The effects of hydrolyzed protein on macronutrient digestibility, fecal metabolites and microbiota, oxidative stress and inflammatory biomarkers, and skin and coat quality in adult dogs

Abstract

Research on protein hydrolysates has observed various properties and functionalities on ingredients depending on the type of hydrolysate. The objective of this study was to evaluate the effects of hydrolyzed chicken protein that was incorporated into diets on digestibility, gut health, skin and coat health, oxidative stress, and intestinal inflammation markers in healthy adult dogs. Five complete and balanced diets were manufactured: (1) CONd: 25% chicken meal diet; (2) 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; (3) CLHd: 25% chicken liver and heart hydrolysate diet; (4) 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; (5) CHd: 25% chicken hydrolysate diet. A replicated 5 × 5 Latin square design was used which included 10 neutered adult Beagles. Each of the 5 periods consisted of a 7-d washout time and a 28-d treatment period. All diets were well accepted by the dogs. Fecal butyrate concentration was higher while fecal isovalerate and total phenol/indole were lower in dogs fed CLHd than CONd (P < 0.05). Dogs fed CHd had higher fecal immunoglobulin A concentration when compared with CLHd (P < 0.05); however, both groups were comparable to the CONd. There was no difference among groups in serum cytokine concentrations, serum oxidative stress biomarkers, or skin and coat health analyses (P > 0.05). Fecal microbiota was shifted by CLHd with higher abundance in Ruminococcus gauvreauii group as well as lower Clostridium sensu stricto 1, Sutterella, Fusobacterium, and Bacteroides when compared with CONd (P < 0.05). There was also a difference in beta diversity of fecal microbiota between CLHd and CHd (P < 0.05). In conclusion, chicken protein hydrolysate could be incorporated into canine extruded diets as a comparable source of protein to traditional chicken meal. The test chicken protein hydrolysates showed the potential to support gut health by modulating immune response and microbiota; however, functional properties of protein hydrolysates are dependent on inclusion level and source.

Protein hydrolysates from poultry-origin as the main protein source in diets showed potential to interact with the immune system and gut microbiota in healthy adult dogs.

Introduction

For decades, protein hydrolysates have been used in human infant formulas for infants with milk protein sensitivity (Anderson et al., 1982; Pahud and Schwarz, 1984). With smaller particle sizes, protein hydrolysates would have lower chances to elicit immunoglobulin E (IgE) mediated allergic reactions (Lehrer et al., 1996). More recently, the pet food industry started incorporating hydrolyzed protein in diets for various reasons, such as increasing digestibility, increasing palatability, decreasing allergenic responses, and/or providing nutraceutical effects (Cave, 2006; Hou et al., 2017). The smaller molecular weight from hydrolysis would require less enzymatic digestion in the small intestine and, thus, potentially increase the digestibility and absorption of the protein (Heimburger et al., 1997; Zhao et al., 1997; Gilbert et al., 2008). The increase in digestibility is especially profound and more extensively utilized in lower-quality proteins (Grazziotin et al., 2006). Some peptides or amino acids could contribute additional flavoring as palatants when incorporated into pet foods (Folador et al., 2006). Some peptides that resulted from hydrolysis could be considered bioactive peptides, demonstrating physiological benefits (Shahidi and Zhong, 2008; Walther and Sieber, 2011; Sánchez and Vazquez, 2017). Specific protein hydrolysates from milk, eggs, aquatic and terrestrial animals, and plants were shown to possess antihypertensive, antioxidant, antimicrobial, hypocholesterolemic, antithrombotic, opioid, immunomodulatory, and/or cytomodulatory properties (Haque et al., 2009; Lasekan et al., 2013; Bhat et al., 2015; Jo et al., 2017; Chai et al., 2021). The protein hydrolysates that have been studied were used in different amounts and delivered in various ways for their particular functions. This resulted in a wide range of functionality and specificity for each type of hydrolyzed protein. Therefore, the current study was conducted to evaluate the effect of two sources of avian hydrolyzed proteins, mechanically separated chicken or chicken liver and heart hydrolysates, as the secondary or primary protein sources of canine diets on digestibility, fecal metabolites and microbiota, and physiological functional properties in healthy adult dogs. It was hypothesized that all treatment diets would be well digested; in addition, the protein hydrolysates would modulate fecal microbiota and exert anti-inflammatory or anti-oxidative stress functionalities.

Materials and Methods

All animal procedures were approved by the University of Illinois Institutional Animal Care and Use Committee. All methods were performed in accordance with the United States Public Health Service Policy on Humane Care and Use of Laboratory Animals.

Experimental Diets

Five treatment diets were formulated to have similar ingredient compositions except for the main protein source (Table 1). The control diet was formulated with chicken meal low ash (CM) as the primary protein source and rice as the primary carbohydrate source. Test hydrolyzed proteins, PROSURANCE CHX Liver.HD (CLH) and PROSURANCE CHX.HD (CH; Kemin Industries, Des Moines, IA), were used to partially or completely substitute CM for the manufacturing of extruded diets for adult dogs. The partial substitution rate was chosen at 5% as a more economical formula to test if any physiological benefits could be observed at a low inclusion rate. The diets were as follows: (1) CONd: 25% chicken meal diet; (2) 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; (3) CLHd: 25% chicken liver and heart hydrolysate diet; (4) 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; (5) CHd: 25% chicken hydrolysate diet. All diets were formulated to meet or exceed the AAFCO (2021) recommendation for adult dog maintenance and were extruded at Wenger Pilot Plant in Sabetha, KS.

Table 1.

Ingredient inclusion rates of treatment diets containing different sources of protein

Ingredient, % as-is	Treatment
	CONd	5% CLHd	CLHd	5% CHd	CHd
Brewer’s rice	38.0	37.3	35.2	36.6	36.3
Chicken meal, low ash	23.9	19.6	——	20.3	—
Chicken liver and heart hydrolysate1	—	5.0	25.9	—	—
Chicken hydrolysate2	—	—	—	5.0	27.0
Corn	10.0	10.0	10.0	10.0	10.0
Chicken fat	7.0	7.0	5.0	7.0	3.0
Whole green pea	5.0	5.0	5.0	5.0	5.0
Cellulose	4.0	4.0	5.0	4.0	5.0
Eggs dried	3.0	3.0	3.0	3.0	3.0
Whole flaxseed	3.0	3.0	3.0	3.0	3.0
Beet pulp	2.0	2.0	2.0	2.0	2.0
Palatant liquid	2.0	2.0	2.0	2.0	2.0
Salt	0.40	0.40	0.40	0.40	0.40
Potassium chloride	0.45	0.45	0.45	0.45	0.45
Calcium carbonate	—	—	1.24	—	1.00
Dicalcium phosphate	—	—	0.82	—	1.00
l-Lysine	0.40	0.40	0.20	0.40	—
dl-Methionine	0.30	0.30	0.30	0.30	0.30
Mineral premix3	0.22	0.22	0.22	0.22	0.22
Vitamin premix4	0.14	0.14	0.14	0.14	0.14
Choline chloride	0.10	0.10	0.10	0.10	0.10
Taurine	0.10	0.10	0.10	0.10	0.10

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

¹PROSURANCE CHX Liver.HD (Kemin Industries, Des Moines, IA): Copper 11.5 ppm, vitamin A 311,000 IU/kg.

