Intramammary lipopolysaccharide challenge in early versus mid-lactation dairy cattle: immune, production, and metabolic responses.

Study objectives were to compare the immune response, metabolism and production following intrama-mmary lipopolysaccharide ( IMM LPS ) administration in early and mid-lactation cows. Early ( E-LPS ; n = 11; 20 ± 4 d in milk [ DIM ]) and mid-( M-LPS ; n = 10; 155 ± 40 DIM) lactation cows were enrolled in an experiment consisting of 2 periods ( P ). During P1 (5 d) cows were fed ad libitum and baseline data were collected, including liver and muscle biopsies. At the beginning of P2 (3 d) cows received 10 mL sterile saline containing 10 µg of LPS from Escherichia coli O111: B4/ mL into the left rear quarter of the mammary gland, and liver and muscle biopsies were collected at 12 h post-LPS. Tissues were analyzed for metabolic flexibility, which measures substrate switching capacity from pyruvic acid to palmitic acid oxidation. Data were analyzed with the MIXED procedure in SAS 9.4. Rectal temperature was assessed hourly for the first 12 h post-LPS and every 6 h thereafter for the remainder of P2. All cows developed a febrile response following LPS, but E-LPS had a more intense fever than M-LPS cows (0.7°C at 5 h after LPS). Blood samples were collected at 0, 3, 6, 9, 12, 24, 36, 48, and 72 h post-LPS for analysis of systemic inflammation and metabolism parameters. Total serum Ca decreased after LPS (26% at 6 h nadir) but did not differ by lactation stage ( LS ). Circulating neutrophils decreased, then increased post-LPS in both LS, but E-LPS had exaggerated neutro-philia (56% from 12 to 48 h) compared with M-LPS. Haptoglobin increased after LPS (15-fold) but did not differ by LS. Many circulating cytokines were increased post-LPS, and IL-6, IL-10, TNF-α, MCP-1, and IP-10 were further augmented in E-LPS compared with M-LPS cows. Relative to P1, all cows had reduced milk yield (26%) and dry matter intake ( DMI; 14%) on d 1 that did not differ by lactation stage ( LS ). Somatic cell score increased rapidly in response to LPS regardless of LS and gradually decreased from 18 h onwards. Milk component yields decreased after LPS. However, E-LPS had increased fat (11%) and tended to have increased lactose (8%) yield compared with M-LPS cows throughout P2. Circulating glucose was not affected by LPS. Nonesterified fatty acids ( NEFA ) decreased in E-LPS (29%) but not M-LPS cows. β-hydroxybutyrate ( BHB ) slightly increased (14%) over time post-LPS regardless of LS. Insulin increased after LPS in all cows, but E-LPS had blunted hyperinsulinemia (52%) compared with M-LPS cows. Blood urea nitrogen ( BUN ) increased after LPS and the relative change in BUN was elevated in E-LPS cows compared with M-LPS cows (36 and 13%, respectively, from 9 to 24 h). During P1, metabolic flexibility was increased in liver and muscle in early lactating cows compared with mid-lactation cows, but 12 h post-LPS, metabolic flexibility was reduced and did not differ by LS. In conclusion, IMM LPS caused severe immune activation and E-LPS cows had a more intense inflammatory response compared with M-LPS cows, but the effects on milk synthesis was similar between LS. Some parameters of the E-LPS metabolic profile suggest continuation of metabolic adjustments associated with early lactation to support both a robust immune system and milk synthesis


INTRODUCTION
Early lactation (EL) cows have a more intense immune response to i.v.LPS than mid-lactation cows (ML) and this is primarily characterized by higher fever, acute phase proteins (APP), cytokines, neutrophilia, and a more severe anorexia response compared with mid-lactation (ML) cows (Opgenorth et al.,202X,202Y).The exacerbated cytokine response in EL cows corroborates ex vivo models of whole blood and isolated monocyte stimulation (Sordillo et al., 1995;Jahan et al., 2015).Despite the aforementioned, the absolute milk yield response to LPS in EL is similar to what occurs during established lactation; an effect likely made possible by marked metabolic adjustments the EL cows employ to prioritize milk synthesis (Opgenorth et al.,202Y).In agreement, other investigators observe a comparable or even ameliorated milk yield response in EL cows in response to stress or immune activation (Perera et al., 1986;Lehtolainen et al., 2003).Thus, it appears the EL mammary gland is more refractory (from a milk synthesis perspective) to intense immune activation than during established lactation.Our objectives were to substantiate this concept with a model that more closely resembles a natural infection via intramammary (IMM) LPS administration.We hypothesized the EL immune system would be more reactive toward IMM LPS, but that milk yield would not differ by lactation stage (LS).
We further aimed to evaluate both systemic changes in nutrient partitioning and skeletal muscle and hepatic intracellular bioenergetics between LS.Cows in EL are more metabolically flexible characterized by their peripheral tissue's affinity for oxidizing lipid derived fuels to spare glucose for the mammary gland (Bauman and Currie, 1980;Ha et al., 2017).Further, the metabolite and endocrine adjustments EL cows enlist during i.v.LPS indicate enhanced metabolic flexibility compared with later lactation (Opgenorth et al.,202Y).Thus, we wanted to directly evaluate hepatic and skeletal muscle metabolic flexibility (increased preference for FA oxidation instead of pyruvate oxidation) both before and after immune activation.

