Effect of glycosaminoglycans on the structure and composition of articular cartilage and bone of broilers

Abstract

This study aimed to assess the influence of glycosaminoglycan (chondroitin and glucosamine sulfates) supplementation in the diet of broilers on the expression of matrix metallopeptidase 9 (MMP-9) and metallopeptidase inhibitor 2 (TIMP-2) genes, the synthesis of proteoglycans, collagen type II and chondrocytes, bone and cartilage macroscopy, bone mineral densitometry, bone breaking strength and mineral profile. A completely randomized design was carried out in a 3 × 3 factorial scheme (3 levels of chondroitin sulfate: 0.00, 0.05, and 0.10%; and 3 levels of glucosamine sulfate: 0.00, 0.15, and 0.30%), totaling 9 treatments. At 21 and 42 d of age, broilers were slaughtered, and tibias and femurs were collected for evaluation. There was an interaction (P < 0.05) of sulfates for the expression of MMP-9 and its inhibitor TIMP-2 in femur articular cartilage, as well as for the number of chondrocytes, collagen type II and proteoglycans in tibia articular cartilage, bone and cartilage macroscopy and mineral profile (P < 0.05), with better results obtained with the inclusion of chondroitin and/or glucosamine sulfates in the feed. In conclusion, chondroitin and glucosamine sulfates can be used in broiler diets in order to favor the development of the structure of the locomotor system (bones and joints), thus preventing locomotion problems.

INTRODUCTION

The incidence of bone fractures and mobility disorders, characterized by alterations in bone and cartilage development, damages the production of broiler chickens. The factors responsible for the occurrence of these problems are related to the high grower rate and high body weight of birds while the bone structure is still forming (Olkowski et al., 2011).

Polysulfated glycosaminoglycans (GAGs) have prevented and/or reduced the progression of pathological changes in bone and joint structures (Kamarul et al., 2011). Among the GAGs, chondroitin and glucosamine sulfates stand out. Chondroitin and glucosamine sulfates can stimulate cartilage anabolic processes, such as proteoglycan and collagen synthesis (Kamarul et al., 2011), chondrocyte proliferation, and bone matrix biosynthesis, which promote longitudinal bone grower (Wolff, 2014). Cartilage degeneration can also be prevented through anti-inflammatory mechanisms (Calamia et al., 2010, 2014; Kantor et al., 2014) and through epigenetic mechanisms, with increased expression of tissue inhibitory genes (TIMPs) and consequent reduction in the expression of matrix metallopeptidase enzymes (MMPs) (Chan et al., 2006), which results in less bone resorption (Maxwell et al., 2016). However, gene expression studies focusing on broiler cartilaginous tissue responses to glycosaminoglycan supplementation have not been addressed to date; these may provide plausible mechanisms for the supposed anti-inflammatory and chondroprotective properties of GAGs in reducing locomotor diseases.

In addition to glucosamine sulfate having an anabolic effect by stimulating chondrocytes to synthesize proteoglycans and collagen, it has been reported that this is a substrate for the new synthesis of chondroitin sulfate (Gouze et al., 2001; Chan et al., 2006). Along with its indirect effect on the metabolism of cartilage, being a precursor of glycosaminoglycans, it is also possible that supplementation with glucosamine can help promote the synthesis of glycosaminoglycans or can reduce their degradation. Another study reported the effectiveness of glucosamine in delaying the degradation of cartilage, and the progression and severity of osteoarthritis (Wen et al., 2010).

Polysulfated glycosaminoglycans, chondroitin and glucosamine, have been studied as nutraceuticals for the prevention and/or treatment of pathologies of bone and joint structures in broilers (Sgavioli et al., 2017; Santos et al., 2018, 2019; Martins et al., 2020a), and consequently better performance (Sgavioli et al., 2017; Martins et al., 2020a,b). However, there is no research in broilers that has analyzed the inclusion of GAGs in the diet on the expression of genes related to the degradation of proteins in cartilage and the synthesis of proteoglycans and collagen type II. Therefore, this research aimed to study the effect of supplemental GAGs in broiler diets the expression of MMP-9 and TIMP-2 genes, the synthesis of proteoglycans, collagen type II and chondrocytes, as well as to bone and cartilage macroscopy, bone mineral densitometry, bone breaking strength and mineral profile.

MATERIALS AND METHODS

All procedures in this study were approved by the Animal Use Ethics Committee, according to protocol 051/16.

Animals, Facilities, and Experimental Design

A total of 1,620 one-day old, male, commercial Cobb 500 broiler chicks with a mean initial weight of 43 ± 0.2 g, were reared until 42 d of age. They were housed in 54 pens, set up in the central area of a 1,500 m² commercial facility, acclimatized by negative pressure. Each pen consisted of 2.88 m², 10 nipple drinkers, and 1 tube feeder for chicks up to the seventh day and 1 adult from 8 to 42 d of age.

Chicks were vaccinated in the hatchery against Marek's disease, infectious bursal disease (IBD), and avian pox. The following vaccination program was completed during the experimental period: Marek and avian pox in the hatchery and IBD (mild strain) on the 14th day in the drinking water, using powdered milk as a carrier (2 g/L). The adopted light regimen was 24:0 h (light:dark) in the first week and 20:4 h (light:dark) from 7- to 42-days old.

The experiment was conducted in a completely randomized design in a 3 × 3 factorial arrangement (3 levels of chondroitin sulfate (0, 0.05, and 0.10%), and 3 levels of glucosamine sulfate (0, 0.15, and 0.30%), both in the broiler chicken diet). The levels of the chondroitin and glucosamine sulfates were obtained from the studies by Martins et al. (2020b) and Sgavioli et al. (2017).

Each treatment was composed of 6 replications of 30 birds, totaling 54 experimental pens. Chondroitin sulfate [(C₁₄H₂₁NO₁₄S)n, Biofac A/S, Englandsvej, Kastrup, Denmark] was 91.27% pure and potassium glucosamine sulfate [(C₆H₁₄NO₅) 2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd., Yuhuan, Zhejiang, China] had 16% sulfate.

The broilers received feed and water ad libitum throughout the experimental period and were reared following the management recommendations of Cobb-Vantress Management Guide (2008). Diets were formulated based on corn and soybean meal and following the recommendations proposed by Rostagno et al. (2011) for the phases preinitial (1–7 d of age), initial (8–21 d of age), grower (22–35 d of age), and finisher (36–42 d of age), all of them with a variable portion of 0.4% of chondroitin sulfate and/or glucosamine sulfate and/or inert, according to treatments (Table 1).

Table 1.

Ingredients and calculated nutritional composition of diets from preinitial (1–7-days old), initial (8–21-days old), growth (22–35-days old), and final (36–42-days old) phases.

Ingredients (%)	Preinitial	Initial	Growth	Final
Corn	54.46	59.80	63.77	69.94
Soybean meal (45.5%)	35.16	30.78	23.82	15.18
Poultry fat	1.13	1.40	1.87	2.07
Meat and bone meal (47%)	3.67	4.40	3.13	6.87
Offal meal (62,5%)	3.00	1.13	3.53	1.80
Feather meal (84,81%)	-	-	1.53	2.00
Limestone	0.53	0.53	0.63	0.23
Salt	0.39	0.35	0.27	0.21
Sodium bicarbonate	0.08	0.05	0.10	0.15
Choline chloride (75%)	0.05	0.08	0.05	0.06
DL-methionine (99%)	0.41	0.35	0.27	0.25
L-lysine HCl (64%)	0.33	0.34	0.28	0.45
L-threonine (98%)	0.07	0.08	0.05	0.09
L-valine (96,5%)	0.02	0.01	-	-
Vitamin supplement1	0.05	0.05	0.05	0.05
Mineral supplement2	0.05	0.05	0.05	0.05
Additives3	0.20	0.20	0.20	0.20
Variable portion4	0.40	0.40	0.40	0.40
Total	100.00	100.00	100.00	100.00
Calculated nutritional composition
Metabolizable energy (kcal/kg)	3000	3050	3150	3200
Crude protein (%)	25.00	22.50	21.60	19.53
Calcium (%)	0.98	0.98	0.95	0.86
Available phosphorus (%)	0.49	0.48	0.46	0.44
Sodium (%)	0.22	0.21	0.20	0.19
Chlorine (%)	0.30	0.27	0.23	0.20
Potassium (%)	0.90	0.82	0.70	0.57
Digestible methionine + cystine (%)	1.03	0.92	0.85	0.75
Digestible methionine (%)	0.73	0.64	0.56	0.49
Digestible lysine (%)	1.36	1.21	1.10	1.00
Digestible threonine (%)	0.87	0.79	0.72	0.68

¹Vitamin supplement (composition per kg product): preinitial and initial (vitamin A 20,000,000 IU; vitamin D3 5,000,000 IU; vitamin E 50,000 IU; vitamin K3 4,000 mg; vitamin B1 5,000 mg; vitamin B2 13,000 mg; vitamin B6 7,000 mg; vitamin B12 36 mg; niacin 84,000 mg; pantothenate 30,000 mg; folic acid 2,400 mg; biotin 160 mg; selenium 600 mg); growth and final (vitamin A 16,000,000 IU; vitamin D3 3,800,000 IU; vitamin E 40,000 IU; vitamin K3 3,600 mg; vitamin B1 3,600 mg; vitamin B2 11,000 mg; vitamin B6 5,200 mg; vitamin B12 30 mg; niacin 70,000 mg; pantothenate 26,000 mg; folic acid 1,800 mg; biotin 100 mg; selenium 600 mg).

