Colostrum increases DNA fractional synthetic rate, villi growth, and cellular proliferation of the jejunum in neonatal gilt piglets

Abstract

Adequate versus low colostrum (COL) intake promotes feed efficiency and greater growth trajectory in swine, indicating that varying COL levels likely impact nutrient absorption. This study’s objective was to determine how diet type (COL vs milk replacer [MR]), and level of intake [20% vs 10% of birth body weight (BBW)] influence jejunum deoxyribonucleic acid (DNA) fractional synthetic rate (FSR), and villi growth and cellular proliferation in neonatal gilt piglets. Gilt piglets were allocated to one of the following treatments: pooled COL fed at 20% (COL20; n = 10) or 10% (COL10; n = 10) BBW, MR fed at 20% (MR20; n = 10) or 10% (MR10; n = 10) BBW, stay-on-sow (SOS) to suckle COL ad libitum (n = 9), or zero hour (ZH), euthanized immediately after birth (n = 8). Following administration of 20 mL/kg of deuterium oxide (D₂O) to metabolically label DNA, dietary treatments were either bottle fed every 2 h or SOS for 24 h, then blood and jejunum were collected. Enrichment of D₂O in serum and deoxyadenosine mass isotopomers in jejunum were analyzed using mass spectrometry to determine DNA FSR of jejunum. Additionally, a separate piece of jejunum was fixed in formalin, paraffin embedded, sectioned, mounted on slides, and stained with hematoxylin and eosin to measure intestinal morphology or immunostained with Ki67 antibody to identify proliferating cells. The responses of SOS and COL20 were similar for all variables, except DNA FSR, which was greatest in SOS compared to all treatments. Villi length, width, crypt depth, total area of villi, lamina propria area of villi, and crypt cell proliferation of MR20 did not differ from SOS or COL20. Cellular proliferation of lamina propria of villi was highest in COL20 compared to MR20, COL10, MR10, and ZH, whereas cellular proliferation of the deep lamina propria was greatest in SOS, which differed from MR20, MR10, COL10, and ZH. Although MR20 increased villi growth and cellular proliferation compared to the 10% treatments, it did not stimulate similar morphological growth and development of the jejunum when compared to response of adequate COL intake (SOS and COL20) groups.

Colostrum induces jejunum development of neonatal piglets.

Graphical abstract

Graphical Abstract.

Introduction

The level of colostrum (COL) intake greatly influences a pig’s short and long-term growth, feed efficiency, development, and fertility (Theil et al 2014; Declerck et al 2016). Inadequate intake of COL, identified as less than 20% of birth body weight (BBW; Quesnel et al 2012; Suárez-Trujillo et al 2020), decreases passive transfer of immunity, leads to weight loss within the first 24 hours (Beckett et al 2026), and poor reproductive competency at puberty (Vallet et al 2015). Multivariable analyses of the impact of birth weight, birth order, and COL intake on piglet survival to 3 days postnatal and preweaning growth, identified COL intake as the predominant factor influencing growth (Quesnel et al 2023).

Upon birth, neonates must transition from continuous intravenous nutrition delivered through the umbilicus in utero to intermittent oral intake of milk as its means of ingesting nutrients. The piglet’s gastrointestinal tract is lined by fetal-type enterocytes at birth (Baintner 2002), which differentiate into mature enterocytes over the first days postnatal to establish a protective barrier and enable efficient absorption and digestion of nutrients (Rezaei et al 2013). Accompanying changes in enterocyte morphology and function, is extensive growth of the gastrointestinal tract, which in neonatal piglets, doubles in size during the first 2 to 4 d postnatal (Xu et al 2002; Zabielski et al 2008). The transition to oral intake of nutrients in the postnatal period drives growth and morphological changes of the gastrointestinal tract. Studies of neonatal piglets demonstrated COL feeding increases fractional synthetic rate of proteins by 3 to 4-fold in the jejunum and ileum compared to ingestion of water alone (Burrin et al 1992). Consumption of diets varying in macronutrient content differentially modify the function and morphology of the gastrointestinal tract (Schokker et al 2025; Wang et al 2025), thus diet composition can affect intestinal tract development and function.

Colostrum serves as a source of nutrients and non-nutritional bioactive factors, which promote immunocompetence and initiate developmental and metabolic pathways. Colostrum-specific factors function in the metabolic transition of neonates to the postnatal environment (Lin et al 2010). Low gastric acid and decreased proteolytic activity of the neonatal pig’s gastrointestinal tract likely preserve bioactivity of colostral proteins enabling their regulation of postnatal gastrointestinal development. Studies demonstrated growth factors in COL like epidermal growth factor, insulin-like growth factor 1 (IGF1), insulin-like growth factor II, and transforming growth factor-β stimulated growth of the gastrointestinal tract as well as other organs with these growth factors being approximately ten times greater in COL than mature milk (Donovan et al 2004).

Given that COL provides both nutrients and colostrum-specific factors that stimulate the growth and development of the gastrointestinal tract, and varying levels of intake relate to differences in long-term growth and feed efficiency of piglets, we aimed to understand the role of nutrient dose versus level of exposure to colostrum-specific factors on deoxyribonucleic acid (DNA) fractional synthetic rate (FSR) measured using deuterium oxide (D₂O) to metabolically label DNA, morphological development, and cellular proliferation quantified using immunostaining with Antigen Kiel 67 (Ki67), a marker of cellular proliferation, of the jejunum in the first 24 h postnatal. To accomplish this goal, female piglets were fed a diet of pooled COL at 20% (COL20) or 10% (COL10) of BBW, or a diet of commercial milk replacer (MR) at 20% (MR20) or 10% (MR10) of BBW. In addition, two other groups of piglets served as either the positive or negative control. Positive control piglets received ad libitum COL by suckling from sow (stay-on-sow; SOS) and negative control piglets were immediately euthanized after birth (zero hour; ZH). Analysis of the effect of diet (pooled COL versus MR) and dose of diet (20% versus 10%) were also conducted.

