Effects of dietary Zanthoxylum bungeanum seed powder on growth performance, immune function, and cecal microbiota in broilers

Abstract

This study was conducted to evaluate the effects of Zanthoxylum bungeanum seed powder (ZBS) on growth performance, immune function, and cecal microbiota in 817 broilers. A total of 260 one-day-old 817 broilers were randomly assigned to 4 treatment groups (CON, ZBS1, ZBS2, and ZBS3) with 5 replicates per treatment and were fed diets containing 0, 1%, 2%, or 3% ZBS, respectively, for 63 d. The results showed that dietary ZBS had no significant effects on average daily feed intake, average daily gain, or gain-to-feed ratio (P > 0.05). However, body weight at 28 and 63 d and average daily gain during d 1–28 and d 1−63 showed significant quadratic responses to increasing dietary ZBS levels (P < 0.05). The bursa of Fabricius index increased linearly with increasing ZBS inclusion (P < 0.05). Serum concentrations of IL-2, IL-6, TNF-α, IgA, and IgG and splenic contents of IgA, IgG, C3, and C4 were significantly increased (P < 0.05), whereas splenic IL-1β content and the expression of TLR4/MyD88/NF-κB pathway-related genes was significantly decreased (P < 0.05). 16S rRNA sequencing showed that ZBS did not affect cecal microbial α-diversity (P > 0.05), but altered microbial community composition. The relative abundances of Faecalibacterium, Bifidobacterium, and Megasphaera were significantly increased in the 3% ZBS group (P < 0.05). PICRUSt2 analysis indicated that the predicted abundances of starch and sucrose metabolism, carbohydrate metabolism, the phosphotransferase system, and pyruvate-formate lyase-related functions were significantly increased in the 3% ZBS group (P < 0.05). Redundancy analysis showed that microbial community variation was significantly associated with dietary ZBS level and immune-related parameters, including IgA, IgG, C3, and C4 (P < 0.05). In conclusion, dietary ZBS improved humoral immunity-related parameters, reduced the expression of inflammation-related genes in the spleen, and affected the composition and predicted functions of the cecal microbiota in broilers. Based on quadratic regression of splenic C4, IL-1β, TRAF2, TLR4, and NFKB2, the optimal dietary ZBS inclusion level was estimated at 1.00−2.55%.

Introduction

Under the background of antibiotic-free production, the development of plant-derived feed ingredients with both nutritional value and biological activity has become an important research focus in poultry nutrition. Phytogenic substances, owing to their antimicrobial, antioxidant, immunomodulatory, and growth-promoting properties, have emerged as promising natural alternatives to synthetic antibiotics. A recent review on poultry production has summarized relevant evidence, indicating that these compounds may improve intestinal health, enhance immune function, and boost growth performance in broilers (Biswas et al., 2024; Obianwuna et al., 2024). In addition, several studies have investigated the impacts of dietary co-supplementation with phytogenic components (e.g., Curcuma longa) and fermentation by-products on broiler carcass performance (Nurhayati et al., 2025). The beneficial effects of phytogenic substances are largely attributed to their diverse bioactive constituents, including phenolic compounds, flavonoids, alkaloids, and essential oils, which exert physiological regulatory functions through antimicrobial, antioxidant, and immunomodulatory pathways (Rafeeq et al., 2023). Zanthoxylum bungeanum is an important economic crop in China, and a large amount of seed by-products is generated during its processing (Hou et al., 2019). Studies have shown that Zanthoxylum bungeanum seeds (ZBS) are rich in bioactive components, including polysaccharides, unsaturated fatty acids, flavonoids, and dietary fiber (Qi et al., 2024), indicating considerable potential for further utilization. However, studies investigating the application of ZBS in broiler production remain limited.

The gut microbiota is an important factor influencing nutrient metabolism, intestinal health, and immune function in poultry. The cecum, as the primary site of microbial fermentation in the hindgut of poultry, harbors a microbial community whose composition and metabolic functions are closely associated with host physiological status (Ali et al., 2022). Because poultry lack an efficient endogenous enzyme system for degrading complex non-starch polysaccharides (Adedokun and Olojede, 2019), some polysaccharides and dietary fiber derived from plant-based feed ingredients can reach the hindgut and be utilized by microorganisms, thereby affecting microbial composition and functional characteristics (Holscher, 2017). Recent studies have demonstrated that various phytogenic substances can modulate gut microbiota composition by increasing the abundance of beneficial bacteria such as Lactobacillus and Bifidobacterium, while reducing potentially pathogenic populations (Mu et al., 2024; Huang et al., 2025). Therefore, ZBS powder, which is rich in polysaccharides and dietary fiber, may regulate host physiological functions through modulation of the gut microbial ecosystem.

Beyond intestinal health, immune function is also an important criterion for evaluating the application value of feed ingredients. The spleen is an important peripheral immune organ in poultry and plays a critical role in humoral immunity and inflammatory regulation (Zhang et al., 2019). The Toll-like receptor 4 (TLR4)/myeloid differentiation primary response 88 (MyD88)/nuclear factor kappa B (NF-κB) signaling pathway is a key molecular pathway involved in the regulation of innate immunity and inflammatory responses, and its excessive activation is often accompanied by increased production of inflammatory cytokines and tissue damage (Block et al., 2018; Gitlin et al., 2024). Previous studies have demonstrate that a wide range of plant-derived bioactive compounds can modulate immune function by regulating the TLR4/MyD88/NF-κB signaling pathway. Farnesol inhibits the TLR4/NF-κB axis against intestinal injury (Li et al., 2025). Macleaya cordata isoquinoline alkaloids regulate the complete TLR4/MyD88/NF-κB pathway to alleviate epithelial damage (Liu et al., 2024a). Callicarpa nudiflora aqueous extract down-regulates MyD88-related inflammatory mediators (Liu et al., 2024b), whereas tannic acid from Galla chinensis suppresses inflammation through NF-κB inhibition (Jing et al., 2022). Additionally, peppermint extract mitigates splenic inflammation provoked by heat stress via restraining NF-κB-dependent innate immune activation (Ma et al., 2024).

Research on the application of Zanthoxylum bungeanum in poultry production has largely focused on its pericarp and related extracts. Studies have shown that pericarp powder or Zanthoxylum powder can improve growth performance and immune function in broilers (Ma et al., 2025), and that leaf powder can enhance antioxidant capacity via the Nrf2/HO-1 signaling pathway (Hu et al., 2024). In contrast, ZBS, which is generated as the primary by-product during processing, has remained largely unexplored. Although a few studies have investigated the effects of ZBS powder on growth performance and immune organ indices in broilers (Chen et al., 2024), its influence on cecal microbial communities and the splenic TLR4/MyD88/NF-κB signaling pathway remains unclear.

Therefore, the objective of this study was to systematically evaluate the effects of dietary supplementation with ZBS powder on growth performance, immune function, cecal microbial composition and predicted functions, and the splenic TLR4/MyD88/NF-κB signaling pathway in 817 broilers, thereby providing a theoretical basis for the utilization of ZBS resources and their application in broiler production.

Materials and methods

Ethics approval

This study was reviewed and approved by the Institutional Animal Care and Use Committee (IACUC) of Southwest University, and is subject to its supervision. The experimental protocol complies with the guidelines of this committee (IACUC-20220920-01).

Zanthoxylum bungeanum seed and animal source

The ZBS meal used in this experiment was provided by the Chongqing Academy of Agricultural Sciences. It was obtained by drying and grinding the residual seeds after the removal of the pericarp (the chemical composition of the ZBS meal is presented in Table 1). A total of 260 healthy 1-day-old 817 broilers were purchased from Sichuan Suining Chia Tai Food Co., Ltd. (Suining, China).

Table 1.

Ingredient composition and nutritional level of Zanthoxylum bungeanum seed.