²PROSURANCE CHX.HD (Kemin Industries, Des Moines, IA).

³Iron (10.0%), copper (1.5%), manganese (5.5%), zinc (20.0%), iodine (1500 mg/kg), selenium (200 mg/kg), and cobalt (1000 mg/kg).

⁴Vitamin A (17,163,000 IU/kg), vitamin D₃ (920,000 IU/kg), vitamin E (79,887 IU/kg), thiamine (14,252 mg/kg), riboflavin (4,719 mg/kg), d-pantothenic acid (12,186 mg/kg), niacin (64,736 mg/kg), vitamin B₆ (5,537 mg/kg), folic acid (720 mg/kg), biotin (70.0 mg/kg), vitamin B₁₂ (22.0 mg/kg).

Chemical Analysis

All treatment diets and dried fecal samples were ground with a Wiley mini-mill (Thomas Scientific, Swedesboro, NJ) through a 2-mm mesh screen. Dry matter (DM), ash, organic matter (OM), acid hydrolyzed fat (AHF), crude protein (CP), total dietary fiber (TDF), and gross energy (GE) were determined in all experimental diets, ingredients, and dried fecal samples. In addition, treatment diets were also analyzed for soluble and insoluble fiber contents. Dry matter and ash content of the diets and ingredients were determined in duplicates according to AOAC (2006; methods 934.01 and 942.05). Total nitrogen values were determined according to AOAC (2006; method 992.15) with CP calculated from Leco (TruMac N, Leco Corporation, St. Joseph, MI). Acid hydrolyzed fat was analyzed according to AACC (1983) and Budde (1952). Gross energy was determined through a bomb calorimeter (Model 6200, Parr Instruments Co., Moline, IL). Total dietary fiber was analyzed according to Prosky et al. (1992).

Canine Study and Experimental Design

Ten neutered adult Beagles (females, mean age 4 ± 0.82 years, mean body weight 9 ± 0.64 kg, mean body condition score 5 ± 0.57) were included in the study with a replicated 5 × 5 Latin Square design. Dogs were randomly assigned to one of the five treatment diets for each period to allow all dogs to receive all diets once during the trial. At the beginning of each period, there was a 7 d washout in which all the dogs consumed CONd. After the washout, there was a 28-d treatment period in which dogs were fed the assigned treatment diet. Fresh fecal collection was conducted at the end of the washout period (day 0). Total fecal collection was conducted during the last 4 d of the treatment period and a fresh fecal sample was also collected during this time. Skin measurements and scoring were performed at days 0, 14, and 28 to determine skin and coat quality. Fasted blood samples were collected at days 0 and 28 to evaluate health status and biomarkers. During collections, 2 mL of blood was placed in EDTA vacutainer tubes, and 10 mL was placed in serum separator tubes (Becton, Dickinson and Company, Franklin Lakes, NJ). Blood samples were sent to the Clinical Pathology Laboratory at the College of Veterinary Medicine at the University of Illinois (Urbana, IL) for serum chemistry and complete blood count analyses. Additional serum was reserved at −80 °C for biomarker analyses.

The dogs were housed individually in kennels at Edward R. Madigan Laboratory (Urbana, IL) in a temperature-controlled room with a 14:10 (L:D) cycle, allowing nose-to-nose interaction with adjacent dogs and visual contact with all dogs. Feeding occurred twice a day at 0800-1000 and 1500-1700 hours. Diets were weighed and recorded for each feeding. Dogs had free access to water at all times and were fed to maintain body weight. Both body weight and body condition score (1 to 9 scale; Laflamme, 1997) were monitored weekly.

Fecal Collection, Preparation, and Fermentative Metabolites Analyses

A fresh fecal sample was collected from each dog within 15 min of defecation on days 0 and 28 during each period. Fresh fecal samples were scored (1 = hard, dry pellets; 2 = hard formed, remains firm and soft; 3 = soft, formed and moist stool; 4 = soft, unformed stool; 5 = watery, liquid that can be poured), weighed, and measured for pH. Aliquots of the fresh fecal sample were collected for branched-chain fatty acids (BCFA), short-chain fatty acids (SCFA), and ammonia analyses by placing 3 g of feces into a 30-mL Nalgene bottle containing 3 mL of 2 N hydrochloric acid. Samples for phenol and indole analysis were collected in duplicate by weighing 2 g of feces into identical plastic tubes and covered with Parafilm. Fecal samples for microbiota analysis were placed in 2 mL cryovials. Fecal samples allocated for fermentative end-product analysis were stored at −20 °C. Fecal samples allocated for inflammatory biomarkers and microbiota were stored at −80 °C until analyses.

Total feces were collected during a 4-d period on days 24 to 28. Each fecal sample was weighed, scored, and stored at −20 °C until analyzed for macronutrient apparent total tract digestibility (ATTD). Fecal samples were composited by dog. The composited fecal samples were dried in a 57 °C forced-air oven. Once dried, fecal samples were ground through a 2-mm screen using a Wiley mill (model 4, Thomas Scientific, Swedesboro, NJ) and analyzed in duplicate according to methods described above under “Chemical Analyses”.

Fecal concentrations of SCFA and BCFA were measured using gas chromatography according to Sunvold et al. (1995). Fecal ammonia concentrations were determined according to Chaney and Marbach (1962). Fecal phenol and indole concentrations were analyzed through gas chromatography according to Flickinger et al. (2003).

Fecal Microbiota

Total DNA was extracted from fresh fecal samples using a DNeasy PowerLyzer PowerSoil Kit (Qiagen, Germantown, MD). Invitrogen Qubit 4 Fluorometer (Thermo Fisher Scientific Inc., Waltham, MA) was used to measure DNA concentration. A Fluidigm Access Array (Fluidigm Corporation, South San Francisco, CA) combined with Roche High Fidelity Fast Start Kits (Roche, Indianapolis, IN) were used to amplify the 16S rRNA gene. The quality of the amplicons’ regions and sizes was confirmed by Fragment Analyzer (Advanced Analytics, Ames, IA). The pooled DNA samples were selected by size on a 1% agarose E-gel (Life Technologies, Grand Island, NY) and extracted using a gel extraction kit (Qiagen). The pooled, size-selected, and cleaned products were then run on an Agilent Bioanalyzer to confirm the appropriate profile and mean size. Full-length 16S PacBio (Pacific Biology, Menlo Park, CA) analysis was performed at Roy J. Carver Biotechnology Center at the University of Illinois. The 16S amplicons were generated with the barcoded full-length 16S primers from PacBio and the 2x Roche KAPA HiFi Hot Start Ready Mix (Roche, Willmington, MA). The amplicons were pooled and converted to a library with the SMRTbell express template prep kit 2.0 (Pacific Biology, Menlo Park, CA) on 1 SMRTcell 8M in the Sequel II using the CCS sequencing mode and a 10-h movie time. The upstream analysis workflow was done according to Ras et al. (2021).