Experimental Design
All procedures were approved by the Iowa State University Animal Care and Use Committee.Twenty one clinically healthy, non-pregnant, multiparous EL (denoted as E-LPS; 20 ± 4 DIM; 50 ± 9 kg daily milk yield; 2.7 ± 0.9 lactations; n = 11) and ML (denoted as M-LPS; 155 ± 40 DIM; 48 ± 7 kg daily milk yield; 2.8 ± 0.8 lactations; n = 10) Holstein cows were moved to individual box-stalls (4.57× 4.57 m) at the Iowa State University Dairy Farm (Ames, IA) before the beginning of the experiment that was conducted in 2 replicates.Regardless of lactation stage (LS), all cows were fed a diet formulated to meet or exceed the predicted requirements (NRC, 2001; Table 1) of energy, protein, minerals, and vitamins for EL cows.Cows were given 4 d to acclimate, and a jugular catheter was implanted.
Immediately before the first milking of acclimation, cows were evaluated for mastitis with the California Mastitis Test (Immucell, Portland, ME).Period (P) 1 lasted 5 d during which cows were fed ad libitum and baseline data were obtained.On the last day of P1, milk collected from rear quarters were evaluated for mastitis with rapid SCC test strips and confirmed to contain < 250,000 cells per mL (PortaCheck Por-taSCC, Moorestown, NJ).At the initiation of P2 (3 d), cows were administered an LPS solution in the left rear quarter of the mammary gland immediately following the last milking of P1 at 0600.Lipopolysaccharide (Escherichia coli O111:B4; Sigma Aldrich, St. Louis, MO) was dissolved in sterile saline at a concentration of 10 µg/mL and passed through a 0.2-µm sterile syringe filter (Thermo Scientific; Waltham, MA).The complete volume of LPS solution administered was 10 mL such that each cow received a total of 100 µg LPS, which was chosen because it had been administered in previous IMM LPS models and successfully induced immune activation in both EL and ML cows (Lehtolainen et al., 2003;Waldron et al., 2006).After aseptically preparing the teat and orifice with 70% alcohol, the LPS solution was administered with a sterile syringe fitted with a teat cannula (Jorgensen Laboratories, Inc., Loveland, CO).The LPS solution was distributed into the gland by massaging the quarter for 30 s after administration.Immediately following LPS infusion, the infused teat end was dipped in iodine.
Throughout the experiment, cows were milked 4 × daily at 0600, 1200, 1800, and 0000 h and yield was recorded.Composite milk samples to evaluate milk components and SCS were collected at the 0600 and 1800 milking during P1 and at every milking during P2.Samples were stored at 4°C with a preservative (bronopol tablet; D & F Control System, San Ramon, CA) until analysis (CentralStar; Kaukauna, WI) using infrared analysis equipment and procedures.Blood samples were collected from the jugular catheter at the beginning of each day during P1 and at 0, 3, 6, 9, 12, 24, 36, 48, and

Tissue Biopsies and Metabolic Flexibility Analysis
Liver and muscle biopsies were collected from all animals at the beginning of P1 d 1 and at 12 h post-LPS.Liver collection followed procedures previously described (Horst et al., 2020).Briefly, biopsy sites were locally anesthetized with 10 mL 2% lidocaine.Following site anesthesia, an incision of approximately 1 cm at liver and 2 cm at longissimus dorsi muscle biopsy sites were made to facilitate biopsy tool insertion.A sterile 2.1 mm x 15 cm hepatic semi-automatic biopsy needle (Argon Medical Devices Inc., Athens, TX) was used to collect liver tissue.One muscle core was collected with a sterile biopsy punch (8 mm, Miltex Inc., York, PA; utilized previously; Horst et al., 2019a).Biopsy sites were wiped clean with 70% ethanol before closing with either absorbable suture or skin stapler (3M, Saint Paul, MN).Tissue samples were immediately rinsed with saline before placing in 500 µL of buffer (0.25 M sucrose, 1 mM EDTA, 0.01 M Tris-HCl, and 2 mM ATP, pH = 7.4), flash frozen, and stored at −80°C until analysis.
