Genome-wide methylation profile following prenatal and postnatal dietary omega-3 fatty acid supplementation in pigs.

The objective of this study was to determine how prenatal and postnatal dietary omega-3 fatty acids alter white blood cell (leukocyte) DNA methylation of offspring. Fifteen gilts (n = 5 per treatment) were selected from one of three treatments: (i) control diet throughout gestation, lactation and nursery phase (CON); (ii) algal omega-3 fatty acid supplementation enriched in EPA and DHA (Gromega™ ) fed throughout gestation, lactation and nursery phase (Cn3); or (iii) Gromega™ supplementation maternally, during gestation and lactation only, and control diet during the nursery phase (Mn3). At 11 weeks of age and after 8 weeks of post-weaning nursery feeding, buffy coat genomic DNA was subjected to methyl CpG binding protein sequencing. The methylation enriched profile mapped to 26% of the porcine genome. On chromosome 4, a 27.7-kb differentially methylated region downstream of RUNX1T1 was hypomethylated in the Mn3 and Cn3 groups by 91.6% and 85.0% respectively compared to CON pigs. Conversely, hypermethylation was detected in intergenic regions of chromosomes 4 and 12. Regulatory impact factor and differential hubbing methods were used to identify pathways that were coordinately regulated by methylation due to feeding EPA and DHA during pregnancy. Despite limited ability to detect differential methylation, we describe methods that allow the identification of coordinated epigenetic regulation that could not otherwise be detected from subtle single locus changes in methylation. These data provide evidence of novel epigenetic regulation by maternal and early life supplementation of omega-3 fatty acids that may have implications to growth and inflammatory processes.

Integrating comparative expression profiling data and association of SNPs with Salmonella shedding for improved food safety and porcine disease resistance.

Salmonella in swine is a major food safety problem, as the majority of US swine herds are Salmonella-positive. Salmonella can be shed from colonized swine and contaminate (i) neighbouring pigs; (ii) slaughter plants and pork products; (iii) edible crops when swine manure is used as a fertilizer; and (iv) water supplies if manure used as crop fertilizer runs off into streams and waterways. A potentially powerful method of addressing pre-harvest food safety at the farm level is through genetic improvement of disease resistance in animals. In this research, we describe a successful strategy for discovering genetic variation at candidate genes associated with disease resistance in pigs. This involves integrating our recent global gene expression analysis of the porcine response to Salmonella with information from the literature about important candidate genes. We identified single-nucleotide polymorphisms (SNPs) in these functional candidate genes and genotyped three independent pig populations that had data on Salmonella faecal shedding or internal burden (total n = 377) at these loci. Of 31 SNPs genotyped, 21 SNPs segregated in at least two populations with a minor allele frequency of 15% or greater. Statistical analysis revealed thirteen SNPs associated with Salmonella faecal shedding or tissue colonization, with an estimated proportion of false positives (PFP) ≤0.2. The genes with associated SNPs included GNG3, NCF2, TAP1, VCL, AMT, CCR1, CD163, CCT7, EMP1 and ACP2. These associations provide new information about the mechanisms of porcine host response to Salmonella and may be useful in improving genetic resistance to this bacterium.

Use of bioinformatic SNP predictions in differentially expressed genes to find SNPs associated with Salmonella colonization in swine.

Asymptomatic Salmonella-carrier pigs present a major problem in preharvest food safety, with a recent survey indicating >50% of swine herds in the United States have Salmonella-positive animals. Salmonella-carrier pigs serve as a reservoir for contamination of neighbouring pigs, abattoir pens and pork products. In addition, fresh produce as well as water can be contaminated with Salmonella from manure used as fertilizer. Control of Salmonella at the farm level could be through genetic improvement of porcine disease resistance, a potentially powerful method of addressing preharvest pork safety. In this research, we integrate gene expression profiling data and sequence alignment-based prediction of single nucleotide polymorphisms (SNPs) to successfully identify SNPs in functional candidate genes to test for the associations with swine response to Salmonella. A list of 2527 genes that were differentially regulated in porcine whole blood in response to infection with Salmonella enterica serovar Typhimurium were selected. In those genes, SNPs were predicted using ANEXdb alignments based on stringent clustering of all publically available porcine cDNA and expressed sequence tag (EST) sequences. A set of 30 mostly non-synonymous SNPs were selected for genotype analysis of four independent populations (n = 750) with Salmonella faecal shedding or tissue colonization phenotypes. Nine SNPs segregated with minor allele frequency ≥15% in at least two populations. Statistical analysis revealed SNPs associated with Salmonella shedding, such as haptoglobin (HP, p = 0.001, q = 0.01), neutrophil cytosolic factor 2 (NCF2 #2, p = 0.04, q = 0.21) and phosphogluconate dehydrogenase (p = 0.066, q = 0.21). These associations may be useful in identifying and selecting pigs with improved resistance to this bacterium.