Sequences were analyzed using DADA2 (version 1.14; Callahan et al., 2016). This analysis imported amplicon sequence variants (ASV) into the phyloseq R package (McMurdie and Holmes, 2013). Sequences were agglomerated based on cophenetic distances, using a threshold of 0.075. A total of 433 taxa were determined after tip agglomeration. Prevalence filtering was performed after agglomeration; ASVs were only retained for further analysis if they were observed in 5 or more samples. After taxonomic agglomeration, the number of assigned taxa was reduced to 133.

Alpha diversity was assessed by observed taxa, Chao1, Faith’s PD, Shannon, Simpson, and Inverse Simpson indexes. Bray-Curtis distance (Bray and Curtis, 1957) and weighted UniFrac distance (Hamady et al., 2010) were calculated between samples after converting ASV abundances to proportions. The DESeq2 R package (Love et al., 2014) was used to identify differentially abundant taxa between treatments. A false discovery rate (Benjamini and Hochberg, 1995) lower than 0.05 were used to declare statistical significance.

Skin and Coat Measurements

The dogs were bathed 10 d before each skin measurement and scoring. The scoring was performed by 3 blinded researchers using 2 systems according to Rees et al. (2001) and Marsh et al. (2000) under the same light settings in the same room with dogs standing on the same platform to minimize movements. The criteria for the 5-point scale from Rees et al. (2001) for hair condition are: 1 dull, coarse, dry; 2 poorly reflective, non-soft; 3 medium reflective, medium soft; 4 highly reflective, very soft; 5 greasy. Skin condition scores from Rees et al. (2001) are: 1 dry; 2 slightly dry; 3 normal; 4 slightly greasy; 5 greasy. The parameters for quantitative descriptive analysis (QDA) of coat condition from Marsh et al. (2000) are: glossiness (1 = very dull to 5 = very shiny), softness (1 = very brittle to 5 = very soft), greasiness (1 = very greasy to 5 = not greasy), and scale/dandruff (1 = very scaly to 5 = no scale). The QDA is a 9-point scale from 1 to 5 with 0.5 increments.

After the scoring, skin condition was measured for: (1) hydration status with probe Corneometer CM 825 (Courage + Khazaka Electronic GmbH, Cologne, Germany), (2) sebum concentrations with Sebumeter SM 815 (Courage + Khazaka Electronic GmbH, Cologne, Germany), and (3) transepidermal water loss (TEWL) with Tewameter TM 300 MDD (Courage + Khazaka Electronic GmbH, Cologne, Germany). There were 4 measuring sites: left pinna, right pinna, inguinal region, and upper back. The testing areas were shaved using standard clipping procedures of a clipper with #40 blade 2 d before the measurements. Three readings per site per measurement were recorded. In addition, the temperature and humidity of the room were monitored.

Oxidative Stress and Inflammatory Biomarker Analyses

Canine serum was used to determine serum malondialdehyde (MDA) and superoxide dismutase (SOD) concentrations with commercially available enzyme-linked immunosorbent assay kits (MyBioSource Inc., San Diego, CA). Serum was also used to determine concentrations of interleukin 8 (IL-8), interferon γ-induced protein 10 (IP-10), keratinocyte chemotactic-like (KC-like), interleukin 18 (IL-18), and monocyte chemoattractant protein 1 (MCP-1) with a commercially available multiplex kit (MilliporeSigma, Burlington, MA). Canine fresh feces were used to determine immunoglobulin A (IgA) and calprotectin concentrations with commercially available enzyme-linked immunosorbent assay kits (MyBioSource Inc.).

Statistical Analysis

The MIXED model procedures of SAS version 9.4 (SAS Institute Inc., Cary, NC) were used to analyze all data. Dietary treatment was set as the fixed effect and animal was included as the random effect. Data normality was checked with the UNIVARIATE procedure. Data with non-normal distributions were transformed and outliers were excluded to reach a normal distribution for the detection of significant differences. Tukey adjustment was used to control experiment-wise error. The α was set at 0.05 for significance. Pooled standard errors of the mean were calculated from the MIXED model procedures.

Results

Chemical Composition of Protein Ingredients and Diets

The macronutrient composition of protein ingredients and five treatment diets are shown in Tables 2 and 3, respectively. All diets had similar nutrient contents of 90.3% to 91.4% DM, 94.2% to 95.2% OM, 15.2% to 18.9% AHF, 25.2% to 29.4% CP, and 13.4% to 14.7% TDF.

Table 2.

Chemical composition of different sources of protein ingredients

	Ingredient
	CM	CLH	CH
Dry matter, %	93.4	96.8	94.1
	% DM basis
Organic matter	89.2	95.2	95.5
Ash	10.8	4.6	4.5
Acid hydrolyzed fat	16.6	25.5	42.4
Crude protein	69.3	67.8	53.9
Gross energy, kcal/g	5.6	5.9	7.0

Abbreviations: CM: chicken meal; CLH: chicken liver and heart hydrolysate; CH: chicken hydrolysate.

Table 3.

Chemical composition of treatment diets containing different sources of protein

	Treatment
	CONd	5% CLHd	CLHd	5% CHd	CHd
Dry matter, %	90.5	90.8	91.4	90.3	91.4
	% DM basis
Organic matter	95.1	95.2	94.2	95.2	94.7
Ash	4.9	4.8	5.8	4.8	5.3
Acid hydrolyzed fat	15.2	15.2	15.8	16.2	18.9
Crude protein	29.4	28.8	26.8	28.1	25.2
Total dietary fibre	14.3	14.2	14.0	14.7	13.4
Soluble dietary fibre	4.4	4.4	4.5	4.6	4.4
Insoluble dietary fibre	9.9	9.7	9.5	10.1	9.0
Gross energy, kcal/g	5.2	5.2	5.1	5.2	5.2

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Apparent Total Tract Digestibility

