Short communication: Evaluation of MALDI-TOF mass spectrometry and a custom reference spectra expanded database for the identification of bovine-associated coagulase-negative staphylococci.

This study evaluated MALDI-TOF mass spectrometry and a custom reference spectra expanded database for the identification of bovine-associated coagulase-negative staphylococci (CNS). A total of 861 CNS isolates were used in the study, covering 21 different CNS species. The majority of the isolates were previously identified by rpoB gene sequencing (n = 804) and the remainder were identified by sequencing of hsp60 (n = 56) and tuf (n = 1). The genotypic identification was considered the gold standard identification. Using a direct transfer protocol and the existing commercial database, MALDI-TOF mass spectrometry showed a typeability of 96.5% (831/861) and an accuracy of 99.2% (824/831). Using a custom reference spectra expanded database, which included an additional 13 in-house created reference spectra, isolates were identified by MALDI-TOF mass spectrometry with 99.2% (854/861) typeability and 99.4% (849/854) accuracy. Overall, MALDI-TOF mass spectrometry using the direct transfer method was shown to be a highly reliable tool for the identification of bovine-associated CNS.

Identification of bovine-associated coagulase-negative staphylococci by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry using a direct transfer protocol.

This study evaluated matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-ToF MS) for the identification of bovine-associated coagulase-negative staphylococci (CNS), a heterogeneous group of different species. Additionally, we aimed to expand the MALDI-ToF MS database with new reference spectra as required to fill the gaps within the existing commercial spectral library. A total of 258 isolates of CNS were used in the study, covering 16 different CNS species. The majority of the isolates were previously identified by rpoB gene sequencing (n = 219), and the remainder were identified by sequencing of 16S rRNA, hsp60, or both rpoB and hsp60. The genotypic identification was considered the gold standard identification. All MALDI-ToF MS identifications were carried out using the direct transfer method. In a preliminary evaluation (n = 32 isolates; 2 of each species) with the existing commercial database, MALDI-ToF MS showed a typeability of 81% (26/32) and an accuracy of 96% (25/26). In the main evaluation (n = 226 isolates), MALDI-ToF MS with the existing commercial Biotyper (Bruker Daltonics Inc., Billerica, MA) database achieved a typeability of 92.0% (208/226) and an accuracy of 99.5% (207/208). Based on the assessment of the existing commercial database and prior knowledge of the species, a total of 13 custom reference spectra, covering 8 species, were created and added to the commercial database. Using the custom reference spectra expanded database, isolates were identified by MALDI-ToF MS with 100% typeability and 100% accuracy. Whereas the MALDI-ToF MS manufacturer's cutoff for species-level identification is 2.000, the reduction of the species level cutpoint to ≥1.700 improved the species-level identification rates (from 64 to 92% for the existing commercial database) when classifying CNS isolates. Overall, MALDI-ToF MS using the direct transfer method was shown to be a highly reliable tool for the identification of bovine-associated CNS.

Short communication: Molecular epidemiology of Streptococcus agalactiae differs between countries.

Group B Streptococcus or Streptococcus agalactiae continue to be challenging for milk quality programs in countries with emerging dairy industries, such as Colombia, where high prevalence has been reported. Molecular typing of isolates is needed to understand the variability and epidemiology of this pathogen and to develop effective control and eradication programs. We characterized the molecular profile of Strep. agalactiae isolated from cows with subclinical mastitis in 21 Colombian dairy herds and measured diversity within and between herds using multilocus sequence typing. Isolates belonged to sequence type 248 [clonal complex (CC) 103; n = 30), ST1 (CC1; n = 6) or ST22 (CC22; n = 4)], whereas members of CC67/61, the dominant type in North America, were not detected. Presence of multiple clonally unrelated sequence type within a herd was common, which contrasts with the situation in European countries and suggests introduction from multiple sources. Our results demonstrate that conclusions from molecular epidemiological studies in 1 region cannot necessarily be extrapolated to other regions, and no single bovine-adapted CC of Strep. agalactiae exists in Colombia. Improvements in internal and external biosecurity will be needed to reduce Strep. agalactiae prevalence in Colombian dairy herds.

Evaluation of the efficacy of intramuscular versus intramammary treatment of subclinical Streptococcus agalactiae mastitis in dairy cows in Colombia.