Tissue samples were shipped on dry ice to Virginia Polytechnic Institute and State University's Integrated Life Sciences Facility to evaluate metabolic flexibility as previously described (Frisard et al., 2010;Zhao et al., 2018).Briefly, samples were homogenized with a Polytron homogenizer and Teflon pestle on glass (295 rpm for ~30 s).Pyruvate oxidation in the tissue homogenates was assessed with and without 100 µM palmitic fatty acid (FA) to measure metabolic flexibility.The larger the difference between pyruvate oxidation with and without palmitic acid indicated increased preference to switch substrate preference toward palmitic acid, which suggests increased metabolic flexibility.Pyruvate oxidation was measured by 14 CO 2 concentrations produced from 1-14 C pyruvate oxidation by pyruvate dehydrogenase complex.Two working buffers were prepared to measure pyruvate oxidation with and without palmitic acid and included, to yield a 10 mL volume: (1) 100 µM palmitic acid (only added to combined pyruvate and palmitic acid buffer), (2) 5 mM pyruvic acid, (3) 0.35 µCi/mL 1-14 C pyruvic acid, (4) 0.5% BSA, (5) 4.13 mL distilled H 2 O, and (6) 5 mL of previously prepared reaction media composed of 125 mM sucrose, 25 mM KH 2 PO 4 , 200 mM KCl, 2.5 mM MgCl, 2.5 mM L-Carnitine, 0.25 mM malic acid, 20 mM Tris-HCl, 2.5 mM dithiothreitol, 0.25 mM NAD+, 4 mM ATP, and 0.125 mM coenzyme A. The reaction media was heated to 37°C, set at a pH of 7.4, and stored at −80°C until use.Ingredients 1-4 were combined first and allowed to incubate in a 37°C water bath for 30 min to allow palmitic acid to bind with BSA.Afterward, ingredients 4 and 5 were added and the solution was vortexed to mix.320 µL of the working buffer was added to a sealed well containing 80 µL of tissue homogenates in duplicate to initiate pyruvate oxidation.The reaction incubated for 1 h at 37°C and 120 rpm.Following incubation, 200 µL of 45% perchloric acid was added to each reaction well to dissociate 14 CO 2 from the reaction media into a gas.Samples were again incubated for 1 h at 37°C at 120 rpm where gaseous 14 CO 2 was absorbed by 400 µL NaOH.After 1 h of trapping 14 CO 2, the NaOH was transferred to a scintillation vial with 5 mL EcoLite scintillation fluid (MP Biomedicals, Santa Ana, CA) and counted for 1 min per sample after settling overnight.Sample 14 CO 2 production was adjusted for total protein measured via the bicinichoninic acid colorimetric assay (Thermo Scientific Pierce BCA Protein Assay Kit; Rockford, IL) and the specific activity of a blank (100 µL working buffer with 5 mL scintillation fluid).

Statistical Analysis
Data were analyzed with PROC MIXED in SAS 9.4 (SAS Institute Inc., Cary, NC).Fixed effects of LS (E-LPS versus M-LPS), time, their interactions, and replicate were analyzed with a spatial power covariance structure for circulating rectal temperature, respiration rate, metabolic parameters, and cell populations and an autoregressive covariance structure for circulating Hp, cytokines, and production parameters.Time relative to LPS administration served as a repeated measure with cow as subject.A logarithmic transformation was performed for IL-6, IL-10, and MCP-1 analysis.All data from P1 and P2 were analyzed separately from each other except for liver and skeletal muscle metabolic flexibility analyses which included period as a repeated measure with cow as subject.Dry matter intake, BUN, and metabolic flexibility in liver and skeletal muscle were additionally analyzed by comparing the magnitude of change (%) from P1 baseline.Data are reported as least squares means ± standard error of the mean and considered significant if P ≤ 0.05 and a tendency if 0.05 < P ≤ 0.10.Logistical constraints and prior research (Lehtolainen et al., 2003) were con-sidered for determining sample size.Post-hoc power analysis (PROC POWER; SAS Institute Inc., Cary, NC) utilizing a primary parameter of interest (rectal temperature) indicated a statistical power of > 99% (α = 0.05) when comparing LS.The experiment originally included 12 cows per LS, but 3 cows developed subclinical mastitis during P1 (in a front quarter), and their data were removed from all analyses (such that E-LPS n = 11 and M-LPS n = 10).

RESULTS
Rectal temperature substantially increased post-LPS in all cows (peaking at 2.9 and 2.4°C above baseline at 5 and 6 h in E-LPS and M-LPS cows, respectively), but E-LPS developed a more intense febrile response than M-LPS cows (0.6°C from 2 to 6 h, respectively; P < 0.01; Figure 1A).Additionally, LPS increased respiration rate, which was primarily observed in E-LPS cows (46% relative to M-LPS at 6 h peak; P < 0.01; data not shown).Circulating Hp was markedly increased post-LPS administration (15-fold) but did not differ by LS (Figure 1B).After LPS administration, the number of circulating neutrophils decreased at 6 h and increased thereafter to reach peak neutrophilia at 24 h (Figure 1C).From 12 to 48 h, E-LPS had increased neutrophilia compared with M-LPS cows (56%; P < 0.01).Serum tCa was decreased in response to LPS regardless of LS and reached a nadir of 2.1 mmol/L at 6 h (P < 0.01; Figure 1D) and progressively increased with time.Circulating IL-6, IL-10, TNF-α, and IP-10 were augmented post-LPS and were further increased or tended to be increased in E-LPS relative to M-LPS cows (21, 3, 91, and 38%; P < 0.01, P = 0.06, P < 0.01, and P = 0.06, respectively; Figures 2A-D); however, E-LPS cows had elevated IL-6 before LPS (3.4-fold; P < 0.01).Monocyte chemoattractant protein-1 increased post-LPS and was augmented at 3 h in E-LPS compared with M-LPS cows (7%; P = 0.02; Figure 2E).Circulating MIP-1α increased post-LPS but did not differ by LS (Figure 2F).Concentrations of IFN-γ, MIP-1β, VEGFA increased and IL-36RA decreased post-LPS and did not differ by LS (Supplementary Figure 1A-D).