²Mineral supplement (composition per kg product): copper 16.25 g; iron 100 g; iodine 2,000 g; manganese 150 g; zinc 125 g.

³Additives: preinitial, initial, and growth [maxiban (narasine + nicarbazine) 0.05 g; enradin (enramycin) 0.01 g; microtech (phytase) 0.01 g; salmex (formaldehyde, propionic acid, terpenes, and surfactants) 0.10 g; endox (ethoxyquin and butylated hidroxyanisole) 0.004 g; copper sulfate 0.03 g];final [maxiban (narasine + nicarbazine) 0.05 g; enradin (enramycin) 0.006 g; microtech (phytase) 0.01 g; salmex (formaldehyde, propionic acid, terpenes, and surfactants) 0.10 g; endox (ethoxyquin and butylated hidroxyanisole) 0.004 g; copper sulfate 0.03 g].

⁴Variable portion: chondroitin sulfate [(C₁₄H₂₁NO₁₄S)n, Biofac A/S, Englandsvej, Kastrup, Denmark] and/or potassium glucosamine sulfate [(C₆H₁₄NO₅) 2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd., Yuhuan, Zhejiang, China] and/or inert (kaulin) according to treatments.

The nutritional composition of the experimental diets was analyzed for dry matter, crude energy, crude protein, calcium, and phosphorus. Dry matter was determined by the gravimetric method using heat and was based on the weight loss of the material subjected to heating at 105°C in a rectilinear oven (model 315/3, Fanen, Guarulhos, São Paulo, Brazil). Crude energy was determined using the calorimetric pump (model 6100, Parr Instrument Company, Moline, IL). The total nitrogen content was analyzed in a nitrogen distiller (model TE-036/1, Tecnal, Piracicaba, São Paulo, Brazil), using the Kjeldahl method (INCT-CA N-001/1), according to Detmann et al. (2012), and the factor of 6.25 was used in the conversion of nitrogen value into crude protein. Calcium and phosphorus were analyzed by an atomic absorption spectrophotometer (model AA-7000, Shimadzu, Barueri, São Paulo, Brazil) and UV/VIS spectrophotometer (model UV-5100, Tecnal, Piracicaba, São Paulo, Brazil), respectively, as proposed by Silva and Queiroz (2002).

The analyzed values for preinitial, initial, grower, and finisher phases were, respectively, 3,755, 3,880, 4,010, and 4,100 kcal/kg for crude energy, 23.85, 22.60, 21.25, and 19.00% for crude protein, 0.99, 0.98, 0.90, and 0.83% for calcium, and 0.75, 0.73, 0.65, and 0.65% for total phosphorus.

Gene Expression of MMP-9 and TIMP-2 in Cartilage

Six broilers per treatment, with similar mean body weights, were selected on d 42 of grower and identified with numbered leg bands. The selected broilers were fasted for 8 h, slaughtered by cervical dislocation, and 2 after that were used to collected 1-cm samples from the proximal epiphysis cartilage (intercondylar eminence), of the right femur. A longitudinal cut of the articular cartilage surrounding the condylar area (contact surface with the femoral condyle) was made. The samples were stored in liquid nitrogen to determine the mRNA gene expression of MMP-9 and TIMP-2, according to the methodology of Bustin et al. (2009).

After obtaining total RNA and its integrity and quality were validated, 1 μg of RNA was treated with RNase-free DNAse-I (Invitrogen, Carlsbad, CA) to remove all contamination with genomic DNA, and DNAse-I was subsequently inactivated by the addition of EDTA and heat (65°C) for 5 min. The first cDNA strand was amplified using the reverse transcriptase enzyme (High Capacity, Invitrogen, Carlsbad, CA) from the RNA treated with DNase-I (Synapse Biotechnology, Tatuapé, Brazil), following the manufacturer's standards.

The primer oligonucleotides used to enlarge the target genes (MMP-9 and TIMP-2) were designed from the sequences obtained from the Gallus gallus genome (assembly bGalGal1.mat.broiler.GRCg7b, in https://www.ncbi.nlm.nih.gov/genome/111?genome_assembly_id=1543395) using the program Gene Runner, version 3.01 (developed by Frank Buquicchio and Michael Spruyt, 1992–2018; freeware, http://www.generunner.net/). The glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) gene was used as endogenous control (Table 2). These oligonucleotides were used to measure gene expression via real-time quantitative PCR (qPCR). qPCR analysis was performed using the fluorescent dye SYBR Green, according to the manufacturer's instructions (SYBR Green PCR Master Mix, Applied Biosystems, Foster City, CA), in the StepOnePlus Real-Time PCR (Applied Biosystems, Foster City, CA). Each reaction, carried out in a final volume of 25 µL, contained 2.5 µM of each oligonucleotide, 12.5 µL of SYBR Green (PCR Master Mix) and 1 µL of cDNA (50 ng). Cycling conditions were an initial step of 2 min at 50°C and 10 min at 95°C, followed by 40 cycles at 95°C for 15 s and 60°C for 1 min. Data normalization was performed using the GenEx qPCR Data Analysis program (MultiD Analyzes AB, Göteborg, Kattegat, Sweden; freeware, http://genex.gene-quantification.info/), and the relative quantification of the genes of interest were calculated by the 2-ΔΔCt method, normalizing the data with an endogenous control.

Table 2.

Reference gene in chickens (F: forward; R: reverse), GenBank ID, and sequence of primers for glycosaminoglycan candidate genes.

Gene	Amplicon (pb)	GenBank ID	Primers sequence (5′–3′)
MMP-9	229	395387	F: GCCATCACTGAGAT CAATGGAG
			R: GATAGAGAAGGC GCCCTGAGT
TIMP-2	169	374178	F: GATGGAGAAGATC GTGGGCGG
			R: TGGGCTTTCCTAC TGGCTACTG
GAPDH	73	374193	F: TCAGCAATGCATCGTGCAC
			R: GCATGGACAGTGGTCATA AGAC

The relative quantification of gene expression is used to describe variation in the expression of the gene of interest in a group of samples in relation to a reference sample, called a calibrator, which in this case was the sample with the lowest ΔCt (cycle threshold). The calculation of this relative expression uses the Ct value of the endogenous control, in this case GAPDH, and then the value obtained is subtracted by the Ct values of the target genes, thus obtaining the ΔCt of each sample. The ΔCt of the calibrator used was the one with the lowest expression value, which is equivalent to the highest value of ΔCt. The ΔCt value of the calibrator was then subtracted from the ΔCt value of the other samples to obtain the ΔΔCt. Finally, relative gene quantifications were expressed as 2-ΔΔCt and converted to log2. The final graphs were obtained from the values of log2.

Histopathological Evaluation of the Cartilage—Chondrocytes, Type II Collagen, and Proteoglycans

At 21 and 42 d of grower, the proximal epiphysis cartilage (intercondylar eminence) was sampled from 12 broilers per treatment. The selected broilers were fasted for 8 h, slaughtered by cervical dislocation, and the right tibia of each broiler were removed and marked for analyses. For histological evaluation of the cartilage, a 1-cm cut of the articular cartilage surrounding the condylar area (contact surface with the femoral condyle) was made.

The cartilage samples were fixed in a formaldehyde solution (10%) for 24 h and decalcified in a formic acid solution (5%) for 14 d at room temperature. Following fixation, the samples were washed in distilled water and processed according to standard methods for light microscopy (Luna, 1968). The number of cartilage chondrocytes, and the concentration of proteoglycans and type II collagen were measured.

