Effect of dry period immunization of Salmonella Dublin latent carriers with a commercial live culture vaccine on intrauterine transmission based on the presence of precolostral antibodies in offspring

Salmonella Dublin latent carrier cows represent a high risk for infection of newborn calves via intrauterine transmission and shedding of bacteria in feces and colostrum at calving. Vaccination of these latent carrier dams during late gestation boosts immunity against S. Dublin. This could reduce the activation of the dormant bacteria during the periparturient immune dysfunction period, thereby reducing the risk of early-life infection in the offspring. Thus, the objective of this study was to evaluate the extent to which vaccinating S . Dublin latent carrier cows at dry-off with a commercial live culture vaccine reduces bacterial shedding at calving and intrauterine infection to calves. To identify latent carriers, we screened 1,084 cows in 4 Michigan commercial dairy farms with a history of S . Dublin. Cows were defined as latent carriers when they showed 3 consecutive positive milk antibody ELISA tests conducted every 2 mo. Subsequently, 148 latent carriers were randomly allocated to the vaccine or control group. Vaccine cows received a commercial live culture vaccine subcutaneously (SC) at dry-off and a booster 2 weeks later. Control cows received saline SC at the same times. At calving, we collected fecal and co-lostrum samples from the dam and a pre-colostral serum sample from the calf. Bacterial shedding was evaluated in feces and colostrum both qualitatively and quantitatively through bacterial culture and qPCR, respectively. Intrauterine transmission was defined when a calf was positive for serum antibody ELISA at birth. Vaccination decreased the likelihood of calves being born S . Dublin seropositive (Relative Risk [95%CI]) = 0.19 [0.04 – 0.84]). However, no S. Dublin positive isolates were identified through either bacteriological culture or qPCR in feces or colostrum. Vaccination of S. Dublin latent-carrier cows at dry-off reduced intrauterine transmission to calves. Further research is warranted into the potential of vaccination to decrease vertical transmission of S. Dublin in dairy farms. Additionally, the absence of S. Dublin positive fecal and colostrum samples warrants further evaluation of the detection methods for identifying latent carriers or S. Dublin isolation, as well as the role of latent carriers in infecting newborn calves in the maternity area at birth.


INTRODUCTION
Salmonella enterica subspecies enterica serovar Dublin (S.Dublin) is a bovine host-adapted serotype with an increasing prevalence of infections in dairy cattle in the US and Canada over the past years (Cummings et al., 2009;2010;NAHMS, 2014;Perry et al., 2023).S. Dublin is now the most isolated serovar in several veterinary diagnostic laboratories, including the Michigan State University Veterinary Diagnostic Laboratory (Erdman et al., 2012;Valenzuela et al., 2017;Holschbach and Peek, 2018;Velasquez-Munoz et al., 2024).Furthermore, most US isolates of S. Dublin are multi-drug antimicrobial resistant (Erdman et al., 2012;Valenzuela et al., 2017;Holschbach and Peek, 2018;Velasquez-Munoz et al., 2024), although with reasonable susceptibility to enrofloxacin (Velasquez-Munoz et al., 2024).This antimicrobial, however, is not labeled as a treatment for S. Dublin infections, and extra-label use of fluoroquinolones such as enrofloxacin is prohibited for food animals in the US (FDA, 1994).Thus, no effective antimicrobial is currently available to treat S. Dublin infections in cattle in the US.Also, S. Dublin is a public health concern as it causes fatal multi-drug resistant infections in humans via contaminated foods or contact with infected animals (Harvey et al., 2017;CDC, 2019;Paudyal et al., 2019), with mortality in humans being 4.4 times higher than in other Salmonella infections (Helms et al., 2003).Establishing reliable control strategies for S. Dublin is, therefore, crucial to the dairy industry.
Infection with the Salmonella host-adapted serovar Dublin can occur in utero or when calves ingest the bacteria shed by infected cows or calves (Richardson, 1973).After colonizing the gastrointestinal tract, S. Dublin can spread through the bloodstream into organs such as the liver, spleen, lung, or lymph nodes.Calves that survive the infection have a high probability of becoming asymptomatic carriers for life (Nielsen et al., 2004), shedding low numbers of bacteria in their secretions for years (Nielsen, 2013a).This potential for asymptomatic infection and latent carriers of disease creates important challenges in managing S. Dublin transmission.Latent carrier cows infect calves in utero and shed bacteria during calving (Counter and Gibson, 1980;Nielsen et al., 2004;Hanson et al., 2016).Thus, controlling the spread of S. Dublin from latent carriers to newborns is needed to reduce disease transmission in affected herds (Velasquez-Munoz et al., 2024).
In other species, the immune response elicited by a Salmonella enterica live attenuated vaccine reduced Salmonella excretion temporarily after challenge (Sharma et al., 2018;Syed et al., 2020).Similarly, vaccination of cattle with a live attenuated recombinant S. Dublin strain expressing an Escherichia coli O157:H7 outer membrane protein induced a transient reduction of shedding of both S. Dublin and E. coli O157:H7 (Khare et al., 2010).Although Salmonella and E. coli show a distinct fecal excretion pattern, the use of vaccination as a tool to reduce S. Dublin excretion warrants further evaluation.In the US, only one commercial vaccine is currently licensed for the vaccination of cattle against S. Dublin (EnterVene ® -d; Boehringer Ingelheim Animal Health).Thus, our objective was to determine the impact of vaccinating S. Dublin latent carrier cows during late pregnancy with this commercial vaccine on S. Dublin in utero transmission and bacterial shedding at calving.Our hypothesis was that vaccinated latent carrier dams will exhibit a lower risk of bacterial shedding and intrauterine transmission, which would, in turn, decrease the infection risk of newborn calves.