Correlating blood immune parameters and a CCT7 genetic variant with the shedding of Salmonella enterica serovar Typhimurium in swine.

The porcine response to Salmonella infection is critical for control of Salmonella fecal shedding and the establishment of Salmonella carrier status. In this study, 40 crossbred pigs were intranasally inoculated with Salmonella enterica serovar Typhimurium (Salmonella Typhimurium) and monitored for Salmonella fecal shedding and blood immune parameters at 2, 7, 14 and 20 days post-inoculation (dpi). Using a multivariate permutation test, a positive correlation was observed between Salmonella Typhimurium shedding levels at 2 and 7dpi and serum interferon-gamma (IFNgamma) levels at 2dpi (p<0.05), with Salmonella being shed in greater numbers from animals with higher IFNgamma levels. A positive correlation was also observed between IFNgamma levels and the number of banded neutrophils (2dpi), circulating neutrophils (7 and 14dpi), monocytes (7dpi), and white blood cells (WBCs) (7, 14 and 20dpi). We have further performed association studies on these immune response parameters as well as shedding status of the Salmonella-infected pigs with a single nucleotide polymorphism (SNP) in the porcine gene CCT7, previously shown by our group to be transcriptionally up-regulated in swine experimentally inoculated with Salmonella Typhimurium. Our analyses with the 40 pigs suggest a positive association (p=0.0012) of SNP genotype A/G at position AK240296.c1153G>A of the CCT7 gene with Salmonella shedding at 7dpi compared to the G/G homozygote genotype. Linking specific genes and genetic polymorphisms with the porcine immune response to Salmonella infection and shedding may identify potential markers for carrier pigs as well as targets for disease diagnosis, intervention and prevention.

Computational integration of structural and functional genomics data across species to develop information on the porcine inflammatory gene regulatory pathway.

We are investigating the porcine gut immune response to infection through gene expression profiling. Porcine Affymetrix GeneChip data was obtained from RNA prepared from mesenteric lymph node of swine infected with either Salmonella enterica serovar Typhimurium (ST) or S. Choleraesuis (SC) for 0, 8, 24, 48 or 504 hours post-inoculation (hpi). In total, 2365 genes with statistical evidence for differential expression (DE; p < 0.01, q < 0.26, fold-change > 2) between at least two time-points were identified. Comparative Gene Ontology analyses revealed that a high proportion of annotated DE genes in both infections are involved in immune and defence responses. Hierarchical clustering of expression patterns and annotations showed that 22 of the 83 genes upregulated from 8-24 hpi in the SC infection are known NF-kappaB targets. The promoter sequences of human genes orthologous to the DE genes were collected and TFM-Explorer was used to identify a set of 72 gene promoters with significant over-representation of NF-kappaB DNA-binding motifs. All 22 known NF-kappaB target genes are in this list; we hypothesize that the remaining 51 genes are un-recognized NF-kappaB targets. Integration of these results and verification of putative target genes will increase our understanding of the porcine response pathways responding to bacterial infection.

Microbial community sequencing analysis of the calf eye microbiota and relationship to infectious bovine keratoconjunctivitis.