There was no difference (P > 0.05) in food intake on either as-is or DM basis among treatment groups (Table 4). Fecal output on both as-is and DM basis was higher (P < 0.05) in dogs fed CLHd (102.8 and 29.8 g/d, respectively) and 5% CHd (98.1 and 29.7 g/d, respectively) when compared with CONd (81.4 and 25.5 g/d, respectively), 5% CLHd (77.7 and 25.3 g/d, respectively), and CHd (77.2 and 25.6 g/d, respectively). Dry matter ATTD were lower (P < 0.05) in CLHd and 5% CHd (78.7 and 78.5%, respectively) compared with CONd, 5% CLHd, and CHd (81.5%, 81.3%, and 80.9%, respectively). Organic matter ATTD were above 80% for all treatment diets while lower (P < 0.05) in CLHd and 5% CHd (81.1% and 81.3%, respectively) in contrast with CONd, 5% CLHd, and CHd (84.1%, 83.7%, and 83.8%, respectively). Crude protein ATTD were comparable between CLHd and 5% CHd (77.4% and 77.6%, respectively) but significantly lower (P < 0.05) than CONd and CHd (81.0% and 82.1%, respectively). The GE ATTD were all higher than 80% and, similarly to DM and OM ATTD, lower (P < 0.05) in CLHd and 5% CHd (82.7% and 82.6%, respectively) compared with CONd, 5% CLHd, and CHd (85.2%, 84.9%, and 85.6%, respectively). For ATTD of TDF, CONd was the highest at 36.0% and CHd was the lowest at 27.1% (P < 0.05).

Table 4.

Food intake, fecal output, and apparent total tract macronutrient digestibility in adult canine consuming treatment diets containing different protein sources

Item	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Food intake, g/d (as-is)	152.4	149.7	152.6	152.5	145.9	5.10
Food intake, g/d (DM)	138.0	135.9	139.4	137.6	133.4	4.63
Fecal output, g/d (as-is)	81.4b	77.7b	102.8a	98.1a	77.2b	6.73
Fecal output, g/d (DM)	25.5b	25.3b	29.8a	29.7a	25.6b	1.34
Digestibility
Dry matter, %	81.5a	81.3a	78.7b	78.5b	80.9a	0.63
	% DM basis
Organic matter	84.1a	83.7a	81.1b	81.3b	83.8a	0.58
Acid hydrolyzed fat	92.6b	92.6b	92.3b	91.8b	94.2a	0.37
Crude protein	81.0a	80.1ab	77.4b	77.6b	82.1a	1.14
Total dietary fibre	36.0a	33.8ab	28.3ab	28.5ab	27.1b	2.11
Gross energy	85.2a	84.9a	82.7b	82.6b	85.6a	0.56

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

^a,bMeans in the same row with different superscript letters are different (P < 0.05).

Fecal Characteristics and Metabolites

There was no difference seen in fecal characteristics and metabolites among dogs fed different treatment diets at day 0 (P > 0.05); therefore, Table 5 shows only the data at day 28. For fecal characteristics, there was no difference (P > 0.05) seen in fecal score (ranging from 2.8 to 3.0) or pH (ranging from 5.9 to 6.4). However, fecal DM from dogs fed CLHd (26.3%) was lower (P < 0.05) than 5% CLHd and CHd (28.8% and 30.3%, respectively). There were also differences seen in fecal metabolites among treatments. Fecal butyrate concentration was higher (P < 0.05) from dogs fed CLHd (226.8 μmol/g) than dogs fed CONd and CHd (149.3 and 127.3 μmol/g, respectively). Total SCFA concentrations were higher (P < 0.05) in CONd and CLHd (611.8 and 646.3 μmol/g, respectively) compared with CHd (506.2 μmol/g). No treatment effect was seen for fecal acetate (ranging from 269.3 to 332.5 μmol/g) or propionate (ranging from 105.9 to 129.9 μmol/g) concentrations (P > 0.05). Fecal isovalerate and total BCFA concentrations were lowest (P < 0.05) in dogs fed CLHd (6.8 and 15.5 μmol/g, respectively) while the other groups were comparable (P > 0.05). There was no difference in fecal isobutyrate (ranging from 6.0 to 8.3 μmol/g) or valerate (ranging from 2.7 to 3.4 μmol/g) concentrations among groups (P > 0.05). There were lower (P < 0.05) fecal concentrations of 4-ethylphenol, indole, and total phenol/indole concentrations in the CLHd group (1.4, 0.7, and 2.8 μmol/g, respectively) when compared with the CONd group (2.5, 1.5, and 5.0 μmol/g, respectively). Fecal ammonia concentration ranged between 62.8 to 89.2 μmol/g (P > 0.05).

Table 5.

Fecal characteristics and fecal metabolites in adult canines consuming treatment diets containing different protein sources

Item	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Fecal dry matter, %	28.4ab	28.8a	26.3b	28.2ab	30.3a	0.78
Fecal score1	2.9	3.0	3.0	2.9	2.8	0.10
Fecal pH	6.0	6.0	5.9	6.1	6.4	0.13
Short-chain fatty acid, μmol/g DMB
Acetate	332.5	297.0	313.7	292.9	269.3	21.93
Propionate	129.9	110.4	105.9	108.3	109.6	7.41
Butyrate	149.3b	167.6ab	226.8a	194.3ab	127.3b	25.43
Total SCFA	611.8a	574.9ab	646.3a	595.6ab	506.2b	30.62
Branched-chain fatty acid, μmol/g DMB
Isobutyrate	8.1	8.3	6.0	6.2	7.5	0.69
Isovalerate	10.8a	10.9a	6.8b	7.8ab	9.7ab	0.93
Valerate	3.1	3.0	2.7	3.3	3.4	0.51
Total BCFA	22.1a	22.2a	15.5b	17.3ab	20.5ab	1.55
Phenol and indole, μmol/g DMB
Phenol	0.5	0.5	0.3	0.3	0.4	0.11
4-Methylphenol	0.2	0.2	0.2	0.2	0.2	0.03
4-Ethylphenol	2.5ab	3.2a	1.4c	1.7bc	2.2abc	0.26
Indole	1.5a	1.5a	0.7b	0.7b	1.7a	0.21
3-Methylindole	0.2	0.2	0.1	0.2	0.3	0.06
Total phenol/indole	5.0ab	5.7a	2.8c	3.2bc	4.7ab	0.47
Ammonia, μmol/g DMB	89.2	87.5	62.8	82.6	83.4	7.73

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

¹Fecal score: 1 = hard, dry pellets; 2 = hard formed, remains firm and soft; 3 = soft, formed and moist stool; 4 = soft, unformed stool; 5 = watery, liquid that can be poured.