Cows in EL had decreased DMI relative to ML cows during P1 (19%; P < 0.01; Figure 3A).Post-LPS, the relative change in DMI decreased (14% on d 1) but did not differ by LS (Figure 3B).Following IMM LPS, milk yield was markedly reduced (37% from P1; Figure 3C) from 12 to 18 h in all cows and it gradually increased with time but did not differ by LS.However, post hoc analysis revealed E-LPS did not have as severe of a reduction in milk yield at 6 h post-LPS (26%; P = 0.05; Figure 3C).Somatic cell score substantially increased post-LPS (4.8-fold at 18 h relative to P1; Figure 3D)  2. During P1, EL cows had increased milk fat and protein content, and had increased ECM compared with M-LPS cows.Milk component yields decreased post-LPS, but E-LPS cows continued to have increased fat yield and tended to have increased lactose production compared with M-LPS cows (11 and 8%; P < 0.01 and P = 0.10, respectively).Further, ECM was decreased post-LPS but remained increased in E-LPS compared with M-LPS cows (P < 0.01).
Glucose concentrations were not decreased post-LPS and reflected pre-existing P1 differences by LS (13% less glucose in E-LPS compared with M-LPS cows; P < 0.01; Figure 4A).Cows became hyperinsulinemic post-LPS but E-LPS cows had blunted insulin response relative to M-LPS cows (52%; P < 0.01; Figure 4B).Concentrations of BUN increased during P2 (13%) and did not differ by LS (Figure 4C).However, the relative change in BUN was increased in E-LPS compared with M-LPS cows from 9 to 24 h (36 relative to 13%, respectively; P = 0.02; Figure 4D).Post-LPS, NEFA concentrations decreased in E-LPS (29%) but not in M-LPS cows; however, E-LPS NEFA remained elevated (3.8-fold; P < 0.01; Figure 4E) compared with M-LPS cows.Circulating BHB increased slightly post-LPS throughout P2 (14%) but was unaffected by LS (Figure 4F).Metabolic flexibility data are presented in Fig-  ures 5A-D.During P1, liver metabolic flexibility was increased in early compared with mid-lactation cows (post hoc analysis revealed a 25% increase in early vs. mid-lactation cows, respectively; P = 0.01; Figure 5A).Before IMM LPS administration, skeletal muscle from early lactating cows had increased metabolic flexibility compared with mid-lactation (41% increase in early vs. mid-lactation cows, respectively; P < 0.01; Figure 5B).Post-LPS, both LS had reduced liver and skeletal muscle metabolic flexibility, but the net change in flexibility did not differ by LS (Figure 5C-D).

DISCUSSION
Cows in EL have a more intensified immune and hypophagic response to i.v.LPS compared with ML cows; however, EL cows maintain similar milk yield compared with ML cows and elicit increased metabolic adjustments that suggest the EL mammary gland's milk synthesizing machinery is more refractory toward the effects of LPS (Opgenorth et al.,202X,202Y).The concept that the EL mammary gland prioritizes milk synthesis amidst increased systemic immune activation and inappetence has implications to our understanding of transition cow biology.Reasons for conducting the current experiment were 3-fold: (1) confirm results of our i.v.LPS model with an approach that more closely resembles a natural route of pathogen exposure (i.e., IMM), (2) assess how an activated immune system coordinates systemic metabolism, and (3) evaluate the metabolic flexibility of liver and skeletal muscle both before and after immune activation to provide an appreciation of cellular bioenergetics.
As expected, all cows developed pyrexia in response to IMM LPS, indicating immune activation was successfully achieved.Further, E-LPS cows had an elevated febrile response and respiration rate relative to M-LPS cows and this agrees with both an Escherichia coli mastitis model and our i.v.LPS model (Shuster et al., 1996;Opgenorth et al., 202X).Fever augments immune system functions (Wrotek et al., 2021) but requires increased metabolic rate and energy expenditure (Kluger, 1979).Although just speculation, the ostensibly increased demand for oxygen may have caused tachypnea, which was predominantly observed in E-LPS cows, an observation corroborating the i.v.LPS model (Opgenorth et al.,202X).Regardless of why, the increased fever and respiration rate in E-LPS cows suggests they had a more robust immune activation and that their response may have been more energetically costly than M-LPS cows.
