1. Introduction
The need to better understand intestinal permeability is growing in the livestock industry as producers seek to prevent or reduce the magnitude of enteric disease. Normally, the intestinal barrier is selectively permeable and controls the passage of water, nutrients, electrolytes, and other essential ions [
1]. Many factors such as age, diet, shifts in intestinal microbiota, infection, disease, environment, and stress can impact the intestinal barrier, leading to increased permeability [
2]. Increasing the rate of non-mediated passive diffusion or intestinal permeability, through or between intestinal epithelial cells, can lead to decreased performance, enteric disease, and systemic disease [
2,
3]. Increased intestinal permeability through different stressors in poultry may lead to decreased production and performance, compromised health, and lameness [
4]. It is known that stressors, such as toxins, parasites, dysbiosis, and environmental stressors (heat stress and feed restriction) can also be causes of intestinal stress leading to disruption of the intestinal tract [
1,
4,
5,
6,
7,
8,
9]. In broilers, it has been observed that increased intestinal permeability can lead to poor feed conversion, lower body weights (BW), and increased incidences of lameness and other skeletal diseases [
2,
4]. Therefore, identifying a consistent method to determine intestinal permeability in non-disease states that can be used on commercial farms is critical for understanding and improving intestinal health and overall performance of poultry.
A variety of in vivo methods for measuring intestinal permeability are available for mammals. In human medicine and a variety of animal species, such as dogs, rats, and pigs, sugars such as lactulose, mannitol, rhamnose, and sucralose are used in dual or multi-sugar tests to measure intestinal permeability in urine [
2,
10,
11,
12]. The variation in size and molecular weight of these sugar molecules allows researchers to better estimate the degree of permeability by measuring the relative amounts of each sugar that has passed through the intestinal barrier into the bloodstream then into the urine. It is thought that larger molecules such as lactulose and sucralose are more likely to move paracellularly through the tight junctions, while monosaccharides such as rhamnose and mannitol move transcellularly through the enterocytes at a more uniform rate [
13]. Because avian species lack the excretion of urine, this method has only been sparsely explored in poultry [
14]. Another in vivo method of measuring intestinal permeability is the use of fluorescein isothiocyanate-dextran (FITC-D), which is a non-digestible, fluorescently labeled dextran molecule that has been validated for measuring intestinal permeability in poultry and is widely used in other species [
15,
16,
17,
18]. The size of FITC-D is too large to pass through the intestinal barrier under normal conditions and moves from the lumen into the bloodstream via passive paracellular transport during times of stress, inflammation or infection [
17]. While both methods have been used as biomarkers to assess intestinal permeability in poultry, it is crucial to directly compare these methods and determine a method that can be used to measure intestinal permeability in a commercial setting. Therefore, the primary objective of this study was to compare these two different methods, serum FITC-D and serum lactulose, mannitol, and sucralose (LMS) in measuring intestinal permeability in poultry in the presence or absence of a known stressor over the 6-week production cycle in broilers.
2. Materials and Methods
All procedures involving live animals were approved by the Iowa State University Institutional Animal Care and Use Committee (IACUC number 18–331). Two independent but consecutive trials were conducted from March to June 2019 at the Iowa State University Poultry Teaching and Research Farm. The experimental design was the same for both trials with the only difference being the genetic strain of bird used due to availability of lines at a local commercial hatchery. Broiler chicks (
n = 500 per trial) were obtained from a local commercial hatchery (Whelp Hatchery, Bancroft, IA, USA) and transported to the Iowa State University Poultry Teaching and Research Farm. Chicks were commonly raised for the first 7 days. At 7 days of age, 480 chicks were weighed and randomly allocated to 40 (1.2 × 1.2 m) floor pens with 12 birds per pen. Pens were the experimental unit and designated to 1 of 2 fed states (fed or fasted) and 1 of 4 sugar treatments (FITC-D only, LMS only, FITC-D+LMS (FITC-LMS), or water). All pens were sampled at 3 time points (14, 28, and 42 days of age (doa)) which resulted in 5 pens per fed state and sugar treatment at each time point. See
Figure 1 for a detailed pen layout of the experimental design. Typical management strategies were followed regarding temperature and lighting schedules. Birds were allowed ad libitum access to feed via hanging feeders, except for pens designated as fasted, which had feeders removed 12 h before the administration of sugars. Diets can be found in Supplemental
Table S1. Birds were allowed ad libitum access to water via a nipple drinker system at all times. The same starter (day 0–14), grower (day 15–28), and finisher (day 29–42) diets were provided to all birds with sampling of the birds occurring on the last day of each diet phase.