For determination of the number of chondrocytes in cartilage, 5-µm-thick semiserial histologic sections were cut with an electronic Rotary microtone (RM2255, Leica Biosystems Inc., Buffalo Grove, IL) and stained with hematoxylin-eosin. In total, 108 slides were made per age group, with 12 replicas per treatment. Five semiserial sections were placed onto each slide. Two images were obtained per section, totaling 120 images per treatment. The pictures were obtained through a system of registration and analysis of images (DM4000B, Leica Microsystems, Wetzlar, Germany) using a 10× objective lens. For cell counts in the images, an area of 86.8 µm² was used (Sgavioli et al., 2017). For the statistical analysis, the mean of the measurements obtained per slide was used for each repetition.

The concentration of type II collagen was quantified by immunohistochemistry. For this, 5-µm-thick semiserial histologic sections were cut with an electronic rotating microtome (RM2255, Leica Biosystems Inc., Buffalo Grove, IL). The sections were placed on polarized slides, deparaffinized, and hydrated in a series of decreasing concentrations of ethanol in distilled water. Subsequently, the slides were transferred to cuvettes containing an SDS solution (1 g of sodium dodecyl sulfate diluted in 100 mL of phosphate-buffered saline (PBS)), and heated in an oven at 37°C for 5 min.

The endogenous peroxidase activity was then blocked for 15 min by incubating the slides in hydrogen peroxide (30 mL), methyl alcohol (70 mL), and Triton X-100 (0.1 mL). Antigen retrieval took place in a citrate solution, pH 6.0, in a water bath (solution preheated to 96°C) for 20 min. Blocking was performed with 3% bovine serum albumin (BSA) for 1 h in a humidified chamber at room temperature to prevent nonspecific binding.

The primary antibody specific for collagen type II (Anti-Collagen II Antibody, code 34712, Abcam, Cambridge, England), diluted in 1.5% BSA (1:500), was instilled on the fragment, and the material was incubated overnight in a humidified chamber at 4°C. Some slides were used as controls, being incubated with only 1.5% BSA. After washing in PBS, the slides were incubated with the secondary antibody (Goat anti-rabbit, code 656140, Life Technologies, Carlsbad, CA), diluted 1:1,000 in 1.5% BSA, for 30 min. Subsequently, the slides were washed with PBS, and the Enzyme Immunoassay Kit (Vector Laboratories, Burlingame, CA) was used for 30 min in a humidified chamber at room temperature. The slides were then washed in PBS. The reaction was revealed with the addition of the chromogen diaminobenzidine-peroxidase (DAB, Abcam, ab64238, GR313581-2, Cambridge, England) for 4 min. After that, the slides were washed with PBS for 5 min, dehydrated in solutions of decreasing alcohol concentration, and diaphonized.

Histochemistry was used to evaluate the concentration of proteoglycans in cartilage. For this, 5-µm-thick semiserial histologic sections were cut with an electronic rotating microtome (RM2255, Leica Biosystems Inc., Buffalo Grove, IL), stained with 1% Alcian Blue in 3% acetic acid, pH 2.5, for 15 min, and counterstained with Aqueous Neutral Red 0.5% for 10 min.

We prepared 108 slides for histochemical analysis and 108 for immunohistochemistry for each age group evaluated, totaling 12 slides per treatment per age group for both. Five semiserial sections were placed onto each slide. Two images were obtained per section, totaling 120 images per treatment. The pictures were obtained through a system of registration and analysis of images (DM4000B, Leica Microsystems, Wetzlar, Germany) using a 40× objective lens. Images were analyzed using ImageJ software (National Institutes of Health, Bethesda, MD; freeware, https://imagej.net/Welcome), with 10 areas selected per image. The program provides the optical density value in the user-selected area in pixels/inch, using a grayscale method. For the statistical analysis, the mean of the measurements obtained per slide was used for each repetition.

Bone and Joint Cartilage Macroscopy

Absolute and relative weight, length, Seedor index, and width of the proximal epiphysis (articulates with the femur), diaphysis and distal epiphysis (articulates with the tarsometatarsus), all from the left tibia, were measured. The length and width of the proximal epiphysis, diaphysis and distal epiphysis were determined using a digital caliper (model 316119, MTX, Vigário Geral, Rio de Janeiro, Brazil) with an accuracy of 0.01 mm. Seedor's index was calculated by dividing the bone weight (mg) by its length (mm), as proposed by Seedor et al. (1991).

The cartilages of the proximal and distal epiphysis were manually removed with the aid of a scalpel and weighed separately on an analytical scale (model AL500C, Marte, Santa Rita do Sapucaí, Minas Gerais, Brazil), with a precision of 0.01 g, to obtain the absolute and relative weight in relation to the broiler weight and the thickness measured with a digital caliper (model 316119, MTX, Vigário Geral, Rio de Janeiro, Brazil) in the transverse and medial portions.

Mineral Densitometry

The bone mineral density (BMD) (g/cm²) and surface area (cm²) were determined for the left tibia, using the dual-energy X-ray absorptiometry (DXA) (Horizon Discovery Dxa Hologic, Marlborough, MA), in the following areas: whole bone, diaphysis, proximal epiphysis, and distal epiphysis. Prior to sample scanning, the appliance was calibrated with bone and tissue equivalents, supplied by the manufacturer (Hologic, MA). For all scans, the small animal software (Hologic, MA) was used to select the region for subsequent densitometric analysis. The clean bones were placed with the cranial face in contact with the scanner table.

Bone Breaking Strength

The left tibia were used for the mechanical bone strength tests (3-point bending). The tests were conducted using an EMIC (DL 10.000, cell Trd 21, software, TescTM 3.13, São José dos Pinhais, PR, Brazil) universal testing machine (UTM). The preload force applied was 10 N, with an adaptation time of 5 s. The load rate was 5 mm/min, using the load cell of 500 N to determine the maximum force (N), the maximum deflection at maximum force (mm), and the rigidity (N/m). The equipment was calibrated to allow the diaphysis length of 6 cm for all bone samples, as this distance was sufficient to ensure that all bones were placed on their diaphysis. The load was then applied at the bone's geometric mean point between the 2 supports (the middle third of the bone) and the equipment recorded the results. This variable determines the maximum force and the maximum deflection at maximum force at the middle third of the bone. The rigidity was calculated automatically by software (TescTM 3.13). All the measurements were made by the same operator (Sgavioli et al., 2017).

Mineral Profile

The left tibia was used to determine bone calcium, phosphorus, and ash content. The soft tissue was removed and the bones boiled in deionized water for 5 min. After drying at room temperature, the samples were immersed in petroleum ether for 48 h, dried in a forced-ventilation oven at 60°C (MA 035/5, Marconi, Piracicaba, Brazil) for 48 h, and then ground in a ball mill. Ash content was determined by burning the samples at 600°C (F-3, Fornitec, São Paulo, Brazil). The calcium and phosphorus were analyzed by atomic absorption spectrophotometer (AA-7000, Shimadzu, Barueri, Brazil) and spectrophotometer UV/VIS (UV-5100, Tecnal, Piracicaba, Brazil). The methods were applied according to Silva and Queiroz (2002), and expressed as a percentage of defatted dry matter.

Statistical Analysis

Supplementation effects with chondroitin (CO) (0, 0.05, and 0.10%) and glucosamine sulfates (GLU) (0, 0.15, and 0.30%) and their interaction (CO × GLU) were analyzed according to the experimental model: Yijk = μ + (CO) i + (GLU) j + (CO × GLU) ij + eijk, where Y is the response variable, μ is the mean of the variable, CO is the chondroitin sulfate, GLU is the glucosamine sulfate, CO × GLU is the interaction of chondroitin and glucosamine sulfates, and eijk is the residual error.

The data of performance and incidence of tibial dyschondroplasia were verified for the presence of outliers (box-and-whisker plot), the normality assumptions of error were studentized (Cramér-von Mises test), and homogeneity of variances (Bartlett's test). After corrections, the data were subjected to analysis of variance using the general linear model (GLM) procedure of SAS (SAS Institute Inc., Cary, NC). An orthogonal analysis was performed when the means differed significantly by the F test at 5% probability to test linear and quadratic effects of the levels of chondroitin and glucosamine sulfates. Statistical analysis of considered individual broilers as experimental units, for histology, bone and cartilage macroscopy, bone densitometry, bone resistance and mineral profile, were used 12 broilers per treatment and 2 broilers per pen and for the gene expression were considered 6 broilers per treatment and 1 broiler per pen.

RESULTS

Gene Expression of MMP-9 and TIMP-2 in Cartilage

Gene expression of MMP-9 and TIMP-2 showed an interaction with the addition of chondroitin sulfate and glucosamine sulfate (P < 0.0001). A linear decrease in MMP-9 expression was observed with the addition for all levels for chondroitin and glucosamine sulfates. In contrast, the inclusion for all levels for chondroitin and glucosamine sulfates in the diet caused a linear increase in the expression of TIMP-2 (Table 3).