Fever is also associated with (or caused by) elevated circulating cytokines that are released in response to leukocyte activation (Conti et al., 2004).Likewise, several cytokines increased post-LPS herein and this pattern agrees with both an i.v.LPS and a mastitis model (Shuster et al., 1996;Opgenorth et al., 202X).Circulating IL-6, TNF-α, IL-10, IP-10, and MCP-1 were further augmented in E-LPS relative to M-LPS cows and this corroborates our i.v.LPS model (Opgenorth et al.,202X).Marked increases in these cytokines in E-LPS compared with M-LPS cows suggest enhanced leukocyte activation (Moore et al., 2003;Deshmane et al., 2009).Both IP-10 and MCP-1 are chemoattractant cytokines that direct leukocytes to inflammation sites (Deshmane et al., 2009;Liu et al., 2011), whereas TNF-α and IL-6 upregulate the inflammatory response through mediating a variety of effects such as fever and APP synthesis (Castell et al., 1989;Netea et al., 2000).While TNF-α and IL-6 promote immune activation, IL-10 has more anti-inflammatory functions to regulate the immune response and protect the host from tissue damage (Saraiva and O'Garra, 2010).Increased IL-10 in E-LPS cows agrees with a numeric increase in i.v.LPS (P = 0.11; Opgenorth et al.,202X), and may indicate E-LPS cows made efforts to control its augmented immune system.Most cytokines are produced by macrophages and lymphocytes (Arango Duque and Descoteaux, 2014), and one explanation for why E-LPS cows had an increased cytokine response could be a result of an expanded mononuclear cell fraction in EL mammary lymph nodes (Sordillo et al., 1995).Interestingly, EL cows had increased IL-6 before LPS administration, but no other LS differences were detected pre-LPS in other inflammatory biomarkers.This suggests the inflammatory tone before LPS was mostly similar between LS and that exaggerated post-LPS inflammation is a direct result of LS differences in immunogenicity.Overall, increased circulating cytokines in E-LPS cows suggest the immune system is more reactive to IMM LPS than in later lactation.
Haptoglobin increased after IMM LPS in all cows and this agrees with other mastitis and i.v.LPS models (Brandão et al., 2016;Chandler et al., 2022).Though EL cows had elevated Hp in response to i.v.LPS compared with ML cows (Opgenorth et al.,202X), herein Hp did not differ by LS, despite E-LPS cows having increased circulating cytokines and a more intense febrile response.Reasons for the inconsistent effects between experiments are not obvious, but during experimental   mastitis, Hp concentrations are not associated with infection severity (Hirvonen et al., 1999), and this may help explain the LS differential Hp responses.
The number of circulating neutrophils decreased in response to IMM LPS, which was followed by a neutrophilic phase before returning to baseline counts.This pattern corroborates other LPS models (Horst et al., 2019b;Chandler et al., 2022).Although the neutropenic response did not differ by LS, E-LPS cows had an earlier, more pronounced, and prolonged neutrophilic phase relative to M-LPS cows and this agrees with previous models (Lehtolainen et al., 2003;Opgenorth et al., 202X), and indicates increased neutrophil release from bone marrow (Stockham and Scott, 2008).The exaggerated neutrophilia is additional evidence suggesting that E-LPS cows had a more intense immune activation than M-LPS cows.
Neutropenia during mastitis coincides with increased SCC in milk (Paape et al., 1974).Likewise, SCS markedly increased post-LPS herein and this agrees with a rise in SCC in previous IMM LPS models (Waldron et al., 2006;Gross et al., 2018), but LS did not influence the SCS response.Some case studies report impaired leukocyte recruitment to the mammary gland during infection in EL cows (Hill et al., 1979;Frost and Brooker, 1986), and several have suggested neutrophil migration is compromised during the transition period (Lee and Kehrli, 1998;Monfardini et al., 2002).However, some evidence indicates EL cows have normal or even improved leukocyte influx into the mammary gland relative to ML cows in mastitis models (Shuster et al., 1996;Lehtolainen et al., 2003), and our results corroborate this.Reasons for the inconsistencies might be due in part to individual animal variation in peri- parturient health status (Galvão et al., 2011).During infection, circulating neutrophils migrate to inflamed sites and are replaced by immature neutrophils released from bone marrow (Stockham and Scott, 2008).Immature neutrophils are less efficient at exerting effector functions, including migration (Pillay et al., 2010;Leliefeld et al., 2016).Therefore, variations in circulating neutrophil maturity could influence how ex vivo neutrophil functions and periparturient immune status are interpreted (Horst et al., 2021).In summary, results herein imply IMM LPS increased SCS in all cows and leukocyte migration toward the mammary gland was seemingly not impaired in EL.
Serum Ca is another barometer of immune activation because hypocalcemia is a species conserved metabolic alteration caused by LPS (Carlstedt et al., 2000;Chandler et al., 2023).Cows in both LS had reduced tCa concentrations post-LPS and this pattern agrees with prior reports (Chandler et al., 2022(Chandler et al., , 2023)).Decreasing the Ca pool is thought to be a protective strategy since maintaining eucalcemia stabilizes LPS aggregates, which is more toxic than LPS monomers (Munford et al., 1981;Mueller et al., 2004).We have previously demonstrated i.v.LPS causes a more severe hypocalcemia (measured as iCa) in transition cows than ML cows (Opgenorth et al.,202X).While tCa and iCa are typically highly correlated, during mastitis or parturition, this association weakens, and iCa becomes a more precise measure of hypocalcemia (Leno et al., 2017;Hisaeda et al., 2020).Thus, it is possible that because we measured tCa we did not accurately capture the potential LS differences.Regardless, several changes in markers of immune activation (i.e., rectal temperature, cytokines, and neutrophilia) were exaggerated in E-LPS cows; modifications reflective of a more engaged inflammatory response toward IMM LPS than M-LPS cows.