2.1. Intestinal Permeability
During each 6-week trial, three sample collection days occurred. Twelve hours before sampling, feed was removed from pens designated as fasted (n = 20). On days of sample collection, pen weights were collected and used to determine gavage-treatment dosage. On average, 4 birds were randomly chosen from each pen and orally gavaged with 1 mL of the designated treatment for the pen: FITC-D only, lactulose, mannitol, and sucralose (LMS) only, FITC-D+LMS (FITC-LMS), or control/no sugar treatment (water). Due to high mortality during the first trial, the number of biological trials was reduced in 11 of the 40 pens, which had less than 4 birds at various sampling days, whereas only 5 of the 40 pens had less than 4 birds in the second trial. These mortalities during the first 7 days were primarily a result of poor starting chicks or yolk sac infections.
2.2. Sugar Treatments
FITC-D with an average molecular weight of 3000 to 5000 (FD4, Sigma Aldrich, St. Louis, MO, USA) was dissolved in distilled water and administered at a rate of 8.32 mg FITC-D/kg BW [
17]. A mixture of lactulose, mannitol, and sucralose (Sigma Aldrich, St. Louis, MO, USA) was dissolved in distilled water and given at rates of 0.25 g lactulose/kg BW, 0.05 g mannitol/kg BW, and 0.05 g sucralose/kg BW [
14]. The combination of FITC-D and LMS was administered at the same rates as stated above, but as a single solution. Water was administered as the control solution. Once orally gavaged, birds were marked with animal-safe paint to denote treatment and returned to their home pen. One hour after oral gavage, birds were euthanized via CO
2 asphyxiation and death was confirmed by checking the nictitating eyelid response. Blood samples were collected from all birds from the femoral artery into collection tubes (BD Medical, Franklin Lakes, NJ, USA) and transported back to the laboratory. Blood was allowed to clot at room temperature followed by centrifugation at 1000×
g for 15 min. Following separation, serum was portioned into 0.5 mL aliquots and stored in amber tubes at −80 °C until analysis.
2.3. FITC-D Analysis
Methods for analysis of FITC-D in blood serum were modeled from a previously described and optimized protocol from Baxter et al. [
18]. Briefly, serum was analyzed at the conclusion of both trials and serum from control birds was combined and diluted at a 1:5 ratio with phosphate-buffered saline (PBS) and used to create a standard curve. Serum samples from birds that received treatments were also diluted at a 1:5 ratio using PBS. An amount of 100 µL of diluted sample from each bird or standard curve was plated in triplicate on a black 96-well plate. Fluorescence was measured using a BioTek Cytation spectrophotometer (BioTek US, Winooski, VT, USA) with excitation and emission wavelengths of 485 and 528 nm, respectively. Average concentration of FITC-D in serum was calculated as ng/mL per bird. Results were averaged per bird for data analysis.
2.4. Lactulose, Mannitol, and Sucralose Analysis
All individual serum samples were analyzed for lactulose and mannitol using a modified method described by Fleming et al. [
19] at the conclusion of both trials. In brief, serum samples were diluted at a rate of 1:4 dilution with ultra-pure water. The solution underwent centrifugal ultrafiltration, using a Nanosep filtration device with a 30 K membrane, removing insoluble particles. Samples were then analyzed for blood sugars on an ion chromatograph (Dionex ICX-6000, ThermoFisher Scientific, Waltham, MA, USA) using carbohydrate disposable working electrodes. The average run time for each sample was approximately 60 min. Several samples of lactulose lacked detectable limits in the serum. For analysis, these samples were assigned the arbitrary value of 1. Additionally, sucralose was not detected in any samples and, therefore, sucralose was not included in any additional analysis.