Table 3.

Gene expression of the metalloproteinase MMP-9 and its TIMP-2 inhibitor in the articular cartilage of the proximal epiphysis of the broiler femur at 42 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
MMP-9	0	10.46	7.37	4.67	L6 (0.0013)	7.50	0.45	ns	ns	< 0.0001
	0.05	9.36	4.57	3.41	L7 (<0.0001)	5.78
	0.10	8.16	1.51	0.67	L8 (<0.0001)	3.45
	Regression	L3 (0.0002)	L4 (<0.0001)	L5 (<0.0001)
	Mean	9.32	4.48	2.91
TIMP-2	0	0.78	4.52	5.23	ns	3.51	0.48	ns	ns	< 0.0001
	0.05	2.54	6.58	7.71	L12 (0.0003)	5.61
	0.10	3.56	10.43	12.11	L13 (<0.0001)	8.70
	Regression	L9 (0.0016)	L10 (<0.0001)	L11 (<0.0001)
	Mean	2.29	7.17	8.35

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression.

³MMP-9 42 dias _{0% GLU} = −23 CO + 10.477; R² = 0.99.

⁴MMP-9 42 dias _{0.15% GLU} = −58.6 CO + 7.4133; R² = 0.99.

⁵MMP-9 42 dias _{0.30% GLU} = −40 CO + 4.9167; R² = 0.96.

⁶MMP-9 42 dias _{0% CO} = −19.3 GLU + 10.395; R² = 0.99.

⁷MMP-9 42 dias _{0.05% CO} = −19.833 GLU + 8.755; R² = 0.89.

⁸MMP-9 42 dias _{0.10% CO} = −24.967 GLU + 7.1917; R² = 0.83.

⁹TIMP-1 42 dias _{0% GLU} = 27.8 CO + 0.9033; R² = 0.98.

¹⁰TIMP-1 42 dias _{0.15% GLU} = 59.1 CO + 4.2217; R² = 0.97.

¹¹TIMP-1 42 dias _{0.30% GLU} = 68.8 CO + 4.91; R² = 0.97.

¹²TIMP-1 42 dias _{0.05% CO} = 17.233 GLU + 3.025; R² = 0.90.

¹³TIMP-1 42 dias _{0.10% CO} = 28.5 GLU + 4.425; R² = 0.89.

Histopathological Evaluation of the Cartilage—Chondrocytes, Type II Collagen, and Proteoglycans

An interaction was found between chondroitin sulfate and glucosamine sulfate at 21- and 42-day old for the chondrocytes (P < 0.0001 and P < 0.0001, respectively), type II collagen (P < 0.0001 and P < 0.0001, respectively) and proteoglycans (P < 0.0001 and P = 0.0007, respectively) in the proximal tibia cartilage (Tables 4 and 5).

Table 4.

Number of chondrocytes, concentration of type II collagen, and proteoglycans in the articular cartilage of the proximal tibial epiphysis of broilers at 21 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Chondrocytes (number in 86.8 μm2)	0	11.33	11.45	13.12	L15 (0.0187)	11.97	0.18	ns	ns	<0.0001
	0.05	11.31	12.92	11.76	Q16 (0.0006)	11.99
	0.10	11.28	11.98	12.76	L17 (0.0022)	12.00
	Regression	ns	Q14 (0.0004)	ns
	Mean	11.31	12.12	12.54
Collagen II (pixel/inch)	0	208,056	207,943	209,504	ns	208,501	0.48	ns	ns	<0.0001
	0.05	211,884	214,252	210,638	Q20 (0.0007)	212,258
	0.10	209,795	212,572	209,912	ns	210,759
	Regression	Q18(0.0057)	Q19(<0.0001)	ns
	Mean	209,911	211,589	210,018
Proteoglycans (pixel/inch)	0	175,716	173,464	166,619	ns	171,933	1.55	ns	ns	<0.0001
	0.05	169,558	186,315	176,094	Q24(<0.0001)	177,322
	0.10	177,132	178,895	171,757	ns	175,928
	Regression	Q21 (0.0361)	Q22(0.0002)	Q23(0.0443)
	Mean	174,135	179,558	171,490

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q: significant quadratic regression.

¹⁴Condrocytes 21 d _{0.15% GLU} = −482 CO² + 53.5 CO + 11.45; R² = 0.99.

¹⁵Condrocytes 21 d _{0% CO} = 5.9667 GLU + 11.072; R² = 0.80.

¹⁶Condrocytes 21 d _{0.05% CO} = −61.556 GLU² + 19.967 GLU + 11.31; R² = 0.99.

¹⁷Condrocytes 21 d _{0.10% CO} = 4.9333 GLU + 11.267; R² = 0.99.

¹⁸Collagen II 21 d _{0% GLU} = −1E + 6 CO² + 135,730 CO + 208,056 R² = 0.99.

¹⁹Collagen II 21 d _{0.15% GLU} = −2E + 6 CO² + 206,070 CO + 207,943 R² = 0.99.

²⁰Collagen II 21 d _{0.05% CO} = −132,933 GLU² + 35,727 GLU + 211,884; R² = 0.99.

²¹Proteoglycans 21 d _{0% GLU} = 3E + 6 CO² + 260,480 CO + 175,716 R² = 0.99.

²²Proteoglycans 21 d _{0.15% GLU} = −4E + 6 CO² + 459,730 CO + 173,464 R² = 0.99.

²³Proteoglycans 21 d _{0.30% GLU} = −3E + 6 CO² + 327,620 CO + 166,619 R² = 0.99.

²⁴Proteoglycans 21 d _{0.05% CO} = −599,511 GLU² + 201,640 GLU + 169,558; R² = 0.99.

Table 5.

Number of chondrocytes, concentration of type II collagen, and proteoglycans in the articular cartilage of the proximal tibial epiphysis of broilers at 42 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				0	0.15	0.30
Chondrocytes (number in 86.8 μm2)	0	8.62	9.79	9.76	Q28(0.0018)	9.39	0.16	ns	ns	<0.0001
	0.05	9.11	10.22	10.42	Q29(0.0028)	9.92
	0.10	9.66	10.93	10.75	Q30(0.0153)	10.45
	Regression	L25(0.0127)	L26(0.0032)	L27(0.0192)
	Mean	9.13	10.32	10.31
Collagen II (pixel/inch)	0	212,331	213,407	209,706	ns	211,815	0.62	ns	ns	<0.0001
	0.05	212,649	211,736	212,458	ns	212,281
	0.10	207,884	210,310	218,994	L33(<0.0001)	212,396
	Regression	Q31(<0.0001)	ns	L32(<0.0001)
	Mean	210,955	211,818	213,719
Proteoglycans (pixel/inch)	0	172,705	182,635	167,584	Q36(<0.0001)	174,308	2.15	ns	ns	0.0007
	0.05	174,790	177,834	186,683	L37 (0.0188)	179,769
	0.10	186,132	179,559	185,548	ns	183,746
	Regression	L34(0.0314)	ns	Q35(0.0019)
	Mean	177,876	180,009	179,938

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q: significant quadratic regression.

²⁵Chondrocytes 42 d _{0% GLU} = 10.4 CO + 8.61; R² = 0.99.

²⁶Chondrocytes 42 d _{0.15% GLU} = 11.4 CO + 9.7433; R² = 0.98.

²⁷Chondrocytes 42 d _{0.30% GLU} = 9.9 CO + 9.815; R² = 0.96.

²⁸Chondrocytes 42 d _{0% CO} = −26.667 GLU² + 11.8 GLU + 8.62; R² = 0.99.

²⁹Chondrocytes 42 d _{0.05% CO} = −20.222 GLU² + 10.433 GLU + 9.11; R² = 0.99.

³⁰Chondrocytes 42 d _{0.10% CO} = −32.222 GLU² + 13.3 GLU + 9.66; R² = 0.99.

³¹Collagen II 42 d _{0% GLU} = −1E + 6 CO² + 57,190 CO + 212,331; R² = 0.99.

³²Collagen II 42 d _{0.30% GLU} = 92,880 CO + 209,075; R² = 0.95.

³³Collagen II 42 d _{0.10% CO} = 37,033 GLU + 206,841; R² = 0.90.

³⁴Proteoglycans 42 d _{0% GLU} = 134,270 CO + 171,162; R² = 0,86.

³⁵Proteoglycans 42 d _{0.30% GLU} = −4E + 6 CO² + 584,320 CO + 167,584; R² = 0.99.