MATERIALS AND METHODS
All animal procedures of the study were approved by the Michigan State University Institutional Animal Care and Use Committee (PROTO202100128), and animals were enrolled with owner consent.This randomized clinical trial took place from June 2022 to June 2023 and was comprised of 2 phases (Figure 1).The screening phase involved the identification of S. Dublin latent carriers, whereas the trial phase included the treatment allocation and sample collection.

Sample Size Determination
The sample size was determined using JMP Pro v. Seventeen (SAS Institute).Our goal was to detect a difference in the proportion of latent carrier cows shedding S. Dublin at calving between vaccinated and control cows.Given the lack of S. Dublin vaccine-specific data, we based our calculation on the reduction in the proportion of S. Dublin shedders observed in a vaccination study that used an attenuated S. Dublin bacterium expressing E. coli O157:H7 outer membrane protein (Khare et al., 2010).With a power of 80%, a precision of 5%, and a reduction of 17%, as seen in the referenced study, the required sample size resulted in 160 latent carrier animals (80 per treatment; 20 cows/farm/treatment) when considering clustering in 4 different farms and a 5% attrition.

Screening Phase
We used a convenience sample of 4 Michigan dairy farms.Farm inclusion criteria consisted of (1) a history of S. Dublin infection in the last 12 mo via isolation of the bacteria from a clinical case, (2) a farm size of 1,000 milking cows or greater, (3) no current vaccination against Salmonella in cows or calves, (4) enrollment in regular Dairy Herd Improvement (DHI) testing, (5) access to cows on farm during the dry period, (6) use of a 60-d long dry period, and (7) willingness to participate in the study.The reported average prevalence of S. Dublin latent carrier within dairy herds with an S. Dublin history was approximately 15% (Nielsen et al., 2004).Thus, to identify the 160 latent carriers needed, a random sample of 250 to 280 cows stratified by lactation group (1st, 2nd, or 3rd and greater) and proximity in expected calving date were screened for S. Dublin carrier status at each of the 4 enrolled farms, resulting in a total of 1,096 cows.S. Dublin, unlike other Salmonella spp., is rarely cultured from feces except in sick animals.Thus, other samples need to be examined to determine latent carrier status.A common approach to determine latent carrier status is to conduct serial testing of animals, as carriers sustain elevated antibodies against S. Dublin for long periods of time (Spier et al., 1990;Nielsen and Dohoo, 2011;Nielsen and Nielsen, 2012).For this, 3 milk samples from each cow were collected at 2-mo intervals (63 to 97, 125 to 155, and 185  instructions, a sample was classified as positive with a percent positivity ≥ 35%.Samples with a negative percent positivity were assigned a value of 0. Cows testing positive in all 3 tests were classified as latent carriers (Nielsen and Dohoo, 2011;Nielsen and Nielsen, 2012).

