Bacterial growth during the early phase of infection determines the severity of experimental Escherichia coli mastitis in dairy cows

The aim of this study was to investigate the importance of bacterial growth for the severity of experimental Escherichia coli mastitis, indirectly expressed as the area under the curve of bacterial counts in milk over time. The association of pre-infusion somatic cell count and post-infusion influx of inflammatory cells in milk with severity of infection was also examined. Bacterial growth was studied through culture in milk samples (in vitro) and through monitoring of bacterial counts in milk during the early phase of infection (in vivo) in 36 cows. Individual variation in bacterial counts was more than 2 x 10(2)-fold after 6 h of in vitro incubation, and more than 8 x 10(2)-fold 6 h after intramammary infusion. In vitro growth in milk was not associated with in vivo growth during the early phase of infection, nor with severity of E. coli mastitis. Somatic cell count before experimental E. coli mastitis was negatively associated with in vivo bacterial growth during the early phase of infection (R2 = 0.28), but was not associated with severity of E. coli mastitis (R2 = 0.06). In vivo bacterial growth during the early phase of infection (positive association; R2 = 0.41), together with influx of inflammatory cells in milk, expressed as mean hourly increase of somatic cell count between 6 and 12 h post-infusion (negative association; R2 = 0.11), are major determinants for the severity of experimental E. coli mastitis (R2 = 0.56).     

Apoptosis of bovine neutrophils during mastitis experimentally induced with Escherichia coli or endotoxin

OBJECTIVE: To determine whether apoptosis of neutrophils was accelerated during mastits experimentally induced by use of Escherichia coli or E coli endotoxin and whether differences were apparent in the response to E coli or endotoxin. ANIMALS: 11 healthy lactating Holstein cows. PROCEDURE: Blood samples were collected from cows at various intervals after intramammary inoculation with E coli or endotoxin. Percentage of apoptotic neutrophils detected after in vitro incubation for 3 hours was determined. Fluorescein isothiocyanate-labeled annexin-V in combination with propidium iodide was used to distinguish apoptosis and necrosis of neutrophils. Total and differential circulating leukocyte counts and rectal temperature were determined at the time of collection of blood samples. Milk yield and milk somatic cell counts were determined at the time of milking. RESULTS: Inoculation of endotoxin did not accelerate in vitro induction of neutrophil apoptosis. However, inoculation of E coli increased the percentage of apoptotic neutrophils. At 18 hours after inoculation, 20% of the neutrophils were apoptotic, compared with 5% before inoculation. Milk somatic cell count and rectal temperature increased, milk production and total leukocyte count decreased, and percentage of immature neutrophils increased after inoculation with E coli or endotoxin. However, kinetics of the responses were more rapid, more severe, and of shorter duration during endotoxin-induced mastitis. CONCLUSIONS AND CLINICAL RELEVANCE: In vitro induction of apoptosis of neutrophils was accelerated only during E coli-induced mastitis and not during endotoxin-induced mastitis. Endotoxin inoculation as a model for studying coliform mastitis in dairy cows should be viewed with caution.     

A clinical trial to evaluate the effectiveness of antibiotic treatment of lactating cows with high somatic cell counts in their milk

OBJECTIVE: To determine the effectiveness of treatment of lactating cows with high somatic cell counts in milk. DESIGN: Randomised clinical trial. METHODS: Single pooled quarter samples of milk were obtained from cows with somatic cell counts above 500,000 cells/mL on fifty farms. Milk samples were cultured for known mastitis bacterial pathogens. Cows were randomly allocated to treated and untreated groups. Treated cows received both intramammary cloxacillin and parenteral erythromycin. Single pooled quarter milk samples were obtained at 6 weeks after treatment and were cultured for the presence of pathogenic bacteria. The percentage of samples with no growth at the post-treatment culture was used as an estimate of the bacteriological cures for each pathogen type and for each treatment group. Somatic cell counts of cows were compared between treatment groups and within pathogen group. The number of cows that completed a full lactation were compared between each treatment group and within each pathogen group. RESULTS: Treatment had no effect upon bacteriological cures, irrespective of pathogen present or the presence of bacteria during the previous lactation. There was no effect of treatment upon somatic cell count except for cows infected with Streptococcus dysgalactiae in which treatment caused a significant lowering of cell counts. This effect was not present in the subsequent lactation. Treatment of chronically infected cows did not alter the probability of a cow completing a full lactation but did improve the probability of newly infected cows being retained for the next lactation. Twenty-eight of 214 treated cows developed clinical mastitis in more than one quarter after treatment, thus indicating a poor technique by farmers for the insertion of intramammary antibiotics. CONCLUSIONS: Treatment during lactation of cows with high somatic cell counts in milk is ineffective in reducing bacterial infections and in reducing somatic cell counts to acceptable numbers.     

