There is a need for robust, non-invasive and simple methods to objectively assess the quality of routine animal cares of commercial poultry flocks as a component of sustainable and ethical animal production practices (Castro et al., 2023; Siegel, 1995). Corticosterone (CORT) is an endogenous glucocorticoid produced in the adrenal glands and serves as the stress hormone in birds (Scanes, 2016). Corticosterone alters metabolic pathways and behavioral responses, redirecting energy reserves to cope with both short- and long-term stressors (Scanes, 2016). Blood, feces, and feathers have been used as sampling sites for CORT concentration measurements (Siegel, 1995, Scanes, 2016). First used by wildlife biologists (Siegel, 1995), feathers are an attractive sample source due to their ease of collection, analyte stability within the feather during room temperature storage (Bortolotti et al., 2008), and ability to reflect CORT concentration over the period of feather growth. The simplicity and non-invasive nature of sampling as well as the ability to collect casual samples as opposed to blood sampling made during complex welfare assessment scenarios further emphasizes the desirability of validating reliable methodology to support use of feather CORTs as part of welfare assessments.
A methanol extraction-based feather CORT assay protocol was originally developed and published by Bortolotti et al., for use with RIA and subsequently adapted and validated for ELISA assay by Carbajal et al. (Bortolotti et al., 2008; Carbajal et al., 2014). These publications reported technical coefficients of variations of 15.7 % using RIA and 17.2 % using ELISA (Bortolotti et al., 2008; Carbajal et al., 2014). In 2019, Alba et al. introduced the use of keratinase to digest the feathers before solid-phase extraction (SPE) of CORT and its subsequent quantification by RIA assay with reported variations of less than 5 % (Alba et al., 2019). In our hands, the method of Bortolotti et al. used in conjunction with a different ELISA assay than was used by Carbajal et al. produced unacceptable average variation of 93 %. Similarly, use of this same ELISA kit to analyze feather CORT extracts obtained using Alba’s protocol generated average variances of 33 % for technical replicates. It was not immediately obvious why the variance was so high as that assay routinely provided standard curves with r2 of 0.999; the unexpected variability prompted the current method improvement study.
A series of studies using a common pooled-feather sample was used to evaluate sample extraction and solubilization steps as likely sources of excess variation. Reduction in the variability of technical replicates when assayed using a validated ELISA CORT assay served to indicate protocol improvement. Subsequently, the improved protocol, which reduced average variation of technical replicates to 2.5 – 7.2 %, was used to determine the variability and evaluate potential correlation of plasma and feather CORT concentrations in different chicken strains as component of a larger data set from a separate study.
MATERIALS AND METHODS
All chemicals were reagent grade or better and obtained from VWR International.
ANIMAL METHODS
Ethical Statement
The experiments on all birds were conducted in accordance with the Institutional Animal Care and Use Committee (IACUC # 2023-0062) of Texas A&M University and in compliance with its Institutional Biosafety Committee (IBC # 2022-061).
Experiment 1: Three Protocol Comparison
A unified feather pool (UFP) was created using equal amounts of finely chopped and well mixed rachis and vanes from the feathers from 10 Hyline W-36 hens, randomly selected from the Texas A&M fertile hen flock. Hens, approximately 60 ± 2 weeks old, were housed in an open-sided house that provided 1.50 m2/hen; breeder recommended 1.24 m2/hen. Feathers were collected during early September when temperatures averaged 29 °C (85 °F). All birds were provided free access to a nutritionally adequate corn/soy-based layer diet.
Experiment 2: Plasma and Feather Corticosterone from Hyline W-36 Pullets and Two Strains of Parent Stock Broiler Pullets
Feathers were harvested from 16-week-old pullets of three different breeds, (n = 15 per breed). All pullets were housed in floor pens that provided 0.093 m2/pullet. At this age, recommended pullet stocking densities were 0.031 m2, 0.125 m2, and 0.10 m2/bird for W-36, broiler parent stock pullets of Breed 1 and Breed 2, respectively. All birds had free access to water. Hyline W36 pullets had free access to a nutritionally adequate corn soy grower-type diet, while breeder pullets were fed amounts of the same diet to meet breeder bodyweight recommendations using a skip-a-day protocol. At the time of feather collection, breeder pullets had been fed using a skip-a-day protocol for 12 weeks. Fifteen birds from each breed were sampled 24 hours after feed removal. Feathers were collected immediately before euthanasia via CO2 asphyxiation, and subsequent collection of blood by cardiac puncture. Feathers from this experiment were extracted using the optimized keratinase protocol (OKP) described below.
