lack of involvement of serum protein interactions in the stabilizing effects of forewarming. It is suggested that the observed effects of forewarming are due to a heat-induced precipitation of colloidal calcium phosphate. This suggestion is supported by measurements of Ca2+ levels in forewarmed and untreated skim-milk as a function of pH and is in agreement with the previously suggested mechanism for the EtOH-induced coagulation of skim-milk. The effects of permitted additives (citrate, phosphate and Ca) on the EtOH stability of forewarmed milk paralleled their effect on untreated milk. A comparison is made of the effects of forewarming on the heat and EtOH stabilities of skim-milk.

A large-scale survey of milks from healthy and mastitic bovine quarters was undertaken to establish the influence of mastitic infection on milk lipase activity and free fatty acid (FFA) level. Mastitic milks tended to have higher FFA levels, but lower lipoprotein lipase activities compared with milk from healthy quarters. These effects became significant at relatively severe levels of infection. The elevated FFA was attributable to higher FFA levels on secretion and to greater lipolysis during storage. Levels of carboxylesterase activity increased with severity of mastitis and showed high positive correlation with mastitis indices.

Marked increases in carboxylesterase, N-acetyl-β-D-glucosaminidase and phospholipase occurred following the induction of mastitis by intramammary infusion of Escherichia, coli endotoxin, in parallel with changes in somatic cell count and other mastitis indices. Relatively little change in lipoprotein lipase activity was observed.