Materials and methods

Ethics statement and power analysis

Prior to beginning animal experiments, the study protocol was reviewed and approved by Purdue University Animal Care and Use Committee (Protocol No. 2110002200). Priori power analysis using an alpha error of 0.05, standard deviation of 0.25-fold, and 1.5-fold (50%) difference of any variable indicated a treatment group size of 10 was adequate to achieve 0.99 power. Notably, a higher power was selected due to an anticipated high level of biological variation.

Animals and study design

The samples and analysis described in this manuscript were part of a larger study, and details regarding sows, litters, and study replicates can be found in the previous publication (Beckett et al 2026). Briefly, 57 gilt piglets born to 35 primiparous and multiparous (parity 2.30 ± 1.43 average ± standard deviation) Yorkshire x Landrace sows terminally bred to Duroc sires were enrolled. Piglets weighing between 1.2 to 1.8 kg were identified immediately after birth, not allowed to suckle from the sow, and were allocated to one of six treatments (Fig. 1). The treatments included: 1) pooled COL bottle-fed every 2 h at 20% of BBW (COL20; n = 10); 2) COL bottle-fed every 2 h at 10% of BBW (COL10; n = 10); 3) Ralco (Marshall, MN) MR bottle-fed every 2 h at 20% BBW (MR20; n = 10); 4) Ralco MR bottle-fed every 2 h at 10% BBW (MR10; n = 10); 5) ad libitum COL intake by staying on the sow (SOS; n = 9); 6) no COL or MR and immediately euthanized at birth to serve as the negative control (ZH; n = 8). After weighing, SOS piglets were immediately returned to sows for ad libitum suckling, and all litters with SOS piglets were standardized to 12 piglets. Eight sows were used for the SOS piglets across both study replicates, and the average parity of the SOS sows was 2.20 ± 1.23 (average ± standard deviation). Piglets on the COL20, COL10, MR20, and MR10 treatments were removed from the farrowing house and transported to a nursery for bottle feeding of treatments. The nursey was equipped with pens with slotted flooring and heat lamps to prevent hypothermia. Piglets in the nursery were fed every 2 h for 24 h using Evenflo Feeding baby bottles (Evenflo, West Chester, OH). The pooled COL and MR diets were warmed to 37°C using a water bath prior to feeding. The COL, MR, and SOS diet nutrient composition can be found in Beckett et al (2026), and through Purdue University Research Repository DOI: doi: 10.4231/KN6Y-7S80 (Beckett et al 2025).

Figure 1.

a) Treatments, experiment timeline, and research questions to determine the impact of colostrum (COL) versus milk replacer (MR) feeding on piglet jejunum DNA fractional synthetic rate (FSR), villi morphology, and cellular proliferation. Treatments were colostrum fed at 20% (COL20) or 10% (COL10) of birth body weight (BBW), milk replacer fed at 20% (MR20) or 10% (MR10) of BBW, piglets stayed on the sow and received ad libitum colostrum (SOS), and piglets that did not receive food and were immediately euthanized after birth (zero hour; ZH). b) Microscopic image taken at 100X magnification to demonstrate villi height measurement in ImageJ. c) Microscopic image taken at 100X magnification to demonstrate villi width measurement in ImageJ. d) Microscopic image taken at 100X magnification to demonstrate crypt depth measurement in ImageJ. e) Microscopic image taken at 100X magnification to demonstrate total area of villi measurement in ImageJ. f) Microscopic image taken at 100X magnification to demonstrate lamina propria of villi measurement in ImageJ. Yellow lines on images represent the structure or area being measured. Created from Biorender.com.

Details regarding pooled COL collection were previously reported in Beckett et al (2026), but are described briefly herein. Pooled COL fed to COL20 and COL10 piglets was collected over a two-week period from ∼100 primiparous and multiparous sows located at a commercial sow farm in Indiana. Approximately 50 to 100 mL of COL was collected per sow, pooled into two separate batches for each study replicate, and frozen at -20°C until use. Then, 72 h prior to the initiation of each study replicate, the pooled COL was thawed at 4°C, and kept at 4°C throughout the study replicates. Prior to freezing and feeding, an aliquot of pooled COL was tested for porcine circovirus and porcine reproductive and respiratory syndrome, and results for both tests were negative. Additionally, there were no adverse health events at the Purdue University Animal Sciences Research and Education Center Swine Farm where animals were housed.

Immediately after birth and four h postnatal, piglets on SOS, COL20, MR20, COL10, and MR10 treatments were administered D₂O in a sterile saline solution (0.9% NaCl in D₂O, 20 mL/kg of BW) intraperitoneally to enrich body water to ∼3% to metabolically label DNA for FSR calculations. Deuterium oxide can be used as an isotopic marker to quantify newly synthesized DNA. Preliminary studies to determine adequate dose of D₂O to achieve ∼3% D₂O in circulation were conducted by Bitsie et al (2021). Zero-hour piglets received a dose of unlabeled sterile saline (0.9% NaCl in H₂O, 20 mL/kg of BW) immediately after birth, then were immediately euthanized. Prior to scheduled euthanasia, piglets were fasted for 2 h, including the SOS group, which were removed from the sow and transported to the nursery for fasting. Piglets were euthanized using an American Veterinary Medical Association approved method using CO₂ gas via a gradual fill method with constant exposure to 80% to 90% CO₂ concentration. Flow of CO₂ was maintained for at least five minutes after respiratory arrest. Brain death was confirmed by lack of eye reflex response.

Sample collection

Immediately after euthanasia, a blood sample was collected from the jugular vein into a serum separator tube (BD Vacutainer, Franklin Lakes, NJ). The sample was allowed to clot, then centrifuged at 10,000 × g for 10 min to separate serum from whole blood. Separated serum was stored at -80°C until extraction for D₂O analysis. Next, the body cavity was flushed with 60 mL of sterile saline (0.9% NaCl) through the abdominal aorta. Then, the entire small intestine was removed from the animal, uncoiled, and the jejunum was identified. A ten-centimeter piece of jejunum was cut from the entire organ, flushed with sterile 1X phosphate buffered saline (PBS) to remove digesta, and then a 2 cm piece of this tissue was placed in 50 mL of 10% neutral buffered formalin for 24 h. After 24 h, the 10% neutral buffered formalin was replaced with sterile 1X PBS. For measurement of DNA FSR, a separate ∼100 mg piece of jejunum tissue was immediately snap frozen in liquid nitrogen, then stored at -80°C until DNA extraction.

