Increased maternal non-esterified fatty acid concentrations during late gestation impair adipose tissue development and metabolic function in Holstein offspring calves

Abstract

In late gestation, dairy cows often experience a negative energy balance (NEB) due to rapid fetal growth, leading to greater circulating levels of non-esterified fatty acids (NEFA). Circulating NEFA are recognized as critical factors of adipose tissue growth, but the impact of greater maternal NEFA during late gestation on offspring subcutaneous white adipose tissue (sWAT) development and remodeling remains unclear. This study aimed to investigate the effects of greater circulating NEFA during late gestation on sWAT development, structural remodeling and metabolic health in calves. A total of 48 pregnant Holstein cows with similar parity (1.95 ± 0.15), body weight (743 ± 8.99 kg), and days in milk (291.4 ± 2.23 d) were categorized into lesser NEFA (DCL-NEFA; n = 24; 246 ± 5.67 μM) and greater NEFA (DCG-NEFA; n = 24; 384 ± 13.36 μM, P < 0.01) groups based on the median of average serum NEFA measured at 1, 3, 5, and 7 wk after dry-off. Calf sWAT was collected at birth and one month to assess tissue growth and metabolic functions in vivo. In vitro, stromal vascular fractions isolated from sWAT of independent donor cattle were induced for adipogenesis to evaluate dose-dependent NEFA effects. Data were analyzed using linear mixed- and fixed-effects models. Although birth weight was similar between groups (P = 0.94), calves born to DCG-NEFA dams exhibited lesser sWAT mass (P = 0.01). Reduced thermogenic capacities (P < 0.01) and increased interleukin 6-mediated inflammatory signaling at birth (P = 0.01) were observed. At one month of age, DCG-NEFA calves showed greater sWAT mass (P = 0.04), impaired glucose sensitivity (P = 0.04), increased insulin resistance (P = 0.02), persistently suppressed thermogenic capacities (P < 0.01), and activated stimulator of interferon genes inflammatory signaling (P < 0.01). In vitro, NEFA addition also consistently impaired white adipogenesis and thermogenic activity (P < 0.01), and promoted inflammatory mRNA expression (P < 0.01). Greater maternal circulating NEFA concentrations during the dry period impaired sWAT development, thermogenic activity and inflammatory regulations, attributing to the dysfunctions of glucose and insulin sensitivity in offspring calves. These findings demonstrate that increased maternal lipid mobilization during late gestation programs offspring sWAT development and metabolic function, highlighting circulating NEFA as a potential target for improving metabolic health in dairy calves.

Increased maternal non-esterified fatty acid concentrations during late pregnancy disrupt offspring-calf adipose tissue development and promotes stimulator of interferon gene-mediated inflammatory signaling.

Graphical abstract

Graphical Abstract.

Introduction

During late gestation, dairy cows often experience greater energy demands due to rapid fetal growth (Sguizzato et al 2020). However, this period often coincides with the reduction of feed intake due to gestational stress and closer to birth (Grummer et al 2004), which resulted in significant negative energy balance (NEB) in cows, driving substantial mobilization of body fat to increased circulating concentrations of non-esterified fatty acids (NEFA) in blood (Adewuyi et al 2005). Excessive blood NEFA not only disrupt metabolic health in cows (Kang et al 2025), but also delay ovulation and impair follicular microenvironment (Giuliodori et al 2011, Van Hoeck et al 2014). Besides, a recent study also suggests that greater maternal NEFA concentrations during pregnancy are also highly associated with the reduced body weight (BW) in calves (Ling et al 2018), showing maternal NEFA as a potential factor in altering offspring body organ development and metabolic functions.

Early-life tissue development and metabolic health are critical determinants of neonatal viability, long-term body growth and milk efficiency (Dallago et al 2024; Meesters et al 2024). Adipose tissue has been recognized as a critical energy reservoir, metabolic and immune regulator (Pond 2017). In particularly, calves rely not only on brown fat and muscle for thermogenesis but also on the browning of subcutaneous fat to generate heat (Silva and Bittar 2019), which facilitates calves in limiting external cold and maintain body temperature. Besides, white adipocytes upon cold exposure can also undergo cellular beiging for thermogenesis, with activation of uncoupling protein 1 (UCP1) within inner mitochondria (Cao et al 2017; Ikeda and Yamada 2020). Moreover, sWAT can also act as an essential endocrine organ to secrete various hormones, including leptin and adiponectin (ADIPOQ), which can regulate the whole-body energy homeostasis (Maurya et al 2012; Al-Thuwaini 2022). Despite this, it is unclear whether greater maternal circulating NEFA concentrations during the late gestation affect subcutaneous white adipose tissue (sWAT) development, thermoregulation, and metabolic functions in fetal and neonatal calves.

Inflammatory signaling plays critical roles in adipose tissue development and remodeling (Itoh et al 2011). Under metabolic stress, excessive lipolysis and increased circulating NEFA concentrations can provoke oxidative stress and activate pro-inflammatory cytokines, such as interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), to impair white adipogenesis (Zhang et al 2018; Häussler et al 2022). In the process, the stimulator of interferon genes (STING) has emerged as a vital regulator triggering inflammatory signaling in adipocytes. Dysregulation of STING or its downstream effector TANK-binding kinase 1 (TBK1) has been demonstrated to be linked to impaired immune homeostasis, white adipocyte formation and even beiging (Bai et al 2020; Bai and Liu 2021). For this signaling, under normal physiological conditions, STING can also serve as a regulator in anti-inflammatory responses to benefit body metabolic health (Wu et al 2020). However, under metabolic stress conditions, the activation of STING-downstream signaling can often trigger chronic inflammation, which can introduce significant disruptions of white adipogenesis and long-term adipocyte hypertrophy (Varga et al 2023). For example, STING-mediated inflammation has been shown to be highly associated with obesity and glucose dysfunctions in mammals (Bai et al 2017; Gong et al 2024). However, the greater maternal circulating NEFA concentrations during the dry period on STING-signaling in prenatal and postnatal calf’s adipose tissue remains unexamined.

In this study, we examined the impacts of greater blood NEFA concentrations in dry cows on sWAT growth and metabolic characteristics in calves at birth and one month of age. We hypothesized that increased maternal NEFA levels during the dry period impaired fetal and postnatal sWAT development, endocrine, and STING-inflammatory signaling, potentially contributing to negative impacts on metabolic health in calves.