³⁶Proteoglycans 42 d _{0% CO} = −555,133 GLU² + 149,470 GLU + 172,705; R² = 0.99.

³⁷Proteoglycans 42 d _{0.05% CO} = 39,643 GLU + 173,823; R² = 0.93.

For the chondrocytes at 21-days old, the treatments with 0.00 and 0.10% chondroitin sulfate had an increasing linear effect of the glucosamine sulfate addition (P = 0.0187, Eq. (15) and P = 0.0022, Eq. (17), respectively), with an increased number of chondrocytes due to the inclusion of glucosamine sulfate. The treatment with 0.05% chondroitin sulfate had a quadratic effect of the glucosamine sulfate addition (P = 0.0006, Eq. (16)), and the inclusion of 0.16% glucosamine sulfate was predicted to increase the number of chondrocytes. A quadratic effect on the chondroitin sulfate levels was observed with the addition of 0.15% glucosamine sulfate (P = 0.0004, Eq. (14)), and the addition of 0.06% chondroitin sulfate was predicted to maximize the number chondrocytes in the proximal tibia cartilage of broilers at 21-day old (Table 4).

At 42-days old, there were a quadratic effect of glucosamine sulfate levels for birds that had received the addition of 0.00, 0.05 and 0.10% chondroitin sulfate (P = 0.0018, Eq. (28), P = 0.0028, Eq. (29) and P = 0.0153, Eq. (30), respectively). It was predicted that adding 0.22, 0.26 and 0.21% glucosamine sulfate would respectively increase the number of chondrocytes in cartilage. An increasing linear effect of chondroitin sulfate was observed for birds that had received 0.00, 0.15 and 0.30% glucosamine sulfate in the diet (P = 0.0127, Eq. (25), P = 0.0032, Eq. (26) and P = 0.0192, Eq. (27), respectively), with an increase in the number of chondrocytes in cartilage due to the inclusion of chondroitin sulfate (Table 5).

For type II collagen at 21 d, the treatment with 0.05% of chondroitin sulfate had a quadratic effect of the glucosamine sulfate addition (P = 0.0007, Eq. (20)), and adding 0.13% glucosamine sulfate was predicted to result in a higher concentration of type II collagen in the tibia cartilage. The treatments without and with 0.15% glucosamine sulfate had a quadratic effect of the chondroitin sulfate addition (P = 0.0057, Eq. (18) and P < 0.0001, Eq. (19)), and the inclusion of 0.07 and 0.05% chondroitin sulfate was predicted to increase type II collagen (Table 4).

At 42 d, there was an increasing linear effect of the glucosamine sulfate addition (P = 0.0001, Eq. (33)) for birds that had received 0.10% chondroitin sulfate in the diet, with an increase in type II collagen due to the inclusion of glucosamine sulfate. A quadratic effect on the chondroitin sulfate levels (P < 0.0001, Eq. (31)) was observed without the addition of glucosamine sulfate, and the addition of 0.07% chondroitin sulfate was predicted to maximize the concentration of type II collagen in the proximal tibia cartilage of broilers. There was an increasing linear effect of the chondroitin sulfate addition (P < 0.0001, Eq. (32)) for birds that received 0.30% glucosamine sulfate in the diet, with an increase in type II collagen due to the inclusion of chondroitin sulfate (Table 5).

For proteoglycans at 21-day old, the treatment with 0.05% chondroitin sulfate had a quadratic effect of the glucosamine sulfate addition (P < 0.0001, Eq. (24)) and the inclusion of 0.16% glucosamine sulfate was predicted to increase proteoglycans. A quadratic effect on the chondroitin sulfate levels was observed with the addition of 0.00, 0.15 and 0.30% glucosamine sulfate (P = 0.0361, Eq. (21), P = 0.0002, Eq. (22) and P = 0.0443, Eq. (23), respectively). Adding 0.06% chondroitin sulfate was predicted to increase proteoglycans in cartilage for the treatments with 0.15 and 0.30% of glucosamine sulfate, and adding 0.04% chondroitin sulfate was predicted to minimize proteoglycans for the treatment without glucosamine sulfate (Table 4).

At 42-day old, there was a quadratic effect of glucosamine sulfate levels for the treatment without chondroitin sulfate (P < 0.0001, Eq. (36)), and the addition of 0.13% glucosamine sulfate was predicted to increase proteoglycans in cartilage. An increasing linear effect of glucosamine sulfate was observed for birds that had received 0.05% chondroitin sulfate in the diet (P = 0.0188, Eq. (37)), with an increase in proteoglycans in cartilage due to the inclusion of glucosamine sulfate. There was an increasing linear effect of the chondroitin sulfate addition (P = 0.0314, Eq. (34)) for birds that had not received glucosamine sulfate in the diet, with an increase in proteoglycans due to the inclusion of chondroitin sulfate. A quadratic effect on the chondroitin levels was observed with the addition of 0.30% glucosamine sulfate (P = 0.0019, Eq. (35)), and the addition of 0.03% chondroitin sulfate was predicted to increase proteoglycans in cartilage (Table 5).

Bone and Joint Cartilage Macroscopy

An interaction was found between chondroitin and glucosamine sulfates at 21-day old for the length and width of the proximal tibial epiphysis (P < 0.0001 and P < 0.0001, respectively) (Table 6). For the length of the tibia, a quadratic effect on the chondroitin sulfate levels was observed with the addition of 0.30% glucosamine sulfate (P = 0.0020, Eq. (38)), and the addition of 0.03% chondroitin sulfate was predicted to decrease the length at 21-day old. The treatment with 0.00 and 0.05% chondroitin sulfate had a quadratic effect of the glucosamine sulfate addition (P = 0.0174, Eq. (39) and P = 0.0084, Eq. (40), respectively), and the inclusion of 0.23 and 0.24% glucosamine sulfate was predicted to increase the length of the tibia. The treatment with 0.10% chondroitin sulfate had an increasing linear effect of the glucosamine sulfate addition for this characteristic (P = 0.0005, Eq. (41)) (Table 6).

Table 6.

Length and width of the proximal tibial epiphysis of broiler at 21 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Length (mm)	0	72.83	76.25	76.33	Q39(0.0174)	75.14	0.53	ns	ns	<0.0001
	0.05	73.33	75.73	75.92	Q40(0.0084)	74.99
	0.10	74.67	75.58	79.25	L41(0.0005)	76.50
	Regression	ns	ns	Q38(0.0020)
	Mean	73.61	75.85	77.17
Width of the proximal (mm)	0	19.70	21.30	20.88	Q43(0.0014)	20.62	0.17	ns	ns	<0.0001
	0.05	20.48	21.16	21.51	L44(0.0118)	21.05
	0.10	21.04	20.76	21.12	ns	20.97
	Regression	L42(0.0002)	ns	ns

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q, significant quadratic regression.

³⁸Legth 21 d _{0.30% GLU} = 748 CO² − 45.6 CO + 76.33; R² = 0.99.

³⁹Legth 21 d _{0% CO} = −74.222 GLU² + 33.933 GLU + 72.83; R² = 0.99.

⁴⁰Legth 21 d _{0.05% CO} = −49.111 GLU² + 23.367 GLU + 73.33; R² = 0.99.

⁴¹Legth 21 d _{0.10% CO} = 15.267 GLU + 74.21; R² = 0.90.

⁴²Width of the proximal 21 d _{0% GLU} = 13.4 CO + 19.737; R² = 0.99.

⁴³Width of the proximal 21 d _{0% CO} = −44.889 GLU² + 17.4 GLU + 19.7; R² = 0.99.

⁴⁴Width of the proximal 21 d _{0.05% CO} = 3.4333 GLU + 20.535; R² = 0.97.

At 21-day old, there was an increasing linear effect of chondroitin sulfate levels for birds that did not receive glucosamine sulfate in the diet (P = 0.0002, Eq. (42)), with an increase in the width of the proximal tibial epiphysis due to the inclusion of chondroitin sulfate. There was a quadratic and increasing linear effect of glucosamine sulfate levels for birds that had received the addition of 0.00 and 0.05% chondroitin sulfate (P = 0.0014, Eq. (43) and P = 0.0018, Eq. (44), respectively). The addition of 0.19% glucosamine was predicted to increase the width of the proximal tibial epiphysis when the birds did not receive chondroitin, and the addition of 0.15% of chondroitin sulfate increased the width of the proximal tibial epiphysis due to the inclusion of glucosamine sulfate (Table 6).