Measurement of enrichment of D₂O in serum

Enrichment of D₂O in serum of piglets was determined using gas chromatography mass spectrometry (GC-MS) in the Metabolite Profiling Facility in Bindley Bioscience Center at Purdue University. The protocol by Yuan et al (2008) was followed. Briefly, 20 μL of serum was combined with 2 μL of 10 N NaOH and 4 μL of 95:5 acetonitrile: acetone and left overnight at room temperature (RT). Next, 500 μL of chloroform and 500 mg of Na₂SO₄ were added, and samples were centrifuged at 14,000 rpm for 1 min. Then, 300 μL of the supernatant was transferred to a glass GC-MS vial for analysis. A serial dilution of external standards of 0%, 0.5%, 1%, 2%, 4%, 8%, and 16% D₂O in distilled water were prepared, extracted, and analyzed concurrently with samples. The intensity of acetone (m/z = 58) and acetone-d (m/z = 59) were monitored using an Agilent 5975 C series GC/MSD system (Agilent Technologies, Santa Clara, CA) with an Agilent 7890 A (GC; Agilent Technologies, Santa Clara, CA) and 7683 B injector equipped with an Agilent Select FAME GC column (50 m x 0.25 mm, fil thickness 0.25 um; Agilent Technologies, Santa Clara, CA). The GC carrier gas was helium with a linear flow rate of 1.0 mL/min. The programmed GC temperature gradient was as follows: 60°C for time 0 to 5 minutes, then ramped to 100°C at a rate of 20°C/minute, then ramped to 220°C at a rate of 50°C/minute with a 1-minute hold. Total run time per sample was 10.5 minutes. The GC inlet was 250°C and samples were injected in a split mode using a split ratio of 20:1. The MS source was set to 230°C, and MS data were collected in select ion mode. All data were analyzed with Agilent Chemstation software (Version E.02; Agilent Technologies, Santa Clara, CA). Acetone had a retention time of 5.1 minutes. Standards were plotted linearly and the enrichment of D₂O in each sample was interpolated using the ratio of the intensity of acetone-d to acetone.

Extraction and hydrolysis of DNA from jejunum tissue

Extraction and purification of DNA was completed using a gMax Mini Genomic DNA kit from IBI Scientific (IBI47281). Manufacturer instructions were followed except an additional wash step was added. DNA concentration, 260:230, and 260:280 ratios were determined using ThermoFisher Nanodrop 2000. The range of 260:230 for all samples was 1.62 to 2.46 and the range of 260:280 was 1.81 to 1.99.

DNA was digested to result in individual deoxyribonucleosides to measure the turnover of DNA using liquid chromatography tandem mass spectrometry (LC-MS/MS). Following the protocol by Quinlivan and Gregory (2008), DNA was diluted to 20 ng/μL with hydrolysis buffer (20 mM Tris-HCl, 100 mM NaCl, 20 mM MgCl₂). Then, 1000 ng of DNA was combined with 50 μL of DNA digestion buffer (hydrolysis buffer plus 250 units of benzonase [Sigma Aldrich E-1014], 300 units of phosphodiesterase [Sigma Aldrich P-3243], and 200 units of alkaline phosphatase [Sigma Aldrich P-7923]). Samples were incubated at 37°C for 6 h, then dried overnight in a speedvac and stored at -80°C until analysis. Samples were reconstituted with 100 μL of 5% methanol in 5 mM ammonium formate and transferred to a high-performance liquid chromatography (HPLC) vial. An Agilent 1290 infinity II ultra-high performance liquid chromatography system coupled to an Agilent 6470 series QQQ triple quadrupole mass spectrometer (Agilent Technologies, Santa Clara, CA, USA) was used to analyze the samples. A Waters Atlantis T3 (2.1mmx75mm, 3um) column (Waters Corporation, Milford, MA) was used for LC separation. Ammonium formate (5 mM) in HPLC grade water was used as mobile phase A. HPLC grade methanol was used as mobile phase B. The linear LC gradient was as follows: time 0 min, 0% B; time 2 min, 0% B; time 11 min, 100% B; time 12 min, 100% B; time 12.1 min, 0% B; time 15 min, 0% B. The flow rate was 0.3 mL/min. Multiple reaction monitoring was employed to identify the precursor ion of deoxyadenosine (m/z = 252.3), its fragment ions, deoxyribose (m/z = 117.1) and adenosine (m/z = 136), and possible deuterium incorporation into deoxyadenosine isotopomers. Data were acquired in positive electrospray ionization (ESI) mode. The jet stream ESI interface had a gas temperature of 330°C, gas flow rate of 7 L/min, nebulizer pressure of 45 psi, sheath gas temperature of 250°C, sheath gas flow rate of 7 L/min, capillary voltage of 4000 V in positive mode, and nozzle voltage of 1000 V. The ΔEMV voltage was 400. All data were analyzed with Agilent Masshunter Quantitative Analysis (Version 10.1; Agilent Technologies, Santa Clara, CA).

Calculation of DNA fractional synthetic rate

Calculation of FSR of DNA was determined following the procedure by Bitsie et al (2021) and Holmes et al (2015). Percent enrichment (p) of D₂O was determined for each pig, and n is the number of hydrogens that can be labeled by deuterium on the ribose portion of adenosine (dA). Fractional synthesis of DNA (f) and FSR were calculated from the mass isotopomer distribution of dA. In short, the following equations were used:

$$E{M}_{0t}=\%M_{0t}-\%M_0$$

First, EM_0t was calculated, which is defined as the difference between the labeled dA [M + 0] and the unlabeled dA [M + 0]. %M_0t is the dA [M + 0] of labeled piglets and %M0 is the average percent [M + 0] of dA for unlabeled animals (ZH animals; n = 8) in the study. Next, %M0* was calculated, which considers both p as the percent enrichment of D₂O and n as the number of hydrogens that can be replaced by deuterium in the ribose portion of adenosine. In this case, n = 7.