A randomized controlled trial was performed in 17 Colombian dairy herds to determine the cure risk among cows subclinically infected with Streptococcus agalactiae exposed to 2 antibiotic therapies. Composite milk samples were collected before milking at the onset of the trial (pretreatment) and 2 subsequent times over a period of approximately 63 d. The intramammary application (IMM) of ampicillin-cloxacillin was compared with the intramuscular application (IM) of penethamate hydriodide, and cure risks after an initial and retreatment application were assessed. Cure risk after the initial treatment was higher (82.4%) for the IMM treatment than for IM therapy (65.8%). However, no difference was observed in the cure risk of refractory cases after retreatment (IMM=52.6% vs. IM=51.2%). The cumulative cure risk (both initial and retreatment) was 90.4 and 82.9% for the IMM and IM products, respectively. A 2-level random effects logistic model that controlled for pretreatment cow-level somatic cell count, indicated that IM treatment (odds ratio=0.37) had a lower cure risk than IMM and a tendency for a lower cure risk with increasing baseline somatic cell count. Our findings suggest that both products and administration routes can reduce the prevalence of S. agalactiae in affected herds, but the IMM product had a better efficacy in curing the infection. In addition to the treatment protocol, the cow somatic cell count should be considered when making management decisions for cows infected with S. agalactiae.

Genetic parameters for paratuberculosis infection and effect of infection on production traits in Israeli Holsteins.

Paratuberculosis (Johne's disease) is an infectious enteric disease in dairy cattle and other species that causes substantial economic loss worldwide. In this study, two recursive Gaussian-threshold models were employed in order to infer the effects of Johne's disease on milk yield, fat yield, and protein yield while simultaneously estimating genetic parameters (i.e. heritability and genetic correlation) in an Israeli Holstein population. Disease diagnosis was based on ELISA serum antibody tests. Data were available for 4694 daughters of 361 sires; 3.5% were positive; and 1.6% were suspect for the disease test. Disease status was coded either as a binary character (negative vs. positive) or as an ordered categorical trait (negative, suspect, and positive) in the two recursive models and as a binary trait in a linear model. Among sires with ≥ 50 daughters, predicted probability of Mycobacterium avium ssp. paratuberculosis (MAP) infection in future daughters ranged from <1% to 16.5%. Heritability estimates for Johne's disease were near 0.15, confirming a genetic contribution to disease susceptibility. Genetic correlation estimates for Johne's disease with the three yield traits were 0.15-0.22. Residual correlations for Johne's disease with the yield traits were between -0.01 and -0.10. For the linear regression model, yield losses associated with a positive disease diagnosis during 305 days of lactation were 300 kg milk and around 10 kg for fat and protein. Yield loss estimates from the recursive models were 25-50% less than linear model estimates. Recursive modeling has theoretical advantages over linear models for these phenotypes, but the estimated genetic parameters in this study did not differ significantly between the two types of models.

Relationship between resistance to complement, virulence and outer membrane protein patterns in pathogenic Escherichia coli O2 isolates.

To establish a possible relationship between resistance to complement, virulence and outer membrane protein banding patterns, ten E. coli O2 strains isolated from chickens with colibacillosis were studied for: (1) resistance to the bactericidal effect of complement by a quantitative microtiter method, (2) virulence, as determined by chicken lethality test, and (3) outer membrane protein banding patterns yielded by SDS-polyacrylamide gel electrophoresis. The ten isolates were classified into three groups: (1) Group 1, consisting of four isolates showed: (a) high resistance to complement, (b) high virulence, and (c) different pattern between 35 and 40 kDa with a weak peptide band at 35 kDa. (2) Group 2, consisting of one isolate showed: (a) high resistance to complement, (b) low virulence, and (c) a weak peptide band at 35 kDa. (3) Group 3, consisting of five isolates showed: (a) low resistance to complement, (b) low virulence, and (c) identical OMP pattern between 35 and 40 kDa exhibiting a strong peptide band at 35 kDa. The results suggest that high resistance to complement may be necessary but no sufficient for virulence and that OMP banding patterns may be a marker for virulence.

Systemic and local immune response of cows to intramammary infection with Staphylococcus aureus.