Trial Phase
We aimed to randomly select 40 latent carrier cows at each of the enrolled farms for inclusion in the vaccination trial.However, one herd was removed from the study due to the low number of latent-carrier animals identified (n = 3), and all the latent carrier cows identified in 2 of the other farms were enrolled to minimize the loss of statisti- Animals in each herd were allocated to the vaccine or control group, stratifying by parity group (1st, 2nd, 3rd or greater), using randomization software (https: / / www .graphpad.com/quickcalcs/ randomize1/ ).Vaccine cows received 2 mL of a commercial live culture S. Dublin vaccine (EnterVene-d; Boehringer Ingelheim Animal Health) subcutaneously (SC) within 7 d of dry-off (215 to 226 actual days pregnant) and a booster 14 d after the primary vaccination per label.Control cows received 2 mL of saline SC at the same times.Treatments were administered by research staff following the Beef Quality Assurance Program TM guidelines (https: / / www .bqa.org/Media/ BQA/ Docs/ nationalmanual .pdf).Farm and laboratory personnel were blinded to treatment allocation.Farm staff were trained to monitor and report adverse reactions in all enrolled cows for 4 d following the administration of treatments.All animals enrolled in the study were managed according to each farm's protocols before, during, and after parturition.Enrolled cows were maintained in the pens with their herd mates.
At calving, feces and colostrum from the enrolled cows were collected by trained farm staff into sterile containers and kept at −20°C pending analysis.A 10 mL blood sample from the calf was also collected before administration of colostrum via jugular venipuncture using vacuum tubes without additives (BD vacutainer) by trained farm staff or research staff.Blood tubes were allowed to clot for 30 min, centrifuged at 1,500 g for 15 min at 4°C, and the serum was harvested, aliquoted, and stored at −20°C pending analyses.

Sample Analysis
Feces (1 g) and colostrum (1 mL) from latent carrier cows were cultured using a pre-enrichment method to qualitatively identify S. Dublin shedding.For this, samples were diluted 1:10 on tetrathionate broth (Beckton Dickinson) and incubated overnight (16-24 h) at 35 ± 2°C in an aerobic atmosphere.Next, the tetrathionate broth culture was sub-cultured on brilliant green agar containing novobiocin (Hardy Diagnostics) and xyloselysine-tergitol (TermoFisher) agar plates.The plates were incubated overnight at 35 ± 2°C in an aerobic atmosphere.Subsequently, plates were examined to identify colonies of Salmonella spp.Isolates were confirmed employing a MicroflexLT MALDI-TOF mass spectrometer (Bruker Daltonics).Whole genome sequencing based serovar prediction was performed on the isolated Salmonella sp. at the Michigan State University Veterinary Diagnostic Laboratory (Lansing, Michigan).In brief, DNA extracted from Salmonella were quantified and library prepared us-ing DNA Prep kit (Illumina San Diego, CA).Short-read sequencing was performed on the NextSeq 1000 P1 cartridge (Illumina San Diego, CA).The raw FASTQ files were used for serovar analysis using SeqSero 1.2 (https: / / cge .food.dtu.dk/services/ SeqSero/ ) (Zhang et al., 2015).
A qPCR assay was utilized to quantify the shedding of bacteria by latent carriers.The qPCR assay was made following the S. Dublin-specific test developed by Persson et al. ( 2012) that targets the VagC gene.For this, DNA was extracted from the fecal and colostrum samples using a commercial kit (QIAamp Fast DNA stool mini kit; Qiagen) following the manufacturer-recommended protocol.The extracted DNA was diluted at 50 μL.A total reaction mix of 20 μL was used, composed of 10 μL of TaqMan Universal PCR Master Mix (ThermoFisher), 0.4 μL (25 μM) of each pair of primers specific for the vagC gene (Forward-GGGTGAGCGAGCTGGAAA; Reverse-CGCCATAAAGTCCGGGTCA), 0.2 μL (25 μM) of probe (FAM-TTTTTCGAGCTGCGCGAACGAGC-BHQ1), 4 μL of nuclease-free water, and 5 μL of the template DNA.DNA from an isolate of S. Dublin ATCC 15480 (American Type Culture Collection) was employed as a positive control and 5 μL of nuclease-free water were used a negative control instead of template DNA.The qPCR was run on an ABI 7500 Fast real-time PCR instrument (Thermo Fisher Scientific Inc.).The construction and standardization of the standard curves for S. Dublin genomic copy number determination was based on a 10-fold serial dilution series of the S. Dublin ATCC 15480 isolate, ranging from 1 × 10 1 to 1 × 10 −3 ng.
A A length dp g mol dp The sensitivity and specificity validation of this qPCR assay did not include several Salmonella serovars commonly isolated from cattle in the US that also express this gene.Thus, the primer annealing specificity was analyzed in silico considering the shotgun genome sequences of S. Dublin (GenBank ID: JF267653.1)and other commonly isolated Salmonella vagC reported in the National Center for Biotechnology Information (NCBI) using the Unipro UGENE software Vo.5.50 (NCBI).vagC sequences included serovars Kentucky (GenBank ID: CALNWF000000000.1),Newport (GenBank ID: AHUG01000037.1),Typhimurium (Gen-Bank ID: CP088137.1),Braenderup (GenBank ID: UFRJ01000003.1,and Bovismorfican (GenBank ID: UGVQ01000001.1).No overlap was identified between To identify intrauterine disease transmission, precolostral serum samples were shipped overnight in dry ice to the Cornell University Animal Health Diagnostic Center (Ithaca, NY) for testing using a commercial S. Dublin antibody ELISA (Applied Biosystems).Samples were defined as positive with a percent positivity ≥35%.Samples with a negative percent positivity were assigned a value of 0. Intrauterine transmission was defined when calves had a positive result on S. Dublin antibody ELISA at birth.