$$\%M0* = {\left(1-\mathit{p}\right)}^{\mathit{n}}\times \%M_0$$

$$EM0*=\%M0*-\%M_0$$

EM0* represents the %M0* that is standardized by the average percent [M + 0] of dA for unlabeled animals. This allows us to calculate the fractional synthesis of DNA (f) in the next equation.

$$f=E{M}_{0t}/ EM0*$$

Then, once we know f, FSR (unit = d⁻¹) can be calculated as the negative natural log of 1-f divided by time (t) in days (1).

$$\mathit{FSR} \left(d^{-1}\right)=-ln \left(1-f\right)/t$$

Processing and microtomy of jejunum tissue

Paraffin embedding, microtomy, and tissue staining were completed in Purdue University’s Histology Research Laboratory. A Sakura Tissue-Tek VIP6 tissue processor (Sakura, Torrence, CA) was used for tissue dehydration through graded ethanol concentrations, clearing in xylene, and infiltration with Leica Paraplast Plus (Leica, Wetzlar, Germany) paraffin for embedding. Tissue was sectioned at a thickness of 4 µm using a Themo HM355S microtome. Sections were mounted on charged slides and dried for 30 min in a 60°C oven.

Hematoxylin and eosin (H&E) staining

After drying, all slides were deparaffinized through three changes of xylene and rehydrated through graded ethanols to water in a Leica Autostainer XL (Leica, Wetzlar, Germany). After rehydration, slides were stained with Gill II hematoxylin for 5 min, then washed in two changes of nanopure water for 2 min each change. To remove excess dye, slides were dipped twice in 0.3% hydrochloric acid in 50% ethanol followed by a rinse in water for 1 min. Slides were blued in 0.2% ammonium hydroxide in water for 30 sec, then rinsed in nanopure water for 1 min. Next, slides were placed in 95% ethanol for 30 sec and then transferred to eosin Y phloxine B for 30 sec. Slides were dehydrated using increasing concentrations of ethanol, 70%, 95% and three changes of 100% ethanol for 2 min each change. Lastly, tissues were cleared in three changes of xylene for 3 min each change. After clearing, tissues were cover slipped using a xylene-based resinous mounting media.

Immunohistochemistry for Ki67, a marker of cellular proliferation

After deparaffinization, antigen retrieval was done with a Tris hydroxymethyl aminomethane (TRIS) and ethylenediaminetetraacetic acid (pH 9) solution in a BioCare decloaking chamber (BioCare Medical, Pacheco, CA) at a temperature of 95°C for 20 min. Slides were cooled for 20 min at RT and transferred to TRIS buffer with Tween 20 detergent (TBST). The rest of the staining was carried out at RT using a BioCare Intellipath stainer (BioCare Medical, Pacheco, CA). Slides were incubated with 3% hydrogen peroxide in water for 5 min to quench endogenous peroxidase activity, then rinsed with TBST, and incubated in 2.5% normal goat serum for 20 min to block non-specific binding of primary antibody. Excess reagent was blown off, and Ki67 primary rabbit monoclonal antibody (Cell Marque Tissue Diagnostics, 275R-16, Rocklin, CA) was applied at a dilution of 1:100 (0.364 µg/mL) for 30 min. The negative control slide was stained with Rabbit IgG (Vector Labs, I-1000, Newark, CA) at a concentration of 1:5000 (1 µg/mL) for 30 min. Slides were rinsed twice in TBST, and a goat anti-rabbit secondary antibody (Vector Labs, MP-7451, Newark, CA) was applied for 30 min. Slides were rinsed twice in TBST, and Vector ImmPACT DAB (Vector Labs, SK-4105, Newark, CA) was applied for 5 min. Slides were rinsed in water, and transferred to a Leica Autostainer XL (BioCare Medical, Pacheco, CA) for hematoxylin counterstain, dehydration, and coverslipping.

Imaging, measuring morphological structures, and quantifying cellular proliferation

Entire tissue image scans of three distinct cross sectioned areas of H&E stained and Ki67 stained jejunum tissue were captured at 200X magnification using an Automated Echo Revolution (Echo, San Diego, CA) microscope equipped with Echo Revolution software for multidimensional imaging. Images resulting from 200X magnified scans were used for quantitative analysis using ImageJ. Images were also captured at 100X magnification to illustrate morphological measures and Ki67 staining in Figs. 1 and 2, respectively. To determine the impact of treatment on jejunum morphological development, H&E images were analyzed using ImageJ software to measure villi length, width, total area, epithelial area, lamina propria area, and crypt depth (Fig. 1). Villi length was defined as the tip of the villus to the villus-crypt junction. Villi width was measured halfway up the villus, which was determined relative to the length measurement for the villus. Total area of the villi was measured using the outline tool in ImageJ and the entire villus was outlined. Lamina propria area of the villi was measured using the same outline tool in ImageJ, then epithelial area was calculated by subtracting the lamina propria area of the villus from the total area of the villus. Crypt depth was measured from the villus-crypt junction to the end of the crypt lining the basement membrane.

Figure 2.

Microscopic images taken at 100X magnification of jejunum tissue immunostained with Ki67, a marker of cellular proliferation. Two representative animals, one from stay-on-sow (SOS) treatment and zero hour (ZH), which serves as the negative control, are shown. Ki67 positive staining is identified as brown cells, whereas blue cells are not positive for Ki67. Images was annotated using ImageJ and Microsoft Powerpoint.

ImageJ IJ-Colour_Deconvolution2 plug-in (https://github.com/landinig/IJ-Colour_Deconvolution2/blob/main/colour_deconvolution2.jar; version 2.1) was used to determine the impact of treatment on cellular proliferation in images of Ki67 stained tissue. The IJ-Colour_Deconvolution2 plug-in quantifies the percent of nuclei positive for Ki67 staining (indicated by brown), and percent of nuclei not positive for Ki67 (indicated by blue). The threshold set to identify positive/brown stained cells was set by the user and then kept constant across all images. Three regions of interests were selected for analysis of percent Ki67 positive staining: lamina propria of villi, deep lamina propria, and crypts with the entire cross section of tissue analyzed in three sections per animal. Measurements were then averaged per region to obtain one value per variable per animal.