The association between Staphylococcus aureus chronic mammary gland infection and the resulting immune response expressed by the production of specific IgG and IgA antibodies in blood and milk was studied in Israeli Holstein cows. Specific antibodies of the IgG class were detected in sera of 82.6 per cent of the cows chronically infected by S aureus, while in 17.4 per cent no such antibodies could be detected. Specific IgG antibodies to S aureus were neither detected in sera of cows free of mammary infection nor in those infected with different coagulase-negative staphylococci (CNS) such as S intermedius, S chromogenes or S haemolyticus. In milk, specific IgG antibodies to S aureus were detected only in cows with positive serology. The end point dilutions in the milk were 5 to 30 per cent of that of blood from the same cow. No significant difference in IgG titres was found in the same cow if the quarter was infected with S aureus or not. Specific antibodies to S aureus of the IgA class could not be detected in the sera of any of the cows included in this study. In milk, a specific IgA antibody was detected only in the samples from the S aureus infected quarters in which S aureus was isolated at the time of the experiment. In the same cow, quarters infected by S aureus were found to have a significantly higher IgA titre (P < 0.0001) than that of the non-infected ones.

Udder disease etiology, milk somatic cell counts and NAGase activity in Israeli Assaf sheep throughout lactation.

Bacterial pathogens causing udder infections in Israeli Assaf dairy sheep were identified and changes occurring throughout lactation were monitored to study the correlation between the contaminant and the severity of the infection, as measured by somatic cell count (SCC) and NAGase tests. A total of 159 Israeli Assaf dairy sheep on one farm, in their first (69), second (13) or third and more (77) lactations were included in this study. Udder halves were tested for bacterial condition, SCC and NAGase activity 2-3 weeks post lambing and every 4 weeks after until drying-off. At first sampling, in 60.7% (193/318 quarters) of the halves no bacterial growth (NBG) was detected. Different species of coagulase negative staphylococci (CNS) were the main pathogen group in infected udders. Streptococci were isolated from 14 halves, most of them in the two udder halves. The percent of udder infection in sheep in their third or further lactations was 2.8 greater (P<0.05) than in that of sheep in their first lactation. During the lactation, 90.6% of the halves did not change their classification status, suggesting that most infections occur before lambing and/or during the following first few days. The arithmetic mean of SCC and NAGase of total half udder milk and samplings (during the lactation) were 1144+/-48x10(3)cells/ml and 49.4+/-2.5, respectively. The average SCC in the milk of halves classified as NBG was 321+/-35x10(3)cells/ml and was not significantly changed during the lactation period. In halves infected with CNS, average SCC was 1371+/-150x10(3)cells/ml at the first testing and increased to 2129+/-347x10(3)cells/ml at drying-off. No significant differences were found in SCC and NAGase activity between the different species of the CNS. The mean SCC over the types of bacteria isolated, lactation number and days in lactation was significantly different (P<0.0001). In 4% of the halves, from all samples, SCC was above 5000x10(3)cells/ml although no bacteria were detected in their milk. The higher SCC in the CNS infected halves contrasted with the more moderate SCC found in dairy cows similarly infected, suggesting that the sheep udder has a lower resistance and an augmented immunological response against this group of bacteria. Thus, this should be considered accordingly in schemes for sheep's milk quality payment.

Quantitative analysis of levels of serum immunoglobulin G against botulinum neurotoxin type D and association with protection in natural outbreaks of cattle botulism.

The recent outbreaks of cattle botulism in vaccinated Israeli dairy cattle prompted us to determine vaccine efficacy and reasons for vaccine failure. Analysis of clinical signs, feeding practice, vaccination history, and epidemic curves enabled us to define a study population in two outbreaks, where high doses of Clostridium botulinum neurotoxin type D (BoNT/D) were evenly consumed by the affected animal groups. Attack rates among unvaccinated 6- to 24-month-old heifers were 96% (55/57) and 85% (53/62). The attack rates in vaccinated parity 1, 2, and >or=3 cows were 40.4% (21/52), 14.3% (4/28), and 5.6% (3/54), respectively. Vaccine efficacies for these cow groups were 52.5%, 83.2%, and 93.4%, respectively. In younger, unvaccinated 2- to 6-month-old calves, presumably protected by maternal antibodies, the attack rate was 24% (17/71). These differences correlated with significant differences in levels of specific anti-BoNT/D antibody in serum by an enzyme-linked immunosorbent assay (ELISA). The ELISA performance for predicting protection was analyzed by receiver operating characteristic analysis and was found to be highly significant, with an area under the curve of 0.941 (standard error, 0.034; 95% confidence interval, 0.875 to 1.008; P < 0.000). No animals with serum ELISA unit levels above 0.33 were affected in these exposed groups. At this cutoff level, the specificity of the ELISA was 100%, sensitivity was 67%, and accuracy was 92%. We concluded that botulinum toxoids can confer adequate protection against natural exposure to lethal doses of BoNT/D; however, the vaccination protocols should be optimized. Our in-house ELISA system will enable us to optimize vaccination protocols in the animal population.