Statistical Analyses
All statistical analyses were completed using JMP Pro v. Seventeen (SAS Institute).Logistic regression models were used to analyze dichotomous data (qualitative shedding and intrauterine transmission).Linear mixed models were used to compare continuous data (quantitative assessment of bacterial shedding).Models included the fixed effect of treatment (vaccine vs. control) and the random effect of farm.The non-parametric Mann-Whitney U test compared the ELISA % positivity between calves born to vaccinated and control dams.Statistical significance was declared at P < 0.05.

RESULTS AND DISCUSSION
Out of 1,096 cows initially enrolled for screening, we obtained samples from all 3 milk tests for 1,084 cows (Figure 1).From these, 158 (14.6%) cows were identified as latent carriers.The ELISA results of the screening phase for the cows identified as carriers are summarized in Table 1.The latent carrier within-herd prevalence varied greatly among the 4 study farms from 1.0 to 23.6% (Figure 2) despite all farms having a recent history of S. Dublin infection.This variation underscores the complex epidemiology of S. Dublin infection dynamics in dairy herds.The overall prevalence of latent carriers as well as the variation in within herd prevalence were consistent with previous reports (Nielsen et al., 2004).
Adverse reactions to the treatments were minimal, including only the development of a 3 to 5 cm of diameter swelling at the injection site (n = 10; vaccine = 8; control = 2) that disappeared without intervention in 4 to 10 d.A total of 127 (vaccine = 62, control = 64) fecal, 126 (vaccine = 64, control = 63) colostrum, and 118 (vaccine = 60, control = 58) serum samples were collected and analyzed (Figure 1).Loss-to-follow-up reasons included death in the dry period (n = 1; vaccine = 0, control = 1), dystocia (n = 4; vaccine = 1, control = 3), stillbirth (n = 6; vaccine = 3, control = 3), misidentification of animals (n = 10; vaccine = 6 and control = 4), and serum samples not collected before colostrum ingestion (n = 9; vaccine = 4 and control = 5).Chi-squared analyses revealed no differences between treatment groups for any loss-to-follow-up reasons (P > 0.62).In line with a previous study (Smith et al., 2015), we found the commercial vaccine to be safe to use in pregnant cattle around the time of dryoff based on the lack of adverse reactions observed.Apart from the loss-to-follow-up reasons described above, the latent carriers did not develop any sign of disease that could be detected by farm staff or the rumination sensors of the farms.A total of 12 (vaccine = 2; control = 10) out of 118 (10.2%) calves tested positive for S. Dublin antibodies at birth (Table 1).Vaccination decreased the likelihood of calves being born S. Dublin seropositive (Relative Risk [95%CI]) = 0.19 [0.04 -0.84]; P = 0.0095).The concentration of S. Dublin antibodies, evaluated by ELISA % positivity, was lower (P = 0.02) in calves born to vaccinated carriers than in calves born to control carriers (Figure 3).Intrauterine transmission of Salmonella spp.has been documented in the literature (Richardson, 1973;Hanson et al., 2016).Hanson et al. (2016) found 50% vertical transmission in infected cattle with various Salmonella enterica serovars when euthanizing calves at birth and culturing various tissues However, the Dublin serovar was not reportedly found among the cultured and serotyped Salmonella isolates in the referenced study.The rate of S. Dublin vertical transmission in the control group of our study (17.2%) was lower than this previous report, but contrary to the Hanson et al. (2016) study that looked at all culturable Salmonella spp., we focused only on the Dublin serovar when evaluating intrauterine transmission.Hanson et al. (2016) also reported that 47.3% of the dams had Salmonella-positive fecal cultures at the calving time.This prevalence of fecal shedding was almost 3 times greater than in our study (Table 3), suggesting that the rate of active infections in their study might have been higher than in ours.
We utilized pre-colostral antibody testing to avoid euthanizing the calves at birth.The bovine fetus starts producing antibodies in the second trimester of pregnancy (Ellis et al., 1978) and, by the third trimester, is able to produce IgG against most antigens (Banks and McGuire, 1989).Nevertheless, it would be possible for in utero infection to occur close to calving, thereby not allowing the fetus sufficient time to mount an immune response that results in measurable antibodies at birth.To our knowledge, this is the first report documenting the prevalence of intrauterine S. Dublin transmission from latent carrier cows, limiting our ability to compare with previous studies.The proportion of calves born with S. Dublin antibodies in the control group further highlights the relevance of the intrauterine transmission route for the persistence of the infection in herds and the importance of controlling S. Dublin latent carrier animals in infected herds.Our approach, however, only assessed the fetus' immune response and indicates intrauterine exposure.Further testing in the calves, such as pre-colostral blood or fecal cultures, would have been required to demonstrate active infection.As more research focuses on the vertical transmission of S. Dublin, it would be useful to elucidate whether these seropositive calves are born bacteremic or shedding bacteria in their feces at birth.