Statistical analysis

The first aim of this study was to determine the effect of treatment on variables of interest. To test the effect of treatment, data were analyzed using the PROC MIXED procedure of SAS 9.4 (SAS, Cary, NC). The following model was used:

$$Y_{ij}=\mu +a_i+b_j+e_{ij}$$

Y_ij is the response variable, µ is the overall mean, a_i is the fixed effect of i^th treatment, b_j is the fixed effect of the j^th study replicate, and e_ij is the residual. Outliers were evaluated using the “influence” command in SAS to generate Cook’s D values. Data points with a Cook’s D value above 10/n were considered an outlier and removed. Tukey’s honest significant difference post hoc analysis was used to separate treatment means.

The second aim of the study was to determine the effects of dose of diet (20% vs 10%), diet type (COL vs MR), and the interaction of dose of diet and diet type on variables of interest. The SOS and ZH treatments were removed from analysis. The following model was used:

$$Y_{\mathit{ijk}}=\mu +a_i+b_j+a{b}_{ij}+c_k+e_{\mathit{ijk}}$$

Y_ijk is the response variable, µ is the overall mean, a_i is the fixed effect of i^th diet, b_j is the fixed effect of the j^th dose, ab_ij is the fixed effect of the interaction of the i^th diet and j^th dose, c_k is the fixed effect of the k^th study replicate, and e_ijk is the residual. Outliers were evaluated using the “influence” command in SAS to generate Cook’s D values. Separation of means of the interaction were generated using Tukey’s honest significant difference post hoc analysis. Data from both analyses were considered significant at P ≤ 0.05 and a tendency was determined at 0.05 < P ≤ 0.10.

Results

The data discussed in this paragraph were previously reported in Beckett et al (2026), but are summarized herein to give context to findings in the current manuscript. Previous analysis of the impact of treatments on 24 h weight gain found piglets on SOS, COL20, and MR20 treatments gained a similar amount of weight (P > 0.05) over this period, increasing on average 0.17, 0.10, and 0.06 kg, respectively (Beckett et al 2026). SOS, COL20, and MR20 treatments were also able to maintain rectal temperature within a euthermic range (38.8 to 38.9°C), whereas piglets on COL10 and MR10 lost weight (-0.03 kg for both treatments) and had lower rectal temperature (38.0 and 38.6°C, respectively). Analysis of the overall effects of diet (COL vs MR) and dose of diet (20% vs 10% of BBW) found no effect of diet, but a significant impact of dose (P < 0.05) on body weight gain and rectal temperature (Beckett et al 2026). These findings indicate the level of intake, irrespective of diet composition, drives weight gain and thermogenesis the first 24 h postnatal. Notably, piglets on the 20% treatments grew and maintained euthermia despite differences in gross energy, protein, lactose, and triglycerides between COL and MR diets. Comprehensive analysis of diets described in Beckett et al (2026) and available through DOI: doi: 10.4231/KN6Y-7S80 (Beckett et al 2025), found COL contained more gross energy and protein, but less lactose and triglycerides than MR. Piglets on SOS and COL20 treatments had the greatest circulating protein (52.4 and 44.4 mg/mL, respectively; treatment P < 0.0001), which was different from COL10 at 31.6 mg/mL, MR20 (15.2 mg/mL), MR10 (13.6 mg/mL), and ZH (17.4 mg/mL), with the latter three treatments not different from one another. Circulating protein responses, our ongoing analysis of effects of treatments on jejunum and liver proteome, and Ito et al (2022) analysis of circulating proteome response to COL intake, indicate colostral proteins remain intact post absorption and are not used as a source for energy in the first 24 h postnatal. Rather, our findings (Beckett et al 2026) indicated that piglets fed MR relied on lactose for energy to support growth and remained in a glycolytic state, whereas COL-fed animals likely used milk lipids as their primary source of energy to support growth in the first 24 h postnatal.

Impact of treatment and diet and dose effects on DNA fractional synthetic rate

To determine the effect of dietary treatments on DNA FSR of the jejunum in the first 24 h postnatal, piglets were metabolically labelled with D₂O to achieve an average enrichment of 2.84 ± 0.46% (average ± standard deviation) in body water across all animals that received D₂O (excludes ZH piglets). The FSR of DNA isolated from the jejunum varied significantly by treatment, with SOS being 23% to 70% greater than all other groups (Treatment P < 0.0001; Fig. 3; Tukey P < 0.05; individual P-values calculated for Tukey’s post-hoc analysis provided in Supplemental Table S1). Piglets on the COL20 treatment had the second highest DNA FSR, but this tended to be not different (Tukey P = 0.08) from MR20, while still being greater (P < 0.05) than COL10 (Tukey P < 0.01) and MR10 (Tukey P = 0.0001) treatments. Analysis of the overall effects of diet (COL versus MR) indicated that COL feeding increased FSR of jejunum DNA over MR (P = 0.002; Fig. 3; Supplemental Fig. S1). Higher dose of intake also had an overall effect on FSR (P = 0.0003; Fig. 3; Supplemental Fig. S1) with piglets fed 20% of BBW having higher DNA FSR than those fed 10%. There was no significant interaction of diet and dose on jejunum DNA FSR (P = 0.58; Fig. 3).

Figure 3.

a) Treatment means for DNA fractional synthetic rate (FSR) of jejunum from piglets fed different levels of colostrum or milk replacer. Treatments were colostrum fed at 20% (COL20) or 10% (COL10) of birth body weight (BBW), milk replacer fed at 20% (MR20) or 10% (MR10) of BBW, or piglets stayed on the sow and received ad libitum colostrum (SOS). b) Effects of diet type (COL vs MR), dose (20% vs 10%), or their interaction on jejunum DNA FSR. Means without a common letter (a–c) differ (P ≤ 0.05).