Application of PCR for detection of Clostridium botulinum type D in bovine samples.

Diagnosis of botulism in cows is obtained by detecting the neurotoxin and/or Clostridium botulinum in the suspected animal. The standard method for detecting the toxin is the mouse bioassay. However, in recent years, the use of mice has become very costly and inconvenient in some facilities, and public pressure has been increasing to find alternatives to live animal bioassays. In this manuscript, we describe the use of the polymerase chain reaction (PCR) procedures in the diagnosis field cases of bovine type D botulism. Bovine samples from clinical cases diagnosed as C. botulinum type D according by clinical symptoms and bioassay resulted in expected PCR product ( approximately 497 bp) similar to the C. botulinum type D NCTC 8265 strain while the gene product was confirmed by sequence data.

An outbreak of Streptococcus canis mastitis in a dairy herd in Israel.

CASE HISTORY: An increase in the bulk somatic cell count (BSCC) of up to 1,000 x 103 cells/ml occurred in a dairy herd in Israel at the end of 2001 and beginning of 2002. CLINICAL FINDINGS: Bacteriological examination of milk from 69 cows revealed a high prevalence of Streptococcus group G bacteria, identified as S. canis, affecting 38% of cows and 20% of all quarters. Isolates were sensitive to cephalothin and moderately sensitive to penicillin G. Infected cows were separated from the herd, treated with intramammary antibiotics, milked last, and strict hygiene practices were introduced to the milking routine. The pathogen was cleared from the herd and BSCC decreased to 250-350 x 103 cells/ml after 6 months. DIAGNOSIS: Streptococcus canis mastitis. CLINICAL RELEVANCE: Streptococcus canis infection may cause subclinical mastitis and high bulk SCC in dairy herds and be resolved by treatment with intramammary antibiotics and the introduction of strict hygiene practices.

Vaccination of turkey poults against pathogenic Escherichia coli.

A vaccine made of bacterial membranes which form vesicles has been developed in our laboratory and found to be effective in protecting chickens against challenge with pathogenic E. coli. In the present study we monitored maternal anti-5. coli antibody in turkey poults and found that although at hatch the antibody titre was relatively high, it decreased gradually to a low level by 14 days of age and remained low up to 31 days of age. Both intramuscular and subcutaneous vaccination with membrane vesicles (MV) resulted in increased antibody titres. Poults with high antibody titres could survive challenge with the pathogenic bacteria. Serum from vaccinated poults reduced E. coli bacterial growth rate when added to the growth medium at a high concentration, whereas serum from non-vaccinated poults did not. Lymphocytes from vaccinated poults responded with an increased rate of blastogenesis when MV was added to the cultures. Lymphocytes from non-vaccinated poults were not stimulated to blastogenesis. We concluded that MV can be used as a vaccine to render poults resistant to E. coli by evoking antibody production, stimulating the bacterial-lysis activity of complement, stimulating T cells to proliferate and stimulating cytotoxic T cells involved in disease resistance.

Milk leucocyte population patterns in bovine udder infection of different aetiology.