Materials and methods

This study was conducted at the HSMD Dairy farm in Jinan City, Shandong Province, China, in accordance with the guidelines approved by the Institutional Animal Care and Use Committee (IACUC) of Nanjing Agricultural University (Approval #NJAULLSC2022006). In this study, all cows enrolled in the trial were dried off at 60 d (± 3 d) prior to their expected calving date.

Experimental design and animal care

After drying, forty-eight Holstein cows with similar BW (mean ± SEM, 743 ± 8.99 kg), parity (1.95 ± 0.15; mode = 1) and days in milk (291.4 ± 2.23 d) were enrolled in this study. To reduce biological variability associated with fetal sex, only cows confirmed to be carrying male fetuses were enrolled in the study, as determined by B-ultrasound examination (Mindray 3300, China). To characterize cows with lesser and greater blood NEFA concentrations during the dry period, blood samples were collected from all dry cows before feeding (0600 h) at 1, 3, 5, and 7 wk after dry-off. The samples were centrifuged at 3,000 × g for 15 min at 4 °C, and serum was aliquoted and stored at −20°C until analysis. Serum NEFA concentrations were further measured by an ELISA kit (#BOFI00145, Genie, China) following the manufacturer’s guideline (Johnson and Peters 1993). Serum NEFA concentrations obtained at the four time points were averaged for each cow to generate an individual mean NEFA value across the dry period (Table S1). Cows were retrospectively classified into lesser or greater blood NEFA groups based on the median of these individual mean NEFA values (320 μM). Accordingly, cows were assigned to a dry cow lesser-NEFA group (DCL-NEFA; n = 24; NEFA < 320 μM; mean ± SEM, 246 ± 5.67 μM) or a dry cow greater-NEFA group (DCG-NEFA; n = 24; NEFA ≥ 320 μM; 384 ± 13.36 μM). Cows were housed in a pen with dry manure bedding, and had ad libitum access to total mixed ration (TMR) according to NRC (2001) (Table 1). Total mixed ration was delivered twice daily (at 0080 and 0170 h), and feed intake was recorded individually twice per week. Body weight, body condition score (BCS), and backfat thickness (BFT) were recorded weekly using a weigh bridge, a 5-point scale (Edmonson et al 1989), and a portable B-ultrasound device (IMV Imaging, UK), respectively, with the imaging depth optimized to 80-100 mm for optimal sWAT visualization in the caudal area. Prior to scanning, the skin surface of back hip was thoroughly cleaned, and ultrasound gel was applied to ensure proper acoustic coupling. All measurements were conducted by one trained individual following standardized protocols to ensure data reproducibility.

Table 1.

Ingredient and chemical composition of diets fed to dry cows.

Item	Content
Ingredients, % DM
Corn silage	44.4
Barley grain	11.1
Wheat straw	16.7
DDGS	16.5
Beet pulp	1.7
Soybean meal	2.6
Corn grain	6.3
Premix1	0.7
Chemical composition, % DM2
DM	46.6
OM	90.2
NEL, Mcal/kg	1.34
CP	14.8
NDF	44.9
ADF	30.1
CF	3.2
Starch	17.8
Ash	9.8
Ca	0.70
P	0.44

DDGS = dry distillers’ grains with soluble; OM = organic matter; NE_L = net energy for lactation; NDF = neutral detergent fiber.

¹Premix contained (per kg of diet on DM basis): vitamin A 200,000 IU, vitamin D 70,000 IU, vitamin E 1,000 IU, Fe 2,000 mg, Cu 600 mg, Zn 2,400 mg, Mn 1,300 mg, I 6 mg, Co 7 mg.

²Chemical composition was on dry matter (% DM) basis. NE_L was predicted based on NRC (2001).

Calf management, sampling, and measurements

After calving, 12 male calves from each maternal group were randomly selected and euthanized within 24 h of birth for sWAT collection by captive bolt stunning, following approved procedures (Edmonson et al 1989). Upon dissection, sWAT from the tailhead region was isolated, with visible connective tissue and blood vessels carefully removed, and immediately weighed using a precision electronic balance. In addition, skin samples were collected from the mid-cervical region, approximately 5–10 cm ventral to the dorsal midline.

The remaining calves (n = 12 per group) were retained and monitored until one month of age. All calves received 6 L of pooled colostrum within the first 12 h after birth and were subsequently fed 3 L pasteurized milk twice daily. Umbilical cord was disinfected by immersion in 10% iodine immediately after birth, followed by twice-daily topical application until detachment (Silva et al 2016).

Calves were individually housed in wooden hutches (1.5 × 2.8 m) bedded with shavings, with hutches spaced 0.5 m apart to prevent animal physical contacts. Ambient temperature and relative humidity averaged 23 ± 2 °C and 56 ± 9.3%, respectively. Water and starters were provided ad libitum. Starter intake was recorded weekly by weighing feed offered and refusals (Table S2); the starter contained 15.7% crude protein (CP), 10.4% acid detergent fiber (ADF), 20.5% neutral detergent fiber (NDF), 28.5% starch and 3.3% crude fat (CF) on a dry matter (DM) basis.

Rectal temperature was recorded using a digital thermometer at 0, 2, 4, 6, 8, 16, and 24 h after birth, and again at 1 month of age to evaluate neonatal thermoregulation and early postnatal physiological adaptation. Body weight, withers height, and body length were obtained weekly to monitor postnatal growth performance and skeletal development, with body weight measured using a digital scale and skeletal dimensions measured using a hipometer.

Dietary chemical analyses

Weekly TMR samples were dried in a forced-air oven at 55 °C to determine DM content. Dry samples were ground using a Wiley mill (Arthur H. Thomas, Philadelphia, PA, USA) to pass through a 1 mm sieve for further chemical analyses. Composite samples were analyzed for CP (AOAC 984.13), CF (AOAC 954.02), ADF (AOAC 973.18), ash (AOAC 942.05), and mineral content following AOAC procedures.

Calf health scores

Calf health conditions were evaluated three times per week until one month of age (n = 12) to monitor overall health status and determine the incidence of common neonatal diseases. The parameters included fecal consistency, cough, nasal discharge, and respiratory rate. Respiratory rate was determined by flank movements during sternal recumbency and recorded in breaths per min (Spain and Spiers 1996). Health status was scored using clinical assessment parameters adapted from the University of Wisconsin–Madison, School of Veterinary Medicine (NRC 2001). Diarrhea was defined as a fecal score ≥ 2 (Monteiro et al 2014).