Impact of treatment on jejunum morphological measurements and Ki67 analysis of cellular proliferation

Treatment strongly impacted all morphological metrics and cellular proliferation, as indicated by Ki67 staining (Treatment P < 0.05; Table 1). The SOS and COL20 treatments had longer, wider villi with greater total and epithelial area of villi compared to COL10 and MR10 (Tukey P < 0.05 for each variable, with individual P-values calculated for Tukey’s post-hoc analysis provided in Supplemental Table S1). The MR20 treatment was numerically intermediate, but not significantly different from SOS/COL20 treatment in terms of villi length, total area of the villi, and lamina propria area of villi (Tukey P > 0.05 Supplemental Table S1). The MR20 treatment was greater, but not different from COL10 and MR10 treatments for villi length, total area of the villi, and lamina propria area of villi, and not different from COL10, MR10, and ZH for villi width (Tukey P > 0.05 Supplemental Table S1).

Table 1.

Effect of treatment on villi morphological measurements and percent of Ki67 positive staining as identified by Ki67 immunostaining of the crypt, lamina propria of villi, and deep lamina propria. Treatments (trt) were colostrum fed at 20% (COL20) or 10% (COL10) of birth body weight (BBW), milk replacer fed at 20% (MR20) or 10% (MR10) of BBW, piglets stayed on the sow and received ad libitum colostrum (SOS), and piglets that did not receive food and were immediately euthanized after birth (zero hour; ZH). Two separate study replicates took place; therefore, study replicate was included in the model (rep). Data are shown as least squares means with standard error of the mean (SEM). Statistical significance was determined at P ≤ 0.05 and a tendency was determined at 0.05 < P ≤ 0.10. Treatment means without a common letter (a–d) significantly differ (P ≤ 0.05) based on Tukey’s separation of means. Tendencies are not represented by superscripts.

Variable	Means							P-value
	SOS	COL20	COL10	MR20	MR10	ZH	SEM	Trt	Rep
Villi Length, μm	651.8a	654.6a	422.4bc	565.8ab	515.1abc	395.2c	36.55	<0.0001	0.01
Villi Width, μm	74.6a	78.4a	70.7ab	69.3ab	59.4b	56.5b	3.36	0.0002	0.01
Crypt Depth, μm	59.2a	54.0a	52.5ab	60.0a	58.4a	43.4b	2.32	0.0002	0.001
Villi Length: Crypt Depth	11.1ab	13.1a	8.99b	9.63b	9.19b	9.63b	0.60	0.0001	0.36
Total Area of Villi, μm2	48,310a	48,573a	30,744bc	38,662ab	30,996c	21,758c	3689	<0.0001	0.02
Lamina Propria Area of Villi, μm2	20,340a	19,699a	13,158ab	18,330a	13,843ab	7691b	1852	0.0002	0.003
Epithelial Area of Villi, μm2	27,969ab	28,874a	17,587c	20,332bc	17,153c	14,066c	1984	<0.0001	0.12
Crypt, % Ki67 positive staining	54.6a	53.1a	49.0a	46.8ab	52.0a	38.3b	2.00	<0.0001	<0.0001
Lamina Propria of Villi, % Ki67 positive staining	29.8ab	32.3a	17.8c	18.5bc	16.3c	7.47c	2.80	<0.0001	0.02
Deep Lamina Propria, % Ki67 positive staining	13.8a	11.7ab	7.07 cd	8.48bc	5.7 cd	4.68d	0.80	<0.0001	<0.0001

Although there was a lack of difference in morphological measurements between MR20 and COL20 and SOS, the MR20 piglets had lower percent of Ki67 positive staining of lamina propria of villi (Tukey P = 0.01 COL20) compared to COL20, indicating lower proliferation of lamina propria cells in stromal areas of villi. When comparing MR20 to SOS, MR20 animals had lower percent of Ki67 positive staining of the deep lamina propria (Tukey P = 0.0004 SOS), and tended to have lower percent of Ki67 positive staining of the lamina propria of villi (Tukey P = 0.07 SOS). At birth (ZH piglets), piglets had numerically shorter, narrower villi with the smaller surface area and least cellular proliferation of lamina propria and deep lamina propria compared to animals that received food. However, villi width, epithelial area of villi, and percent of Ki67 positive staining of lamina propria of villi of ZH piglets did not statistically differ from MR20, MR10, or COL10 (Tukey P > 0.05; Supplemental Table S1). Response of ZH piglets was distinctly different from SOS and COL20 for these variables (Tukey P < 0.05; Supplemental Table S1), as well as for villi length, total area of villi, lamina propria area of villi, and lamina propria of villi and deep lamina propria measured for percent of Ki67 positive staining (Tukey P < 0.05; Supplemental Table S1).

Despite these differences in morphological development of the villi due to treatment, the crypts did not demonstrate divergent responses. Crypt depth was similar for all treatments, except when SOS, COL20, MR20, and MR10 were compared ZH (Tukey P < 0.05; Supplemental Table S1) which indicated the intake of food, except COL10 treatment, increased crypt depth. The percent of Ki67 positive staining of crypts were similar for all treatments (Tukey P > 0.05; Supplemental Table S1), except ZH piglets (Tukey P < 0.05; Supplemental Table S1). Crypt depth at birth (ZH piglets), greatly differed from SOS, COL20, COL10, and MR10 (Tukey P < 0.05; Supplemental Table S1) with ZH piglets having the least proliferation, indicating oral intake of nutrients, except MR20 treatment, induces crypt proliferation.

Overall effect of diet, dose, and diet by dose interaction on villi morphological measurements and Ki67 analysis of cellular proliferation

Diet (COL vs MR) had an overall effect on villi width (P = 0.006; Fig. 4; Supplemental Fig. S2), crypt depth (P = 0.003; Fig. 4; Supplemental Fig. S2), epithelial area of villi (P = 0.05; Fig. 4; Supplemental Fig. S2) and percent of Ki67 positive staining of the lamina propria of villi (P = 0.0009; Fig. 5; Supplemental Fig. S3) and deep lamina propria (P = 0.01; Fig. 5; Supplemental Fig. S3). Piglets fed COL had wider villi with more epithelial surface area, shallower crypts, and more proliferation compared to piglets fed MR. The response of diet was primarily driven by the magnitude of difference between COL20 and COL10 being greater than the magnitude of difference between MR20 and MR10. The other variables were not significantly affected by diet (P > 0.10; Figs. 4 and 5; Supplemental Figs. S2 and S3).