Kinetics of local and systemic isoforms of serum amyloid A in bovine mastitic milk

The aim of the present study was to characterise the serum amyloid A (SAA) response to intramammary inoculation of Escherichia coli and to examine the distribution of hepatically and extrahepatically produced SAA isoforms in plasma and milk from cows with mastitis. Milk and plasma SAA concentrations were determined before and after experimental induction of E. coli mastitis in six dairy cows. The milk SAA response was characterised by low or undetectable levels before inoculation, very rapid and large increases in concentration after inoculation, and rapid decline towards baseline levels after resolution of disease. In plasma from cows with experimentally induced E. coli mastitis, four hepatically derived SAA isoforms with apparent isoelectric point (pI) values of 5.8, 6.2, 6.8 and 7.4 were demonstrated by denaturing isoelectric focusing. In milk three highly alkaline isoforms with apparent pI values above 9.3 appeared 12 h post-inoculation. These isoforms were not present in any of the plasma samples, and it therefore seems likely that they were locally produced, tissue-specific isoforms. At 24-36 h post-inoculation one or more acidic isoforms corresponding to those found in plasma appeared in the milk samples. The isoforms demonstrated in plasma from cows with E. coli mastitis were also present in serum obtained from three cows with clinical Streptococcus uberis mastitis. In conclusion, experimentally induced E. coli mastitis is accompanied by a prominent SAA response. The results of the present study indicate that SAA accumulation in mastitic milk is the result of both local synthesis of SAA and of hepatically derived SAA gaining access to the milk due to increased permeability of the blood-milk barrier.     

High milk neutrophil chemiluminescence limits the severity of bovine coliform mastitis

Polymorphonuclear neutrophil (PMN) function changes during mastitis. To investigate the contribution of milk PMN to the severity of Escherichia coli (E. coli) mastitis, chemiluminescence (CL) of blood and milk PMN and their efficiency to destroy coliform bacteria in the mammary gland were examined following the induction of E. coli mastitis in early lactating cows. To better assess and define the degree of mastitis severity, cows were classified as moderate and severe responders according to milk production loss in the non-infected quarters at post-infection hour (PIH) 48. There was an inverse relationship between pre-infection milk PMN CL and colony-forming units at PIH 6. In moderate cows, the pre-infection blood and milk PMN CL was approximately 2-fold higher than that of severe cows. The probability of severe response increased with decreasing pre-infection PMN CL. At the beginning of the infection blood and milk PMN CL was consistently higher, and milk PMN CL increased faster after infection in moderate cows. At PIH > 48 milk PMN CL in severe cows exceeded that of moderate cows. The somatic cell count (SCC) in moderate cows increased faster than colony-forming units, whereas in severe cows the results were reversed. The kinetics of CL activity for blood and milk PMN before and during the early phase of infection confirmed an impairment in PMN CL activity for severe responding cows. High pre-infection blood and milk PMN CL and the immediate increase of milk PMN CL and SCC after infection limited bacterial growth thereby facilitating the recovery of E. coli mastitis in moderate cows. Our study strengthens the idea that pre-existing milk PMN (a static part of the udder's immune defense) functions as a "cellular antibiotic" before and during infection, and low milk PMN CL is a risk factor for bovine coliform mastitis.     

Influence of estradiol administration to lactating cows on bacterial growth in milk whey

Bacterial growth (E. coli) in whey was studied by turbidometric technique during estradiol benzoate administration (0.02 mg/kg of body weight/day for 12-19 days) to 13 ovariectomized cows at three stages (early, mid and late) of the 1st- and 3rd-lactations. Whey samples from cows at early stage (60-90 days) of 1st-lactation promote the growth of E. coli during the estradiol treatment. This included a significant increase in the maximum turbidity and a decrease in the generation time. Bacterial growth was inhibited in whey from cows at other stages of lactation during the hormone treatment. The degree of inhibition varied at different stages of lactation. No significant alterations in the ability of whey to support bacterial growth were observed in 3 ovariectomized cows treated with the drug vehicle (arachis oil) alone. Majority of the quarters included in the present study were bacteriologically negative throughout the study.     