Blood biochemical, glucose, and insulin sensitivity assessments

Serum stored at −20°C was thawed on ice and subsequently analyzed using bovine- specific ELISA kits to measure the concentrations of NEFA (#BOFI00145, Genie, China), β-hydroxybutyrate (BHB) (#MM-5100301, Meimian, China) and insulin concentrations (#RD-RX77142, Henghuibio, China) in serum according to manufacturer’ protocols, individually (Johnson and Peters 1993; Muniyappa et al 2008). Before measurements, these assays were also validated for analyzing accuracy, with inter- and intra-assay coefficients as follows: NEFA (1.9% and 1.7%), BHB (2.5% and 1.5%), and insulin (1.4% and 1.5%). Blood glucose concentrations were measured using test strips (#07124112, Roche Care GmbH, Germany).

Calves at one month of age were subjected to glucose tolerance test (GTT). For blood collection and glucose administration, a sterile 16-gauge indwelling catheter (14 cm length, Li-heparin coated) was placed in the jugular vein. Before intravenous infusion of glucose, calves fasted for 12 h overnight with free access to water. A dose of glucose solution (250 mg/kg BW dissolved in 20 mL sterile 0.9% saline) was administered intravenously at 0700 h (Yohe et al 2021). Blood glucose was measured at 0, 15, 30, 45, 60, and 90 min post-infusion using a calibrated medical-grade glucometer (Roche Care GmbH, Germany). Besides, insulin sensitivity was also evaluated by calculating the homeostasis model assessment of insulin resistance (HOMA-IR), following the established methodology: HOMA-IR = fasting glucose (mM) × fasting insulin (μIU/mL)/22.5 (De Koster and Opsomer 2013; Hasegawa et al 2019; Nikolaevich et al 2024). Given the lack of a validated HOMA correction factor for adult ruminants, results derived from HOMA-IR should be interpreted cautiously and primarily as indicators of relative differences in insulin sensitivity among experimental groups.

Hematoxylin and eosin and immunofluorescence staining

The hematoxylin and eosin (H&E) staining of sWAT was conducted as previously described (Saraf et al 2016). Tissues were fixed in 4% paraformaldehyde at 4 °C for 24 h, following paraffin embedding and microtome sectioning (HM 355 S; SLEE Medical, Germany). Sections were deparaffinized in xylene, rehydrated through graded ethanol, stained with hematoxylin for 1 min, followed by counterstaining with eosin for 3 min. The stained sections were dehydrated in ethanol, cleared in xylene, and mounted for microscopic analysis. Adipocyte number and size were quantified using the Deconvolution2 plugin in ImageJ (v1.53c, NIH, MD, USA). For measurements, at least three tissue sections per sample were analyzed at 40 × magnification, with approximately 89–142 adipocytes counted per slide.

For immunofluorescent staining, sWAT sections were incubated in 3% hydrogen peroxide to block endogenous peroxidase activity, followed by blocking with bovine serum albumin to limit nonspecific binding. Primary antibodies were incubated overnight at 4 °C, and secondary antibodies were applied at R.T. for 1 h. Mitochondrial density was analyzed by MitoSPY staining (1:1000; #C1048, Beyotime, China). Thermogenic protein UCP1 was stained using a primary antibody (1:200; #A5857, ABclonal, China), and Cy3-conjugated Goat anti-Rabbit immunoglobulin G (IgG) (1:1000; #AS007, ABclonal, China) was used as a secondary antibody. Nuclei were stained using 4′,6-diamidino-2-phenylindole (DAPI) (5 μg/mL; #RM02978, ABclonal, China). Images were captured using a fluorescence inverted microscope (Leica DMi8, USA). Fluorescent signal intensity was analyzed using ImageJ software (version 1.53c, National Institutes of Health, MD, USA), and intensity was normalized to background.

NEFA treatments and adipogenic differentiation in vitro

Stromal vascular fractions (SVFs) were isolated from sWAT (150 mg) of healthy calves at one month of age. Tissue was minced and digested in 1 mg/mL collagenase type I (#BS163; Biosharp, China) for 1 h at 37 °C under constant gentle shaking. Enzymatic digestion was halted with low-glucose Dulbecco’s modified Eagle medium (DMEM; #A4192001, Gibco, USA) containing 10% fetal bovine serum (FBS; #U11-059A, Yobibio, China), filtered through 100 μm cell strainers, and centrifuged at 1,000 × g for 5 min. The collected SVFs were resuspended in DMEM with 10% FBS, 100 U/mL penicillin, and 1,000 U/mL streptomycin (#15240062, Gibco, USA). Cells were cultured at 37 °C in 5% CO₂ (Wci-260; Wiggens, Germany), with media refreshed every 2 days. Cell morphology was monitored daily using an inverted light microscope (XD-202, Jiangnan, China). For white adipocyte differentiation, the isolated SVFs were seeded in 6-well plates and cultured until 90% confluence was reached. Stromal vascular fractions were induced to white adipogenesis using DMEM supplemented with 125 μM indomethacin (#I811784; Macklin, China), 0.5 mM 3-isobutyl-1-methylxanthine (IBMX, #I811775, Macklin, China), 1 μM dexamethasone (#BL171B; Biosharp, China), and 5 μg/mL insulin (#I828365; Macklin, China) for 4 days, followed by additional 4 days of 5 μg/mL insulin. During differentiation, SVFs were also treated with PBS (control), 100 μM NEFA, or 200 μM NEFA throughout differentiation. The NEFA cocktail contained oleic acid (22.9 mM; #SC9320, Servicebio, China), linoleic acid (2.6 mM; #SL8520, Servicebio, China), palmitic acid (16.8 mM; #SP8060, Solarbio, China), stearic acid (7.6 mM; #SS8520, Servicebio, China), and palmitoleic acid (2.8 mM; #P794587, Macklin, China) according to a previous study (Du et al 2018).