Figure 4.

The impact of colostrum versus milk replacer and different doses of diet on a) villi length, b) villi width, c) total area of villi, d) lamina propria area of villi, e) epithelial area of villi, and f) crypt depth. Treatments are colostrum fed at 20% (COL20) or 10% (COL10) of birth body weight (BBW) or milk replacer fed at 20% (MR20) or 10% (MR10) of BBW. Data are presented as least squares means of the interaction of diet and dose. The P-value for the effect of diet (diet; COL vs MR), dose of diet (dose; 20% vs 10%), and the interaction (D x D) for each response variable is shown. Statistical significance was determined at P ≤ 0.05, and a tendency was determined at 0.05 < P ≤ 0.10.

Figure 5.

The impact of colostrum versus milk replacer and different doses of diet on percent of Ki67 positive staining of a) lamina propria of villi, b) deep lamina propria, and c) crypts. Treatments are colostrum fed at 20% (COL20) or 10% (COL10) of birth body weight (BBW) or milk replacer fed at 20% (MR20) or 10% (MR10) of BBW. Data are presented as least squares means of the interaction of diet and dose. The P-value for the effect of diet (diet; COL vs MR), dose of diet (dose; 20% vs 10%), and the interaction (D x D) for each response variable is shown. Statistical significance was determined at P ≤ 0.05 and a tendency was determined at 0.05 < P ≤ 0.10.

Dose (20% vs 10%) of diet also impacted jejunum morphological measures, and cellular proliferation measured using Ki67 immunostaining. Dose significantly influenced all morphological and proliferation variables (maximum P < 0.02 for villi width; Figs. 4 and 5; Supplemental Figs. S2 and S3), except crypt depth and percent of Ki67 positive staining of the crypts (P > 0.10; Figs. 4 and 5). Piglets that consumed 20% of their BBW had longer (P < 0.0001), wider villi (P = 0.02) with greater total (P = 0.0002), epithelial (P < 0.0001), and lamina propria (P = 0.006) surface area compared to piglets fed 10% of BBW. Piglets that consumed 20% of BBW also had a greater percent of Ki67 positive staining of the lamina propria of villi (P = 0.0005) and deep lamina propria (P < 0.0001) indicating greater cellular proliferation compared to piglets fed 10% of BBW.

Significant interactions between diet and dose were found for villi length (P = 0.002; Fig. 4), epithelial area of villi (P = 0.007; Fig. 4), and the percent of Ki67 positive staining of the lamina propria of villi (P = 0.004; Fig. 5) and crypts (P = 0.03; Fig. 5). There was a tendency for a diet by dose interaction for total area of villi (P = 0.06; Fig. 4). For all significant variables, the interaction was driven by the COL20 treatment. The COL20 treatment had the longest villi with the greatest surface area and greatest amount of proliferation in the lamina propria and crypts. For all variables, except for percent of Ki67 positive staining of crypts, MR20 was intermediate. The other variables: villi width, crypt depth, lamina propria area, and percent of Ki67 positive staining of the deep lamina propria were unaffected by the interaction of diet and dose (P > 0.10; Figs. 4 and 5).

Discussion

Higher levels of COL intake are associated with greater growth trajectory and feed efficiency in swine, indicating that varying COL intake levels likely impact nutrient absorption and metabolism. Our aim was to determine the impact of level of COL and MR feeding on DNA FSR and villi growth and cellular proliferation of the jejunum in the first 24 h postnatal because the jejunum is the primary site for nutrient absorption. Findings demonstrated that COL relative to MR potentiated the maturation of the jejunum by increasing villi size and surface area, along with inducing cellular proliferation and DNA FSR, but this response was only realized with an adequate dose of COL (COL20 or SOS). Although MR20 resulted in similar weight gain as SOS and COL20 in the first 24 h postnatal, and increased villi growth and cellular proliferation compared to the 10% treatments, it did not stimulate similar morphological growth and development of the jejunum when compared to response of adequate COL intake (SOS and COL20) groups. Together, findings add to the body of literature that support components unique to COL function to potentiate the postnatal development of the jejunum.

Feeding COL at an adequate level (COL20 or SOS) induced villi growth (length and width) and surface area of villi (total, epithelial, and lamina propria) compared to all other treatments. Growth of the jejunum within the first 3 d postnatal is dependent upon an increase in gastrointestinal tract blood flow, disappearance of transverse furrows (Skrzypek et al 2010), and epithelial cell turnover (Zabielski et al 2008). Increased blood flow enables elongation of the villi and promotes disappearance of transverse furrows present on the apical-most portion of villi, which appear to inhibit growth of villi (Skrzypek et al 2010). Colostrum and transition milk consumption in the first 3 d postnatal likely play a role in the disappearance of transverse furrows as COL mediates epithelial turnover (Skrzypek et al 2010). Epithelial cell turnover is a function of mitotic and apoptotic mechanisms, and COL contains factors, such as growth factors and cytokines, that mediate these processes. Proteome analysis of the pooled COL diet fed to neonates found high levels of IGF1 and IGF-binding proteins (Beckett et al 2026). Studies of neonatal calves found that factors specific to COL increase IGF1 systemically (Fischer-Tlustos et al 2020). Although IGF1 alone cannot replace the effects of COL in stimulating gastrointestinal tract development of neonatal pigs, IGF1 functions to stimulate gastrointestinal tract cell proliferation (Xu et al 1994). Greater DNA FSR and proliferation index of cells within the lamina propria of villi and deep lamina propria of the COL20 and SOS piglets, reflect the growth-promoting aspects of COL on the gastrointestinal tract and likely underlie the greater surface area of the villi of these groups, as indicated by increased lamina propria area and total area of villi.