A quadratic effect of glucosamine sulfate was observed on the tibia length at 42-day old (P = 0.0320, Eq. (45)), with an increase in tibia length with the addition of 0.15% glucosamine sulfate (Table 7). An interaction was found between chondroitin and glucosamine sulfates at 42-day old for the width of the proximal and distal epiphysis of the tibia (P = 0.0133 and P < 0.0001, respectively) (Table 7). There was a quadratic effect of glucosamine sulfate levels for birds that had received the addition of 0.05% chondroitin (P = 0.0085, Eq. (46)). Addition of 0.12% chondroitin sulfate was predicted to increase the width of the proximal tibial epiphysis (Table 7).

Table 7.

Length and width of the proximal epiphysis and width of the distal epiphysis of the tibia of broiler at 42 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Length (mm)	0	102.50	104.83	103.25	ns	103.53	0.59	ns	Q45(0.0320)	ns
	0.05	102.83	104.77	102.65	ns	103.42
	0.10	103.00	106.33	102.83	ns	104.06
	Regression	ns	ns	ns
	Mean	102.78	105.31	102.91
Width of the proximal epiphysis (mm)	0	30.45	30.15	30.45	ns	30.35	0.19	ns	ns	0.0133
	0.05	30.72	31.07	30.25	Q46(0.0085)	30.68
	0.10	30.66	31.00	30.45	ns	30.70
	Regression	ns	ns	ns
	Mean	30.61	30.74	30.38
Width of the distal epiphysis (mm)	0	18.44	19.83	20.41	L49(<0.0001)	19.56	0.15	ns	ns	<0.0001
	0.05	18.53	20.42	20.08	Q50(0.0013)	19.68
	0.10	19.28	20.35	20.02	Q51(0.0036)	19.88
	Regression	L47(0.0184)	Q48(0.0485)	ns
	Mean	18.75	20.20	20.17

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q: significant quadratic regression.

⁴⁵Leght 42 d = −109.56 GLU² + 33.3 GLU + 102.78; R² = 0.99.

⁴⁶Width of the proximal epiphysis 42 d _{0.05% CO} = −26 GLU² + 6.2333 GLU + 30.72; R² = 0.99.

⁴⁷Width of the distal epiphysis 42 d _{0% GLU} = 8.4 CO + 18.33; R² = 0.83.

⁴⁸Width of the distal epiphysis 42 d _{0.15% GLU} = −132 CO² + 18,4 CO + 19,83; R² = 0,99.

⁴⁹Width of the distal epiphysis 42 d _{0% CO} = 6.5667 GLU + 18.575; R² = 0.95.

⁵⁰Width of the distal epiphysis 42 d _{0.05% CO} = - 49.556 GLU² + 20.033 GLU + 18.53; R² = 0.99.

⁵¹Width of the distal epiphysis 42 d _{0.10% CO} = −31.111 GLU² + 11.8 GLU + 19.28; R² = 0.99.

For the of width of the distal tibial epiphysis at 42-day old, there were increasing linear and quadratic effects of chondroitin sulfate levels for birds that had received the addition of 0.00 and 0.15% glucosamine, respectively (P = 0.0184, Eq. (47) and P = 0.0485, Eq. (48), respectively), with an increase in the distal epiphysis width in birds that did not receive glucosamine sulfate, and the addition of 0.07% chondroitin sulfate was predicted to increase the distal epiphysis width for birds that had received 0.15% glucosamine sulfate (Table 7).

There was an increasing linear effect and 2 quadratic effects of glucosamine sulfate levels for birds that had received the addition of 0.00, 0.05 and 0.10% chondroitin sulfate, respectively (P < 0.0001, Eq. (49), P = 0.0013, Eq. (50) and P = 0.0036, Eq. (51), respectively), with an increase in the distal epiphysis width due to the inclusion of glucosamine sulfate when the birds did not receive chondroitin sulfate, and the addition of 0.20 and 0.19% glucosamine sulfate was predicted to increase the distal epiphysis width when the birds received 0.05 and 0.10% chondroitin sulfate in the diet, respectively (Table 7).

No effects of treatments (P > 0.05) were observed for the absolute and relative weights or the Seedor index of the tibia at 21 and 42 d of age, for the diaphysis and distal epiphysis width of the tibia at 21 d of age, or for the diaphysis width at 42 d of age.

An interaction was found between chondroitin and glucosamine sulfates at 21-day old for the transverse and medial thickness of the articular cartilage of the proximal tibial epiphysis (P < 0.0001 and P < 0.0001, respectively) (Table 8). For the transverse thickness, the treatment without glucosamine sulfate had an increasing linear effect of the chondroitin sulfate addition (P < 0.0001, Eq. (52)) at 21-day old. The treatments without and with 0.05% chondroitin sulfate had a quadratic effect and an increasing linear effect of the glucosamine sulfate addition, respectively (P = 0.0002, Eq. (53) and P = 0.0119, Eq. (54), respectively). The addition of 0.19% glucosamine sulfate was predicted to increase the transverse thickness when the birds received 0.05% chondroitin sulfate, and when the birds did not receive chondroitin, there was increase in the transverse thickness of the proximal epiphysis due to the inclusion of glucosamine sulfate (Table 8).

Table 8.

Transverse and medial thickness of the articular cartilage of the proximal tibial epiphysis of broiler at 21 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Transverse thickness (mm)	0	19.67	21.82	21.26	Q53 (0.0002)	20.92	0.17	ns	ns	<0.0001
	0.05	20.82	21.67	22.01	L54 (0.0119)	21.50
	0.10	21.63	21.09	21.61	ns	21.44
	Regression	L52 (< 0.0001)	ns	ns
	Mean	20.70	21.53	21.62
Medial thickness (mm)	0	14.95	16.80	16.37	Q56 (0.0002)	16.04	0.12	ns	ns	<0.0001
	0.05	16.03	16.69	16.95	L57 (0.0119)	16.56
	0.10	16.65	16.24	16.64	ns	16.51
	Regression	L55 (0.0001)	ns	ns
	Mean	15.88	16.58	16.65

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q: significant quadratic regression.

⁵²Transverse thickness 21 d _{0% GLU} = 19.6 CO + 19.727; R² = 0.99.

⁵³Transverse thickness 21 d _{0% CO} = - 60.222 GLU² + 23.367 GLU + 19.67; R² = 0.99.

⁵⁴Transverse thickness 21 d _{0.05% CO} = 3.9667 GLU + 20.905; R² = 0.94.

⁵⁵Medial thickness 21 d _{0% GLU} = 17 CO + 15.027; R² = 0.98.

⁵⁶Medial thickness 21 d _{0% CO} = −50.667 GLU² + 19.933 GLU + 14.95; R² = 0.99.

⁵⁷Medial thickness 21 d _{0.05% CO} = 3.0667 GLU + 16.097; R² = 0.94.

At 21-day old, there was an increasing linear effect of the glucosamine sulfate addition (P = 0.0001, Eq. (55)) for birds that did not receive glucosamine sulfate in the diet, with an increase in the medial thickness of the proximal epiphysis cartilage due to the inclusion of chondroitin sulfate. A quadratic effect and an increasing linear effect on glucosamine sulfate levels (P = 0.0002, Eq. (56) and P = 0.0119, Eq. (57)) was observed without and with 0.05% chondroitin sulfate, respectively. The addition of 0.20% glucosamine sulfate was predicted to increase the medial thickness when the birds did not receive chondroitin sulfate, and when the birds received 0.05% chondroitin sulfate, there was increase in the medial thickness due to the inclusion of glucosamine sulfate (Table 8).

An increasing linear effect of glucosamine sulfate addition (P = 0.0322, Eq. (58)) was observed on the cartilage weight of the distal epiphysis at 42-day old, with an increase in cartilage weight due the addition glucosamine sulfate (Table 9). An interaction was found between chondroitin and glucosamine sulfates at 42-day old for the transverse and medial thickness of the articular cartilage of the proximal tibial epiphysis (P < 0.0001 and P < 0.0001, respectively) (Table 9).

Table 9.

Weight, transverse, and medial thickness of the articular cartilage of the distal epiphysis of the tibiae of broiler at 42 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Weight (g)	0	3.64	3.82	3.76	ns	3.74	0.07	ns	L58(<0.0322)	ns
	0.05	3.75	3.79	3.77	ns	3.77
	0.10	3.57	3.61	3.95	ns	3.71
	Regression	ns	ns	ns
	Mean	3.65	3.74	3.83
Transverse thickness (mm)	0	19.44	20.83	21.24	L60(<0.0001)	20.50	0.15	ns	ns	<0.0001
	0.05	19.53	21.50	21.33	Q61(<0.0001)	20.79
	0.10	20.11	21.39	21.10	Q62(0.0002)	20.87
	Regression	ns	Q59 (0.0293)	ns
	Mean	19.69	21.25	21.22
Medial thickness (mm)	0	14.67	15.62	15.97	L65(<0.0001)	15.42	0.11	ns	ns	<0.0001
	0.05	14.65	16.13	16.10	Q66(<0.0001)	15.63
	0.10	15.25	16.04	15.82	Q67(0.0026)	15.70
	Regression	L63 (0.0371)	Q64 (0.0293)	ns
	Mean	14.86	15.93	15.96

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression; Q: significant quadratic regression.