We additionally found that latent carriers vaccinated at dry-off with a live culture S. Dublin commercial vaccine were 5 times less likely to give birth to a seropositive calf (Table 2).Previous research showed that the administration of the same commercial vaccine used in this study to late-gestation cows resulted in a measurable immune response until calving (Smith et al., 2015).Because S.   ELISA-positive samples had a percent positivity ≥35%.
Dublin-infected animals carry the bacteria in their lymph nodes (Hall et al., 1978;Steinbach et al., 1996), we speculate that the activation of the immune system against S.
Dublin by immunization limited the spread of the bacteria from the lymph nodes to other tissues, resulting in a lower vertical transmission of the infection.This strategy could be used in combination with other management practices to limit the transmission of S. Dublin in cattle herds.However, more research is warranted before this tool can be broadly applied in the field.
In feces, only 21 of the 127 samples (in 2 of the 3 farms) were positive for Salmonella spp. on culture (Table 3), but none of the isolates was S. Dublin (Table 4).Similarly, Salmonella spp.were only identified in 2 colostrum samples (in the same 2 farms).However, serotyping excluded both as not Dublin (Table 4), as did the qPCR results where none of the feces or colostrum samples was positive.Given the absence of S. Dublin bacterial identification in the latent carriers enrolled in this study, a comparison of bacterial shedding between vaccinates and controls was not undertaken.Thus, we were unable to test the portion of our hypothesis relative to the effect of vaccination on bacterial shedding at calving.
Contrary to our expectations, we were not able to identify S. Dublin in feces or colostrum via traditional bacterial culture or molecular biology diagnostic methods.The recovery of S. Dublin in feces of apparently healthy cows is usually low.However, we anticipated most of the cows in the control group to shed the bacteria at calving because we only enrolled latent carriers in the trial phase, and the stress associated with parturition increases the probability of shedding the bacteria in feces (Nielsen, 2013a).These results make us question the identification of latent carriers based on repeated milk antibody ELISA testing, the sensitivity of the methods available to detect S. Dublin via bacteriology, and/or the relevance of bacterial shedding from latent carriers at the time of calving on disease transmission.Nielsen (2013b) also reported a low prevalence of fecal shedders among cattle with persistently high S. Dublin antibodies, although these animals were not sampled around a stressful event.Repeated sampling could have increased our ability to identify shedders by fecal culture.However, repeated culture before or after calving was logistically not viable under the conditions of this study and, to a certain degree, shedding at different times than calving would have limited impact on disease transmission to the newborn in systems where dams and calves are separated shortly after birth (discussed below).
Identification of latent carrier cows is based on serology because the intermittent shedding of S. Dublin by these animals makes bacteria culture and identification methods unreliable (Smith et al., 1992).However, the classification of animals based on antibody concentration results varies in the literature.Most researchers recommend serial testing of animals to differentiate between recent exposure and carrier status because the antibody response of transiently infected animals may take months to decrease to undetectable levels using ELISA (Smith et al., 1989;Spier et al., 1990).However, the number and frequency of tests required to define an animal as a carrier through serial testing remains controversial, with some authors suggesting that 2 positive ELISA tests taken 60 d apart are sufficient to confidently classify S. Dublin latent carriers (Spier et al., 1990).In our study, however, we took a more conservative approach and required 3 positive tests 2 mo apart each to define latent carrier status, in line with other recommendations (Smith et al., 1992;Nielsen et al., 2004).