Lipid droplet staining in adipocytes

Following adipocyte differentiation, the culture medium was removed, and the plates were washed with PBS. The differentiated adipocytes were fixed in 4% PFA for 10 min at room temperature, following Oil Red-O staining (#AC11564; Acmec, China). Lipid content was quantified by eluting Oil Red-O with 100% isopropanol, and the absorbance was measured at 520 nm using a microplate reader (Synergy HTX microplate reader, BioTek Instruments) (Kwan et al 2017). Besides, the neutral lipids in adipocytes were also labeled with BODIPY (#121207-31-6; Med Chem Express, USA) for 30 min at room temperature, with DAPI counterstaining. Fluorescence images were captured using a Leica DMi8 microscope (Wetzlar, Germany), and the signal intensity was adjusted relative to background.

Real-time qPCR analysis

Total RNA was extracted from sWAT and differentiated adipocytes using TRIzol (Invitrogen, NY, USA) as previously described (Cao et al 2017). RNA concentration and purity were assessed using a NanoDrop 2000 (Thermo Fisher, USA), with acceptable A260/A280 ratios from 1.9 to 2.0. cDNA was synthesized from 1 µg RNA using an iScript cDNA synthesis kit (#TSK302S; Tsingke, China) according to the manufacturer’s protocol (initial denaturation at 60 °C for 1 min followed by 42 °C for 2 min; reverse transcription at 50 °C for 15 min; and enzyme inactivation at 85 °C for 5 s). qPCR analysis was conducted in a total volume of 20 µL using iQ SYBR Green Supermix (#TSE201; Tsingke, China) on an applied 7500 system (Thermo Fisher Scientific, Waltham, MA, USA). All primers were designed with an optimal annealing temperature of 60 °C. Primer specificity was initially validated by melt curve analysis, confirming a single amplification peak for each primer pair. In addition, the specificity of qPCR products was further verified by 1.5% agarose gel electrophoresis (#TSJ001; Tsingke, China), in which a single band of the expected size was observed for each target gene (Figure S1; see online supplementary material for a color version of this figure). For each qPCR assay, standard curves were generated using five 10-fold serial dilutions of pooled cDNA to determine amplification efficiency and linearity, in accordance with the MIQE (Minimum Information for Publication of Quantitative Real-Time PCR Experiments) guidelines. Only primer sets with amplification efficiencies between 90% and 110% and correlation coefficients (R²) greater than 0.985 were used for further analysis. Relative mRNA expression was determined using Bio-Rad CFX Manager 3.1, and mRNA expression was calculated using 2^−ΔΔCT method, with 18S rRNA or GAPDH as reference genes (Ontsouka et al 2004; Gai et al 2025). Primer sequences are listed in Table 2.

Table 2.

Primers used in qPCR analyses.

Gene	Primers (5′-3′)	GeneBank number	Product length, bp	Melting temperature, °C
UCP1	For: TAGGGAGCGCTATCCAGGG	NM_001166528.1	121	60.2
	Rev: GGCAGCCCACTGTAGAGTTT			60.0
PPARGC1α	For: TGATTAGTTGAGCCCTTGCCG	NM_177945.3	200	60.6
	Rev: TGCCAGGAGTTTGGTTGTGAT			60.1
DIO2	For: CCTCTTCCTGGCGCTCTATG	NM_001010992.7	174	60.0
	Rev: TCTTCACCCAGTTTCACCTGT			59.2
PPARγ	For: AGACGACAGACAAATCACCGT	NM_181024.2	252	59.7
	Rev: CGTGCACGCCGTATTTTAGG			60.0
LEPTIN	For: TGACATCTCACACACGCAGTC	NM_173928.2	114	60.6
	Rev: ATCGCCAATGTCTGGTCCAT			59.4
ADIPOQ	For: CCTTACCCAGTCATGGCGAG	NM_174742.2	221	60.2
	Rev: CCCTTCAGCTCCTGTCATTCC			60.4
CIRBP	For: TGGTTTCTCCAGAGGAGGTGG	NM_001434956.1	273	61.1
	Rev: GGCGTGGTCACTACACACTAA			60.0
HSBP1	For: CGTGCAGGATCTCACCTCAG	NM_001113316.2	224	60.2
	Rev: CGAAGACAAGCAACCTTCAGC			60.1
HSP70	For: GGGGAGGACTTCGACAACAG	NM_203322.3	192	60.0
	Rev: GAAGTCGATGCCCTCGAACA			60.1
RBM3	For: TGGCTACCCTAGAGGTGGTG	NM_001303480.1	124	60.3
	Rev: CTGGCTGCCACCATAGTCTC			60.1
HSPB5	For: CTCAGCCAGAGGGCATGTTT	KU497666.1	294	60.3
	Rev: CAGCTACATACCTGGCGCT			59.6
IL-6	For: TGGGTTCAATCAGGCGATTTG	NM_173923.2	196	59.2
	Rev: AGTGTTTGTGGCTGGAGTGG			60.5
CCL2	For: CCAGCTCAGAGACAAATGCCC	NM_001034349	227	61.3
	Rev: GCACTTGTTTCCGGAGATGTTC			60.1
TNF-α	For: GGTCAACATCCTGTCTGCCA	NM_173966.3	120	60.1
	Rev: CACAGTGCGATGATTCCAAAGT			60.0
IL-10	For: AGCACTACTCTGTTGCCTGG	NM_174088.1	230	59.7
	Rev: GGCAACCCAGGTAACCCTTAAA			60.5
18S	For: GTAACCCGTTGAACCCCATT	DQ222453	151	58.1
	Rev: CCATCCAATCGGTAGTAGCG			57.9
GAPDH	For: GGCGTGAACCACGAGAAGTATAA	NM_001034034.2	119	60.7
	Rev: CCCTCCACGATGCCAAAGT			60.0

UCP1 = uncoupling protein 1; PPARGC1α = peroxisome proliferator-activated receptor gamma coactivator 1 alpha; DIO2 = type 2 iodothyronine deiodinase; PPARγ = peroxisome proliferator-activated receptor gamma; ADIPOQ = adiponectin; CIRBP = cold inducible RNA-binding protein; HSBP1 = heat shock factor binding protein 1; HSP70 = heat shock protein 70; RBM3 = RNA-binding motif protein 3; HSPB5 = heat shock protein beta-5; IL-6 = interleukin-6; CCL2 = C-C motif chemokine ligand 2; TNF-α = tumor necrosis factor-alpha; IL-10 = interleukin-10; 18S = 18S ribosomal RNA; GAPDH = glyceraldehyde-3-phosphate dehydrogenase.