2.5. Statistical Analysis
Serum data for each sugar were analyzed using the PROC GLIMMIX procedure in SAS Cary, NC, USA; [
20] and analyzed for the fixed effects of day (14, 28, or 42), fed state (fed or fasted), and sugar treatment (FITC-D, LMS, FITC+LMS, or control). All 2- and 3-way interactions between day, fed state, and sugar treatment were analyzed. The replicate (trial 1 and trial 2) was fit as a random effect and the pen (1 through 40) was fit as a repeated measure. The model distribution was set as a lognormal distribution. Residuals were tested for normality and homoscedasticity using Proc Univariate when following a lognormal distribution [
20]. Least-squares (LS) means were calculated for all effects using the LSmeans statement and all LSmeans presented were back-transformed for publication.
p-values for differences of the LSmeans were generated using the pdiff option and the Tukey procedure was performed to account for multiple comparisons. Statistical significance was set at
p < 0.05.
4. Discussion
Intestinal health is a concern in the poultry industry as it encompasses anything from enteric disease to reduced performance. Because poultry producers aim to maximize production and efficiency, new methods for monitoring and measuring sub-clinical departures in intestinal health are crucial. Unlike other species, the lack of urine excretion has limited non-lethal methods of determining intestinal health, particularly intestinal permeability from directly translating across other species. This study aimed to examine the effectiveness of commonly used methods to determine intestinal permeability in livestock across the broiler production cycle.
First, the use of FITC-D to measure intestinal barrier permeability has been widely used in poultry research and can detect extreme changes in intestinal permeability. This relatively simple to perform assay requires the use of a fluorescently labeled maker that can immediately be read using a fluorescent plate reader and has great potential for use in a commercial facility due to ease of administering and measuring. The present study identified that birds which were gavaged with FITC-D exhibited the highest relative serum fluorescence compared to birds that did not receive FITC-D. However, it should be noted that control birds had relatively high fluorescence levels demonstrating a high level of autofluorescence of the serum. This has been observed in other studies and was an underlying need for optimization [
17,
18,
21]. As expected, broilers that were fasted also had higher serum fluorescence compared to fed broilers. The ability to detect these changes was supported by Baxter et al. [
17] who used fasting as a method to optimize the FITC-D protocol. When comparing the two-way interaction of fed state and sugar provided, we observed that birds provided FITC-D had higher fluorescence regardless of fed state, indicating that the assay could not detect a state of expected intestinal permeability. Additionally, both LMS (containing no FITC-D) and FITC-LMS gavaged birds had elevated fluorescence in the fasted birds compared to fed birds. While these results disagree with other reports [
3,
4,
14,
17,
18], where serum FITC-D concentrations were elevated in fasted broilers provided FITC-D, it should be noted that these studies did not have the two-way interaction of fed state and sugar treatment. These studies only compared fed versus fasted with all birds receiving FITC-D [
3,
4,
17]. This is comparable to the main effect of dietary treatment in this study where a difference was observed statistically. While a numerical increase in FITC-D in the serum was observed with fasted birds compared to control fasted birds, the inability of this study to detect these differences is likely a statistical issue. The standard error of the means in this study ranged between 6.84 and 20.96 ng/mL despite a minimum of 40 birds used per treatment within the three-way interaction. Therefore, the inability of FITC-D to detect differences in fed state and sugar treatment for the FITC-D birds and the high variation makes it a less appealing method to be used in commercial settings as small differences could not be detected. Additionally, this method would not provide consistent differences unless a flock is undergoing a major enteric challenge.