Feed intake was decreased in response to IMM LPS and this agrees with other mastitis models (Waldron et al., 2006;Swartz et al., 2023).The hypophagic response was unaffected by LS and DMI recovered by d 2 post-LPS, indicating that E-LPS and M-LPS appetites were similarly influenced by IMM immune activation.In contrast, EL cows were much more anorexic following i.v.LPS than ML cows (Opgenorth et al.,202Y).Differences in how i.v.versus IMM LPS administration influence feed intake are unclear, although i.v.LPS perturbs ruminal contractions more than IMM LPS (Lohuis et al., 1988) and LPS reduces alimentary track motility (Turner and Berry, 1963).Lipopolysaccharide causes anorexia, at least in part, by directly interacting with the central nervous system (Wisse et al., 2007) as the hypothalamus contains TLR-4 receptors (Iwasa et al., 2015).Clearing circulating LPS is mediated partly by lipoproteins (Weinstock et al., 1992), and some lipoprotein components are reduced in EL (Raphael et al., 1973;Takahashi et al., 2003).Following i.v.LPS, clearance in circulation might be delayed in EL due to differences in lipoprotein compounds, and this could increase the opportunity for LPS to influence appetite.Presumably, negligible amounts of IMM LPS crosses from the mammary gland lumen into circulation (Lohuis et al., 1988;Dosogne et al., 2002).Thus, i.v.LPS may directly affect appetite whereas IMM LPS may influence appetite indirectly and to a lesser extent via cytokines.Regardless, in the current study the DMI response was mostly similar between LS despite a more intensely activated immune system in EL.
Milk yield post-IMM LPS was likewise decreased but not different by LS, which agrees with our previous results after i.v.LPS (Opgenorth et al.,202Y).In fact, the E-LPS cows herein had a less severe decrease in milk yield at 6 h compared with M-LPS cows.This is perplexing especially considering the higher fever and exaggerated cytokine response at 6 h in E-LPS cows, and this could indicate the EL mammary gland milk synthesizing machinery is less responsive to IMM LPS.This general concept agrees with prior research indicating that EL cows recover quicker than ML cows after IMM LPS (Lehtolainen et al., 2003).Further, IMM LPS reduced lactose yields, but E-LPS cows tended to produce more lactose compared with M-LPS cows.This is notable because glucose transporters are internalized in mammary epithelial cells (MEC) upon LPS stimulation (Kobayashi et al., 2013) and glucose is the primary fuel utilized by an activated immune system (Warburg 1927).Since only cumulative milk from all glands was collected, it is unclear if an attenuated lactose deficit in E-LPS cows is due to less perturbed MEC in the LPS-infused gland or if uninfused glands were less susceptible to neighboring inflammation.Both infused and neighboring glands are negatively influenced by LPS (Shuster et al., 1996), but it would be of interest to understand how LS might influence differences in milk yield between infected and uninfected glands during mastitis.In summary, the mammary gland in E-LPS cows maintained similar milk yield relative to M-LPS cows and tended to produce more lactose during a heightened immune activation, which might suggest a more refractory mammary gland (from an anabolic viewpoint) in EL to LPS or its ensuing pathology.
In healthy cows, systemic nutrient utilization and metabolism are coordinated to support lactogenesis and galactopoiesis (Bauman and Currie, 1980).Therefore, it was of interest to understand metabolic adjustments reflecting the initial response and recovery from immune activation.As expected (Bell and Bauman, 1997) there were basal LS differences in circulating glucose (decreased in EL).Post-IMM LPS, glucose concentrations fluctuated over time but neither LS developed hypoglycemia as observed in some i.v.LPS challenges (Kvidera et al., 2017;Horst et al., 2020), and the relatively unperturbed glucose response to IMM LPS agrees with other mastitis models (Waldron et al., 2006;Gross et al., 2020).Overall, regardless of LS, cows managed to mostly maintain euglycemia during IMM LPS and this suggests that the route of LPS administration influences circulating carbohydrate homeostasis.
Before LPS, EL cows had decreased insulin and this is a fundamental endocrine strategy that healthy cows utilize to homeorhetically partition nutrients toward the mammary gland (Bauman and Currie, 1980;Bell and Bauman, 1997).Despite a lack of meaningful glucose changes, insulin acutely increased in response to IMM LPS, which corroborates other IMM models (Waldron et al., 2006;Gross et al., 2018Gross et al., , 2020) ) and i.v.LPS models (Kvidera et al., 2017;Horst et al., 2019b).Importantly, insulin enhances glucose uptake by activated leukocytes to supply the demand for increased substrates (Lang and Dobrescu, 1991).However, E-LPS cows had blunted hyperinsulinemia relative to M-LPS cows, and this corresponds with the homeorhetic drive in EL to spare glucose (Bell and Bauman, 1997) and agrees with our i.v.LPS model (Opgenorth et al.,202Y).Insulin also has anti-inflammatory functions (Chang et al., 2021), and reduced insulin in EL cows could additionally evince augmented inflammation.Results suggest that amid intensified immune activation, EL cows continue to prefer a hypoinsulinemic state, which enables glucose uptake for a magnified immune response and milk synthesis.