Western blot analysis

Approximately 10 mg of sWAT from neonatal or one month of age calves was homogenized in 1 mL ice-cold RIPA lysis buffer (#BL504A; Biosharp, China) with protease inhibitors (#BL507A; Biosharp, China). Protein concentration was determined by a BCA assay kit (#P0012; Beyotime, China). Proteins were separated by SDS-PAGE using a 5% stacking gel and 10% resolving gel at 100 V for 90 min, and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore, #IPVH10100) at 150 mA for 2 h. Membranes were blocked with 5% skim milk, washed, and incubated overnight at 4 °C with primary antibodies, including UCP1 (1:500; #A5857, ABclonal, China), PGC1α (1:500; #A22113, ABclonal, China), STING (1:500; #A21051, ABclonal, China), p-TBK1 (1:500; #AP1026, ABclonal, China), TBK1 (1:500; #A3458, ABclonal, China) and β-tubulin (1:1000; #AF2835, Beyotime, China). Cy3-conjugated goat anti-rabbit IgG (1:2000; #AS007, ABclonal, China) was used as a secondary antibody, and membranes were incubated for 2 h at room temperature. Proteins were measured using enhanced chemiluminescence substrate (ECL, #RM00021P; ABclonal), and imaged with a ChemiDoc system (ProteinSimple, USA). Quantification was performed with ImageLab software (Bio-Rad), and protein contents were normalized to β-tubulin as an internal control.

Statistical analysis

The nlme package (v.3.1-168) in R 4.4.0 was applied to data analysis (Team 2024). To determine the minimum number of cattle required for the maternal NEFA treatment experiment, a power analysis was conducted using the simr package in R. Based on our previous body fat data in a calf study, a minimum sample size of nine animals per group was sufficient to achieve a biological difference with 80% statistical power at an alpha level of 0.05. Individual cows (for dry-cow analyses) or calves (for calf analyses) were considered as experimental units. Continuous outcome residuals normality was evaluated using the Shapiro–Wilk test from the stats package and UNIVARIATE function from the car package (Fox and Weisberg 2018). Homogeneity of variance was assessed via Levene’s test and Q-Q plots (Levene 1960). For fecal score and cough score, log transformation improved normality; however, transformation did not alter the significance of treatment, time, or treatment × time effects. Therefore, raw-scale data are presented to facilitate biological interpretation. For dry cows, metabolic and body condition variables measured repeatedly during the dry period, including blood metabolites, feed intake, and body composition, were analyzed to characterize maternal metabolic status associated with maternal NEFA grouping using a linear mixed-effects model. These models included maternal NEFA level (greater vs. lesser), time, and their interaction as fixed effects, and cow as a random effect to account for repeated measures within individuals. For calves, longitudinal health and growth variables, including respiratory and digestive health scores, and morphometric traits, were analyzed using mixed-effects model, with maternal NEFA level, time, and their interaction as fixed effects, and calf as a random effect. Single time point measurements obtained at birth or one month of age, including BW, height, length, mRNA expression, and in vitro outcomes, were analyzed using linear fixed-effects model with maternal NEFA level as the fixed effect. Bonferroni corrections were applied to confidence intervals. For repeated-measures analysis, four covariance structures (heterogeneous autoregressive type 1, autoregressive type 1, compound symmetry, and Toeplitz) were compared, and the optimal structure was selected based on the lowest Akaike and Bayesian information criteria. Diarrhea was recorded as a binary health outcome for each calf at each observation time. For statistical analysis, diarrhea data were summarized at the group level as the number of calves exhibiting diarrhea divided by the total number of calves assessed at each time point within each treatment group. Because this outcome represents proportional count data rather than a continuous variable, diarrhea occurrence was analyzed using a generalized linear model with a binomial distribution and logit link function. The model included treatment, time, and their interaction as fixed effects. The binomial response was specified using the number of calves with diarrhea and the number of calves without diarrhea at each time point (cbind events, non-events). Estimated marginal means on the probability scale were obtained using the emmeans package, and pairwise comparisons were adjusted using Bonferroni correction. Model fit for the binomial analysis was evaluated using simulation-based residual diagnostics implemented in the DHARMa package, including assessment of residual dispersion and uniformity. No evidence of overdispersion was detected (P = 0.30), confirming the appropriateness of the binomial model. Because diarrhea occurrence represents discrete count data, normality test and variance homogeneity tests were not applied to this variable. Statistical significance was set at P ≤ 0.05, and trends were defined as 0.05 < P ≤ 0.10.

The repeated-measures model was as follows: y_ijk=μ + α_i+ β_j + (αβ)_ij + b_k + ϵ_ijk.

Where y_ijk is the observation for experimental unit k in treatment i at time j, μ is the overall mean, α_i is the fixed effect of treatment i (maternal NEFA level), β_j is the fixed effect of time j, (αβ)_ij is the fixed interaction effect between treatment i and time j, b_k is the random effect of the individual animal k (cow for dry-period measurements and calf for offspring measurements), where b_k ∼ N(0,σ_b²), and ϵ_ijk is the residual term, with ϵ_ijk ∼ N(0, R) and R represents the covariance structure selected for the analysis.

The equation of the single time point model was as follows: y_ik = μ + α_i+ϵ_ik, where y_ik is the observation for calf k in treatment i, μ is the overall mean, α_i is the fixed effect of treatment i (maternal NEFA level), ϵ_ik is the residual term, where ϵ_ik ∼ N (0, σ²).

The equation for diarrhea occurrence model was as follows: logit(p_ij) = u  +  α_i+b_j+(αb)_ij. Where is p_ij the probability of diarrhea in treatment at time, α_i is the fixed effect of treatment i, b_j is the fixed effect of time j, and (αb)_ij is their interaction.

Results

sWAT and thermogenic activity in calves at birth

Cows in the DCG-NEFA group exhibited persistently greater serum NEFA concentrations during the dry period (P < 0.01; 246 ± 5.67 vs 384 ± 13.36 μM, Table S3). β-hydroxybutyrate tended to differ between groups (P = 0.07), whereas glucose levels (P = 0.52) in cows were not significantly different between groups (Table S3). Dry matter intake (P = 0.59), BW (P = 0.94), BCS (P = 0.47) and BFT (P = 0.61) were also similar during the dry period (Table S4).