Although COL increased cellular proliferation relative to MR, indicating that colostrum-specific factors promote growth and development of the jejunum, the level of intake of both the nutrients and unique bioactive factors are needed for realization of this potential. The low dose of COL, COL10, did not match any of the responses of COL20 piglets for any of the variables used to evaluate jejunum growth and development. A potential mechanism for colostrum stimulated growth may be through mechanistic target of rapamycin complex 1 (mTORC1), which mediates cell growth, functioning to integrate environmental and intracellular cues. Activation of mTORC1 requires converging cell signaling pathways, indicative of both adequate nutrients, like amino acids and glucose, and endocrine signals, such as insulin and IGF1 (Condon and Sabatini 2019). Enterically feeding a bolus meal of MR to young piglets (5 to 7 d old) stimulated mTORC1 activity in the muscle, liver, jejunum, and pancreas, which was associated with increased protein synthesis (Gazzaneo et al 2011). Comparison of neonatal piglets that received only water with those fed COL, nutrient-matched formula, or mature sow’s milk, found feeding formula or sow’s milk induced a 2.5- to 3-fold increase in skeletal muscle protein synthesis compared to water, with an accompanying accretion of ribosome and total polyadenylated RNA, markers of mTORC1 activation (Fiorotto et al 2000). Notably, COL feeding induced 28% more protein synthesis than either the formula or mature milk (Fiorotto et al 2000). Thus, we interpret our findings in the context of work of others, that the dose of the diet must be adequate to promote growth, potentially mediated through mTORC1, and non-nutritional factors within COL play a critical role in mediating a maximal growth response in the jejunum.

Although there were no differences in crypt depth and proliferation of crypt cells between COL-fed animals and MR-fed, total epithelial area was greater in COL20 and SOS piglets compared to COL10 and MR-fed animals. Lack of a detected difference in percent of proliferating crypt cells may indicate the greater area was due to hypertrophy, without hyperplasia. A potential explanation for this phenomenon is an increase in enterocyte size due to vacuolization of nutrients (Baintner 2002). Unlike mature enterocytes, the fetal enterocytes lining the villi of neonatal piglets have the endocytic capacity to uptake macromolecules, including whole proteins, from the lumen of the gastrointestinal tract. Vacuoles can occupy half of the cell volume during colostral protein uptake ultimately increasing the total volume and height of the cell (reviewed by Skrzypek et al 2007; Zabielski et al 2008).

In some capacity, all the variables were influenced by diet, dose, or the interaction of diet and dose. Simply feeding piglets at 20% of BBW increased length and area of the villi and cellular proliferation of lamina propria of villi and deep lamina propria compared to feeding at 10% of BBW, but this response was further elevated due to feeding COL at 20% of BBW. The elevated response of COL20 compared to MR20 was likely due to factors discussed earlier, like growth factors and cytokines, that mediate cell division and apoptosis, and vacuole-mediated uptake of macromolecules influencing cell size, and in turn, epithelial area and total area of villi. These responses demonstrate that MR feeding can provide nutrients needed for growth, but cannot replicate the colostrum-specific factors that advance development.

Also notable to discuss was the finding that DNA FSR of SOS was significantly greater than COL20, whereas there was no difference in Ki67 markers of cell proliferation between these groups. Measuring DNA FSR and Ki67 immunostaining to identify proliferating cells of interest represent distinct approaches for capturing growth response of the jejunum to dietary treatments. Included in this difference was Ki67 focused on specific cell populations and locations, whereas DNA FSR measured rate over the first 24 h postnatal in all cells within jejunum, which includes, epithelial cells, stromal fibroblasts, endothelial cells, immune cells, smooth muscle cells that encase the tract, and cells of the enteric nervous system. The difference in the groups may reflect that we did not measure all cell populations in the Ki67 approach, as well as biological or technical factors that differed between SOS and COL20 diets. The differences between DNA FSR and Ki67 responses of SOS and COL20 piglets could be due to the origin of COL, as COL came from different farms and sows with different diets, with pooling of COL aiming to minimize some variation due to sow. Alternatively, or additionally, different responses of DNA FSR and the Ki67 approach may be due to the handling of diets with COL20 being subject to freeze-thaw that can affect the activity of COL components, like proteins and cells.

Conclusion

Nutrient ingestion alone affects growth and development of the jejunum in the first 24 h postnatal, with a higher level of intake inducing more changes. However, maximal growth and development appear to be dependent on colostrum-specific components. Relative to MR, COL induces jejunum DNA FSR, villi growth, and cellular proliferation, but maximal stimulation was only realized with higher doses of COL (COL20 and SOS). Higher doses of MR (MR20) induced growth of villi, but not to the extent as the equivalent dose of COL (COL20). Future studies should focus on identifying factors in COL that regulate development of the jejunum.

Glossary

LIST OF ABBREVIATIONS: 

BBW =

birth body weight

COL =

Colostrum

COL20 =

colostrum fed at 20% of birth body weight

COL10 =

colostrum fed at 10% of birth body weight

dA =

ribose portion of adenosine

DNA =

deoxyribonucleic acid

D₂O =

deuterium oxide

ESI =

electrospray ionization

f =

fractional synthesis

FSR =

fractional synthetic rate

GC-MS =

gas chromatography mass spectrometry

H&E =

hematoxylin and eosin

HPLC =

high performance liquid chromatography

Ki67 =

Antigen Kiel 67

LC-MS/MS =

liquid chromatography tandem mass spectrometry

MR =

milk replacer

MR20 =

milk replacer fed at 20% of birth body weight

MR10 =

milk replacer fed at 10% of birth body weight

n =

number of hydrogens that can be replaced by deuterium

p =

percent enrichment of D2O

PBS =

phosphate buffered saline

RT =

room temperature

SOS =

stay-on-sow

TRIS =

Tris hydroxymethyl aminomethane

TBST =

tris buffer with tween 20 detergent

ZH =

zero hour

Supplementary data

Supplementary data is available at Translational Animal Science online.

Funding

This work is supported by the Agriculture and Food Research Initiative grant no. 2022–67015-36199 from the U.S. Department of Agriculture’s National Institute of Food and Agriculture. This work is also supported by the Agriculture and Food Research Initiative-Education and Workforce Development Postdoctoral Fellowship grant no. 2024–67034-42236 from the U.S. Department of Agriculture.

Disclosures

The authors have no disclosures.

Conflicts of interest

The authors have no conflicts of interest.