SAMPLING METHODS
Feather Collection and Processing
Feathers were collected by cutting all primary and secondary feathers from the right wing approximately 0.635 cm (0.25 inches) from the point of attachment. Primary wing feathers were used for analysis. Cut feathers were stored in plastic bags at room temperature until analyzed. Harvested feathers were initially processed by a common protocol to remove external contaminants (Häffelin et al., 2020). The calamus of individual feathers was removed prior to coarsely chopping the vane and rachis in a glass holding vessel. Coarsely chopped samples were then finely chopped using a razor blade to increase the surface area available for extraction and uniformly distribute pieces. Reproducibility studies used the UFP and were conducted in parallel using both the methanol (Bortolotti et al., 2008) and keratinase (Alba et al., 2019) extraction protocols. In experiment 2, feathers from individual birds were harvested and initially processed using the common protocol prior to extraction and sample preparation using OKP.
Blood Collection
Cardiac blood samples were collected and stored in heparin containing tubes (Vacutainer) immediately post euthanasia. The blood was then centrifuged at 2,000 x G for 15 min, after which the plasma was removed, labeled, and stored at -80°C prior to analysis. At the time of extraction samples were thawed on ice. Plasma samples were combined with CORT ELISA kit (K014-H1/K014-H5, Arbor Assays) dissociation reagent and assay buffer to create 1:100 dilution of sample prior to assay.
EXTRACTION METHODS
Extraction Protocols
The UFP was divided into 30 equal replicates, with 10 replicates allocated to the methanol protocol (MP), 10 to the keratinase protocol (KP), and 10 to the OKP.
The MP extraction procedure followed published methods (Bortolotti et al., 2008). Briefly, samples were placed into glass vials containing 10 ml of HPLC grade methanol and sonicated in a water bath for 30 min (Bortolotti et al., 2008). After sonication, the feather mince was incubated overnight in a shaking water bath at 50 °C (Bortolotti et al., 2008). The following day, the feather mince in methanol was filtered with vacuum filtration (Whatman, #4) into new glass tubes, and any remaining mince remnants and original vials were rinsed with 2-5 ml of HPLC grade methanol (Bortolotti et al., 2008). The filtered samples were concentrated by evaporation using vacuum centrifugation at 40 °C and 1200 rpm (Jouan RC 10-10).
KP was performed as published (Alba et al., 2019). The digestion medium was prepared with a ratio of 1 gram of keratinase (Cibenza: IND900, Novus) per 30 mL of phosphate buffered saline at a pH of 9 (Alba et al., 2019). A ratio of 2 mg of feather mince per mL of digestion medium was used with 5 mL volumes for digestion. Extraction tubes were capped and incubated at 45 °C, and digestion progress monitored over the next 5 days. On day 5, when samples had liquefied into a slurry with slight particulates, extracts were subjected to SPE (C-18 Bond Elute, Varian) (Alba et al., 2019). Solid-phase extraction cartridges were conditioned with 1 mL of methanol and water. The total feather extracts, including solid and aqueous components, were then loaded into sterile syringes, and passed through attached 25 mm, 0.45 µm syringe filters into extraction cartridges. Samples were eluted with two 0.5 mL washes of 2 % acetic acid in acetone into glass tubes. The original KP method took samples to complete dryness via heat and vacuum-based centrifugal evaporation (40 °C and 1200 rpm), respectively. At dryness a waxy adherent material remained in the sample tubes.
The OKP was performed identically to KP except for the drying step. For the OKP, SPE was followed by heat and vacuum-based centrifugal evaporation (40 °C and 1200 rpm), respectively, that was stopped when a 5-10 µL convex droplet of extract remained. The droplet was then utilized for analysis by ELISA by combination with 1X assay buffer from the kit to a final volume of 500 µL.