At birth, calves from the DCG-NEFA group had greater NEFA levels (P < 0.01; Table 3). No differences were observed in BW, withers height or body length in calves at birth (Table 3). Interestingly, calves born to the DCG-NEFA group displayed significantly lesser rectal temperatures at birth (P = 0.01, 38.11 ± 0.18 vs. 37.23 ± 0.39 °C, Table 3), showing lesser thermogenic activation in DCG-NEFA calves.

Table 3.

Growth traits and blood NEFA of newborn calves from dry cows with lesser or greater blood NEFA concentrations.

	Treatment		SEM	P-value
Items	DCL-NEFA	DCG-NEFA
NEFA, μM	291.9	527.0	39.08	<0.01
Body weight, kg	38.6	36.0	0.94	0.17
Withers height, cm	79.7	76.7	1.02	0.14
Body length, cm	77.8	75.6	1.26	0.40
Body temperature, °C	38.1	37.2	0.23	0.01

DCL-NEFA = dry cows with lower concentration of non-esterified fatty acids in blood.

Calves born from the DCG-NEFA group had lesser sWAT mass (P = 0.01), and larger adipocyte size (Figure 1A and B). Moreover, calves from the DCG-NEFA group showed a markedly decreased of thermogenic and endocrine related mRNA expressions in sWAT, including UCP1, PPARGC1α, type 2 iodothyronine deiodinase (DIO2), peroxisome proliferator-activated receptor gamma (PPARγ), LEPTIN, and ADIPOQ (P < 0.01, Figure 1C). Western-blot analyses also confirmed the reduced levels of thermogenic protein UCP1 and PGC1α (P < 0.01, Figure 1D). MitoSpy analysis showed a significant decrease in mitochondrial density in white adipocytes from the DCG-NEFA group (P < 0.01, Figure 1E). In addition, calves from this group exhibited significantly greater expression of cold stress-related genes in the skin (P < 0.05, Figure 1F), including Cold-Inducible RNA-Binding Protein (CIRBP), Heat Shock Protein 70 (HSP70), and Heat Shock Factor Binding Protein 1 (HSBP1), RNA-Binding Motif Protein 3 (RBM3), Heat Shock Protein Beta-5 (HSPB5), indicating increased thermogenic stress in DCG-NEFA calves.

Figure 1.

Subcutaneous white adipose tissue (sWAT) characteristics and thermogenic markers in neonatal calves at birth. A) sWAT and its mass at birth (n = 12). B) Hematoxylin and eosin (H&E) staining of sWAT and quantification of adipocyte size (scale bar = 50 μm). C) mRNA expression of thermogenic genes in sWAT. D) Immunoblot analysis of uncoupling protein 1 (UCP1) and peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1α) protein levels in sWAT, normalized to β-tubulin. E) Immunofluorescence staining of mitoSpy, UCP1, and 4′,6-diamidino-2-phenylindole (DAPI) stained nuclei in sWAT (scale bar = 100 μm). Fluorescence intensity quantified using ImageJ. F) mRNA expression of stress-related genes, including CIRBP (Cold Inducible RNA-Binding Protein), HSP70 (Heat Shock Protein 70), HSBP1 (Heat Shock Factor Binding Protein 1), RBM3 (RNA-Binding Motif Protein 3), and HSPB5 (Heat Shock Protein Beta-5), in neck skin at birth. Data are presented as mean ± SEM.

Glucose and insulin sensitivity, and sWAT thermogenic functions in 1-month-old calves

Throughout the first month, BW, body height and body length in calves remained comparable between groups (Table 4). Health scores, including fecal, cough, nasal discharge scores and respiratory rate, were also similar between groups, but calves born to the DCG-NEFA group experienced a greater numerical incidence of diarrhea (Table 4). At 1 month of age, rectal temperatures in calves did not differ between groups (P = 0.98, Figure 2A). Calves from the DCG-NEFA group showed reduced glucose sensitivity (P = 0.04) from the glucose tolerance test (Figure 2B). Besides, calves in DCG-NEFA group also had greater HOMA-IR (P = 0.02), showing impaired insulin sensitivity (Figure 2C). For sWAT, 1-month-old calves from the DCG-NEFA group had greater sWAT mass (P = 0.04) and larger white adipocytes (Figure 2D and E). Besides, the mRNA expression of thermogenic and endocrine-associated markers (UCP1, P < 0.01; PPARGC1α, P < 0.01; DIO2, P < 0.01; PPARγ, P < 0.01; LEPTIN, P < 0.01; ADIPOQ, P < 0.01) and proteins (UCP1, P < 0.01; PGC1α, P < 0.01) were significantly reduced in DCG-NEFA calves (Figure 2F and G), suggesting a persistently impaired thermogenic and metabolic capacity of white adipocytes.

Table 4.

The growth performance and health indicators of one-month-old calves born to dry cows with lesser or greater blood non-esterified fatty acids (NEFAs) concentrations.

	Treatment			P-value
Items1	DCL-NEFA	DCG-NEFA	SEM	Trt	Week	Trt × Week
Body weight, kg	49.9	52.0	1.45	0.49	<0.01	0.89
Withers height, cm	86.7	87.8	1.00	0.57	<0.01	0.80
Body length, cm	83.2	82.4	0.92	0.69	<0.01	0.70
Fecal score (0–3)	0.92	1.17	0.18	0.17	0.01	0.94
Cough score (0–3)	0.83	1.17	0.20	0.53	0.01	0.84
Nasal discharge score (0–3)	0.72	0.76	0.15	0.86	0.03	0.93
Diarrhea, %2	12.3	26.9	6.70	0.14	0.34	0.67
Respiration rate, breaths/min	37.9	38.4	1.88	0.30	0.07	0.76

DCL-NEFA = dry cows with lesser concentration of non-esterified fatty acids in blood.

¹Adapted from School of Veterinary Medicine, University of Wisconsin–Madison (2011).

²Fecal score ≥2 is defined as diarrhea.

Figure 2.

Subcutaneous white adipose tissue (sWAT) development and metabolic parameters in one-month-old calves. A) Rectal temperature of calves at one month of age (n = 12). B) Blood glucose levels during glucose tolerance tests (GTT). C) Homeostatic Model Assessment of Insulin Resistance (HOMA-IR) scores (n = 12). D) sWAT mass (n = 12). E) Hematoxylin and eosin (H&E) staining and adipocyte size quantification (scale bar = 50 μm). F) mRNA expression of adipogenic and browning genes, including UCP1 (uncoupling protein 1), PPARGC1α (PPAR gamma coactivator 1 alpha), PPARγ (peroxisome proliferator-activated receptor gamma), DIO2 (type 2 iodothyronine deiodinase), LEPTIN, ADIPOQ (adiponectin). G) Immunoblot analysis of uncoupling protein 1 (UCP1) and peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1α) protein levels, normalized to β-tubulin. Data are presented as mean ± SEM.