This study compared the different leucocyte populations in milk from udders infected with different mastitic pathogens and in different stages of infection. Milk samples were collected from quarters free of intramammary infection, acutely infected with Escherichia coli or Staphylococcus aureus and chronically infected with S. aureus, coagulase-negative staphylococci (CNS) or Streptococcus dysgalactiae. Udder bacteriological status was confirmed after three consecutive bacteriological examinations from weekly quarter milk samples. At the time of the trial, milk samples were tested for somatic cell count (SCC) and differential cell count by both light microscopy (LM) and flow cytometry. Monoclonal antibody (mAb) CD11a/CD18 was used in order to differentiate between leucocytes and epithelial cells when tested by flow cytometry. Udder quarters free of intramammary infection had a mean SCC lower than 107 x 10(3) cells/ml in which the epithelial cells were the main cell type followed by polymorphonuclear cells (PMNs), while macrophages and lymphocytes had a lower concentration. Only 56% of the cells were labelled with the mAb anti-CD11a/CD18. In either acute E. coli- or S. aureus-infected quarters, SCC were significantly higher (P < 0.0001) than in samples from the time of inoculation, with over 90% of the cells labelled with the mAb anti-CD11a/CD18. The main cell type was neutrophils. In chronically infected cows, differences in SCC and in leucocyte patterns were found between infecting pathogens as well as between quarters harbouring the same pathogen. In all the chronically infected quarters, SCC was significantly higher (P < 0.05) than in uninfected ones. The distribution of the leucocyte patterns in the quarters infected with S. dysgalactiae did not differ from that in quarters with acute infection with both E. coli and S. aureus. In the cows chronically infected with S. aureus or CNS, the proportion of PMN was higher but not significantly different from quarters free of intramammary infection, while epithelial cells were significantly lower (P < 0.05). The T lymphocytes bearing CD4+ or CD8+ were significantly higher in quarters chronically infected with S. aureus than in quarters free of intramammary infection and in quarters acutely infected with either E. coli or S. aureus. In all samples B cells were negligible.

Milk leucocyte populations in heifers free of udder infection.

Improvement of udder health through a process of genetic selection is related to heritability and the role of the specific trait in the probability of an individual cow developing an infection. It was suggested that different patterns of leucocyte population of the healthy gland are a significant factor in mastitis. Thus, in order to analyse the heritability of a trait and its correlation with udder health, the present study examined the leucocyte populations of uninfected mammary glands, their variability among quarters in a particular cow, and the changes that occur during lactation. Each one of the 20 cows examined was tested on average 3.06 times during lactation. The somatic cell count (SCC)/ml ranged from 12,000 to 151,000, the coefficients of determination (R2) were higher than 0.5 for SCC. No significant differences were found in the dependent variables between the sampled times (test) nor any interaction between the slopes calculated for the cows over time. No significant differences were found among quarters within a cow for any of the dependent variables including SCC. The effect of the cow trait was found to be significant for polymorphonuclear (PMN), macrophage (MO), and T-lymphocyte-bearing CD4+. The number of lymphocytes labelled with the anti-B monoclonal antibodies was negligible. In conclusion the patterns of leucocyte populations in milk together with the variance among cows should enable an analysis of the heritability of this trait and its correlation with udder health in a future study.

Coagulase-negative staphylococci and mammary gland infections in cows.

Coagulase-negative staphylococci (CNS) are the most frequently isolated bacteria from bovine mammary gland milk samples. The objective of this study was to determine the type of inflammation evoked by CNS in the mammary gland of cows during their first lactation. Twenty-four Israeli-Holstein heifers in their first lactation were tested for bacteriological status, somatic cell count (SCC) and differential leucocyte count in milk 60-120 days postparturition and every 50-60 days after until drying off. Following the first testing, the 96 quarters of the 24 heifers were classified as follows: 69.8% as no bacterial growth (NBG), 27.1% infected with CNS and 3.1% infected with Staphylococcus aureus. During lactation, 84.5% quarters had no change in their classification, 6.2% were newly infected with other pathogens, 3.1% were classified as self-cured and in 6.2% sporadic bacteria were isolated. Among the CNS, S. intermedius, S. chromogenes and S. haemolyticus were the most frequently isolated. Milk from CNS-infected quarters had significantly higher SCC than milk from NBG quarters. An analysis of the leucocyte pattern in milk from CNS vs. NBG quarters revealed a significant increase in polymorphonuclears and a significant decrease in the percentage of total lymphocytes and lymphocytes bearing CD4+ or CD8+. The high percentage of CNS-infected quarters that remained unchanged in their bacterial status during the first lactation, indicates that those CNS have the ability to elude the immune system and persist in the mammary gland for a long time. The persisting infection, resulting to some extent from an increase of SCC by some CNS strains, suggests that in the near future control steps will have to be taken into consideration, in order to enhance the improvement of milk quality.