⁵⁸Weight 42 d = 0.6 GLU + 3.65; R² = 0.99.

⁵⁹Transverse thickness 42 d _{0.15% GLU} = −156 CO² + 21.2 CO + 20.83; R² = 0.99.

⁶⁰Transverse thickness 42 d _{0% CO} = 6 GLU + 19.603; R² = 0.91.

⁶¹Transverse thickness 42 d _{0.05% CO} = −47.556 GLU² + 20.267 GLU + 19.53; R² = 0.99.

⁶²Transverse thickness 42 d _{0.10% CO} = −34.889 GLU² + 13.767 GLU + 20.11; R² = 0.99.

⁶³Medial thickness 42 d _{0% GLU} = 5.8 CO + 14.567; R² = 0.72.

⁶⁴Medial thickness 42 d _{0.15% GLU} = −120 CO² + 16.2 CO + 15.62; R² = 0.99.

⁶⁵Medial thickness 42 d _{0% CO} = 4.3333 GLU + 14.77; R² = 0.93.

⁶⁶Medial thickness 42 d _{0.05% CO} = −33.556 GLU2 + 14.9 GLU + 14.65; R² = 0.99.

⁶⁷Medial thickness 42 d _{0.10% CO} = −22.444 GLU2 + 8.6333 GLU + 15.25; R² = 0.99.

For the transverse thickness, the treatment with 0.15% glucosamine sulfate had a quadratic effect of the chondroitin sulfate addition (P = 0.0293, Eq. (59)) at 42-day old, and the addition of 0.07% chondroitin sulfate was predicted to increase the transverse thickness. There was an increasing linear effect and 2 quadratic effects of glucosamine sulfate levels for birds that had received the addition of 0.00, 0.05 and 0.10% chondroitin sulfate, respectively (P < 0.0001, Eq. (60), P < 0.0001, Eq. (61) and P = 0.0002, Eq. (62), respectively), with an increase in the transverse thickness due to the inclusion of glucosamine sulfate when the birds did not receive chondroitin sulfate, and the addition of 0.21 and 0.20% glucosamine sulfate was predicted to increase the transverse thickness when the birds received 0.05 and 0.10% chondroitin sulfate in the diet, respectively (Table 9).

For the medial thickness of distal epiphysis cartilage at 42-day old, there was an increasing linear effect and a quadratic effect of chondroitin sulfate levels for birds that had received the addition of 0.00 and 0.15% glucosamine sulfate, respectively (P = 0.0371, Eq. (63) and P = 0.0293, Eq. (64), respectively), with an increase in the medial thickness when the birds did not receive glucosamine sulfate, and the addition of 0.07% chondroitin sulfate was predicted to increase the thickness for birds that had received 0.15% glucosamine sulfate (Table 9).

There was an increasing linear effect and 2 quadratic effects of glucosamine sulfate levels for birds that had received the addition of 0.00, 0.05 and 0.10% chondroitin sulfate, respectively (P < 0.0001, Eq. (65), P < 0.0001, Eq. (66), and P = 0.0026, Eq. (67), respectively), with an increase in the medial thickness due to the inclusion of glucosamine sulfate when the birds did not receive chondroitin sulfate, and the addition of 0.22 and 0.19% glucosamine sulfate was predicted to increase the medial thickness when the birds received 0.05 and 0.10% chondroitin sulfate in the diet, respectively (Table 9).

No effects of treatments (P > 0.05) were observed for the absolute and relative weights of the cartilage of the proximal tibial epiphysis at 21 d of age, and the absolute and relative weights of the transversal and medial thickness of the cartilage of the proximal tibial epiphysis at 42 d of age. No effects of treatments (P > 0.05) were observed for the absolute and relative weights or the transversal and medial thickness of the cartilage of the distal epiphysis of the tibia at 21-day old, or for the relative weight of the cartilage of the distal epiphysis of the tibia at 42-day old.

Mineral Densitometry and Bone Breaking Strength

No interaction (P > 0.05) and isolated effects of chondroitin and glucosamine sulfates were observed for densitometry (bone mineral density, bone mineral content and surface area) and bone breaking strength (maximum force, maximum deflection at maximum force and rigidity) (P > 0.05) at 21 and 42-day old.

Mineral Profile

An increasing linear effect of glucosamine sulfate was observed on calcium and phosphorus at 21-day old (P = 0.0009, Eq. (68) and P = 0.0007, Eq. (69), respectively), with an increase in the percentage of calcium and phosphorus in the tibia due to the inclusion of glucosamine sulfate (Table 10).

Table 10.

Percentage of calcium and phosphorus in the tibia ashes of broilers at 21 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Calcium (%)	0	36.44	36.72	36.96	ns	36.72	0.28	ns	L68 (<0.0009)	ns
	0.05	35.50	37.14	37.31	ns	36.65
	0.10	37.13	37.10	37.35	ns	37.19
	Regression	ns	ns	ns
	Mean	36.35	36.99	37.21
Phosphorus (%)	0	17.27	17.62	17.67	ns	17.52	0.25	ns	L69 (<0.0007)	ns
	0.05	17.57	17.24	18.03	ns	17.61
	0.10	17.59	18.15	17.90	ns	17.88
	Regression	ns	ns	ns
	Mean	17.48	17.67	17.87

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression.

⁶⁸Calcium 21 d = 2.8667 GLU + 36.42; R² = 0.93.

⁶⁹Phosphorus 21 d = 1.3 GLU + 17.478; R² = 0.99.

An interaction was identified between the treatments (P < 0.0001) and the amount of calcium in the tibia at 42-day old, with an increasing linear effect on glucosamine sulfate levels for birds that had received 0.00, 0.05 and 0.10% chondroitin sulfate in the diet (P = 0.0003, Eq. (71), P < 0.0001, Eq. (72) and P < 0.0001, Eq. (73), respectively), and an increasing linear effect of chondroitin sulfate levels for birds that had received 0.30% glucosamine sulfate in the diet (P < 0.0001, Eq. (70)). Therefore, in all treatments with chondroitin sulfate, increasing levels of glucosamine sulfate favored the increase of calcium in the tibia of chickens; the same occurred for the treatment with 0.30% glucosamine sulfate, where increasing levels of chondroitin sulfate favored the increase of calcium in the tibia (Table 11). No effects of the treatments with ash (P > 0.05) were observed for the tibia at 21 d of age, or with ash and phosphorus for the tibia at 42-day old (Tables 10 and 11).

Table 11.

Percentage of calcium and phosphorus in the tibia ashes of broilers at 42 d of age, supplemented with chondroitin and glucosamine sulfates in the diet.

Evaluated characteristics	Chondroitin1 (CO, %)	Glucosamine2 (GLU, %)			Regression	Mean	SEM	Probability
		0	0.15	0.30				CO	GLU	CO × GLU
Calcium (%)	0	36.98	39.58	39.64	L71 (0.0003)	38.73	0.29	ns	ns	<0.0001
	0.05	37.11	39.49	40.01	L72 (<0.0001)	38.87
	0.10	37.64	41.34	42.52	L73 (<0.0001)	40.50
	Regression	ns	Ns	L70 (<0.0001)
	Mean	37.24	40.14	40.72
Phosphorus (%)	0	18.45	18.92	18.98	Ns	18.78	0.17	ns	ns	ns
	0.05	18.64	19.56	19.15	ns	19.11
	0.10	18.84	19.29	19.37	ns	19.16
	Regression	ns	Ns	ns
	Mean	18.64	19.25	19.16

¹[(C₁₄H₂₁NO₁₄S)n, Biofac A/S] purity of 91.27%.

²[(C₆H₁₄NO₅)2SO₄ × 2KCl, Zhejiang Golden-Shell Pharmaceutical Co. Ltd.] sulfate content: 16%.

SEM: standard error of the mean; ns: not significant; L: significant linear regression.

⁷⁰Calcium 42 d _{0.30% GLU} = 28.8 CO + 39.283; R² = 0.84.

⁷¹Calcium 42 d _{0% CO} = 8.8667 GLU + 37.403; R² = 0.77.