Comparison of L-selectin and Mac-1 expression on blood and milk neutrophils during experimental Escherichia coli-induced mastitis in cows

OBJECTIVE: To evaluate L-selectin (CD62L) and Mac-1 (CD11b) expression at the surface of blood and milk neutrophils during the early inflammatory response to Escherichia coli-induced mastitis in cows. ANIMALS: 6 healthy Holstein heifers in early lactation. PROCEDURE: Blood and milk samples were collected before and after intramammary administration of 10(4) CFU's of E coli in the left mammary gland quarters. Bacterial counts and electrolyte concentrations in milk, rectal temperature, differential blood leukocyte counts, milk somatic cell counts, neutrophil viability, and the expression of CD62L and CD11b on blood and milk neutrophils were determined longitudinally. RESULTS: Bacteria grew during the first 6 hours after inoculation with a pronounced leukocytic influx. Coincident with neutrophil influx was an increase in CD62L+ and CD11b+ milk neutrophils, as well as an improved viability of milk neutrophils. The peak of the inflammatory reaction was reached approximately 12 hours after E coli inoculation. From that time forward, changes in CD62L and CD11b expression were opposed to each other, with a decrease in CD62L expression and an increase in CD11b expression on blood and milk neutrophils; the magnitude of the differences in CD62L and CD11b expression between blood and milk neutrophils decreased. Percentages of CD62L+ and CD11b+ milk neutrophils increased to percentages that were similar to blood neutrophils (ie, approx 92%). CONCLUSIONS AND CLINICAL RELEVANCE: The presence of adhesion molecules on a large percentage of milk neutrophils during the acute inflammatory response, together with the changes in receptor density, suggest a major role for CD62L and CD11b in neutrophil function during coliform mastitis.     

Expression of CD3 and CD11b antigens on blood and mammary gland leukocytes and bacterial survival in milk of cows with experimentally induced Staphylococcus aureus mastitis.

OBJECTIVES: To differentiate early (1 to 8 days) from late (9 to 14 days) inflammatory phases and assess relationships between leukocyte phenotype and bacterial recovery in cows with Staphylococcus aureus-induced mastitis. ANIMALS: 10 first-lactation Holstein cows. PROCEDURE: Blood and milk samples were collected from 4 or 6 cows before and after intramammary infusion of sterile broth or S. aureus, respectively. Flow cytometric expression of CD3 and CD11b antigens on blood and milk leukocytes, leukocyte differential counts, bacterial counts in milk, and somatic cell counts were determined longitudinally. RESULTS: Density of CD3 molecules decreased on blood lymphocytes and increased on milk lymphocytes after infusion of bacteria. Density of CD11b molecules on lymphocytes and phagocytes and percentage of CD11b+ lymphocytes in milk increased significantly after infusion; maximum values were achieved during the early inflammatory phase. Density of CD3 and CD11b molecules on milk lymphocytes and macrophages, respectively, 1 day after inoculation were negatively correlated with bacterial recovery on day 1 and days 9 to 14, respectively. Density of CD11b molecules on milk macrophages and the ratios of phagocyte to lymphocyte percentages and polymorphonuclear cell to macrophage percentages in milk differentiated the early from the late inflammatory phase. CONCLUSIONS AND CLINICAL RELEVANCE: Activation of bovine mammary gland macrophages and T cells in response to intramammary infusion of S. aureus was associated with an inability to culture this bacterium from milk. Identification of specific inflammatory phases of S. aureus-induced mastitis in cows may allow for the design of more efficacious treatment and control programs.     

The opsonic activity of bovine milk whey for the phagocytosis and killing by neutrophils of encapsulated and non-encapsulated Escherichia coli

Encapsulated strains of Escherichia coli were found to be more resistant to phagocytosis and killing by bovine neutrophils; requiring in the order of 100 times more serum opsonins than non-encapsulated strains. Mid-lactation pooled whey from cows with no history of mastitis was opsonic for non-encapsulated strains, but had no effect on encapsulated organisms. In contrast, early lactation pooled whey (5-10 days post-partum) was opsonic for all strains of E. coli. It is concluded that since early lactation milk contains sufficient opsonins, severe E. coli mastitis at this stage of lactation is not due to opsonic deficiency.     