Nutrients componenta		Amino acidsa, %		Amino acidsa, %
Moisture, %	11.39	Asp	0.84	Phe	0.30
CP, %	10.93	Ser	0.21	Met	0.10
CF, %	22.32	Glu	0.67	Lys	0.49
CA, %	9.70	Gly	0.40	Thr	0.23
Ca, %	1.04	Ala	0.33	Leu	0.25
Total P, %	0.45	Cys	0.00	Ile	0.52
Chemical composition2		Tyr	0.05	Val	0.28
Total flavonoids, mg/g	7.02	His	0.02	EAAb	2.17
Flavonoid, mg/g	3.12	Arg	0.39	NEAA	3.22
Polyphenol, μmol/g	19.25	Pro	0.31	TAA	5.39
Polysaccharide, mg/g	11.48

^aAll data are measured values;

^bThe total amount of the essential amino acids, except Tryptophan.

Experimental design, dietary composition, and management

A single-factor completely randomized design was adopted in this experiment, which was divided into a starter period (d 1−28) and a grower period (d 29−63), to evaluate the phase-specific effects of ZBS supplementation and to capture the long-term cumulative effects of ZBS bioactive compounds on immune and microbiota parameters. A total of 260 1-day-old 817 broilers were randomly assigned to 4 treatments. During the brooding period, each treatment consisted of 5 replicates with 13 birds per replicate, while during the growing period, each treatment contained 5 replicates with 12 birds per replicate. The experimental diets consisted of a control group (CON, a basal corn-soybean meal diet) and three ZBS-formulated groups, where ZBS was incorporated into the basal diet at inclusion levels of 1% (ZBS1 group), 2% (ZBS2 group), and 3% (ZBS3 group), respectively. The basal diets were formulated to meet or exceed the nutrient requirements recommended by the NRC (1994) and the Chinese Chicken Feeding Standard (NY/T 33-2004). The composition and nutrient levels of the basal diets are listed in Table 2, and the ingredient compositions of the experimental diets are presented in Table 3. All experimental diets were offered in mash form, and the total feeding trial lasted for 63 d.

Table 2.

Ingredient composition and nutrient levels of basal diets.

Ingredient (%)	Starter perioda (1–28 d)	Grower periodb (29–63 d)	Nutrients contentc	Starter perioda (1–28 d)	Grower periodb (29–63 d)
Corn	55.87	64.64	ME, MJ/kg	12.40	12.60
Soybean meal (44% CP)	32.04	27.00	CP, %	21.75	18.52
Expanded soybean	4.00	0.00	CF, %	1.05	1.12
Fish meal (60.2% CP)	2.50	2.00	Ca, %	0.48	0.39
Calcium hydrophosphate	1.40	1.10	Total P, %	0.54	0.48
Limestone	1.00	1.60	Available P, %	0.25	0.20
Soybean oil	2.00	2.50	Lys, %	1.40	1.17
Sodium chloride	0.22	0.22	Met, %	0.54	0.48
Sodium bicarbonate	0.20	0.20	Trp, %	0.25	0.20
HCl-Lysine	0.25	0.24
DL-Methionine	0.22	0.20
Vitamin-mineral Premix	0.30	0.30
Total	100.00	100.00

^aSupplied per kg of diet:16 mg Cu, 20 mg Fe, 1.25 mg I, 120 mg Mn, 110 mg Zn, 0.3 mg Se, 11 000 IU vitamin A, 4 750 IU vitamin D₃, 72.5 IU vitamin E, 3.1 mg vitamin K₃, 7.55 mg vitamin B₂, 2.85 mg vitamin B₁, 3.75 mg vitamin B₆, 0.017 mg vitamin B₁₂, 19 mg pantothenic acid, 62.5 mg niacin, 0.2 mg biotin, 2.05 mg folic acid, 250 mg chloride, 50 mg mold inhibitor, 200 mg antioxidant.

^bSupplied per kg of diet on finisher period:16 mg Cu, 20 mg Fe, 1.25 mg I, 120 mg Mn, 110 mg Zn, 0.3 mg Se, 9 000 IU vitamin A, 4 000 IU vitamin D₃, 55 IU vitamin E, 2.2 mg vitamin K₃, 5.4 mg vitamin B₂, 2.2 mg vitamin B₁, 2.2 mg vitamin B₆, 0.011 mg vitamin B₁₂, 15 mg pantothenic acid, 45 mg niacin, 0.15 mg biotin, 1.6 mg folic acid, 250 mg chloride, 50 mg mold inhibitor, 200 mg antioxidant.

^cNutrients contents are theoretical values.

Table 3.

Ingredient composition and nutrient levels of experimental diets.

Items	CON	ZBS1	ZBS2	ZBS3
Starter period (1–28 d)
Basal diet, %	97	97	97	97
Rice bran, %	2.3	1.55	0.8	0
Soybean meal, %	0.7	0.45	0.2	0
ZBS, %	0	1	2	3
Total, %	100	100	100	100
Nutrients contenta
ME, MJ/kg	12.21	12.23	12.25	12.27
CP, %	21.50	21.47	21.44	21.43
Grower period (29–63 d)
Basal diet, %	97	97	97	97
Rice bran, %	2.3	1.55	075	0
Soybean meal, %	0.7	0.45	0.25	0
ZBS, %	0	1	2	3
Total, %	100	100	100	100
Nutrients contenta
ME, MJ/kg	12.41	12.43	12.45	12.47
CP, %	18.36	18.36	18.37	18.36

^aNutrients contents are theoretical values.

The feeding trial was conducted at the Experimental Poultry Farm of Southwest University. One week prior to the arrival of the chicks, the poultry house and all facilities (including ventilation, heating, drinking systems, and cages) were thoroughly inspected, cleaned, and disinfected. The ambient temperature was initially maintained at 33 to 35 °C during the first week, and then decreased by 3 °C weekly until reaching a constant temperature of 24 °C. The relative humidity was controlled at 65% to 70% during weeks 1 to 2, and then reduced to 60% to 65% from week 3 onward. A continuous lighting regime was applied initially, with the dark period gradually extended as the birds aged. At 2 weeks of age, all broilers were immunized against Newcastle disease via drinking water using a commercial live vaccine (LaSota strain; YEBIO Bio-engineering Co., Ltd., Qingdao, China).

Diets were offered twice daily at 8:00 and 14:30 to ensure ad libitum access to feed and clean water. Bird health, behavior, and excreta characteristics were monitored daily, while heating, ventilation, and sanitation were strictly managed, including floor sweeping and disinfection every morning and evening. All broilers were reared in multi-layer wire cages of identical dimensions (40 cm × 175 cm × 155 cm). To maintain appropriate stocking density, each replicate was housed in one cage (13 birds per cage) during the first phase (d 1−28), and then split into three adjacent cages (4 birds per cage, totaling 12 birds per replicate) during the second phase (d 29−63). Feed allocation and orts (residual feed) were recorded daily on a replicate basis to determine feed consumption. The body weight of individual broilers was measured at d 1, 28, and 63 to evaluate growth performance.

Sample collection and processing

Serum sample collection

On d 63 of the feeding trial, one broiler close to the average body weight was randomly selected from each replicate (totaling 5 birds per treatment). Blood samples (10 mL) were collected from the wing vein into procoagulation tubes. After standing at room temperature for 30 min, the blood samples were centrifuged at 3000 rpm for 15 min to separate the serum. The isolated serum was then aliquoted and stored at −80 °C for subsequent determination of serum immunoglobulins and cytokines.

Tissue and intestinal digesta collection

Immediately following blood collection, the same broilers were euthanized and exsanguinated. The abdominal cavity was opened to expose and identify the target segments and tissues. The spleen was rapidly excised, wrapped in sterile aluminum foil, properly labeled, and immediately frozen in liquid nitrogen for the determination of splenic signaling pathway parameters. Subsequently, the cecal segments were ligated, and the cecal digesta was collected into sterile microcentrifuge tubes on ice. The digesta samples were labeled and stored at −80 °C for 16S rRNA gene sequencing.