Infectious bovine keratoconjunctivitis (IBK) is an important production limiting disease in cattle. Moraxella bovis has historically been considered the primary causal agent; however, vaccines have not been consistently shown as effective in controlling disease incidence. The purpose of this study was to examine the bacterial community of calf eyes prior to disease onset using high-throughput sequencing of 16S ribosomal RNA and determine if it was associated with IBK occurrence. The study was designed as a case-control nested within a randomized controlled trial (RCT). Eye swabs were collected from all spring-born calves without clinical signs of IBK (t0 swabs) on a research farm with a previous history of IBK disease outbreaks. At follow-up or weaning, calves were diagnosed as IBK positive or negative. The lag time between enrollment swabs (t0) and IBK diagnosis ranged from approximately one to three months. Cases were randomly selected from IBK positive calves and controls were selected from IBK negative calves (i.e. calves that did not exhibit clinical signs of IBK throughout the course of the RCT). Analysis of the fold-change differences between cases and controls did not reveal large-scale distinctions in bacterial composition. However, principal component analysis suggested bacterial composition differences between calf management groups, which were based on dam parity. Moraxella was found to be among the top ten most abundant genera in our population; however, the difference in abundance was not significant between the cases and controls. No large-scale differences in the bacterial communities of calves that did or did not develop IBK were observed in our population. Nevertheless, it remains unclear whether the "natural" bacterial population of the calf might ultimately impact disease status. Further study is warranted to examine bacterial taxa that were observed to be significantly more abundant in the cases or controls as potential vaccines/therapeutic targets.

Distribution of 5-doxylstearic acid in the membranes of mammalian cells.

Concentration-dependent spin broadening of ESR spectra of the nitroxide 5-doxylstearic acid has been used to evaluate the distribution of 5-doxylstearic acid in the membranes of intact mouse thymus-bone marrow (TB) and Chinese hamster ovary (CHO) cells. TB cells, CHO cells, erythrocytes, and isolated plasma membranes from CHO cells were labelled with 5-doxylstearic acid and the peak to peak linewidths of the central line of the resulting ESR spectra were measured. The measured line widths were linearly dependent on the amount of 5-doxylstearic acid incorporated into the sample over the range of 0-0.18 mol nitroxide per mol lipid. In erythrocytes, the relationship between linewidths approximated a linear function at lower concentrations of 5-doxylstearic acid, up to 0.07 mol nitroxide per mol lipid. The amount of broadening of the central line for a given amount of 5-doxylstearic acid was far less for intact cells than for either erythrocytes or plasma membrane, indicating that the 5-doxylstearic acid samples a much larger lipid pool in the intact cells. With the broad assumption that the mobility of the 5-doxylstearic acid is similar in different membranes, the size of the lipid pool sampled by 5-doxylstearic acid is approximately equal to the total cellular lipid in intact cells. If a given concentration of 5-doxylstearic acid sampled only the plasma membrane of TB or CHO cells, we would expect to see a linewidth corresponding to a 12-20-fold greater local concentration of 5-doxylstearic acid than was observed, since the plasma membranes of CHO and TB cells represent only 5-8 percent of the total cellular lipid. Therefore, the 5-doxylstearic acid must distribute into most or all cellular membranes of intact cells and is not localized in the plasma membrane alone.

Controlling the proportion of false positives in multiple dependent tests.

Genome scan mapping experiments involve multiple tests of significance. Thus, controlling the error rate in such experiments is important. Simple extension of classical concepts results in attempts to control the genomewise error rate (GWER), i.e., the probability of even a single false positive among all tests. This results in very stringent comparisonwise error rates (CWER) and, consequently, low experimental power. We here present an approach based on controlling the proportion of false positives (PFP) among all positive test results. The CWER needed to attain a desired PFP level does not depend on the correlation among the tests or on the number of tests as in other approaches. To estimate the PFP it is necessary to estimate the proportion of true null hypotheses. Here we show how this can be estimated directly from experimental results. The PFP approach is similar to the false discovery rate (FDR) and positive false discovery rate (pFDR) approaches. For a fixed CWER, we have estimated PFP, FDR, pFDR, and GWER through simulation under a variety of models to illustrate practical and philosophical similarities and differences among the methods.

Investigation of the efficacy of albumin removal procedures on porcine serum proteome profile.

Improving the ability to predict livestock performance using biomarkers will provide a benefit for livestock genetic evaluation and improvement. The most practical biological sample to screen for development of biomarkers is serum due to the ease of collection. However, protein profiles in serum are complex and dynamic. Strategies are needed to manage variation in serum proteins used for biomarker identification. Albumin is the most abundant protein in serum, comprising over 50% of the overall protein content, and has historically been depleted from serum before biomarker identification. The objective of this study was to investigate the use of gel-based proteomic techniques to evaluate the need for porcine albumin depletion in biomarker identification. Albumin is known to bind many proteins in the blood, thus potential biomarkers could be removed during albumin depletion. Using two-dimensional difference in gel electrophoresis (2D-DIGE), we show whole serum can be used for biomarker discovery. The data obtained show that albumin removal methods are effective for porcine sera. Over 85% of the protein spots resolved on at least half of the gels were changed in abundance between whole and albumin depleted sera. Of the 204 protein spots significantly altered in abundance, 59 were changed over 400%. However, albumin removal also altered the serum proteome in an unpredictable manner; in the depleted sera, 86 protein spots were increased in abundance and 118 were decreased. Furthermore, the abundance of 59.4% of the protein spots in the albumin depleted samples had a larger standard error than whole sera. However, the resolution of albumin in 2D-DIGE analysis of whole sera permitted the detection and quantification of substantial numbers of proteins. Thus, it is proposed that whole serum can be used in a gel-based proteomics system for the identification of porcine biomarkers.