Bacteriostatic activity of bovine lactoferrin in mastitic milk

The antibacterial activity of milk against a virulent strain of Escherichia coli was investigated using milk fractions from normal or inflamed glands. Mastitic whey exhibited either bactericidal or bacteriostatic activities, depending on whether bacteria were enumerated by the pour plate technique or by surface plating onto sheep blood agar. The former activity was not due to lactoferrin (Lf), which never exerted bactericidal activity, even when assayed in distilled water. Milk whey ultrafiltrate (UF) (mol. wt. less than 5000 d) was used to assay the ability of normal and mastitic milk to support the antibacterial activities of Lf against a strain of E. coli. The addition of purified Lf to UF from mastitic whey resulted in bacteriostasis, whereas Lf was without effect in UF from normal whey. It was concluded that Lf can actually slow down the growth of Lf-sensitive bacteria during mastitis, provided that plasma exudation takes place.     

奶牛临床型乳房炎Nisin抗菌肽治疗后日产奶量及主要乳成分的变化

选取临床型乳房炎自然发病病例9头,乳房内灌注乳酸链球菌素(Nisin)进行治疗,通过检测日产奶量以及停药后2,7,14,21 d的主要乳成分指标,确定Nisin乳房灌注剂对临床型乳房炎病例泌乳性能的影响情况。结果显示,乳房内灌注Nisin后患病乳区乳腺组织得到一定程度的修复,日产奶量逐渐恢复,停药后8 d的日产奶量与发病前6 d相比,平均降幅2.5 kg。治愈后患病乳区牛奶乳脂肪、乳蛋白、乳糖和非脂乳固体含量都呈现上升趋势,但与同期相比,略低于非用药乳区混合牛奶;其中,乳糖含量增加幅度较快,除停药后2 d的治疗乳区外,乳糖含量均在4.7%以上,提示奶牛乳房炎病例经Nisin乳房灌注治疗后,受损乳腺组织修复较快,乳腺上皮细胞合成乳成分的能力大幅度提高。由此可见,Nisin对奶牛临床型乳房炎具有良好的治疗作用。     

Endotoxin-induced bovine mastitis: arachidonic acid metabolites in milk and plasma and effect of flunixin meglumine

Arachidonic acid metabolites (AAM) were measured in milk and plasma during the course of acute endotoxin-induced mastitis in 12 lactating cows. Mastitis was induced by intramammary challenge exposure with 10 micrograms of Escherichia coli (026:B6) endotoxin. Endotoxin was injected into the teat cistern via the teat canal of a single randomly selected rear quarter of each cow. Concentrations of prostaglandin (PG) F2 alpha and thromboxane (Tx) B2 in fat-free unextracted milk and of 15-keto-13,14-dihydro-PGF2 alpha in plasma were measured by radioimmunoassay. Total production of AAM in milk was determined by measuring quarter milk production. The AAM were compared in 6 cows administered flunixin meglumine (1.1 mg/kg of body weight) and in 6 cows administered saline solution. Concentrations of TxB2 in milk were significantly (P less than 0.001) increased during the early course of acute mastitis in endotoxin-treated quarters of cows not administered flunixin meglumine. Peak concentrations of TxB2 in milk occurred at 8 hours after endotoxin inoculation. Flunixin meglumine treatment produced significant (P less than 0.05) reductions in milk TxB2 and plasma 15-keto-13,14-dihydro-PGF2 alpha concentrations. Concentrations of PGF2 alpha in milk and total PGF2 alpha and TxB2 production per quarter per milking were not significantly influenced by endotoxin challenge or by flunixin meglumine treatment.     

Effects of two anti-inflammatory drugs on physiologic variables and milk production in cows with endotoxin-induced mastitis

OBJECTIVE: To determine the effects of 2 anti-inflammatory drugs in lactating Holstein cows with endotoxin-induced mastitis. ANIMALS: 30 multiparous Holstein cows that had been lactating for 30 to 60 days. PROCEDURE: Bacterial culture of milk samples and physical examinations established that study cows were in good health and free of mastitis. Mastitis was induced in 1 front mammary gland by intramammary administration of purified bacterial endotoxin. Cows were allocated into 1 of 3 treatment groups: untreated endotoxic mastitis (n = 9), endotoxic mastitis plus flunixin meglumine (9), and endotoxic mastitis plus isoflupredone acetate (10). Heart rate, rectal temperature, mammary surface area, and rumen motility were recorded hourly for 14 hours following endotoxin administration. Flunixin meglumine or isoflupredone acetate was administered after mammary swelling and rectal temperature > or = 40 degrees C had developed. Milk production was evaluated from 5 days before to 10 days after induction of mastitis. RESULTS: Neither drug ameliorated loss of milk production or swelling of the affected mammary gland. Both drugs reduced mean heart rate during the 14 hours following endotoxin administration, compared with untreated control cows. Cows treated with flunixin meglumine had increased rumen motility and decreased rectal temperature during the same period, compared with all other cows. CONCLUSIONS AND CLINICAL RELEVANCE: Neither drug enhanced recovery of milk production following endotoxin-induced mastitis. Flunixin meglumine decreased rectal temperature, whereas isoflupredone did not; however, it has not been established that reduction of fever is beneficial to cows with naturally occurring mastitis.     