Optimized Protocol Validation
To ensure results of the OKP were accurate representations of actual feather CORT concentrations a spike and recovery test was conducted using UFP as the sample. Prior to start of enzymatic digestion sufficient CORT standard (Corticosterone ELISA Standard, Cayman Chemicals) was added to UFP to create a sample set that contained 4 replicates of the 5 most central points of the standard curve for the chosen ELISA assay kit, namely 0.3125 pg CORT/mg feather, 0.625 pg CORT/mg feather, 1.25 pg CORT/mg feather, 2.5 pg CORT/mg feather and 5 pg CORT/mg feather. The four replicate preparations were assayed in duplicate for each concentration to create 4 averaged values at each concentration. Four replicates of a blank, un-spiked, UFP sample were included to establish basal UFP CORT concentration. All feather/keratinase solution tubes contained a 2 mg feather per 1 mL of keratinase digestion solution with 5 mL used for each digestion. Enzymatic digests used the OKP extraction protocol in which extracts were taken to incomplete dryness, and the resultant 5-10 µL extract droplet was resuspended to a total volume of 500 µL in 1 X assay buffer provided within the manufacturer’s kit and used for ELISA assay. After ELISA analysis, the average CORT concentration of the unspiked UFP sample was subtracted from each spiked sample and measured concentrations were calculated against theoretical yield to determine percent return.
Corticosterone Concentration Quantification
Corticosterone determinations, regardless of the extraction method, were made using the K014-H1/K014-H5 CORT ELISA kit from Arbor Assays. Assay plates were read using a Synergy H1 Microplate reader at the manufacturer’s specified wavelength of 450 nm.
Statistical Analysis
Values are expressed as means ± SEM. Mean comparisons used Student’s T-test. Simple correlation was used to assess the relationship between plasma and feather CORT. Observed differences were considered significant at p <0.05. Assay validity was assessed using a spike/recovery analysis in which the amount spiked was correlated to the amount detected by the assay following subtraction of the baseline amount detected in the UFP test matrix. All statistical analysis has been completed using JMP version 17.1.
RESULTS AND DISCUSSION
Three Protocol Comparison
Evaporated extracts prepared using MP were white, brown, yellow, or some combination thereof, while evaporated KP had a more consistent translucent amber appearance. Further, the appearance of a properly evaporated OKP extract was that of a clear to faintly yellow glossy droplet. Resuspension of dried extracts proved another source of variation. It was discovered that sample resuspension in kit buffer required the addition of ≤ 5 % organic solvent to fully solubilize the sample for antibody binding. The 5-10 µL of remaining 2 % acetic acid in acetone recovered from the evaporation step in OKP proved to serve as such an organic solvent.
Each extraction protocol was evaluated using 10 replicates assayed in duplicate. The MP and OKP each possessed two samples that were omitted due to errors in processing and thus resulted in an n = 8. The MP found the UFP to contain 5.40 ± 3.93 pg CORT/mg feather and an intra assay CV of 61 %. The KP found the UFP to contain 4.10 ± 1.35 pg CORT/mg feather and an intra assay CV of 33 %. The OKP found the UFP to contain 1.42 ± 0.075 pg CORT/mg feather and an intra assay CV of 5.4 %. The reduction in variation was significant both for KP and OKP when compared to MP (Bartlett, p = 0.0054 and p < 0.0001, respectively).
The substantially lower feather CORT concentration found using OKP compared to MP or KP was validated using a spike/recovery analysis. Fig. 1 shows the recovery of the CORT spikes from UFP samples extracted using OKP. The average recovery was found to be ≥ 90 % and the recovery response was linear, y= 0.97x + 36.8, with an R2 = 0.97, p = 0.002. it is unknown why the MP and KP extraction methods resulted in overestimates of feather CORT. Common causes for ELISA assay overestimates include inadequate washing, overly long incubation times, contamination due to improper plate sealing, or edge effects such as uneven incubation temperature. We do not believe that contamination or edge effects occurred as systematic errors such as these would have also occurred with the OKP assay method. We found no reports that the kit used in this study had been used with feathers extracted using MP or KP and perhaps additional washing or shorter incubation times could reduce observed overestimates although these protocol modifications were not attempted.
Fig. 1. Optimized Keratinase Protocol Recovery Test. Mean pg CORT/mg feather for n = 4 averaged values for duplicated replicates per target concentration of 5 pg CORT/mg feather, 2.5 pg CORT/mg feather, 1.25 pg CORT/mg feather, 0.625 pg CORT/mg feather, & 0.3125 pg CORT/mg feather.
Plasma and Feather Corticosterone from Hyline W-36 Pullets and Two Strains of Parent Stock Broiler Pullets
The CORT concentrations of plasma and feather OKP extracts from Experiment 2 are displayed in Fig. 2. Mean values and CVs of each sample group are given in the legend of Fig. 2 along with sample value ranges for plasma and feathers. The average CV of the three plasma samples groups was 22.43 % ± 2.28, some 3-fold higher than the 6.97 % ± 0.12 CV of the three feather sample groups (p < 0.02). Average intra-assay CVs for each sample group were derived from averaged duplicates of the 15 individual samples used in each test group. No significant correlations could be drawn between plasma and feather CORT concentrations in any group.