Inflammation and STING-signaling in sWAT of calves at birth and 1 month old

At birth, the mRNA expression of pro-inflammatory cytokines, including IL-6 and TNF-α, was markedly activated (P < 0.01) in sWAT of DCG-NEFA calves (Figure 3A), while STING protein and its downstream activators TBK1 and p-TBK1 were barely altered (Figure 3B). At one month of age, the inflammatory cytokines including IL-6 (P < 0.01), TNF-α (P < 0.01) and interleukin-10 (IL-10, P < 0.01) were significantly increased in sWAT of DCG-NEFA calves (Figure 3C), aligned with the marked activation of STING (P < 0.01) and p-TBK1 proteins (P < 0.01, Figure 3D).

Figure 3.

Stimulator of interferon genes (STING) signaling and inflammatory markers in calf subcutaneous white adipose tissue (sWAT) at birth and 1 month of age. A) mRNA expression of inflammatory markers in sWAT of calves at birth, including IL-6 (interleukin-6), CCL2 (C-C motif chemokine ligand 2), TNF-α (tumor necrosis factor-alpha), and IL-10 (interleukin-10). B) immunoblot analysis of STING, TBK1, and p-TBK1 protein levels, normalized to β-tubulin, in sWAT at birth. C) mRNA expression of inflammatory genes in sWAT of calves at 1 month of age. D) Immunoblot analysis of STING, TBK1, and p-TBK1 protein levels, normalized to β-tubulin, in sWAT at 1 month of age. Data are presented as mean ± SEM.

Effects of dosed NEFA addition on white adipogenesis and metabolic functions in vitro

To further examine the dose effects of NEFA on white adipogenesis, SVFs isolated from the sWAT of 1-month-old healthy calves were induced for adipocyte formation under lesser (100 μM) or greater (200 μM) NEFA inclusion in vitro. Relative to the control, NEFA addition significantly impaired white adipocyte formation, as evidenced by the reduced lipid contents in Oil Red O (P < 0.01) or BODIPY staining, even under the lesser NEFA concentration (P < 0.01, Figure 4A and B). NEFA inclusion also significantly suppressed the mRNA expression of thermogenic and adipogenic endocrine-related markers (UCP1, PPARGC1α, DIO2, LEPTIN, ADIPOQ), while activated inflammatory markers (IL-6, TNF-α, IL-10, and STING, Figure 4C). Consistently, the contents of UCP1 and PGC1α proteins were markedly reduced with NEFA treatment, aligned with the increased STING and p-TBK1 proteins (P < 0.01, Figure 4D), showing that NEFA impaired white adipocyte formation and metabolic functions.

Figure 4.

Effects of non-esterified fatty acids (NEFA) supplementation on white adipocyte differentiation and inflammatory signaling in vitro. A) Oil Red O staining and lipid content quantification in white adipocytes treated with 100 or 200 μM NEFA during differentiation (phosphate-buffered saline (PBS) as the control). Images captured at 10× (scale bar = 50 μm) and 40× (scale bar = 200 μm) magnification. B) Boron-dipyrromethene (BODIPY) and 4′,6-diamidino-2-phenylindole (DAPI) staining of lipid droplets and nuclei, respectively (scale bar = 200 μm). Fluorescence quantified using ImageJ. C) mRNA expression of genes related to browning of sWAT, endocrine function, and inflammation, including UCP1 (uncoupling protein 1), PPARGC1α (PPAR gamma coactivator 1 alpha), DIO2 (type 2 iodothyronine deiodinase), LEPTIN, ADIPOQ (adiponectin), IL-6 (interleukin-6), TNF-α (tumor necrosis factor-alpha), and IL-10 (interleukin-10). D) Immunoblot analysis of uncoupling protein 1 (UCP1), peroxisome proliferator-activated receptor gamma coactivator 1 alpha (PGC1α), stimulator of interferon genes (STING), phosphorylated TANK-binding kinase 1 (p-TBK1) and TANK-binding kinase 1 (TBK1) proteins, normalized to β-tubulin. Data are presented as mean ± SEM.

Discussion

Maternal metabolic status during late pregnancy plays a crucial role in shaping offspring tissue development and metabolic health (Levin 2006; Amrom and Schwartz 2023). During this period in dairy cows, accelerated fetal growth and mammary gland development substantially increase metabolic requirements, whereas decreased feed intake promotes lipolysis, resulting in increased circulating concentrations of NEFA (Overton and Waldron 2004; Contreras et al 2017). Among metabolic indicators, excessive increases in NEFA have been shown to not only disrupt insulin signaling and energy partitioning (Pereira et al 2016, Qiao et al 2024), but also cause oxidative stress in cows (Li et al 2012, Ster et al 2012). Besides, increased NEFA concentration before calving is also highly associated with lesser birth weight (Ling et al 2018). Despite this, the role of maternal increased NEFA concentrations within physiological range during the dry period in affecting fetal and postnatal sWAT development and metabolic functions remains unclear. In this study, we did not observe the impacts of DCG-NEFA concentrations on gross body growth traits in calves, but the calves born to the DCG-NEFA group had significantly impaired glucose and insulin sensitivity. Besides, the calves from dams with increased NEFA concentrations also showed impaired sWAT growth and thermogenesis, which was aligned with the activation of STING-inflammatory signaling, indicating dysregulated sWAT development and metabolic functions.

Persistent increased NEFA concentrations during the transition period have been linked to increased BHB concentrations (Ospina et al 2013; Barletta et al 2017). Although a numerical trend in BHB was observed, its lack of statistical significance and inconsistency with NEFA responses suggest that it does not represent a biologically relevant treatment effect. At birth, calves born to the maternal DCG-NEFA group also exhibited increased NEFA levels, supporting that maternal NEFA could directly across the placenta to exert direct impacts on fetal growth (Herrera et al 2006; Gil-Sánchez et al 2012). Late gestation is an essential period for sWAT development (Louveau et al 2016), which can be affected by maternal nutrition and metabolic status (Budge et al 2005). A previous study had showed that maternal undernutrition during pregnancy can impair fetal sWAT growth (Birtwistle 2016). In this study, we also observed the increased NEFA concentrations during the dry period, an indicator of energy deficiency in cows, also associated with significantly reduced sWAT mass at birth. Interestingly, in line with the increased adipocyte size, sWAT mass was paradoxically increased in the DCG-NEFA group at 1 month of age, which potentially reflects the excessive adipocyte hypertrophy due to limited adipocyte numbers.