Parameters and analytical methods

Determination of Zanthoxylum bungeanum seed components

The contents of total flavonoids and polysaccharides in the ZBS meal were determined using ultrasound-assisted extraction, according to the methods described by Cui et al. (2022) and Wei and Zhang (2023). Specifically, total flavonoids were quantified by colorimetry, while total phenolics were determined by spectrophotometry. The amino acid profiles were analyzed using a Hitachi L-8800 amino acid analyzer (Hitachi Ltd., Tokyo, Japan) in accordance with the National Standards of the People's Republic of China (GB/T 18246-2000). The crude fiber content was determined following the official procedure specified in GB/T 6438-2007.

Growth performance

On d 1, 28, and 63 of the experiment, all broilers were weighed individually after an overnight fast (feed was withdrawn at 20:00 h the previous evening, while water remained accessible ad libitum). Feed consumption and remaining orts were recorded daily on a cage basis. The average daily gain (ADG), average daily feed intake (ADFI), and feed-to-gain ratio (F/G) per broiler were calculated on a replicate basis for each periods (d 1 to 28, d 29 to 63, and d 1 to 63) using the following formulas:

$$\text{ADG}(\mathrm{g}/\mathrm{d})=\frac{\text{Total}\text{body}\text{weight}\text{gain}\text{per}\text{replicate}\left(\mathrm{g}\right)}{\text{Number}\text{of}\text{broilers}\times \text{Number}\text{of}\text{experimental}\text{days}\left(\mathrm{d}\right)}$$

$$\text{ADFI}(\mathrm{g}/\mathrm{d})=\frac{\text{Total}\text{feed}\text{consumption}\text{per}\text{replicate}\left(\mathrm{g}\right)}{\text{Number}\text{of}\text{broilers}\times \text{Number}\text{of}\text{experimental}\text{days}\left(\mathrm{d}\right)}$$

$$\mathrm{F}/\mathrm{G}=\frac{\text{Total}\text{feed}\text{consumption}\text{per}\text{replicate}\left(\mathrm{g}\right)}{\text{Total}\text{body}\text{weight}\text{gain}\text{per}\text{replicate}\left(\mathrm{g}\right)}$$

Immune organ indices

Immediately following euthanasia and weighing on d 63, the spleen and bursa of Fabricius were excised from the selected broilers (one bird per replicate, totaling 5 birds per treatment). Individual organ weights were recorded after removing adherent fat and tissues. The immune organ indices were expressed as organ weight relative to live body weight using the following formula:

$$\text{Organ}\text{index}(\mathrm{g}/\text{kg})=\frac{\text{Organ}\text{weight}\left(\mathrm{g}\right)}{\text{Live}\text{body}\text{weight}\left(\text{kg}\right)}$$

Serum and splenic immune and cytokine parameters

The total protein content in serum and splenic tissue homogenates was determined by the Coomassie brilliant blue method using a BioTek ELx800 microplate reader (BioTek Instruments, Inc., Winooski, VT, USA). For splenic parameters, the frozen spleen samples were thawed on ice, weighed, and homogenized in 9 volumes of ice-cold sterile physiological saline (1:9, w/v). The homogenates were then centrifuged at 3000 rpm for 10 min at 4 °C, and the supernatants were collected for subsequent analysis and normalized to the total protein concentration of the tissue.

Commercial enzyme-linked immunosorbent assay (ELISA) kits were employed to quantify the concentrations of immunoglobulins (IgA, IgG, and IgM), complement proteins (C3 and C4), and cytokines (IL-1β and TNF-α) in both serum and splenic supernatants; additionally, serum IL-2 and IL-6 levels were also determined using corresponding ELISA kits. All ELISA measurements were performed using a Rayto RT-6100 microplate reader (Rayto Life and Analytical Sciences Co., Ltd., Shenzhen, China) according to the manufacturer's protocols (Quanzhou Ruixin Biotechnology Co., Ltd., Quanzhou, China).

Splenic gene expression analysis

Specific primers for broiler splenic glyceraldehyde-3-phosphate dehydrogenase (GAPDH), TNF receptor-associated factor 2 (TRAF2), Toll-like receptor 4 (TLR4), MyD88 innate immune signal transduction adaptor (MyD88), nuclear factor kappa B subunit 2 (NFKB2), inhibitor of nuclear factor kappa B kinase subunit alpha (IKKA), and NFKB inhibitor alpha (NFKBIA) were designed and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China), as detailed in Table 4. Total RNA was extracted from the frozen spleen samples using the SteadyPure Universal RNA Extraction Kit (Accurate Biotechnology Co., Ltd., Changsha, China) following the manufacturer's instructions and previous protocols (Hu et al., 2024). The purity and concentration of the extracted RNA were quantified spectrophotometrically. Subsequently, complementary DNA was synthesized from the isolated RNA using the Evo M-MLV RT Premix Kit (Accurate Biotechnology Co., Ltd., Changsha, China).

Table 4.

Sequences of PCR primers.

Gene name	Primer sequence (5′→3′)	Accession number
TRAF2	F:GGTTGTCGGCTGTGCTGAGATG R:TGCTTCAGGTCTCCAGATCCACTC	XM_025141427.3
TLR4	F:GCCATCCCAACCCAACCACAG R:CCACTGAGCAGCACCAATGAGTAG	NM_001030693.2
MyD88	F:TGAGAAGCCTGTGGACATGAGAG R:GGCAAGAGGACACGAAGATGAAC	XM_046910878.1
NFKB2	F:ACCGTCCTGTCACCGTATTCC R:GCTTCCGCTCCACCTCCTC	XM_046919839.1
IKKA	F:CCTTTGCTTCTCGGAGTCTGTCTG R:AACTGCTTCTGCCCATACCTTGC	NM_001012904.2
NFKBIA	F:GGCTGTTGTCGAATACCTGC R:TCTACGGCCAAGTGTAGTGC	NM_001001472.3
GADPH	F:CAGAACATCATCCCAGCGTCCAC R:CGGCAGGTCAGGTCAACAACAG	NM_204305.2

Abbreviations: GAPDH, glyceraldehyde-3-phosphate dehydrogenase; TRAF2, TNF receptor-associated factor 2; TLR4, Toll-like receptor 4; MyD88, MyD88 innate immune signal transduction adaptor; NFKB2, nuclear factor kappa B subunit 2; IKKA, inhibitor of nuclear factor kappa B kinase subunit alpha; NFKBIA, NFKB inhibitor alpha.

Real-time quantitative PCR was performed using the SYBR Green Premix Pro Taq HS qPCR Kit (Accurate Biotechnology Co., Ltd., Changsha, China) on a LightCycler 480 Real-Time PCR System (Roche Diagnostics GmbH, Mannheim, Germany). Each sample was analyzed in duplicate, and GAPDH was employed as the endogenous reference gene to normalize the expression of target genes. The relative mRNA expression levels were calculated using the comparative C_T(2^-△△CT) method (Livak and Schmittgen, 2001), The final results were expressed as the fold-change in target gene expression in the experimental treatments relative to the CON group. where:

$$▵▵{\mathrm{C}}_{\mathrm{T}}=({\mathrm{C}}_{\mathrm{T}}\text{target}-{\mathrm{C}}_{\mathrm{T}}\text{reference})\text{treatment}-({\mathrm{C}}_{\mathrm{T}}\text{target}-{\mathrm{C}}_{\mathrm{T}}\text{reference})\text{control}$$

Cecal microbiota 16S rRNA gene sequencing and bioinformatics

The cryopreserved cecal digesta samples were dispatched to LC-Bio Technology Co., Ltd. (Hangzhou, China) for microbial genomic DNA extraction, 16S rRNA gene PCR amplification, and library construction. The hypervariable V3–V4 regions of the microbial 16S rRNA gene were amplified and subjected to paired-end sequencing on an Illumina sequencing platform (Illumina, Inc., San Diego, CA, USA). Raw sequencing reads were assembled, quality-controlled, and filtered to obtain clean data. Amplicon sequence variants (ASVs) were then resolved from the refined dataset using the QIIME2 pipeline. Alpha diversity indices and beta diversity metrics were calculated to evaluate the overall microbial community structure, and taxonomic classification was performed based on the SILVA database. Potential metabolic and functional profiles of the cecal microbiota were predicted using the PICRUSt2 tool. Redundancy analysis (RDA) was further implemented to determine the correlation between the microbiota composition and host systemic or splenic immune phenotypes.