Swine as a model of skin inflammation. Phospholipase A2-induced inflammation.

A predictive animal model of skin inflammation is needed for the development of potential therapeutic agents. The existing models of inflammation rely on animals whose skin physiology or biochemistry differs significantly from human. The objective of this investigation was to evaluate the swine as a potential model of inflammation, because its skin has been recognized to exhibit morphologic and functional similarities to human skin. In the swine, an inflammatory response was produced following intradermal injection of snake venom phospholipase A2 (PLA2). This response was characterized by transient erythema (2-3 h) and microscopic changes of cell infiltration, epidermal hyperplasia, and dermal damage, which were apparent two days after PLA2 and peaked by day 7. In general, these microscopic changes persisted up to 21 days. Treatment with the antiinflammatory steroid, betamethasone dipropionate (Diprolene), gave a significant reduction of the inflammatory responses. Heat-inactivated PLA2, ovalbumin, or saline did not provoke this reaction, although PLA2 inactivated by bromophenacyl bromide alkylation did produce an inflammatory response. The alkylated PLA2 was also able to provoke an inflammatory response in the mouse paw edema assay. These results demonstrate that PLA2 can stimulate an inflammatory response in the swine skin, but that phospholipid hydrolytic activity is not required.

New mitomycin analogs produced by directed biosynthesis.

When the normal fermentation medium for the production of mitomycin C with Streptomyces caespitosus is supplemented with a number of primary amines, two new types of mitomycin analogs described as Type I and Type II are produced. Type I analogs are related to mitomycin C with the amine substitution at position C7 on the mitosane ring. Type II analogs also contain the same substitutions at C7 but the conformation of the mitosane ring is related to mitomycin B having an OH at positions C9a and a methyl substituted aziridine. The products obtained from the supplementation of the medium with methylamine, ethylamine, propylamine, propargylamine and 2-methylallylamine were isolated and characterized. In all cases the Type I analogs are more active in a prophage induction test and against L1210 lymphatic leukemia in mice. A number of other amines have been tested and shown to yield new products that have not yet been isolated. No secondary amines are incorporated.

Evolutionary algorithms and a fractal inverse problem.

Over the past 30 years, algorithms that model natural evolution have generated robust search methods. These so-called evolutionary algorithms have been successfully applied to a wide range of problems. This paper discusses two types of evolutionary algorithms and their application to a problem in shape representation. Genetic algorithms and evolutionary programming, although both based on evolutionary principles, each place different emphasis on what drives the evolutionary process. While genetic algorithms rely on mimicking specific genotypic transformations, evolutionary programming emphasizes phenotypic adaptation. Results presented show the success of evolutionary programming in solving an example of a fractal inverse problem, but indicate that a genetic algorithm is not as successful. Reasons for this disparity are discussed.

A Limited International Intercomparison of Responsivity Scales at Fiber Optic Wavelengths.

We report here on a recent limited international intercomparison of responsivity scales at wavelengths of interest to the optical communications community. Participants in the comparison were the national laboratories in the United States, the United Kingdom, Germany, and Australia. The wavelengths tested were 1300 and 1550 nm. Data taken at 850 nm are only briefly discussed. The disagreement between the national laboratories' responsivity scale is comfortably within the uncertainty claimed by each laboratory.

Complementary DNA macroarray analyses of differential gene expression in porcine fetal and postnatal muscle.