For calves, sWAT plays a vital role in thermoregulation as a body insulator and fatty acids reservoir to generate heat, particularly in the cold season (Li et al 2023). White adipocytes in subcutaneous fat have browning ability, which can activate UCP1 protein in the inner mitochondria to generate non-shivering heat (Tehrani et al 2023). Besides, browning of white adipocytes in sWAT can also significantly improve glucose and insulin sensitivity, benefiting whole body metabolic health in animals (Kaisanlahti and Glumoff 2019; Machado et al 2022). In this study, calves born to dams with increased NEFA concentrations had reduced mRNA expression of thermogenic protein UCP1, PPARGC1α, and DIO2 in sWAT at birth and one month. Consistently, calves showed reduced body temperature as well as impaired glucose and insulin sensitivity, supporting the presence of impaired white adipocyte browning in sWAT.

Beyond thermoregulation, sWAT also serves as a pro- and anti-inflammatory regulator (DiSpirito and Mathis 2015). The STING pathway, a critical mediator of innate immunity, has recently emerged as a key regulator of adipose tissue inflammation by modulating macrophage polarization and cytokine production (Huang et al 2022). Our results showed increased STING and p-TBK1 proteins in calves from DCG-NEFA dams, indicating increased activation of the STING-inflammatory pathway in sWAT. Under excessive stress in adipocytes, such as lipid-induced hypertrophy, white adipocytes can generate reactive oxygen species and secrete pro-inflammatory cytokines, including IL-6 and TNF-α (Čolak and Pap 2021). In our study, neonatal and postnatal sWAT in maternal greater NEFA group increased the mRNA expression of inflammatory genes including IL-6, and TNF-α, showing significant activation of inflammatory responses.

To further elucidate the direct effects of NEFA on white adipogenesis and thermogenic response, we conducted an in vitro study to assess the effects of increased NEFA concentrations on white adipocyte formation. Aligned with the observations in vivo, NEFA addition significantly impaired white adipocyte formation and reduced thermogenic protein expression. Besides, the STING-TBK1 inflammatory signaling was also increased by NEFA. These outcomes reinforced the observations that increased maternal NEFA concentrations during the dry period negatively affect fetal and postnatal sWAT development and functions.

To minimize biological variability associated with fetal sex, only cows carrying male fetuses were enrolled in this study, given that fetal sex is known to influence fetal growth trajectories and metabolic regulation. Accordingly, the findings should be interpreted within the context of male fetal pregnancies, and future studies are needed to determine whether similar effects are observed in female offspring.

Conclusions

In summary, this study showed that DCG-NEFA concentrations within physiological ranges had no effect on calf gross body growth indices from birth to one month of age. However, calves born to the DCG-NEFA group exhibited impaired sWAT development, thermogenic activity, and immune function, which were associated with reduced thermogenesis and compromised glucose and insulin sensitivity during early life. These alterations were accompanied by enhanced inflammatory responses and increased STING-related signaling in sWAT. Consistent with these in vivo findings, in vitro NEFA exposure directly impaired white adipocyte differentiation and metabolic function while increasing inflammatory signaling pathways. Together, these results indicate that DCG-NEFA programs offspring adipose tissue dysfunction independently of overall growth, highlighting a critical role of maternal lipid metabolism in shaping early-life metabolic health.

Glossary

Abbreviations

ADIPOQ

adiponectin

BCS

body condition score

BFT

back fat thickness

BHB

β-hydroxybutyrate

BW

body weight

CIRBP

cold-inducible RNA-binding protein

CPT1B

carnitine palmitoyltransferase 1B

DAPI

4′,6-diamidino-2-phenylindole

DCG-NEFA

dry cows with greater blood non-esterified fatty acids

DCL-NEFA

dry cows with lesser blood non-esterified fatty acids

DDGS

dry distillers’ grains with soluble

DIO2

type 2 iodothyronine deiodinase

DM

dry matter

DMEM

Dulbecco’s modified Eagle medium

DMI

dry matter intake

FBS

fetal bovine serum

GTT

glucose tolerance test

H&E

hematoxylin and eosin

HOMA-IR

homeostasis model assessment of insulin resistance

HSBP1

heat shock factor binding protein 1

HSP70

heat shock protein 70

HSPB5

heat shock protein beta-5

IBMX

3-isobutyl-1methylxanthine

IgG

immunoglobulin G

IL-6

interleukin-6

IL-10

interleukin-10

NEB

negative energy balance

NEFA

non-esterified fatty acids

NE_L

net energy for lactation

OM

organic matter

PGC1α

peroxisome proliferator-activated receptor gamma coactivator 1 alpha

PPARγ

peroxisome proliferator-activated receptor gamma

PVDF

polyvinylidene fluoride

RBM3

RNA-binding motif Protein 3

STING

stimulator of interferon genes

SVFs

stromal vascular fractions

sWAT

subcutaneous white adipose tissue

TBK1

TANK-binding kinase 1

TMR

total mixed ration

TNF-α

tumor necrosis factor-α

UCP-1

uncoupling protein 1

WAT

white adipose tissue

Author contributions

Yang Gai (Data curation, Formal analysis, Investigation, Writing—original draft), Guiling Ma (Conceptualization, Funding acquisition, Methodology, Supervision), Rui He (Data curation, Formal analysis, Investigation, Writing—original draft), Zhaobing Gu (Conceptualization, Methodology), Manhong Wang (Conceptualization, Methodology, Supervision, Writing—review & editing), Shengyong Mao (Conceptualization, Methodology, Supervision, Writing—review & editing), and Yanting Chen (Conceptualization, Funding acquisition, Methodology, Supervision)

Supplementary data

Supplementary data are available at Journal of Animal Science online.

Conflict of interest statement. The authors declare no real or perceived conflicts of interest.

Data availability

All data analyzed in this study are included in the tables and figures of this manuscript.