Changes in milk composition as affected by subclinical mastitis in sheep.

The mechanism of the effects of glandular-level subclinical mastitis in dairy sheep on milk yield and on its composition as expressed in curd yield was studied. Thirty-six Israeli-Assaf dairy sheep with one udder half infected with identified coagulase-negative staphylococci and the contralateral gland free of bacteria were chosen. The milk yield of the infected halves was significantly lower than that of the uninfected ones (0.36 vs. 0.76 kg/milking). The somatic cell count and N-acetyl-beta-D-glucosaminidase activity were significantly higher in the infected halves than in the uninfected ones. The plasminogen activator and plasmin (PL) activities were significantly higher in the infected glands than in the uninfected ones, whereas plasminogen (PLG) activity and the ratio PLG:PL were significantly lower in the infected glands. Concentrations of Ca2+ did not differ, whereas Ca2+ activity was significantly lower and proteose peptone concentration was 2.4 times as high in the infected glands than in the uninfected ones. Curd yield was significantly lower in the infected glands than in the uninfected ones.

Experimental vaccination of young chickens with a live, non-pathogenic strain of Escherichia coli.

A non-pathogenic, piliated strain of Escherichia coli (BT-7; Frommer et al., 1990), isolated from a meat-type chicken flock, was studied as a candidate for a live vaccine to protect chickens from E. coli infection. Active immunization provided substantial protection of chicks vaccinated at 14 or 21 days of age, resulting in better resistance to challenge than in those vaccinated at 1 or 7 days. Chicks vaccinated at 21 days of age and challenged 1 week later with pathogenic E. coli strains 01-.K1, 02:K1 or 078:K80, exhibited good protection for at least 2 weeks against all strains. Three vaccination routes were found to give the highest resistance to challenge with pathogenic E. coli strain 078:K80. Intramuscular (i.m.) at 7 and 21 days of age, i.m. at 21 days of age and spray at 7, followed by per os at 21 days of age. Vaccination per os once at 7 or twice at 7 and 21 days resulted in good protection. Chicks exhibiting high antibody titres by ELISA were well-protected against challenge.

Phenotypic characteristics of Staphylococcus aureus isolated from bovine mastitis in Israeli dairy herds.

A study of the characterization of the phenotypic patterns of Staphylococcus aureus strains isolated from bovine subclinical mastitis in Israeli dairy herds and their correlation with the severity of the disease was undertaken. A total of 400 chronically S. aureus-infected Israeli-Holstein cows, from 15 dairy herds were included in this study. Based on the results of the biochemical reactions, of the anti-biogram and phage typing, one major type of S. aureus was determined in each herd, its prevalence being between 54 and 100% of the total isolates from that same herd. The majority of the isolates were found to be non-haemolytic (62.7%). The most common phage type was 3/A,3/C,55,71, which was predominant in five herds. In two herds none of the isolates (24) were typable by this set of phages. All isolates were susceptible to methicillin, erythromycin, cephalotin, norfloxacin, trimethoprin-sulphamethoxazole and novobiocin. Most isolates were resistant to penicillin (96.6%) and 52% to oxytetracyclin. Differences in protein patterns between 50 and 36 kDa were found by one-dimensional sodium dodecyl sulphate-polyacrylamide gel electrophoresis. No correlation between any combination of the phenotypic characteristics was found when correlation was done with milk yield and somatic cell count, corresponding to the 6 months before sampling. Otherwise, a positive correlation was found between type of haemolysis and the N-acetyl-beta-glucosaminidase (NAGase) values. In milk from quarters infected with the-non-haemolytic strains, the level of NAGase was significantly lower (P < 0.05) than that from quarters infected with the haemolytic strains (69.7 and 105.9, respectively). However, the level of NAGase activity in the milk of the quarters infected with the non-haemolytic strains was significantly higher (P < 0.05) when compared to the milk of quarters infected with coagulase-negative staphylococci (43.5).

Natural Clostridium botulinum type C toxicosis in a group of cats.

Clinical signs of botulism were observed in a group of eight cats, four of which died, after being fed pelican carrion. Clostridium botulinum type C was isolated from one cat. The microorganism and its toxin were found in the pelican. This is apparently the first report of natural botulism in cats.