To study differential gene expression in porcine skeletal muscle, a porcine complementary DNA (cDNA) macroarray was produced that contained 327 expressed sequence tags (EST) derived from whole embryo and adult skeletal muscle, and differential display PCR products from fetal and postnatal muscle. Total RNA from four muscle samples, 75- and 105-d fetal hind limb muscles, and 1- and 7-wk postnatal semitendinosus muscle was used to make radiolabeled targets for duplicate hybridization to the macroarray membranes in an initial screen for expression. All EST that gave clear signals (n = 238) were then re-arrayed, and hybridization was conducted with additional biological replication of samples in the 75-d and 1-wk ages. Signal intensity for each gene was normalized to signal intensity measured at control spots on each membrane, which consisted of total cDNA from liver, lung, spleen, and skeletal muscle. Both normalized ratio levels and a mixed linear model analyses were used to identify genes differentially expressed among the muscle samples. Results showed 28 genes had differences in expression level greater than twofold between the 75-d fetal and 1-wk muscle RNA samples. All 28 genes were also identified as genes with significantly different (P < 0.01) expression using a mixed linear model analysis. Nineteen of these 28 genes had significant matches (basic local alignment search tool [BLAST] score > 100; P < 0.01) to known genes, two matched genes encoding human hypothetical proteins, and seven had no significant matches to Genbank nonredundant and dbEST (database of expressed sequence tags) entries. These results were confirmed for representative genes with RNA blot analysis of seven developmental time points, including RNA from the same muscle samples tested previously in the macroarray. The RNA blot results confirmed the macroarray results for all selected genes, demonstrating that the macroarray technique used in this study is accurate and reproducible. An unknown muscle clone (M218) with a slightly less than twofold increase in expression from the 75-d to the 1-wk age (1 wk/75 d = 1.94; P = 0.0114) was also shown to differ between these two ages using RNA blot analysis, demonstrating the methods used to identify differentially expressed genes may be conservative. The association between expression patterns of vimentin and desmin was also investigated. Results indicate the switch in intermediate filament protein from vimentin to desmin occurs primarily at the level of transcription and/or RNA processing.

Processing and representation of meta-data for sleep apnea diagnosis with an artificial intelligence approach.

In this article, we revise and try to resolve some of the problems inherent in questionnaire screening of sleep apnea cases and apnea diagnosis based on attributes which are relevant and reliable. We present a way of learning information about the relevance of the data, comparing this with the definition of the information by the medical expert. We generate a predictive data model using a data aggregation operator which takes relevance and reliability information about the data into account to produce a diagnosis for each case. We also introduce a grade of membership for each question response which allows the patient to indicate a level of confidence or doubt in their own judgement. The method is tested with data collected from patients in a Sleep Clinic using questionnaires specially designed for the study. Other artificial intelligence predictive modeling algorithms are also tested on the same data and their predictive accuracy compared to that of the aggregation operator.

The effect of selection for residual feed intake on general behavioral activity and the occurrence of lesions in Yorkshire gilts.

The objectives of this study were to determine the effect of selection for improved residual feed intake on behavior, activity, and lesion scores in gilts in their home pen. A total of 192 gilts were used, 96 from a line that had been selected for decreased residual feed intake over 5 generations (LRFI) and 96 from a randomly bred control line. Gilts were housed in 12 pens (16 gilts/pen; 0.82 m(2)/gilt) containing 8 gilts from each line in a conventional grow-finish unit. Twelve hours of video footage were collected on the day of placement and then every 4 wk for 3 more observational periods. Video was scored using a 10-min instantaneous scan sampling technique for 4 postures (standing, lying, sitting, and locomotion) and 1 behavior (at drinker). Categories of active (standing, locomotion, and at drinker) and inactive (sitting and lying) were also created. Lesion scores were collected 24 h after behavior collection had begun. The body of a gilt was divided into 4 regions, with each region receiving a score of 0 (0 lesions) to 3 (5+ lesions). All statistical analyses used Proc Mixed of SAS. Data were analyzed separately for the day of placement and the subsequent 3 rounds. General activity was summarized on a percentage basis by each posture and behavior and subjected to an arcsine square root transformation to normalize data and stabilize variance. Analysis was performed on each behavior and posture. Lesion scores for each region of the body were analyzed as repeated measures. There were no differences (P > 0.05) between genetic lines for all postures and the behavior at drinker on the day of placement. However, over subsequent rounds it was observed that LRFI gilts spent less (P = 0.03) time standing, more time sitting (P = 0.05), and were less active (P = 0.03) overall. Gilts from the LRFI line had decreased (P < 0.045) lesion scores on the day after placement. However, over subsequent rounds there were no (P > 0.05) differences between the genetic lines. In conclusion, on the day of placement there were no postural, behavior, or general activity differences between genetic lines, but LRFI gilts had decreased lesion scores. Behavioral differences were observed between genetic lines over subsequent rounds, with LRFI gilts becoming less active, but there were no differences in lesion scores.