Statistical analysis

Data were analyzed on a replicate basis using SPSS 21.0 software. The statistical differences among the experimental treatments were evaluated by one-way ANOVA, followed by Tukey's multiple comparison test. Orthogonal polynomial contrasts were implemented to detect the linear and quadratic effects of dietary ZBS inclusion levels. Furthermore, a quadratic regression model (Y = aX² + bX + c) was established to determine the continuous response of host phenotypes to different dietary ZBS levels, with a coefficient of determination (R² > 0.50) used as the screening threshold for mathematical fit. The optimal ZBS inclusion level for the specific targeted parameter was calculated using the inflection point formula (-b/2a). Results are expressed as means ± standard error of the mean (SEM), and statistical significance was declared at P < 0.05.

Results

Growth performance and immune organ indices

As presented in Table 5, dietary ZBS inclusion levels exerted no statistically significant effects on the overall ADFI, ADG, and F/G of broilers throughout the experimental periods (P > 0.05). However, BW at d 28 and 63 and ADG during d 1−28 and 1−63 showed quadratic responses (quadratic, P < 0.05), peaking in the ZBS2 group and exceeding CON at most levels.

Table 5.

Effect of dietary Zanthoxylum bungeanum seed (ZBS) on growth performance in broilers.

Item	Groupc				SEMa	P-Valueb
	CON	ZBS1	ZBS2	ZBS3		A	L	Q
1 d BW, g	29.28	29.31	29.54	29.60	0.33	0.701	0.270	0.970
28 d BW, g	275.55	286.54	291.57	258.83	13.40	0.112	0.303	0.035
63 d BW, kg	1.25	1.31	1.31	1.27	0.04	0.184	0.626	0.037
1–28 d
ADFI, g/d	20.68	20.67	20.37	19.93	1.32	0.933	0.550	0.819
ADG, g/d	8.80	9.19	9.36	8.19	0.48	0.109	0.293	0.034
F/G, g/g	2.35	2.26	2.19	2.44	0.17	0.520	0.691	0.187
29–63 d
ADFI, g/d	75.97	79.49	78.71	78.74	4.12	0.837	0.571	0.557
ADG, g/d	27.72	29.38	29.19	28.76	0.79	0.198	0.261	0.081
F/G, g/g	2.75	2.70	2.69	2.74	0.13	0.966	0.957	0.622
1–63 d
ADFI, g/d	51.39	53.35	52.78	52.60	2.43	0.876	0.696	0.544
ADG, g/d	19.31	20.41	20.37	19.62	0.58	0.184	0.635	0.037
F/G, g/g	2.66	2.61	2.59	2.68	0.10	0.762	0.923	0.315

^a–dMeans within the same row with no common superscript differ significantly (P < 0.05).

Abbreviations: BW, body weight; ADFI, average daily feed intake; ADG: average daily gain;G/F, gain-to-feed ratio.

^aStandard error of the means (n=5).

^bOrthogonal polynomials were used to evaluate linear and quadratic responses to the levels of ZBS treatment.

^cCON, basal diet; ZBS1, ZBS2 and ZBS3 group, basal diet contains 1, 2, and 3% Zanthoxylum bungeanum seed powder, respectively.

The relative immune organ indices are summarized in Table 6. The bursa of Fabricius index increased linearly with the elevation of dietary ZBS inclusion levels (linear, P < 0.05). Regarding the spleen index, no statistically significant differences were observed among all experimental treatments (P > 0.05), although a numerical quadratic trend was noted, with the highest value observed in the ZBS1 group followed by a subsequent decrease as the ZBS inclusion further increased.

Table 6.

Effect of dietary ZBS on organ index in broilers (g/kg).

Item	Groupc				SEMa	P-Valueb
	CON	ZBS1	ZBS2	ZBS3		A	L	Q
Spleen index	2.03	2.51	2.35	2.34	0.40	0.682	0.529	0.402
Bursal index	1.80	2.42	2.73	3.55	0.62	0.075	0.012	0.834

^a–dMeans within the same row with no common superscript differ significantly (P < 0.05).

^aStandard error of the means (n=5).

^bOrthogonal polynomials were used to evaluate linear and quadratic responses to the levels of ZBS treatment.

^cCON, basal diet; ZBS1, ZBS2 and ZBS3 group, basal diet contains 1, 2, and 3% Zanthoxylum bungeanum seed powder, respectively.

Serum immunoglobulins and cytokines

As presented in Table 7, compared with the control group, dietary inclusion of 2% and 3% ZBS significantly increased the concentrations of IL-2, IL-6, TNF-α, IgA, and IgG in the serum of broilers (P < 0.05). Notably, increasing the dietary ZBS inclusion levels led to a linear increase in the serum concentrations of IL-2, IL-6, IL-1β, TNF-α, IgA, and IgG (linear, P < 0.05).

Table 7.

Effect of dietary ZBS on serum immune parameters broilers.

Item	Groupd				SEMb	P-Valuec
	CON	ZBS1	ZBS2	ZBS3		A	L	Q
IL-2, pg/mL	184.90a	225.86a	259.40a	252.69a	19.07	0.005	0.001	0.096
IL-6, pg/mL	16.07a	18.91a	22.63a	22.80a	2.06	0.013	0.002	0.374
IL-1β, pg/mL	319.96a	341.92a	422.47a	466.43a	47.79	0.024	0.003	0.749
TNF-α, pg/mL	33.93a	50.70a	51.11a	57.31a	4.52	<0.001	<0.001	0.118
IgA, μg/mL	190.31a	203.67a	266.96a	264.62a	18.32	<0.001	<0.001	0.553
IgG, μg/mL	2413.64a	2321.52a	2926.13a	2923.36a	115.43	<0.001	<0.001	0.592
IgM, μg/mL	564.57a	635.12a	699.60a	799.07a	51.34	0.002	<0.001	0.696

^aMeans within the same row with no common superscript differ significantly (P < 0.05).

Abbreviations: IL-2, Interleukin-2; IL-6, Interleukin-6; IL-1β, Interleukin-1 beta; TNF-α, tumor necrosis factor-alpha; IgA, immunoglobulin A; IgG, immunoglobulin G; IgM, immunoglobulin M;

^bStandard error of the means (n=5).

^cOrthogonal polynomials were used to evaluate linear and quadratic responses to the levels of ZBS treatment.

^dCON, basal diet; ZBS1, ZBS2 and ZBS3 group, basal diet contains 1, 2, and 3% Zanthoxylum bungeanum seed powder, respectively.