Fecal Microbiota

A total of 2371 ASVs were identified for the fresh fecal samples. At the phylum level, Firmicutes, Bacteriodota, Fusobacteriota, Actinobacteriota, and Proteobacteria were the top 5 most abundant phyla (Fig. 1). At day 28, CHd group had a higher relative abundance of Bacteriodota and Fusobacteriota while CLHd group had a higher abundance of Actinobacteria when compared with the other groups. At the class level, samples from dogs fed the 3 diets which contained CM (CONd, 5% CHd, and 5% CLHd) had similar compositions (Fig. 2). Samples from CHd group had higher relative abundances of Bacteriodia and Fusobacteria as well as lower Bacilli. Conversely, CLHd group had higher abundances of Bacilli and Actinobacteria with lower Fusobacteria. At the level of order, CHd group had higher Bacteriodales and Fusobacteriales while CLHd group had higher Bifidobacteriales (Fig. 3). At the family level, CHd group showed a higher abundance of Muribaculaceae and a lower abundance of Bifidobacteriaceae when compared with the other groups (Fig. 4). Alternatively, CLHd group demonstrated a higher abundance of Bifidobacteriaceae together with lower abundance of Fusobacteriaceae, Muribaculaceae, Sutterrellaceae, and Ruminococcaceae. At the genus level, CLHd group had a more pronounced difference in composition with a higher abundance of Bifidobacterium, lower Bacteroides, and lower Fusobacterium when compared with the other groups (Fig. 5).

Figure 1.

Phylum composition of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 2.

Class composition of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 3.

Order composition of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 4.

Family composition of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 5.

Genus composition of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

There was no difference (P > 0.05) in alpha diversity among treatment groups at either day 0 or 28 according to any of the analyses (Fig. 6). There was a difference (P < 0.05) in beta diversity between CLHd and CHd but both groups were comparable to CONd (P > 0.05). Principal coordinates analysis plots using Bray-Curtis distance and weighted Unifrac matrix showed that CLHd clustered together, separated from the other groups (Figs. 7 and 8). The differential abundances of Clostridium sensu stricto 1, Sutterella, Fusobacterium, and Bacteroides were lower while Ruminococcus gauvreauii group was higher in dogs fed CLHd compared with CONd (P < 0.05; Table 6). There was no difference in differential abundance between CHd and CONd (P > 0.05).

Figure 6.

Alpha diversity of fecal microbiota from dogs fed treatments diets containing different sources of protein. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 7.

Beta diversity of fecal microbiota from dogs fed treatments diets containing different sources of protein using Bray-Curtis distance. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 8.

Beta diversity of fecal microbiota from dogs fed treatments diets containing different sources of protein using weighted Unifrac matrix. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Table 6.

Differential abundance of fecal microbiota of dogs fed a chicken liver and heart hydrolysate diet compared with a chicken meal diet

Phylum	Class	Order	Family	Genus	Species	Log2 fold change	P-adjusted value
Firmicutes	Clostridia	Lachnospirales	Lachnospiraceae	Ruminococcus gauvreauii group	NA	2.65	0.0229
Firmicutes	Clostridia	Clostridiales	Clostridiaceae	Clostridium sensu stricto 1	NA	−4.71	0.0002
Proteobacteria	Gammaproteobacteria	Burkholderiales	Sutterellaceae	Sutterella	NA	−3.17	0.0034
Fusobacteriota	Fusobacteriia	Fusobacteriales	Fusobacteriaceae	Fusobacterium	NA	−2.27	0.0093
Fusobacteriota	Fusobacteriia	Fusobacteriales	Fusobacteriaceae	Fusobacterium	NA	−2.11	< 0.0001
Bacteroidota	Bacteroidia	Bacteroidales	Bacteroidaceae	Bacteroides	NA	−1.31	0.0222

The correlations between fecal metabolites and microbiota taxa are shown in Fig. 9. Fecal butyrate concentration was negatively correlated with the genera Blautia, Phascolarctobacterium, Sutterella, Faecalibacterium, and Bacteriodes. Conversely, Allobaculum and Bifidobacterium were positively correlated with butyrate concentration. Another fecal metabolite that showed a stronger correlation with microbiota was valerate. Fecal valerate concentration was negatively correlated with the genera Ruminococcus gnavus group, Blautia, Faecalibacterium, Peptoclostridium, and Bacteriodes. Alternatively, valerate was positively correlated with Peptostreptococcus, Allobaculum, and Coriobacteriaceae UCG-002.

Figure 9.

Spearman correlation of fecal metabolites and microbiota taxa in dogs fed treatments diets containing different sources of protein.

Skin and Coat Measurement

There was no difference observed among treatment groups for any parameters of skin and coat analyses at day 0 (P > 0.05). At day 28, no treatment effect (P > 0.05) was observed for coat condition by QDA (Table 7) or skin and coat condition score (Table 8) for dogs fed extruded diets containing different protein sources. In addition, no differences (P > 0.05) were observed in any of the skin areas for hydration status or water loss among dogs fed CONd or diets containing hydrolyzed proteins at day 28 (Table 9). Sebum concentration measured on the right pinna, inguinal, and scapular regions did not differ (P > 0.05) by treatment at day 28 (Table 9). Dogs fed CHd had greater (P < 0.05) sebum concentration of the left pinna (71.7 μg/cm²) than dogs fed CONd (46.7 μg/cm²) at day 28. However, there was no difference in change from the baseline for any measurements of anybody regions (P > 0.05).

Table 7.

Quantitative descriptive analysis (QDA) of coat condition of adult canines consuming treatment diets containing different protein sources

Item1	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Gloss	3.9	3.8	3.9	3.8	3.8	0.10
Softness	3.4	3.5	3.4	3.5	3.5	0.10
Coat feel	3.0	3.1	3.1	2.9	3.0	0.10
Scale/dandruff	4.1	4.2	4.0	4.3	4.4	0.18

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

¹Glossiness (1 = very dull to 5 = very shiny); softness (1 = very brittle to 5 = very soft); greasiness (1 = very greasy to 5 = not greasy); scale/dandruff (1 = very scaly to 5 = no scale).

Table 8.

Skin and coat condition score of adult canines consuming treatment diets containing different protein sources

Item	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Hair1	3.2	3.3	3.2	3.3	3.3	0.09
Skin2	3.0	3.1	3.0	3.2	3.2	0.10

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

¹Hair condition score: 1 dull, coarse, dry; 2 poorly reflective, non-soft; 3 medium reflective, medium soft; 4 highly reflective, very soft; 5 greasy.

²Skin condition score: 1 dry; 2 slightly dry; 3 normal; 4 slightly greasy; 5 greasy.

Table 9.