The commercial ELISA assay used targets antibodies directed against the Salmonella spp LPS O-antigens 1, 9, and 12. Thus, there is some possibility for cross-reaction among Salmonella serovars (Konrad et al., 1994), which could have impacted our results and the classification of latent carriers.However, lack of cross-reactivity has also been reported (Agren et al., 2016), and none of the isolates identified in our study expressed any of the antigens Results are expressed as number (%) of total isolates in each sample type.
included in the ELISA assay (Table 4).We, therefore, speculate that ELISA cross-reactivity might have had a limited impact in our results.Furthermore, the serial testing approach used is intended to minimize the impact of potential ELISA cross-reactivity.Also, we took a more conservative approach by using the ELISA assay's manufacturer-recommended cut-off of percent positivity of 35% despite some evidence in the literature that a 20% cut-off could be sufficient to detect latent carriers (Nyman et al., 2013).We utilized common Salmonella culture methods for the isolation of S. Dublin in feces and colostrum that require pre-enrichment and subsequent selective media culture to increase bacterial detection (ISO, 2017).Nevertheless, only 16.5 and 1.6% of fecal and colostrum samples, respectively, had a positive Salmonella culture (Table 3) despite all enrolled cows having sustained Salmonella antibody titers before calving.Individual animal fecal culture for S. Dublin is thought to have a sensitivity of 16-20% (Nielsen, 2013a).This relatively low sensitivity, coupled with the sporadic nature of shedding and the fact that latent carrier cattle shed lower numbers of organisms than clinically ill or acutely infected animals, along with only sampling cows once, might explain the low detection observed in this study via culture.Even manure samples spiked with known amounts of S. Dublin and S. Typhimurium show limited detection via culture or PCR at low bacterial concentrations (Jensen et al., 2013).Furthermore, cattle can be co-infected with more than one Salmonella serovar (Hanson et al., 2016).In those instances, the intraspecies gut competition may increase the longitudinal shedding and colony-forming unit counts from the non-host adapted serotypes over those host-adapted like Dublin (Kirchner et al., 2012).Ultimately, based on our one-time sampling, the lack of S. Dublin detection on direct fecal/colostrum PCR indicates that S. Dublin shedding was non-existent or below the PCR detection limit at the time of calving.
We also conducted qPCR analyses to (1) identify the bacteria based on detection of genetic material, and (2) estimate the number of bacteria being shed as the required pre-enrichment stage in the culture protocol precludes a direct quantification of colony forming units.Molecular biology techniques can complement bacteria culture in Salmonella spp.field studies to increase diagnostic ability (Jensen et al., 2013).Thus, we used PCR specific for the Dublin serovar (Persson et al., 2012).Despite these efforts, we were unable to identify S. Dublin in fecal or colostrum samples.Overall, the limitations of the bacterial identification techniques available, the shedding pattern of the bacteria, and only sampling cows at one point precluded us from evaluating the extent to which dry cow vaccination could have decreased S. Dublin shedding.
Ultimately, our results also question the role of S. Dublin latent carriers in transmitting the disease to newborn calves in the maternity area.Latent carriers are believed to play a role in the transmission of the disease to newborn calves via excretion of the bacteria in feces and colostrum (Nielsen, 2013a;Velasquez-Munoz et al., 2024).However, we were unable to isolate S. Dublin in any of the fecal or colostrum samples collected at the time of calving.Although it is possible that shedding occurred later than when samples were collected, 56.2% of US dairy farms separate calves and cows within 6 h of birth, and 24.5% of farms with 500 or more cattle within 1 h (NAHMS, 2014).Thus, delayed bacterial shedding relative to calving would have a limited impact on the transmission of disease to newborns provided the maternity area is maintained clean between animals because calves would be removed from maternity in most farms before cows start shedding bacteria.This is consistent with a recent cross-sectional study that found that increasing the frequency of adding bedding to the maternity pen was associated with decreased odds of the farm being positive for S. Dublin (Perry et al., 2023).