Cecal Microbiota

Alpha and beta diversity and composition of cecal microbiota

As illustrated in Fig. 1 A–C, α-diversity analysis revealed that variations in dietary ZBS inclusion levels exerted no statistically significant effects on the richness (Chao1 index) and diversity (Shannon and Simpson indices) of the cecal microbiota (P > 0.05). However, β-diversity analysis via principal component analysis (PCA) demonstrated a distinct separation in the microbial community structure between the ZBS1 group and the CON and ZBS2 groups (P < 0.05; Fig. 1D). This indicated that dietary ZBS inclusion primarily modulated the microbial community composition rather than its overall alpha diversity. Taxonomic composition analysis at the phylum and genus levels is presented in Fig. 1 (E, F). At the phylum level, Firmicutes, Bacteroidota, Desulfobacterota, Proteobacteria, and Verrucomicrobiota constituted the top 5 dominant phyla across all treatments. Among these, the relative abundance of Firmicutes showed a quadratic response—initially decreasing and subsequently increasing—with the elevation of dietary ZBS levels, whereas Bacteroidota exhibited the opposite trend. Notably, the ZBS1 group possessed the lowest relative abundance of Firmicutes and the highest abundance of Bacteroidota. At the genus level, Faecalibacterium, Bacteroides, Alistipes, Barnesiella, and UCG-005 were identified as the top 5 dominant genera across all groups. The abundance of Faecalibacterium displayed a quadratic pattern of an initial decline followed by a subsequent rise as dietary ZBS levels increased. Remarkably, the abundance of Faecalibacterium in the ZBS3 group increased by 160.80% compared with the CON group, whereas the abundances of Alistipes, Barnesiella, and UCG-005 in the ZBS-supplemented treatments were substantially lower than those in the control group.

Fig. 1.

Effects of dietary ZBS levels on alpha and beta diversity of the cecal microbiota in broilers. A-C: represent the boxplots of Chao1, Shannon, and Simpson indices of cecal microbiota, respectively; P-values are calculated based on the Kruskal-Wallis test; D: the Principal Component Analysis (PCA) biplot based on Bray-Curtis distance; R- and P-values are determined by ANOSIM/Adonis test (P < 0.01 indicates a highly significant difference in community structure among groups); E and F: represent the relative abundance bar plots of cecal microbiota at the phylum and genus levels, respectively; CON, control group (0% ZBS); ZBS1, 1% ZBS group; ZBS2, 2% ZBS group; ZBS3, 3% ZBS group. The same applies below.

Identification of key differential cecal microbiota via LEfSe analysis

The LEfSe method (LDA score > 3.0, P < 0.05) was employed to systematically screen the differential taxa among the experimental groups, generating an LDA score bar chart (Fig. 2A) and a cladogram (Fig. 2B). The results revealed that a total of 15 taxonomically distinct groups with statistical significance were enriched across the different treatments. Specifically, the CON group exhibited a significant enrichment of Peptococcales and its corresponding family Peptococcaceae. The cecal microbiota of broilers in the ZBS1 group was characterized by a prominent enrichment of Veillonellales_Selenomonadales and Alistipes_senegalensis, whereas the ZBS2 group demonstrated a significant enrichment of Bacteroides uniformis. Notably, the ZBS3 group harbored the largest number of specific biomarker taxa; at the phylum level, it was primarily enriched with Actinobacteriota and its subordinate class Actinobacteria, while at the genus level, core beneficial genera such as Faecalibacterium, Megasphaera, and Bifidobacterium displayed prominent enrichment advantages in the 3% ZBS inclusion group.

Fig. 2.

Selection and verification of significantly differential cecal microbiota among different dietary ZBS inclusion levels based on LEfSe analysis. A: the horizontal bar plot of LDA scores for significantly differential taxa (threshold: LDA Score > 3 and P < 0.05), with different colors representing significant enrichment in the respective groups; B: the cladogram generated from LEfSe analysis; circles from inside to outside represent phylum, class, order, family, and genus, respectively; C–E: the relative abundance verification of key differential taxa—p_Actinobacteriota, f_Peptococcaceae, and g_Faecalibacterium—via one-way ANOVA. Data are expressed as Mean ± SEM; "*" above the bars indicates a significant difference between groups (P < 0.05).

One-way ANOVA was further implemented to analyze the relative abundances of the key differential taxa identified by LEfSe. As illustrated in Fig. 2C, the ZBS3 group significantly increased the relative abundance of the phylum Actinobacteriota compared with the CON, ZBS1, and ZBS2 groups (P < 0.05), while no statistically significant differences were observed among the latter three groups. Fig. 2D demonstrated that the relative abundance of Peptococcaceae decreased with the increase of dietary ZBS inclusion levels, with the ZBS3 group being significantly lower than the CON group (P < 0.05). Furthermore, Fig. 2E showed that the relative abundance of Faecalibacterium in the ZBS3 group reached over 20%, which was significantly higher than that in all other experimental groups (P < 0.05).

Functional prediction of cecal microbiota via PICRUSt2 analysis

PICRUSt2 functional prediction revealed that COG categories related to carbohydrate and lipid metabolism, including pyruvate-formate lyase (PFL), predicted phospholipase, isocitrate/isopropylmalate dehydrogenase, and acyl-CoA hydrolase, were maintained at high baseline abundances across all groups (Fig. 3A). Notably, PFL abundance in the ZBS3 group showed a marked increase, consistent with the enrichment of butyrate-producing Faecalibacterium in this group. Additionally, the ZBS3 group exhibited increasing trends in genes involved in energy metabolism and redox processes, such as GTPase (G3E family) and FMN-dependent dehydrogenase.

Fig. 3.

Potential functional prediction of the cecal microbiota in broilers across different dietary ZBS inclusion levels based on PICRUSt2. A: the relative abundance proportion of major functional enzymes/proteins (e.g., KOs or orthologous groups) in the cecal microbial community predicted by PICRUSt2. B: the gene abundance proportion of level-2 or level-3 KEGG pathways (e.g., carbohydrate and energy metabolism) annotated via the KEGG database. The vertical axis represents functional descriptions, and the horizontal axis represents the mean proportions of predicted genes. Colors of the bars denote different dietary treatment groups, and error bars indicate the standard error of the mean (SEM).

KEGG pathway prediction (Fig. 3B) showed that cecal microbial functional genes across all groups were predominantly enriched in starch and sucrose metabolism, photosynthesis proteins, and carbohydrate metabolism. The ZBS3 group exhibited significantly higher abundances in starch and sucrose metabolism and carbohydrate metabolism pathways, as well as in the phosphotransferase system (PTS) pathway, compared with the other groups. Conversely, the ZBS1 group showed slightly higher or comparable abundances in branched-chain amino acid degradation pathways (valine, leucine, and isoleucine degradation) relative to the CON group.

Splenic immune parameters, TLR4 signaling pathway, and RDA analysis

The immune parameters in splenic tissues were determined and summarized in Table 8. Compared with the CON group, dietary inclusion of 3% ZBS significantly increased the contents of C3, C4, IgA, and IgG in the splenic tissues of broilers (P < 0.05), whereas the splenic IL-1β concentration was significantly decreased across all ZBS inclusion groups. Notably, with increasing dietary ZBS levels, the splenic contents of C3, C4, and TNF-α exhibited clear quadratic responses (quadratic, P < 0.05), while the IL-1β content decreased linearly, and the IgA, IgG, and IgM contents increased linearly (linear, P < 0.05).

Table 8.

Effect of dietary ZBS on spleen immune parameters in broilers.

Item	Groupc				SEM1	P-Valueb
	CON	ZBS1	ZBS2	ZBS3		A	L	Q
C3, μg/mgprot	150.19a	114.50a	160.21a	181.52a	14.07	0.002	0.006	0.011
C4, μg/mgprot	52.35a	46.72a	52.91a	66.34a	4.41	0.003	0.003	0.008
IL-1β, pg/mgprot	133.13a	94.09a	87.30a	85.76a	11.62	0.003	<0.001	0.037
TNF-α, pg/mgprot	11.56a	8.05a	8.37a	8.93a	1.21	0.040	0.065	0.030
IgA, μg/mgprot	72.71a	76.24a	81.78a	99.99a	7.42	0.010	0.002	0.181
IgG, μg/mgprot	319.30c	419.55a	450.21a	569.40a	44.54	<0.001	<0.001	0.768
IgM, μg/mgprot	132.57	125.50	160.21	181.52	20.28	0.050	0.012	0.337

^aMeans within the same row with no common superscript differ significantly (P < 0.05).