Measurement of skin and hair of adult canines consuming treatment diets containing different protein sources

Item	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Hydration
Left ear	36.2	33.9	31.8	27.3	36.4	4.56
Right ear	30.4	35.4	32.2	30.8	35.2	3.62
Inguinal region	14.8	15.9	12.8	17.1	17.3	2.48
Back	4.6	5.9	6.5	4.7	6.1	1.51
Sebum, μg/cm2
Left ear	46.7b	63.6ab	55.5ab	56.5ab	71.7a	5.89
Right ear	56.9	61.7	50.9	57.0	76.3	7.31
Inguinal region	6.2	4.6	5.2	7.1	6.2	1.76
Back	20.8	20.5	9.7	13.7	27.2	5.00
Water loss, g/h/m2
Left ear	9.9	10.9	10.9	9.5	9.7	0.95
Right ear	10.4	10.4	9.6	10.3	10.8	1.22
Inguinal region	11.5	11.6	12.7	11.2	9.9	0.84
Back	18.4	16.0	16.7	17.1	15.2	2.34

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

^a,bMeans in the same row with different superscript letters are different (P < 0.05).

Oxidative Stress and Inflammatory Biomarker Analyses

There was no difference seen at the baseline of each period comparing among treatments for either fasted serum MDA or SOD (P > 0.05). As shown in Table 10, both MDA (ranging from 25.72 to 31.90 nmol/mL) and SOD (ranging from 1.88 to 2.32 ng/mL) concentrations were also comparable at day 28 for all dogs (P > 0.05). For the fecal concentrations of IgA and calprotectin, there was no difference at the baseline of each period among treatments (P > 0.05). At day 28, fecal IgA concentration shown in Fig. 10 was significantly higher in dogs fed CHd (2.86 mg/g) compared with dogs fed CLHd (0.91 mg/g; P < 0.05). There was also a trend for fecal calprotectin concentration at d 28 being higher in dogs fed CHd (0.76 μg/g) than dogs fed 5% CLHd and CLHd (0.33 and 0.31 μg/g, respectively; P < 0.1). There were no differences (P > 0.05) in serum concentrations of IL-8, IP-10, KC-like, IL-18, and MCP-1 at day 0 or 28 (Table 11 shows the concentrations on day 28).

Table 10.

Oxidative stress biomarkers of adult canine consuming treatment diets containing different protein sources

Item	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Malondialdehyde, nmol/mL	31.48	31.12	32.16	25.72	31.90	3.177
Superoxide dismutase, ng/mL	1.98	2.32	2.19	1.88	2.07	0.259

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

Figure 10.

Inflammatory biomarkers of adult canine consuming treatment diets containing different protein sources. CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet. ^a,bMeans on the same day with different superscript letters are different on a given day (P < 0.05).

Table 11.

Serum cytokine concentration in adult canines consuming treatment diets containing different protein sources

Item, pg/mL	Treatment					SEM
	CONd	5% CLHd	CLHd	5% CHd	CHd
Interleukin 8	2310	2230	2365	2186	2331	437.9
Interferon γ-induced protein 10	4.4	5.8	5.7	4.8	4.9	4.16
Keratinocyte chemotactic-like	119	125	111	103	151	29.6
Interleukin 18	254	238	260	187	240	190.6
Monocyte chemoattractant protein 1	252	228	262	232	296	63.6

Abbreviations: CONd: 25% chicken meal diet; 5% CLHd: 5% chicken liver and heart hydrolysate plus 20% chicken meal diet; CLHd: 25% chicken liver and heart hydrolysate diet; 5% CHd: 5% chicken hydrolysate plus 20% chicken meal diet; CHd: 25% chicken hydrolysate diet.

There was no difference (P > 0.05) among treatment groups in serum cytokine concentrations.

Discussion

Some differences were observed for ATTD among the treatment groups. Fecal output was higher in dogs fed CLHd and 5% CHd which corresponded with the macronutrient digestibility. Higher moisture content in fecal samples from the CLHd group could be one of the explanations. Another reason could be the modulations of gut microbiota that could have resulted in different amounts of microbial protein excretion in the feces, resulting in a lower value for ATTD. A previous study using the same ingredients and treatment diets showed higher than 80% of standardized amino acid digestibility for all test ingredients with a precision-fed rooster assay using cecectomized roosters (Hsu et al., 2023). In addition, the standardized amino acid digestibility of all indispensable amino acids for CLHd was also above 80% and comparable to CONd (Hsu et al., 2023). Since a cecectomized rooster model minimizes the effect of fermentation, it could be more sensitive than a canine ATTD when considering amino acid digestibility. Therefore, it could be hypothesized that the observed CP ATTD difference could be due to altered microbial protein output as fecal microbiota showed differences in composition and fermentative activity in the present study. Regarding the numerical values, previous studies using chicken-based protein in extruded diets reported CP ATTD to be around 80% in dogs, ranging from 79.2% to 89.9%, which corresponded with the current study (Dust et al., 2005; Maria et al., 2017; Sieja et al., 2023).

Fecal butyrate concentration was higher while isovalerate, 4-ethyphenol, and 4-methylphenol concentrations were lower in dogs fed CLHd when compared with CONd. A higher butyrate concentration could indicate CLHd shifted the microbiota and supported gut health as butyrate had been correlated to beneficial physiological functions, such as maintaining healthy colonocytes and reducing cancer cell growth (Topping and Clifton, 2001). Previous research had conflicting results on the effect of protein sources on hindgut fermentation. Some studies also showed differences in fermentative activity in dogs fed different protein sources or concentrations (Kuzmuk et al., 2005; Nery et al., 2012; Beloshapka et al., 2016; Maria et al., 2017; Pinna et al., 2018; Reilly et al., 2021). However, the fermentative activity differences seen in the study from Beloshapka et al. (2016) may be due to the different fiber contents of the diets which lead to a variable abundance of carbohydrate supply for the gut microbiota. Similarly, research on plant protein based diets and animal protein based diets suggested that the types of fibers could contribute to differences in fermentation even if the TDF content was similar (Kuzmuk et al., 2005; Nery et al., 2012; Maria et al., 2017; Reilly et al., 2021). Nery et al. (2012) found lower fecal total BCFA and indole concentrations in dogs fed a high-protein diet made with wheat gluten when compared with a high-protein diet made with poultry meal. They hypothesized the difference in fermentation products came from the different digestibility of poultry meal and wheat gluten. The study by Pinna et al. (2018) found that higher protein concentration in diets contributed to higher fecal ammonia concentration in dogs. The higher fecal ammonia could be a result of undigested protein entering the large intestine to serve as a substrate for fermentation which produced ammonia as the end product. The lower fecal proteolytic fermentative end products from highly digestible protein sources corresponded to the finding of the current study; therefore, the lower fecal BCFA and phenol/indole concentrations could indicate a higher digestibility of the CLHd. Even though there were conflicting results between the indications from ATTD and fecal metabolites, ATTD does not take endogenous losses and microbial effects into consideration. Hence, data from ATTD may not truly represent the hydrolytic and enzymatic digestibility of the diets. Nonetheless, some studies found no changes in fecal characteristics and fermentative end products with dogs fed different proteins (Urrego et al., 2017; Venturini et al., 2018). Regarding specific protein sources, one recent canine study examined the effect of hydrolyzed chicken liver on fermentative end products on healthy dogs (Pinto et al., 2023). The authors found lower fecal concentrations of proteolytic fermentative end products, such as isovalerate and ammonia in dogs fed the hydrolyzed chicken liver diet when compared with the control poultry byproduct meal diet. This finding was in accordance with the current study that could support the hypothesis of highly digestible hydrolyzed protein would provide less protein substrate for hindgut fermentation. A few human studies have been conducted to determine the effect of protein hydrolysate supplementation on gut health, showing no differences in fecal fermentative end products (Moreno-Pérez et al., 2018; Dale et al., 2019).