CONCLUSIONS
In this study, vaccinating S. Dublin latent carrier cows at dry-off with a commercial live culture vaccine reduced intrauterine transmission to calves by 81% based on the presence of precolostral S. Dublin antibodies.Further research is needed into the role of vertical transmission on disease dynamics and the extent to which vaccination could contribute to decreasing the transmission of S. Dublin in dairy farms.Additionally, the absence of S. Dublin positive fecal and colostrum samples warrants further evaluation of the traditional laboratory methods for identification of latent carriers or S. Dublin isolation, as well as the role of latent carriers in infecting newborn calves through the fecal-oral route in the maternity area at birth.

NOTES
We thank CentralStar Cooperative (East Lansing, MI) for their assistance in accessing the DHI milk samples for testing.This trial was financially supported by Boehringer Ingelheim Animal Health USA Inc. and the company employed J.N. Roberts as study monitor during the period of this study.The company and J.N. Roberts had no role in the study design, data collection and analysis, publication decision, or manuscript preparation.Boehringer Ingelheim Animal Health USA Inc. has filed a provisional patent application regarding the use of the EnterVene ® -d vaccine to prevent intrauterine transmis- Figure 1.Diagram representing the number of animals included at each stage of the study and reasons for their removal.Created with BioRender.com Castro-Vargas et al.: SALMONELLA DUBLIN LATENT CARRIER VACCINATION the primers and the other serovars, confirming the specificity of the primers for S. Dublin detection.

Figure 2 .
Figure 2. Number of milk tests testing positive for Salmonella Dublin antibodies in each farm.Results are expressed as the number (%) of cows with positive milk antibody ELISA tests.Milk samples were collected every other month for a total of 3 consecutive tests.Farm B had 10 (3.6%), 6 (2.2%), and 3 (1.0%)cows with one, 2, and 3 positive tests, respectively.Cows with all 3 positive tests were considered S. Dublin latent carriers.
Figure 3.Comparison of sera Salmonella Dublin antibodies at birth between calves born to Vaccine (n = 60) and Control (n = 58) cows.Results are expressed as the % positivity of the ELISA.Samples with a negative % positivity result were assigned a value of 0. Results of % positivity >35% are indicative of antibodies present in the sample, and the greater % positivity is indicative of a greater antibody concentration.The results were compared statistically with a 2-tailed Mann-Whitney U test.
Castro-Vargas et al.: SALMONELLA DUBLIN LATENT CARRIER VACCINATION Castro-Vargas et al.: SALMONELLA DUBLIN LATENT CARRIER VACCINATION sion of S. Dublin.The authors have not stated any other conflicts of interest.
Castro-Vargas et al.: SALMONELLA DUBLIN LATENT CARRIER VACCINATION

Table 1 .
Castro-Vargas et al.: SALMONELLA DUBLIN LATENT CARRIER VACCINATION Descriptive results of the ELISA % positivity in each sampling of the screening phase among the 158 Salmonella Dublin latent carriers identified

Table 3 .
Results of bacterial shedding analyses Results are expressed as number (%) of positive/negative samples in each test.

Table 2 .
Contingency table of results of antibody ELISA results in precolostral serum samples of calves.Results are expressed as the number (%) of samples in each category

Table 4 .
Whole genome sequencing based serovar prediction of all Salmonella isolates