Effects of ad libitum and restricted feed intake on growth performance and body composition of Yorkshire pigs selected for reduced residual feed intake.

Residual feed intake (RFI), defined as the difference in the observed and expected feed intake while accounting for growth and backfat, has gained much attention, but little is known about why pigs selected for reduced RFI are more efficient. To this end, a line of Yorkshire pigs selected for reduced RFI was developed. The objective of this study was to evaluate the 5th generation of this select line against a randomly selected control line for performance, carcass and chemical carcass composition, and overall efficiency toward the later part of the growth phase. Eighty barrows, 40 from each line, were paired by age (~132 d, P < 0.60) and BW (74.8 ± 9.9 kg, P < 0.49) and randomly assigned to 1 of 4 feeding treatments in 10 replicates: 1) ad libitum, 2) 75% of ad libitum, 55% of ad libitum, and BW stasis, with weekly adjustments in intake to keep BW constant for each pig. Pigs were individually penned (group housing was used for selection) and on treatment for 6 wk. Initial BW did not differ between the lines (P < 0.49). The ad libitum select pigs consumed 10% less feed (P < 0.09) than the ad libitum control with no significant difference in BW (P < 0.80) and slight differences in carcass fat composition (P < 0.20) and backfat (P < 0.11), which resulted in significantly less carcass energy (P < 0.03). Under restricted feeding, the select line had an increase in BW (P = 0.10) while consuming the same ration of feed as the control line with no significant difference in chemical carcass composition and lighter visceral weights, which was significant for the 75% of ad libitum treatment (P < 0.01). Under BW stasis feeding the select line consumed 7.6% less feed overall (P = 0.21) and 18% less feed at the end of the 6 wk (P < 0.08), to maintain static BW with no significant difference in chemical carcass composition compared with the control line. Overall, the select line had lighter visceral weight (P < 0.02) and a greater dressing percentage (P < 0.03) compared with the control line. Using regression, the select line had reduced energy retention (P < 0.04) and feed energy utilization (P < 0.34); however, the select line appeared to have reduced maintenance requirements (P < 0.13). In conclusion, selection for reduced RFI decreases feed intake with no significant difference (P > 0.05) in growth performance, reduced backfat, increased dressing percentage, and reduced maintenance requirements. All of these traits are appealing to the producer and result in increased profits in the production setting.

Effects of ad libitum and restricted feeding on early production performance and body composition of Yorkshire pigs selected for reduced residual feed intake.

Residual feed intake (RFI), defined as the difference between observed and expected feed intake based on growth and backfat, has been used to investigate genetic variation in feed efficiency in cattle, poultry and pigs. However, little is known about the biological basis of differences in RFI in pigs. To this end, the objective of this study was to evaluate the fifth generation of a line of pigs selected for reduced RFI against a randomly selected Control line for performance, carcass and chemical carcass composition and overall efficiency. Here, emphasis was on the early grower phase. A total of 100 barrows, 50 from each line, were paired by age and weight (22.6 ± 3.9 kg) and randomly assigned to one of four feeding treatments in 11 replicates: ad libitum (Ad), 75% of Ad (Ad75), 55% of Ad (Ad55) and weight stasis (WS), which involved weekly adjustments in intake to keep body weight (BW) constant for each pig. Pigs were individually penned (group housing was used for selection) and were on treatment for 6 weeks. Initial BW did not significantly differ between the lines (P > 0.17). Under Ad feeding, the low RFI pigs consumed 8% less feed compared with Control line pigs (P < 0.06), had less carcass fat (P < 0.05), but with no significant difference in growth rate (P > 0.85). Under restricted feeding, low RFI pigs under the Ad75 treatment had a greater rate of gain while consuming the same amount of feed as Control pigs. Despite the greater gain, no significant line differences in carcass composition or carcass traits were observed. For the WS treatment, low RFI pigs had similar BW (P > 0.37) with no significant difference in feed consumption (P > 0.32). Overall, selection for reduced RFI has decreased feed intake, with limited differences in growth rate but reduced carcass fat, as seen under Ad feeding. Collectively, results indicate that the effects of selection for low RFI are evident during the early grower stage, which allows for greater savings to the producer.