Protein sources did not affect α diversity but resulted in a shift in beta diversity according to previous studies which corresponded to the findings of the current study (Martínez-López et al., 2021; Pinto et al., 2022). Since α diversity could be considered a health indicator, it could be argued that using the test hydrolyzed protein sources as the main protein source in the present study did not pose a negative health impact on the gut microbiome (Suchodolski, 2011). The increase of Muribaculaceae in CHd group could indicate CH may support gut health as the Muribaculaceae family has been positively correlated with propionate production, efficient lipid metabolism, and longevity while negatively correlated with obesity in previous studies (Sibai et al., 2020; Smith et al., 2021; Wang et al., 2021; Zhang et al., 2021; Li et al., 2022; Lai et al., 2023). The higher composition of the Bifidobacteriaceae family, especially genus Bifidobacterium, in dogs fed CLHd could also indicate a positive change in host health since Bifidobacterium have been associated with physiological benefits and used as probiotics (Vaughan et al., 2005; O’Callaghan and van Sinderen, 2016; White et al., 2017; Ciaravolo et al., 2021; Cecilia et al., 2022). Even though bifidobacteria do not produce butyrate themselves, it has been shown that the cross-feeding between bacteria results in increased butyrate production with more abundant bifidobacteria (De Vuyst and Leroy, 2011). The positive correlation between Bifidobacterium and fecal butyrate concentration was also seen in the present study. This could be the explanation for the higher butyrate concentration in fecal samples from the CLHd group. Previous studies had conflicting results on the effect of protein sources on microbiota composition (Maria et al., 2017; Pinna et al., 2018; Martínez-López et al., 2021; Pinto et al., 2022). An in vitro study using canine fecal samples as inoculant observed increased counts of Clostridium perfringens with low protein digestibility but no difference in Bifidobacterium after 24 h of incubation (Pinna et al., 2018). A study by Maria et al. (2017), however, found no difference in the fecal microbial count of Clostridium, Lactobacillus, Bifidobacterium, and Escherichia coli with healthy Beagles fed different sources of protein (chicken byproduct meal and soybean meal) for 28 d. A study by Pinto et al. (2022) also found no changes in differential abundance at the genus level with healthy Beagles fed a hydrolyzed chicken liver diet and a control diet with intact protein sources at day 45.

Regarding the differential abundance, the increase of Ruminococcus gauvreauii group of the Lachnospiraceae family in dogs fed CLHd could be another reason that there was higher butyrate concentration in the fecal samples as Lachnospiraceae has been considered as butyrate producers (Ma et al., 2020). Since butyrate is believed to be a beneficial fermentative end product, the higher abundance of Ruminococcus gauvreauii group could imply supported gut health. Some species in the genus Clostridium sensu stricto 1 are considered pathogenic (Yang et al., 2019). Similarly, some species in the genus Sutterella showed pro-inflammatory properties and correlations to diseases (Hiippala et al., 2016). Therefore, lower abundances of Clostridium sensu stricto 1 and Sutterella in the CLHd group could indicate improved gut health. Additionally, both Fusobacterium and Bacteroides are protein fermenting bacteria that are commonly seen in healthy dogs and, thus, the lower abundance correlated with the lower fecal proteolytic metabolite concentrations in the CLHd group (Pilla and Suchodolski, 2020).

Previous studies were conducted to examine the effect of hydrolyzed protein diets on skin and coat health in dogs with adverse food reactions (Weemhoff et al., 2021; Szczepanik et al., 2022). Szczepanik et al. (2022) found decreased severeness of pruritis after the 13 client-owned dogs switched from their original diets to the hydrolyzed protein diet for 28 d as pruritus visual analog scale and canine atopic dermatitis extent and severity index both decreased after treatment. Weemhoff et al. (2021) also used similar measurements of pruritus visual analog score and canine atopic dermatitis lesion index for skin and coat health. After 42 d of treatment, the two measurements of skin and coat health did not differ between the client-owned dogs fed hydrolyzed protein (n = 18) and the dogs fed a positive control commercial diet (n = 14). However, the authors mentioned that both groups had low scores for pruritus indicating the conditions of all dogs in the study were well maintained. Similar scoring systems adopted in the present study were also used by Geary et al. (2022) for examining the skin and coat quality of healthy adult dogs fed extruded kibbles and human-grade foods; there was no difference in change from baseline between the two dietary groups. Another study by Wilson et al. (2022) showed no differences in skin and hair scores in healthy adult dogs with or without a yeast fermentation product supplementation. These results corresponded to the current study not observing differences in skin and coat health scoring among treatment groups. The comparability of scores among groups could be due to the fact that all dogs in the present study were healthy without any pre-existing skin conditions. The same studies by Geary et al. (2022) and Wilson et al. (2022) also measured sebum concentration, hydration status, and transepidermal water loss in healthy adult Beagles receiving different dietary treatments with no difference in change from baseline for most measurements was shown among treatment groups. This finding matched with the measurements of hydration, sebum, and water loss from the current study as healthy dogs may not demonstrate a great change in these parameters when fed different diets. Other studies that used similar techniques to measure these parameters suggested that there were high day-to-day and dog-to-dog variation in healthy dogs (Lau-Gillard et al., 2010; Hester et al., 2014; Jeong et al., 2017). The high variation of measurements may have contributed to the overall similarity of skin hydration, sebum, and water loss among groups in the current study.

Studies have shown that different types of stressors could increase oxidative stress levels in animals (Kearns et al., 1999; McMichael, 2007; Head, 2009). On the other hand, the addition of antioxidants in diets or supplements could decrease oxidative stress (Jewell et al., 2000; Head, 2009; Murai et al., 2019; Wilson et al., 2022). Some peptides from animal muscle, liver, and/or blood hydrolyzation were shown to have antioxidant properties in rodent or in vitro studies (Sun et al., 2012; Lafarga and Hayes, 2014; Fukada et al., 2016; Chen et al., 2017; Xiao et al., 2022). There was no difference in serum oxidative stress biomarkers among treatment groups in the current study. One explanation for this finding could be that no challenge was placed among the healthy animals. Therefore, the serum MDA and SOD concentrations would be stable at the baseline level throughout the experiment. The serum MDA concentration from the current study was numerically comparable to a study by Rummell et al. (2022) which measured serum MDA in healthy non-challenged adult dogs with or without yeast supplementation; they did not find serum MDA difference between the 2 groups. Conversely, Murai et al. (2019) reported improvement in oxidative stress with antioxidant supplementation in old dogs but not in younger adult dogs. Future studies could include a challenge or stressor for healthy dogs in the experiment to further examine whether the protein hydrolysates possess antioxidant properties and could help reduce oxidative stress.