Second, mannitol is a small monosaccharide used in combination with the disaccharide, lactulose, to determine intestinal permeability in urine of many mammalian species. In using the lactulose to mannitol ratio, it is expected that mannitol will pass at a constant rate across the intestinal epithelium of all birds administered mannitol. However, this was not observed in this study. In this study, a greater concentration of mannitol in fed birds was observed compared to fasted. Additionally, mannitol concentrations were observed in broilers that did not receive mannitol. Comparison of differences in collection of urine versus blood highlight the fact that urine is a filtered product where many substrates, particularly glucose, are reabsorbed into the blood [
22]. Therefore, glucose and other small molecules would be included in blood but not urine. Furthermore, comparisons of the molecular size of glucose (180 mol/g) to mannitol (183 mol/g) and the chemical composition of glucose (C
6H
12O
6) to mannitol (C
6H
14O
6) demonstrate that the differences are relatively minimal. Additionally, an ion chromatograph run spiked with glucose showed glucose elutes near the same time as mannitol, confirming this suspicion (analysis not shown). This, along with the poor chromatogram peak separation of mannitol, might indicate that the elevated “mannitol” concentrations likely include glucose. Based on these data, the use of mannitol and the subsequent lactulose to mannitol ratio would not be advised in birds unless methods are used that improve the elution of mannitol from other substances.
Third, lactulose is a disaccharide molecule, is larger than mannitol, and is expected to pass through the intestinal barrier when intestinal stress is elevated, unlike mannitol. Lactulose is expected to pass through the intestinal barrier at a similar rate to FITC-D. Therefore, it is not surprising that this method had similar results to the results observed with FITC-D administration, with both being elevated in serum during fasting. Unlike with the serum concentration of FITC-D, this study observed the majority of samples from control broilers having no detectible concentrations of lactulose in serum, demonstrating the selectivity of the assay. This is essential when determining if the assay can be used for determination of intestinal permeability in commercial broiler production. The lack of detection of lactulose in birds not administered lactulose establishes a clean background unlike the FITC-D assay which has a high background signal. Additionally, the ability to observe an average 73.8% increase in lactulose in fasted and fed birds across the sugar treatments is a greater percent increase compared to the FITC-D assay (19.5%). Unlike the FITC-D assay, the lactulose assay produced consistent differences between fasted and fed birds across different ages, indicating the potential robustness of this assay for a commercial setting. However, this study consisted of two replications in non-commercial settings. Additional studies would need to be conducted to determine if other known intestinal stressors can elicit a similar result and if a range can be established to determine optimal, sub-optimal, and detrimental concentrations of lactulose. Lastly, it should be mentioned that when comparing differences in the methods, determination of serum lactulose does require significantly more time, expertise, and equipment to determine serum concentrations due to the use of the ion chromatograph.
While age was not a major focus of this experiment, age remains a consideration when trying to understand intestinal permeability and determine optimal methods to measure intestinal permeability. In this study, we observed a significant and consistent decrease in all markers of intestinal permeability at 28 doa in both trials. The concentration of sugars observed at 14 doa indicated a higher degree of intestinal permeability in younger birds. This may be due to the fact that the gastrointestinal tract and intestinal microbiota are not fully matured and developed, respectively, during the earlier weeks of life. Interestingly, serum concentrations of lactulose increased at 42 doa compared to 28 doa. These trends were observed over both trials leading to questions about what might be occurring in the broiler between 28 and 42 doa that leads to these differences. These results were not expected but may be explained by shifts in the intestinal microbiota. A study by Lu et al. [
23] identified changes in intestinal microbiota at various ages in broilers. They discovered that microbial populations and microbial community structure were the most stable during 14 to 28 days of age and then had significant shifts around 49 days of age. An additional study by Liao et al. [
24] confirms changes in intestinal morphology in growing broilers. Increased permeability at day 42 may also result from general stress or heat stress caused by heavy body weights during the finisher period [
25,
26].