Abbreviations: C3, Complement 3; C4, Complement 4; IL-1β, Interleukin-1 beta; TNF-α, tumor necrosis factor-alpha; IgA, immunoglobulin A; IgG, immunoglobulin G; IgM, immunoglobulin M;

¹Standard error of the means (n=5).

^bOrthogonal polynomials were used to evaluate linear and quadratic responses to the levels of ZBS treatment.

^cCON, basal diet; ZBS1, ZBS2 and ZBS3 group, basal diet contains 1, 2, and 3% Zanthoxylum bungeanum seed powder, respectively.

The mRNA relative expression levels of key genes involved in the splenic TLR4/MyD88/NF-κB signaling pathway were evaluated via RT-qPCR (Table 9). Compared with the CON group, dietary ZBS inclusion significantly down-regulated the relative expression levels of splenic TRAF2, TLR4, MyD88, NFKB2, IKKA, and NFKBIA genes in a distinct linear manner (linear, P < 0.05).

Table 9.

Effect of dietary ZBS on the TLR4/MyD88/NF-κB pathway in the spleen in broilers.

Item	Groupd				SEMb	P-Valuec
	CON	ZBS1	ZBS2	ZBS3		A	L	Q
TRAF2	1.05a	0.38a	0.48a	0.24a	0.14	<0.001	<0.001	0.048
TLR4	1.04a	0.51a	0.45a	0.51a	0.14	0.003	0.002	0.030
MyD88	1.04a	0.80a	0.73a	0.44a	0.19	0.046	0.007	0.873
NFkB2	1.02a	0.51a	0.51a	0.50a	0.14	0.003	0.002	0.019
IKKA	1.07a	0.46a	0.49a	0.34a	0.16	0.001	<0.001	0.055
NFkBIA	1.07a	0.56a	0.61a	0.35a	0.17	0.005	0.001	0.315

^aMeans within the same row with no common superscript differ significantly (P < 0.05).

Abbreviations: TRAF2, TNF receptor-associated factor 2; TLR4, Toll-like receptor 4; MyD88, MyD88 innate immune signal transduction adaptor; NFKB2, nuclear factor kappa B subunit 2; IKKA, inhibitor of nuclear factor kappa B kinase subunit alpha; NFKBIA, NFKB inhibitor alpha.

^bStandard error of the means (n=5).

^cOrthogonal polynomials were used to evaluate linear and quadratic responses to the levels of ZBS treatment.

^dCON, basal diet; ZBS1, ZBS2 and ZBS3 group, basal diet contains 1, 2, and 3% Zanthoxylum bungeanum seed powder, respectively.

Furthermore, based on the quadratic regression models (Table 10), the quadratic regression models revealed that the optimal dietary ZBS inclusion response values for the C4 and IL-1β concentrations in the splenic tissues of 817 broilers were modeled at 1.00% and 2.29%, respectively. Concurrently, the optimal dietary ZBS inclusion response values for the mRNA relative expression of TRAF2, TLR4, and NFKB2 in the splenic tissues were determined to be 2.55%, 2.00%, and 2.07%, respectively.

Table 10.

Estimation of the maximum response for dietary ZBS levels based on quadratic regressions in broilers (n=5).

Dependent variablesa	Regression equation	R2	P-value	Extremum response to ZBS levels (%)
C4 (μg/mgprot)	Y=4.76 X2−9.48 X+52.12	0.569	0.008	1.00
IL-1β (pg/mgprot)	Y=9.38X2–43.02 X+131.78	0.567	0.037	2.29
TRAF2	Y=0.11 X2–0.56 X+1.00	0.596	0.048	2.55
TLR4	Y=0.15 X2–0.60 X+1.02	0.563	0.030	2.00
NFKB2	Y=0.13 X2–0.54 X+1.00	0.535	0.019	2.07

^aSpleen;

Extremum was the maximum or minimum response to dietary Zanthoxylum bungeanum seeds levels according to each regression equation (%); R², determination coefficient; P, P value of quadratic effect; Y was the dependent viable; X was the dietary ZBS level (%).

Abbreviations: C4, Complement 4; IL-1β, Interleukin-1 beta; TRAF2, TNF receptor-associated factor 2; TLR4, Toll-like receptor 4; NFKB2, nuclear factor kappa B subunit 2.

RDA was performed to elucidate the correlations between ZBS-induced cecal microbial shifts and host immune parameters (serum cytokines and splenic immune status), using the top 30 abundant genera and host parameters as environmental vectors. In the RDA model (Fig. 4A and D), RDA1 and RDA2 explained 7.901% and 6.168% of the total variance, respectively. The envfit test showed that dietary ZBS level (Contents) strongly explained microbial structural variation (R² = 0.598, P = 0.002; Table 11). Among serum parameters, IL-6 (R² = 0.489, P = 0.004), IL-1β (R² = 0.378, P = 0.011), and IL-2 (R² = 0.319, P = 0.031) significantly drove community shifts, whereas TNF-α did not (P > 0.05). The biplot revealed obtuse angles between the Contents vector and IL-6, IL-1β, and IL-2 vectors, indicating a negative correlation between ZBS inclusion and these inflammatory cytokines, which separated the high-dose groups (ZBS2 and ZBS3) from the inflammatory clusters.

Fig. 4.

Redundancy analysis biplots illustrating the correlations between cecal microbiota and host physiological phenotypes across different dietary ZBS levels. A-C: the complete RDA biplots of cecal microbiota (at the genus level of top 30) with serum inflammatory cytokines, splenic immune indicators, and splenic signaling molecules, respectively; D, E, and F: the corresponding simplified biplots focusing on samples and environmental factors; Blue arrows/texts represent cecal bacterial genera; red arrows represent environmental factors (Content, dietary ZBS level; others, physiological indicators). The angle and length of the red vectors reflect their explanatory power on microbiota variation. Acute and obtuse angles between red vectors indicate positive and negative correlations, respectively. Colored dots represent individual biological replicates from different groups.

Table 11.

Significance test (envfit) of the effects of dietary ZBS levels and serum cytokines on cecal microbiota structure.

Item	RDA1	RDA2	R2	P-value
Contents	–0.51	0.86	0.598	0.002
IL-2	–0.44	0.90	0.319	0.031
IL-6	0.21	0.98	0.489	0.004
IL-1β	–0.38	0.93	0.378	0.011
TNF-α	–0.97	0.24	0.231	0.119

Abbreviations: Contents, dietary Zanthoxylum bungeanum seed (ZBS) inclusion levels; RDA, redundancy analysis; IL, interleukin; TNF-α, tumor necrosis factor-alpha;

Values for RDA1 and RDA2 represent the directional cosines of the environmental vectors;

R², the goodness-of-fit statistic determined by the envfit function based on 999 permutations;

P < 0.05 indicates a statistically significant effect of the environmental vector on the microbial community variation.

Regarding the splenic immune parameters RDA model (Fig. 4B and E), RDA1 and RDA2 explained 14.67% and 10.55% of the microbial community variation, respectively. The envfit test demonstrated that besides the Contents factor, splenic IgA, IgG, C3, and C4 were identified as key drivers significantly shaping the cecal microbiota structure (P < 0.05; Table 12), while the inflammatory cytokines (IL-1β and TNF-α) exerted negligible explanatory power. Spatial vector distribution revealed that the Contents vector was tightly aligned in the same direction with the splenic IgA, IgG, IgM, C3, and C4 vectors, demonstrating a strong positive correlation and driving the aggregation of high-dose samples along these pathways.