Fig. 2. Plasma and Feather Corticosterone Concentrations. All plasma and feather values are based on 15 duplicated replicates per test group for Breed 1 = layer strain and Breeds 2 & 3 = broiler breeder strains. Mean plasma CORT, shown in solid bars, was 4.9, 8.2 and 7.8 pg CORT / µL plasma, for Breeds 1, 2, & 3, respectively. Mean feather CORT, shown in striped bars, were 2.9, 2.5 & 2.6 pg CORT/mg feather for Breeds 1, 2, & 3, respectively. The coefficient of variation (CV) for plasma CORT was 18.2 %, 26.0 % and 23.1 % for Breeds 1, 2, & 3, respectively. Sample values ranged from 7: 4.9 – 8.2 pg CORT/µl plasma. The CVs for feather CORT were 7.2 %, 6.8 % and 6.9 % for Breeds 1, 2, & 3, respectively Sample values ranged from 2.5 – 2.9 pg CORT/mg feather.
Optimized Keratinase Protocol: Comparing Feather to Plasma Corticosterone values
The OKP was used to assay feather CORT concentrations in three different chicken strains and the same ELISA kit used to measure plasma CORT of these same birds per manufacturer’s instructions. Variability in plasma CORT concentration was greater than that of feather CORT concentration supporting the perspective that the invasive sampling required for plasma determination is a source of sample variability (Chloupek et al., 2011) and that the relatively non-invasive method of feather sampling used in this study provided a more reliable indicator of long-term stress (Romero and Fairhurst, 2016). Differences were observed among the different breeds and ages of birds that likely reflect genetic and/or developmental differences. The validated assay provides the means to more precisely study such variables and may support more specific “normal” range values for different types or ages of birds. For example, while samples were not collected as part of a longitudinal study and so cannot be statistically compared in this report, feather CORT concentrations of 60 wk old W-36 hens (1.42 pg CORT/mg feather) was lower compared to 16 wk old pullets of the same strain (2.9 pg CORT/mg feather)
Present Issues and Future Directions
The availability of a validated feather CORT assay with optimized extraction conditions allows for studies specifically designed to assess genetic, developmental and environmental variables that influence bird stress. Our initial efforts to measure feather CORT showed unacceptable sample variation when a UFP was extracted using MP. Variation was reduced with KP but remained unacceptable despite consistently obtaining standard curves that possessed R2 = 0.999 for the ELISA kit used here. All extraction protocols met manufacturer’s stated sample requirements and resultant values were resemblant of those reported by Carbajal’s for feathers in broilers via ELISA (Carbajal et al., 2014) and supported a conclusion that the assay itself was an unlikely source of sample variation. We determined that the variation stemmed from a combination of extraction method and sample resuspension. Using the validated OKP method, feather CORT differed among different strains of birds, and differences appeared to be influenced by bird developmental stage, albeit the study was not designed to rigorously test for either genetic or developmental variables. However, now that a result can be acquired consistently with low variation within and between tests properly designed studies to test such variables can be undertaken and could result in the definition of values of normal/acceptable stress using feather CORT as an index. Future studies would thus need to work toward defining “normal” ranges for each strain and species of clinically normal/healthy poultry of interest. Additionally, it would be of interest to determine whether the OKP reduced variability of measurements when other commercial CORT ELISA or RIA kits are used.
The primary objectives of our study were twofold. First, we aimed to assess the effectiveness of feather CORT extraction protocols and test modifications with the aim of reducing the sample variation encountered with initial trials using MP as the extraction protocol. The secondary aim was to explore any relationships between feather CORT and plasma CORT in three commercially relevant poultry strains. The modifications resulting in OKP produced significant reductions in assay variability in all feather types tested. Lack of correlation between plasma and feather CORT is not unexpected for while spikes in plasma CORT over time are believed responsible for elevated feather CORT concentrations, the short half-life of plasma CORT suggests that a complex relationship exists between these two pools. Overall, our findings suggest that feather CORT concentrations can be reliably analyzed and provide a useful tool in assessing poultry welfare.
DISCLOSURES
Data presented as a poster at the 2024 American Association of Avian Pathologists meeting (pg 75).
Source: Science Direct