Concentrations of BUN increased over time after IMM LPS and returned to baseline as P2 progressed and this pattern agrees with others (Horst et al., 2020;Opgenorth et al., 202Y).During immune activation, AA are mobilized from skeletal muscle to provide precursors for gluconeogenesis and APP synthesis (Iseri and Klasing, 2014).However, the AA profile coming from muscle is different to that of APP (Reeds et al., 1994) and thus unneeded AA are deaminated (similar to that of the gluconeogenic AA), and the amino groups enter into the urea cycle (Horst et al., 2019a).Further, the net increase in BUN from baseline was aggravated in E-LPS relative to M-LPS cows from 9 to 24 h and suggests enhanced skeletal muscle proteolysis.The inflated increase in BUN in E-LPS cows likely stems from reduced insulin sensitivity in skeletal muscle which would allow for enhanced AA mobilization; a metabolic scenario that presumably allows EL cows to meet the energetic and nitrogen needs of both an excessively activated immune system and milk production.
As anticipated, EL cows had elevated NEFA concentrations relative to ML cows before LPS, which reflects increased reliance on FA oxidation in EL peripheral tissues.Nonesterified FA decreased post-LPS and this corroborates other IMM LPS models (Pires et al., 2019;Swartz et al., 2023) and is ostensibly caused by insulin's antilipolytic action (Vernon, 1992).However, decreased NEFA post-LPS in E-LPS cows does not fully agree with our i.v.LPS model and some IMM models, where EL cows temporarily have increased NEFA post-LPS (Lehtolainen et al., 2003;Graugnard et al., 2013;Opgenorth et al., 202Y).Reasons for the inconsistencies are not clear, but may be dependent on differences in insulin resistance of adipose tissue or lipolytic hormones (i.e., cortisol and catecholamines; Wasyluk and Zwolak, 2021) or the amount of circulating LPS (which presumably increases adipocyte lipolysis in vitro; Chirivi et al., 2022).Nevertheless, IMM LPS decreased NEFA in both EL and ML, but E-LPS maintained increased NEFA relative to M-LPS cows consistent with their LS.Interestingly, plasma BHB did not meaningfully change post-LPS and this disagrees with many reports, including our i.v.LPS model indicating that LPS actually decreases BHB (Graugnard et al., 2013;Gross et al., 2020;Opgenorth et al., 202Y).Decreased post-LPS BHB may stem from decreased hepatic ketogenesis as NEFA uptake by the liver is ostensibly dependent upon NEFA plasma concentrations (Danfaer, 1994) or it could be caused by decreased alimentary BHB production.Regardless, how immune activation coordinates whole body ketone metabolism is relatively unknown but having a better understanding would have pragmatic benefits to the interpretation of periparturient bioenergetics.
The ability of the liver to adapt to alterations in glucose and FA metabolism is a critical component of successfully transitioning dairy cow.As anticipated, hepatic metabolic flexibility was increased in EL compared with ML cows before LPS, and indicated EL cows have an enhanced capacity to sense and switch preference to FA oxidation (Smith et al., 2018;Palmer and Clegg, 2022).In corroboration with our results, the EL liver has a higher affinity for FA oxidation than ML liver (Gross et al., 2013;Kennedy and Kuhla, 2023), and they upregulate hepatic FA oxidation proportional to circulating concentrations (Danfaer, 1994;Ha et al., 2017).Through an increased preference for FA in transition cow hepatocytes, gluconeogenic precursors can be spared for glucose synthesis (Bell and Bauman, 1997).Overall, enhanced hepatic metabolic flexibility in EL is a homeorhetic adjustment healthy cows engage to prioritize milk synthesis.
Twelve hours after IMM LPS, liver from both E-LPS and M-LPS cows had decreased metabolic flexibility.

Opgenorth et al.: EARLY VS MID-LACTATION RESPONSE TO INTRAMAMMARY LPS
Immune activation-induced decreased hepatic metabolic flexibility agrees with other lines of evidence demonstrating reduced reliance on FA in both cows and rodents (Jiang et al., 2008;Maitra et al., 2009).However, Waldron et al. (2003) observed increased liver palmitate oxidation after LPS.Reasons for the inconsistencies are not clear, but in vivo the liver would have less opportunity to oxidize FA because circulating NEFA decreases following LPS administration.Further, hepatic uptake of lactate and glucogenic AA (both increased in circulation during immune activation; Giri et al., 1990;Su et al., 2015) would flux through pyruvate for complete oxidation to CO2.This is consistent with maintenance of euglycemia post-LPS in this study, making the excess uptake of lactate and glucogenic AA dispensable for gluconeogenesis.Collectively, the hepatic carbon flow through pyruvate is enhanced following immune activation and this is LS independent.