Table 12.

Significance test (envfit) of the effects of dietary ZBS levels and splenic immune parameters on cecal microbiota structure.

Item	RDA1	RDA2	R2	P-value
Contents	–0.22	0.98	0.318	0.026
C3	0.10	0.99	0.431	0.006
C4	–0.01	1.00	0.331	0.033
IL-1β	1.00	–0.08	0.030	0.769
TNF-α	0.85	–0.52	0.151	0.250
IgA	–0.05	1.00	0.494	0.007
IgG	–0.60	0.80	0.409	0.011
IgM	–0.11	0.99	0.221	0.107

Abbreviations: Contents, dietary Zanthoxylum bungeanum seed (ZBS) inclusion levels; RDA, redundancy analysis; IL, interleukin; TNF-α, tumor necrosis factor-alpha; IgA, immunoglobulin A; IgG, immunoglobulin G; IgM, immunoglobulin M;

Values for RDA1 and RDA2 represent the directional cosines of the environmental vectors;

R², the goodness-of-fit statistic determined by the envfit function based on 999 permutations;

P < 0.05 indicates a statistically significant effect of the environmental vector on the microbial community variation.

For the RDA model of the splenic TLR4/NF-κB pathway molecules (Table 13; Fig. 4C and F), the results demonstrated that only the dietary ZBS inclusion level (Contents) served as a prominent driver reshaping the microbiota structure (R² = 0.470, P = 0.007), whereas individual downstream signaling molecules (TRAF2, TLR4, etc.) exerted no standalone significance. Crucially, the Contents vector formed distinct obtuse angles with the vectors of splenic inflammatory signaling molecules, confirming a clear negative correlation. This verified that dietary ZBS intervention represents the primary driver in reshaping the cecal microbiota while simultaneously modulating the downstream systemic down-regulation of the splenic TLR4/NF-κB inflammatory signaling cascade.

Table 13.

Significance test (envfit) of the effects of dietary ZBS levels and splenic signaling pathway molecules on cecal microbiota structure.

Item	RDA1	RDA2	R2	P-value
Contents	–0.96	–0.27	0.470	0.007
TRAF2	0.97	–0.25	0.123	0.317
TLR4	0.43	–0.90	0.037	0.718
MyD88	0.81	0.59	0.110	0.372
NFKB2	0.94	–0.35	0.050	0.650
IKKA	0.98	–0.19	0.072	0.519
NFKBIA	0.93	–0.37	0.230	0.107

Abbreviations: Contents, dietary Zanthoxylum bungeanum seed (ZBS) inclusion levels; RDA, redundancy analysis; TRAF2, TNF receptor-associated factor 2; TLR4, Toll-like receptor 4; MyD88, MyD88 innate immune signal transduction adaptor; NFKB2, nuclear factor kappa B subunit 2; IKKA, inhibitor of nuclear factor kappa B kinase subunit alpha; NFKBIA, NFKB inhibitor alpha;

Values for RDA1 and RDA2 represent the directional cosines of the environmental vectors;

R², the goodness-of-fit statistic determined by the envfit function based on 999 permutations;

P < 0.05 indicates a statistically significant effect of the environmental vector on the microbial community variation.

Discussion

With increasing global attention to reducing antibiotic use and promoting sustainable livestock and poultry production, the development of safe and effective natural plant-derived feed ingredients has become an important research focus (Zheng et al., 2025). The present results showed that dietary supplementation with ZBS had no significant effects on average daily feed intake, average daily gain, or feed conversion ratio in broilers. However, body weight at 28 and 63 d and average daily gain during d 1−28 and d 1−63 exhibited significant quadratic responses, indicating dose-dependent differences in growth performance among broilers fed different levels of ZBS. Similar quadratic dose-response effects have been reported in studies evaluating plant-derived by-product feed ingredients (Ma et al., 2023; Zhang et al., 2023). Nurhayati et al. (2025) reported dose-dependent effects of Phyllanthus amarus leaf extract on growth performance in broilers, further supporting the biphasic dose-response pattern of phytogenic compounds. ZBS are rich in polysaccharides, unsaturated fatty acids, and various bioactive compounds (Qi et al., 2024). In the present study, the ZBS used were determined to contain 7.02 mg/g total flavonoids, 3.12 mg/g flavonoids, 19.25 μmol/g polyphenols, and 11.48 mg/g polysaccharides. The quadratic responses indicated that moderate inclusion levels corresponded to the highest body weight and average daily gain, whereas higher inclusion levels did not further improve these parameters. This pattern suggests a biphasic dose-dependent effect, in which moderate inclusion levels may provide sufficient bioactive compounds to support growth, while higher levels may exceed the optimal range without conferring additional benefits. Although the bioactive compounds in ZBS were quantified, their actual concentrations in the experimental diets were not directly measured, and their specific contributions to the observed effects were not determined, and the underlying mechanisms require further investigation. Notably, the overall production performance indicators (ADFI and F/G) were not significantly altered in this study, therefore, the biological significance of the quadratic responses should be interpreted with caution, and the growth-promoting effects should not be overstated. The underlying reasons remain unclear and require further investigation.

Immune organ indices are commonly used to evaluate the development of immune organs in poultry, and changes in the relative weights of immune organs such as the bursa of Fabricius and spleen are generally associated with immune status (Wang et al., 2025a). In the present study, the bursa of Fabricius index increased linearly with increasing dietary ZBS levels, indicating that ZBS promoted the development of the bursa of Fabricius. In contrast, the spleen index was not significantly affected. Although it showed an increasing trend followed by a decline, the differences among treatments were not significant, suggesting that the effects of ZBS differed among immune organs.

The serum and splenic immune-related parameters further indicated that ZBS affected the immune status of broilers. As the dietary level of ZBS increased, serum IgA and IgG concentrations increased significantly, and the splenic contents of IgA, IgG, C3, and C4 also increased significantly. Previous studies have shown that IgA and IgG play important roles in mucosal immunity and humoral immunity, respectively (Zugmaier et al., 2014), whereas complement components C3 and C4 are central effectors of the complement system, mediating opsonization, inflammatory cell recruitment, and immune complex clearance, and their levels are commonly used to evaluate humoral immune function and complement activation status (Dunkelberger and Song, 2010). Therefore, these results indicate that ZBS improved several immune parameters associated with humoral immunity. Aljohani (2024) reviewed that phenolic compounds possess immunomodulatory and anti-inflammatory properties. Meanwhile, serum concentrations of IL-2, IL-6, and TNF-α also increased significantly with increasing dietary ZBS levels. Previous studies have demonstrated that IL-2 is mainly involved in lymphocyte proliferation and immune activation (Spolski et al., 2018), whereas IL-6 and TNF-α function in both immune regulation and inflammatory mediation (Schumertl et al., 2025; Lalaoui and Masters, 2026). Therefore, changes in these cytokines reflect alterations in immune response status, suggesting that feeding diets containing ZBS may activate immune responses and enhance immune potential in broilers. Unlike the changes observed in serum cytokines, splenic IL-1β (a classic pro-inflammatory cytokine that regulates acute inflammation) content showed a decreasing trend, indicating that the responses of serum and spleen to ZBS were not entirely consistent. Previous studies have shown that polyphenols and flavonoids in plant-derived materials can participate in immune regulation and inflammatory responses (Yahfoufi et al., 2018; Sharma et al., 2025). Considering that ZBS are rich in polysaccharides, flavonoids, and other bioactive compounds (Qi et al., 2024), the immune-related changes observed in the present study may be associated with these components.