Secretory IgA plays an important role in gut barrier function and the prevention of inflammatory responses as the predominant mucosal antibody (Russell et al., 1989; McKay and Perdue, 1993). A study by Peters et al. (2004) examined fecal immunoglobulin concentrations in different dog breeds also indicated that there were breed differences and the normal range of fecal immunoglobulin concentrations had yet to be established for healthy dogs. A study by Verlinden et al. (2006) found a decrease in fecal IgA concentration in dogs fed a hydrolyzed protein diet (10.4 mg/g DM feces) when compared with dogs fed a commercial duck and rice limited ingredient diet (15.8 mg/g DM feces). It was suggested that hydrolyzed protein could impact immunoglobulin secretion and lower allergenic responses. Alternatively, other studies found increased fecal IgA (ranging from 2.53 to 37.8 mg/g DM feces) in healthy dogs fed prebiotics, probiotics, or a blend of prebiotics and probiotics when compared with the control (Delucchi et al., 2014; Panasevich et al., 2021; Lee et al., 2022), stating this as a sign of increased immune defense system. Similarly, a study by Pinto et al. (2022) found an increase in plasma IgA concentration in dogs fed a diet with hydrolyzed chicken liver as the only protein source for 45 d. The increase in fecal IgA concentration in dogs fed CHd in the present study could be an indication of supported gut health as the current results were within the reported range from healthy dogs in previous studies (Hosono et al., 2003; Delucchi et al., 2014; Ma et al., 2017; Panasevich et al., 2021; Lee et al., 2022).

Calprotectin has been used as a biomarker for inflammation and disease severity in dogs (Heilmann et al., 2018). Canine studies that used supplementations of milk oligosaccharide biosimilar or a yeast fermentation product found no difference in fecal calprotectin concentration when compared with the control groups (Lee et al., 2021; Wilson et al., 2023). A study that used cod protein hydrolysate as supplementation found no difference in fecal calprotectin after 6 weeks of treatment, in comparison with the placebo group, in humans with irritable bowel syndrome (Dale et al., 2019). A study by Lee et al. (2022) found a slightly higher fecal calprotectin concentration in dogs fed a prebiotic and probiotic enriched diet when compared with the control; however, the numerical difference was small and the authors suggested it may not be physiologically relevant. Similar to previous findings, the current study did not find a significant difference in fecal calprotectin content among treatments. The consistently lower value of calprotectin throughout the experiment in all dogs could be an indication of the intestine being maintained in a healthy state.

Other inflammatory biomarkers that have been used in research are cytokines. Most studies examining the effect of poultry-originated protein hydrolysates on cytokine responses as an indicator of inflammation were executed in rodent models (Chen et al., 2017; Lin et al., 2017; Yu et al., 2018; Bjørndal et al., 2020; Fan et al., 2022). Previous studies on rats with liver fibrosis and mice with fatty liver found decreases in some of the tested hepatic pro-inflammatory cytokines, tumor necrosis factor α (TNF-α), interleukin 6, and interleukin 1β, after daily chicken liver hydrolysate supplementation (Chen et al., 2017; Lin et al., 2017). Bjørndal et al. (2020) reported that feeding a high-fat diet with chicken protein hydrolysate as the main protein source to mice could decrease plasma MCP-1 concentration when compared with casein. Similarly, a study by Fan et al. (2022) found reduced plasma TNF-α and MCP-1 in hypertensive rats with orally administered chicken muscle hydrolysate. Yu et al. (2018) observed higher interleukin 10 as well as lower interleukin 1β and TNF-α in spleen cells stimulated with pokeweed mitogen in rats fed a diet using hen muscle protein hydrolysate as the protein source when compared with casein; however, the differences in cytokine concentrations were not present without the pokeweed mitogen stimulation. A canine study reported no differences in any plasma pro- or anti-inflammatory cytokines in healthy adult dogs after being fed a diet with chicken liver hydrolysate as the only protein source when compared with intact protein (Pinto et al., 2022). Overall, previous studies that showed alteration of cytokine concentrations used animals under stress or disease (Chen et al., 2017; Lin et al., 2017; Yu et al., 2018; Bjørndal et al., 2020; Fan et al., 2022). Meanwhile, when animals were healthy, there was no difference in inflammatory signs from cytokines after consuming chicken hydrolysates (Yu et al., 2018; Pinto et al., 2022). Those studies were in agreement with ours, in which no difference in the serum cytokine concentration was detected among treatment groups as all animals in the study were healthy.

Conclusions

Protein hydrolysate from either mechanically separated chicken or chicken liver and heart could serve as a compatible protein source with potential physiological benefits when compared with the traditional chicken meal. Increased fecal butyrate concentration, lower BCFA, and lower phenol/indole were observed in the CLHd group while increased fecal IgA was observed in the CHd group, indicating the possible anti-inflammatory and immunomodulatory properties of protein hydrolysates. Nevertheless, more research is needed to confirm the effect. Dietary treatment did not affect the alpha diversity of fecal microbiota; however, there was a shift in microbiota in dogs fed CLHd with higher Bifidobacterium and Ruminococcus gauvreauii group, both considered beneficial microorganisms. Overall, avian protein hydrolysates may support gut health and sustain anti-inflammatory and immunomodulatory effects in adult dogs. However, their effects are dependent on inclusion level and source-specific.

Glossary

Abbreviations

AHF

acid hydrolyzed fat

ATTD

apparent total tract digestibility

BCFA

branched-chain fatty acid

CP

crude protein

DM

dry matter

DMB

dry matter basis

GE

gross energy

IgA

immunoglobulin A

IgE

immunoglobulin E

IL-18

interleukin 18

IL-8

interleukin 8

IP-10

interferon γ-induced protein 10

KC like

keratinocyte chemotactic like

MCP-1

monocyte chemoattractant protein 1

MDA

malondialdehyde

OM

organic matter

QDA

quantitative descriptive analysis

SCFA

short-chain fatty acid

SOD

superoxide dismutase

TDF

total dietary fiber

TNF-α

tumor necrosis factor α

Conflict of Interest Statement

C. H., N. V., and M. R. C. G. have no conflict of interest to declare. F. M. and R. G. are employed by Kemin Industries, Inc.