In vivo effects of a thymosin alpha 1-containing colostral whey product on neutrophils and lymphocytes from lactating cows without and with experimentally induced Staphylococcus aureus mastitis

Two separate experiments evaluated ID-1 (a commercial bovine whey product containing 5200 pg of thymosin alpha 1/ml) as an immunotherapeutic agent in lactating cows. In the first experiment, cows without mastitis were evaluated for blood leukogram, milk production, total and differential milk cell counts, lymphocyte (Lc) blastogenesis, and neutrophil (PMN) functions (random and directed migration under agarose, chemiluminescence, ingestion of bacteria, iodination, cytochrome C reduction, antibody-independent neutrophil-mediated cytotoxicity, and antibody-dependent cell-mediated cytotoxicity) before and after ID-1 therapy. ID-1 treatment resulted in a significant treatment group by time period interaction for the relative proportion of mononuclear cells (MNC) in milk (P less than 0.009) and for PMN random migration (P less than 0.01). Based on these interactions, ID-1 treatment appeared to slightly increase the proportion of small MNC in milk and to increase random migration from pretreatment levels by 73% more than increases observed in controls. No significant effect of ID-1 treatment on milk production, total milk somatic cell counts, Lc blastogenesis, or other PMN functions was observed. In cows with experimental Staphylococcus aureus intramammary infections, ID-1 treatment resulted in a significant decline in blood leukocyte count (P less than 0.001) and blood PMN count (P less than 0.02), and maintained PMN random migration (P less than 0.01) while controls declined and abrogated a depression in the ability of Lc to respond to mitogens (P less than 0.05) that developed in controls as a result of S. aureus mastitis. Injection of ID-1 into cows had no adverse effect on their overall health or level of milk production, but did cause subtle and potentially favorable changes in several in vitro immune parameters. In spite of these subtle changes which might indicate increased resistance to mastitis, cows actually developed a more severe S. aureus intramammary infection based on a 9% increase in log 10 bacterial shedding in milk.     

Mastitis Whey — a Good Medium for Bacteria?

Growth of mastitis pathogenic bacteria was measured in bovine whey samples by a turbidometric microtechnique. Whey from mastitis cows supported growth as compared with whey prepared from normal milk. Blood proteins leak into milk during mastitis. A study was undertaken to analyze which molecules from blood would promote bacterial growth in whey Fractions containing hemoglobin showed a distinct growth-promoting effect. An inadequate iron supply is one of the restricting growth factors for bacteria in milk. By utilizing heme-compounds the pathogens can by-pass the effect of antimicrobial iron-binding present in milk in the form of lactoferrin.     

Efficacy of specific egg yolk immunoglobulin (IgY) to bovine mastitis caused by Staphylococcus aureus

The objective of this study was to estimate the efficacy of specific egg yolk immunoglobulin (IgY) to bovine mastitis caused by Staphylococcus aureus. Eighteen lactating cows with clinical mastitis and 18 lactating cows with experimental mastitis (1 quarter per cow) were randomly assigned to three treatments: IgY (20mg/ml) infusion, penicillin (100mg/ml) infusion and no infusion. Treatments for clinical mastitis and experimental mastitis were performed by a 6-day course of intramammary infusion with a dosage of 10ml at an interval of 12h. Milk samples were collected at morning milking time for testing color, clot, somatic cell counts (SCC) and bacterial count. For most of the cows treated with IgY and penicillin, the milk color and clot recovered to normal form during the therapy course. The milk SCCs and bacterial counts of treated cows decreased compared to those of untreated cows (p<0.05). The cure rates by IgY for experimental and clinical mastitis were 83.3% and 50%, respectively, and those by penicillin were 66.7% and 33.3%, respectively. These results showed the potential of specific IgY to be an alternative therapy for mastitis caused by S. aureus.     