Changes in gut microbiota composition are closely associated with host immune status, and microbial dysbiosis is often accompanied by inflammatory responses and alterations in host physiological functions (Wang et al., 2023; Zhang et al., 2025). Dysbiosis, characterized by an expansion of opportunistic pathogens and a reduction of beneficial commensals, disrupts intestinal homeostasis and promotes local inflammation. Wang et al. (2023) demonstrated that specific cecal bacterial genera, such as Synergistes, negatively affect growth performance by inducing the expression of the inflammatory cytokine IL-6. Zhang et al. (2025) further reported that Salmonella infection in chickens activated the TLR4/MyD88/NF-κB pathway, leading to increased pro-inflammatory cytokine expression and altered microbial balance. In the present study, feeding diets containing ZBS did not significantly affect the α-diversity of the cecal microbiota in broilers, whereas β-diversity analysis revealed clear separation of microbial communities between the ZBS1 group and the CON and ZBS2 groups, indicating that the effects of ZBS were mainly reflected in microbial composition rather than overall diversity. At the phylum level, the relative abundances of Firmicutes and Bacteroidota showed different trends with increasing dietary ZBS levels. At the genus level, the relative abundance of Faecalibacterium in the ZBS3 group was significantly higher than that in the CON group, representing an increase of 160.80%. LEfSe and differential abundance analyses further showed that Faecalibacterium, Megasphaera, Bifidobacterium, and Actinobacteriota were significantly enriched in the ZBS3 group. Among these taxa, Faecalibacterium and Bifidobacterium are regarded as important beneficial microorganisms in the animal gut and are capable of metabolizing complex dietary carbohydrates to produce a variety of biologically active metabolites (Biagi et al., 2020; Wang et al., 2025b). Megasphaera also possesses strong organic acid metabolic capacity and plays an important role in maintaining intestinal microbial homeostasis (Cabral and Weimer, 2024). Liu et al. (2024c) reported that hesperidin, thymol, and rosmarinic acid increased the relative abundance of beneficial bacteria; (El-Hamid et al., 2024) found that essential oil blends significantly increased the count of beneficial Lactobacillus while reducing pathogenic bacterial loads, which is consistent with the microbial changes observed in the present study.

Consistent with the changes in microbial composition, PICRUSt2 functional prediction showed that the predicted abundances of pyruvate-formate lyase (PFL), starch and sucrose metabolism, carbohydrate metabolism, and the phosphotransferase system (PTS) were markedly increased in the cecal microbiota of the ZBS3 group compared with the control group. PFL is an important enzyme involved in the conversion of pyruvate to acetyl-CoA during bacterial anaerobic fermentation, and enhanced PFL activity is generally associated with metabolic processes related to short-chain fatty acid production (Anand et al., 2016). The PTS is an important pathway for bacterial carbohydrate uptake and utilization, and increased PTS activity is consistent with an enhanced capacity of the microbiota to utilize dietary carbohydrates (Deutscher et al., 2006). It should be noted that PICRUSt2 predictions are computational inferences based on 16S rRNA data rather than direct measurements of microbial metabolic activity (Douglas et al., 2020), and should therefore be interpreted as suggestive evidence warranting further validation through targeted metabolomics or metagenomic approaches. Therefore, considering that ZBS are rich in polysaccharides and dietary fiber, these results indicate that ZBS not only altered microbial composition but also affected carbohydrate utilization and potential fermentation-related metabolic functions of the cecal microbiota.

Moreover, The changes in microbial structure were accompanied by increases in serum IgA, IgG, IL-2, IL-6, and TNF-α levels, as well as increased splenic contents of IgA, IgG, C3, and C4, together with a reduction in splenic IL-1β content. Redundancy analysis showed that dietary ZBS level and immune-related parameters, including serum IL-2, IL-6, and TNF-α and splenic IgA, IgG, C3, and C4, were significantly associated with microbial structural variation. Previous studies have shown that the gut microbiota not only participates in nutrient metabolism but is also closely involved in the maintenance of host immune function (Boza et al., 2016). Obianwuna et al. (2024)reviewed that phytogenic substances can improve poultry health by modulating gut microbiota and immune function. Based on the observed changes in microbial structure, predicted microbial functions, and immune-related parameters, it can be considered that ZBS-induced alterations in the cecal microbiota were closely associated with changes in host immune status. However, given the observational nature of this study, the concurrent changes in microbiota and immune parameters should be interpreted as associations rather than causal relationships. Establishing causality would require targeted intervention approaches such as fecal microbiota transplantation or gnotobiotic models in future studies.

To investigate the effects of ZBS on inflammation-related signaling pathways in the spleen, the expression levels of key genes in the TLR4/MyD88/NF-κB pathway were examined. The results showed that the relative expression levels of TRAF2, TLR4, MyD88, NFKB2, IKKA, and NFKBIA decreased linearly with increasing dietary ZBS levels. Previous studies have demonstrated that the TLR4/MyD88/NF-κB signaling pathway is a key regulator of innate immunity and inflammatory responses, and its activation is generally accompanied by increased production of pro-inflammatory cytokines and enhanced inflammatory responses (Block et al., 2018; Gitlin et al., 2024). Liu et al. (2024c) demonstrated that hesperidin, thymol, and rosmarinic acid down-regulated the mRNA expression of TLR4/MyD88/NF-κB in broilers, which is consistent with our observation of decreased expression of these pathway-related genes in the spleen. In the present study, the downregulation of these pathway-related genes occurred concurrently with the reduction in splenic IL-1β content, indicating that ZBS suppressed inflammation-related signal transduction in the spleen. Although changes in microbial composition and predicted functions, immune-related parameters, and inflammation-related gene expression were observed simultaneously, RDA did not identify significant associations between microbial structure and key inflammatory pathway genes such as TLR4, MyD88, and NFKB2. Therefore, the present findings support an association between ZBS-induced microbial structural changes and alterations in humoral immunity-related parameters, but are insufficient to demonstrate that microbial changes directly mediated regulation of the TLR4/MyD88/NF-κB signaling pathway.

Conclusion

ZBS affected the composition and predicted functions of the cecal microbiota, enhanced humoral immunity-related parameters, and downregulated the expression of genes associated with the splenic TLR4/MyD88/NF-κB signaling pathway in broilers. Based on quadratic regression of splenic C4, IL-1β, TRAF2, TLR4, and NFKB2, the optimal dietary ZBS inclusion level was estimated at 1.00−2.55%, suggesting its potential as an unconventional plant-derived feed ingredient for broilers.

Funding

This study was supported by Fundamental Research Funds for National Key R&D Program of China (2022YFD1300501), the National Natural Science Foundation of China (32372894), Key Project of Science and Technology Research Program of Chongqing Municipal Education Commission (KJZD-K202300209) and Technical research and development projects (M2022069, F2025650).

Informed consent statement

Not applicable.

Author contributions

All authors listed below meet all four ethical authorship criteria: substantial contribution, drafting or critical review, final approval, and accountability. Zhijin Hu: collected and managed the data; analyzed the data; contributed to the investigation and methodology; drafted the original manuscript; and participated in writing, review, and editing; Huatao Liu: analyzed the data; contributed to the investigation; and contributed to the methodology; Yuxi Shen and Fu Feng: contributed to the experimental work and study methods; Zhihong Sun: contributed to the experimental work, provided resources, and critically reviewed the manuscript for important intellectual content; Weizhong Sun, Yetong Xu and Jiaman Pang: contributed to the design of the study and the development of the methods used; Zhiru Tang: obtained funding; contributed to the investigation and methodology; administered and supervised the project; and participated in writing, review, and editing.

Disclosures

The authors declare no conflicts of interest.

Data availability

• The raw sequencing data for 16S rDNA sequencing generated in this study have been deposited in the NCBI BioProject database and are publicly accessible under the accession number PRJNA1355354 (). Additional raw data supporting the findings of this study are available from the first author and the corresponding author upon request.http://www.ncbi.nlm.nih.gov/bioproject/1355354