Due to its sheer mass and because it consumes a considerable amount of energy (Zurlo et al., 1990) skeletal muscle's metabolic flexibility has important implications for nutrient partitioning.Like liver, metabolic flexibility was increased pre-LPS in EL skeletal muscle compared with ML.This was expected because bovine myocytes have limited glycogen stores, especially in EL (Kuhla et al., 2011), EL muscle has reduced insulin sensitivity (De Koster and Opsomer, 2013), and muscle oxidizes FA at a rate proportional to circulating FA concentrations (which are increased in EL; Danfaer et al., 1994).These factors allow periparturient muscle to rely on ancillary fuels and are classic homeorhetic "glucose sparing" adjustments healthy cows employ to prioritize milk synthesis.
Similar to the liver, cows in both LS had reduced skeletal muscle metabolic flexibility during immune activation.Increased carbon flux through pyruvate (decreased metabolic flexibility) does not necessarily indicate increased reliance on circulating glucose for ATP production.Fuels that traverse pyruvate include glycogen, lactate and some AA (Gray et al., 2014) and immune activation markedly increases circulating lactate (Giri et al., 1990) and specific AA (Freund et al., 1978;Su et al., 2015).Further, the metabolic flexibility assay utilizes a tissue homogenate and both the liver and skeletal muscle contain multiple cell types in addition to hepatocytes and myocytes.Importantly, both tissues have resident leukocytes (Yoshioka et al., 1997;Brigitte et al., 2010) and circulating immune cells can infiltrate both tissues during infection (McNamara and Cockburn, 2016;Nakanashi et al., 2022).Almost all antigen activated leukocytes switch metabolism from oxidative phosphorylation to aerobic glycolysis (Warburg, 1927;Palsson-McDermott and O'Neill, 2013) and become ravenous glucose consumers (Lang and Dobrescu, 1991;Kvidera et al., 2017).Consequently, interpreting metabolic flexibility in whole tissue homogenates during immune activation is challenging because some cells are markedly decreasing their dependence on glucose while others are substantially increasing their reliance on glucose as a fuel.
Immunosuppression is traditionally considered a surmised hallmark of periparturient cows (Kehrli et al., 1989;Goff and Horst, 1997).However, data from the current study corroborate inferences from our i.v.LPS model that EL cows mount a robust immune response to LPS and are not universally immune-suppressed (Opgenorth et al.,202X).Appending to results herein, other investigators report many functions of the immune system are actually upregulated in EL (Mann et al., 2019;Minuti et al., 2020).Burton et al. (2005) postulated that transition cows have an altered, but not suppressed, immune status that is designed to support postpartum physiological changes (e.g., uterine involution, lactogenesis).Accordingly, our data reinforces the assertion that the immunosuppression descriptor oversimplifies the intricate complexity of periparturient immunity.

CONCLUSION
Despite increased immune sensitivity to a mammary infection (i.e., higher fever, cytokines, and neutrophilia), EL cows had similar milk production and DMI responses compared with ML cows.This suggests the EL mammary gland's milk synthesizing pathways might be more refractory to intense immune activation.The evolutionary drive to prioritize milk synthesis in the face of zealous immune activation is associated with the metabolic consequences of a blunted insulin response.Collectively, the ipsedixitism of transition cow immunosuppression requires adjustments as using it as the foundation for strategies to improve cow health, welfare and productivity is likely preventing critical pharmaceutical, nutritional, genetic and management advancements 72 h relative to LPS bolus administration in tubes containing K 2 EDTA (BD, Franklin Lakes, NJ) to harvest plasma and analyze complete blood count (CBC) or in tubes containing clot activating factor (BD, Franklin Lakes, NJ) to harvest serum.Serum tubes were allowed to clot for at least 15 min before centrifugation.Plasma and serum were harvested after centrifugation at 1,500 × g for 15 min at 4°C before storing at −20°C until analysis.Whole blood (~3 mL) was stored at 4°C for up to 1 d until submission to the Iowa State University's Department of Veterinary Pathology for CBC analysis.Rectal temperature and respiration rate were recorded hourly for the first 12 h of P2.Rectal temperatures were measured using a digital Opgenorth et al.: EARLY VS MID-LACTATION RESPONSE TO INTRAMAMMARY LPS thermometer (GLA M900 Digital Thermometer, San Luis Obispo, CA).Respiration rates were determined by counting flank movements for 15 s and multiplying by 4 to obtain breaths per minute (bpm).
Figure 1.Effects of 100 µg Escherichia coli O111:B4 intramammary lipopolysaccharide (LPS) on lactation stage (LS): early (E-LPS; n = 11) or mid-lactation (M-LPS; n = 10) cow (A) rectal temperature, (B) haptoglobin (Hp), (C) circulating neutrophil count, and (D) total serum calcium (tCa).Results are expressed as least squares means ± SEM.Baseline values are reported to the left of the data set and were analyzed separately.Results are expressed as least squares means ± SEM.

Table 1 .
Opgenorth et al.: EARLY VS MID-LACTATION RESPONSE TO INTRAMAMMARY LPS Ingredients and composition of diet 1 Opgenorth et al.: EARLY VS MID-LACTATION RESPONSE TO INTRAMAMMARY LPS and slowly decreased during the remainder of P2 but this pattern was unaffected by LS.Milk component parameters during P1 are displayed in Supplementary Table 1 whereas P2 parameters are shown in Table