Cow milk yield and composition before development of Escherichia coli mastitis

High milk yield, low milk fat and low milk protein were considered as possible predisposing factors to bovine Escherichia coli mastitis. Morning and afternoon milk yields were recorded in 46 Friesian cows later developing E coli mastitis and compared with 92 uninfected controls. Animals developing E coli mastitis gave a significantly higher milk yield than controls. The overall morning: afternoon ratio was (mean +/- se) 1.66 +/- 0.41, with no difference in ratio for the two groups. Further studies on 85 animals later developing E coli mastitis, and 192 controls, in four Friesian herds did not reveal differences in milk fat content (except as related to yield), milk protein or in the interrelationship of days of lactation, milk protein or in the interrelationship of days of lactation, milk fat and milk protein in the two groups. Again there was a correlation between high milk yield and a tendency to develop E coli mastitis but this may have been an age effect in both investigations. No correlation between milk yield and mastitis severity was detected. High yielders which succumbed to E coli mastitis in three herds were producing less milk than mastitis-free controls in the fourth herd which suggests that the correlation is not with yield per se.     

Virulence genes of Escherichia coli strains isolated from mastitic milk

Escherichia coli, a Gram-negative environmental pathogen associated with bovine mastitis was isolated from the milk of 34 symptomatic cows that had been diagnosed with clinical mastitis. Eighty isolates were obtained over a 17-month period and these isolates were screened by DNA amplification for the following E. coli virulence genes: cnf1, cnf2, eaeA, eagg, einv, ltx1, stx1, stx2 and vt2e. Thirty of the bacterial isolates, obtained from 23 different cows, had toxin genes identified in their DNA. The most common virulence gene detected was stx1, with a prevalence of 31%, followed by cnf2 (7.5%), vt2e (6.25%) and eaeA (4%). The possession of different virulence genes by the bacterial isolates had no discernable impact on the health status of the cows as there was no correlation between the potential for toxin production by the E. coli isolates and the systemic clinical condition of the respective infected cows.     

Effect of antibiotic induced bacterial clearance in the udder on L-selectin shedding of blood neutrophils in cows with Escherichia coli mastitis

Bacterial clearance, L-selectin adhesion receptor expression and neutrophil recruitment during experimentally induced Escherichia coli mastitis were investigated. Acute mastitis was induced by injection of 10(4) cfu E. coli into both left quarters of 12 clinically healthy lactating cows, 2-6 weeks after parturition. Half the cows were treated intravenously 10 h after infection, and subcutaneously 30 h after infection with enrofloxacin. In both groups, bacterial count, milk serum albumin, milk somatic cell count, circulating leukocyte count and L-selectin expression on neutrophils were determined. Both groups responded to challenge with udder inflammation and fever. Treatment with enrofloxacin affected the elimination rate of bacteria at hours +14, +18, +24, +48, and +72 after challenge, but not the bacteriological cure rate (five out of six for treated and three out of six for nontreated cows). The drop in L-selectin expression found following challenge did not differ between groups throughout the study. No effect of treatment was found on SCC. Based on these findings, it may be concluded that when treatment is administered 10 h after infection, the number of bacteria in milk is not correlated with L-selectin expression on circulating PMN during experimentally induced E. coli mastitis. The initial bacterial load probably dictates the extent of the decrease in L-selectin expression and milk somatic cells.     

Effect of intramammary injection of RbIL-8 on milk levels of somatic cell count, chemiluminescence activity and shedding patterns of total bacteria and S. aureus in Holstein cows with naturally infected-subclinical mastitis

Summary The effect of intramammary injection of recombinant bovine interleukin-8 (rbIL-8, 1 mg/10 ml of saline) on quarter milk levels of somatic cell count (SCC), chemiluminescence (CL) activity and counts of total bacteria and Staphylococcus aureus (S. aureus) was investigated, using 10 Holstein cows with an early stage or a late stage of subclinical mastitis naturally infected with S. aureus. In the late-stage group, milk SCC and CL activity had significant rises with maximum levels at 6 h, following maintained high levels thereafter post-cytokine injection. The counts in milk total bacteria and S. aureus were insignificantly decreased, being increased back on day 7 post-cytokine injection. Thus, the cytokine was inefficient for the late-stage subclinical mastitis. However, in the early-stage group milk SCC and CL activity declined to under pre-injection levels on day 7 after marked and significant rises at 6 h and day 1 post-cytokine injection. The milk total bacterial count decreased significantly on days 0.25 and 2. Furthermore, the milk S. aureus count was decreased significantly on days 1, 2, 3 and 7 by the cytokine injection. These results suggest that the rbIL-8 has a potential as a therapeutic agent of the subclinical mastitis of dairy cows, if the cytokine is applied at an initial stage of infection.