⁷²Calcium 42 d _{0.05% CO} = 9.6667 GLU + 37.42; R² = 0.88.

⁷³Calcium 42 d _{0.10% CO} = 16.267 GLU + 38.06; R² = 0.92.

DISCUSSION

The results show that the addition of glycosaminoglycans in the diet of broiler chickens can reduce MMP-9 expression and increase TIMP-2 expression and, in parallel, increase the synthesis of proteoglycans and collagen type II. This is the first time in the literature that a linear reduction in the expression of the MMP-9 gene and increase in the expression of the TIMP-2 gene was observed via the inclusion of chondroitin and glucosamine sulfates in diets for broiler chickens; these results are important, as they are able to justify the results of reduced locomotor problems and increased performance found in previous studies by Martins et al. (2020a,b).

A linear reduction in the expression of the MMP-9 gene and an increase in the expression of the TIMP-2 gene were shown via the inclusion of chondroitin and glucosamine sulfates in diets for broilers. These results are important because matrix metalloproteinases (MMPs) constitute a family of zinc-dependent enzymes with predominant proteolytic activity in cartilage. They differ structurally and, in their ability to degrade a particular group of extracellular matrix proteins and, together, they can degrade all of their protein components. The gelatinase class of MMPs, such as types 2 (MMP-2) and 9 (MMP-9), have a peculiar ability to degrade the collagen that makes up the basal lamina. On the other hand, the expression of tissue inhibitors (TIMPs) of MMPs is observed during physiological tissue remodeling, contributing to the maintenance of the metabolic and structural balance of the extracellular matrix. Type 1 (TIMP-1) and 2 (TIMP-2) TIMPs represent well-characterized members of this family of inhibitors and exhibit inhibitory activity against active forms of the entire MMP family (Ribeiro et al., 2008).

Chondroitin and glucosamine sulfates were shown to be able to inhibit the synthesis of MMP-9; the hypothesis from this result is that the degradation of the articular cartilage matrix can be avoided, which could be beneficial to prevent diseases such as femoral degeneration. According to Olkowski et al. (2011), there is a decline in protein content in the bones of affected birds, which may be associated with inadequate synthesis or excessive matrix destruction. This hypothesis is supported by several studies in other species that have identified similar effects of glycosaminoglycans. Taniguchi et al. (2012) reported that long-term administration of glucosamine sulfate and chondroitin sulfate reduced cartilage degeneration in a guinea pig model of spontaneous osteoarthritis due to inhibition of MMP-3 metalloproteinase expression.

In general, the association of chondroitin and glucosamine sulfates at different levels favored the synthesis of chondrocytes, collagen type II and proteoglycans in the articular cartilage of the tibias of broiler chickens at 42-day old. Chondrocytes are highly specialized, and their main function is to produce and maintain the biomechanical properties of cartilage by synthesizing extracellular matrix components (Junqueira and Carneiro, 2017). According to Varghese et al. (2007), glucosamine increases cartilage extracellular matrix components, such as aggrecan and collagen type II, which contribute to better cartilage structure and repair.

According to Wen et al. (2010), glucosamine modulates chondrocyte metabolism, possibly by inhibiting the expression of mitogen-activated protein kinases, which negatively affect the mitotic process of chondrocytes. Kamarul et al. (2011) reported that these glycosaminoglycans provided high expression of proteoglycans and type II collagen in repair areas, and concluded that the combination of sulfates may be beneficial for the healing of damaged cartilage.

These results reinforce and justify the beneficial effect of chondroitin and glucosamine sulfates as chondroprotective and chondrostimulating nutraceuticals, which may act in the prevention of locomotor problems in broilers, as it was previously verified by Martins et al. (2020a) that the inclusion of sulfates promoted a reduction in the incidence of valgus and varus angular deviations, tibial dyschondroplasia and femoral degeneration.

Regarding the bone macroscopy of the tibia and cartilage of birds, there were several effects that demonstrate the importance of chondroitin and glucosamine sulfates on bone and cartilage development in birds, having positive results on thickness, width, weight, and length of the tibia, from its epiphyses and from the cartilages of the tibia. Sgavioli et al. (2017) also verified the effect of including glucosamine sulfate in the diet of broiler chickens on the tibia weight, demonstrating the efficiency of this exogenous glycosaminoglycan in broiler bone development. The association of these polysulfated glycosaminoglycans increases the proliferation of chondrocytes in the epiphyseal grower plate, thus promoting longitudinal bone grower, which can result in greater weight (Wolff, 2014).

Pecchi et al. (2012) demonstrated that, in an inflammatory context, chondroitin sulfate inhibits the production of prostaglandin and metalloproteinases not only in cartilage, but also in subchondral bone. Kim et al. (2007) found that glucosamine sulfate can increase alkaline phosphatase activity, collagen synthesis, osteocalcin secretion, and mineralization in osteoblastic cells in vitro. In addition, they observed an anti-inflammatory effect that reduced catabolic processes, which can promote cell differentiation in osteoblastic cells and, therefore, promote bone grower.

The increase in tibia length correlated positively with tibia width, indicating a larger bone in both length and width (Van Wyle et al., 2012). Yalçin et al. (1998) found a positive correlation between bone strength, weight and bone length to support a greater load during the grower process, where narrow bones had a greater occurrence of fractures. These results justify the use of chondroitin and glucosamine in the diet of fast-growing broilers, as both can contribute to preventing the development of structural lesions or reducing the progression of pathological changes in the bone and joint structure while it is still forming.

Polysulfated glycosaminoglycans, chondroitin and glucosamine, influenced cartilage macroscopy, as the association of these nutraceuticals has a synergistic chondroprotective and chondrostimulating effect, as mentioned above, due to the anabolic processes of cartilage, which stimulated the synthesis of proteoglycans, collagen (Kamarul et al., 2011) and chondrocyte proliferation (Wolff, 2014), preventing cartilage degeneration through anti-inflammatory and inhibitory mechanisms (Chou et al., 2005; Chan et al., 2006; Calamia et al., 2010, 2014; Kantor et al., 2014).

Chondroitin and glucosamine sulfates also demonstrated positive effects for increasing the percentage of bone calcium and phosphorus. Calcium and phosphorus, in the form of hydroxyapatite crystals, are the most abundant minerals in bones (Goff, 2017; Junqueira and Carneiro, 2017). Both are transported to the epiphyseal grower plate, needing to be in adequate concentrations to precipitate into hydroxyapatite crystals and complete the bone calcification process. If there is a reduction in calcium availability, this process can be impaired, favoring an increase in the area of the hypertrophic zone, due to the noncalcification of the chondrocytes, which would remain prehypertrophic, leading to the accumulation of a noncalcified cartilaginous mass in this area, indicating the occurrence of injury (Franco et al., 2004). Therefore, increasing bone minerals can prevent the appearance of deformities and/or bone diseases in broilers (Martins et al., 2020a).

The increase in the amount of these minerals with the addition of sulfates to the feed can be explained by the ability of glycosaminoglycans to complex with ions, due to the high density of negative charges conferred by sulfate and carboxylic groups (Hernandez et al., 2015; Kim et al., 2017). By binding to these ions, glycosaminoglycans could alter their availability and bioactivity in precalcified bone tissues.

The increase in the amounts of calcium and phosphorus in the ash was not accompanied by an increase in the percentage of total mineral matter. This characterizes a change in the distribution of mineral constituents in bone ash in relation to other minerals present. According to Müller et al. (2012), the composition of minerals in bones is not fixed; however, it reflects the chemical equilibrium state of the animal organism. Therefore, in cases of severe disturbances, there will be mobilizations of minerals, which alter their concentrations.

Despite the increase in calcium and phosphorus in the tibia of birds, there was no change in density and bone strength. These results corroborate the work of Sgavioli et al. (2017), who also found no differences for densitometry or bone strength in the tibia of broilers with the supplementation of chondroitin sulfates and glucosamine in the diets. Variations in the ratios of calcium to phosphorus in bone have little effect on mechanical competence, with the ratio of collagen to total minerals being more important than the exact crystal shape in determining bone strength (Williams et al., 2000). Resistance to tibial breakage is not correlated with individual mineral concentrations, but is negatively correlated with collagenous and noncollagenous protein concentrations (Araújo et al., 2011).

As evidenced by the results, the association of chondroitin and glucosamine sulfates has positive synergistic effects on the composition, cartilage and bone structure of broilers, which justifies their combined use, as has been demonstrated in several studies (Chou et al., 2005; Chan et al., 2006; Calamia et al., 2010, 2014; Kamarul et al., 2011; Taniguchi et al., 2012; Kantor et al., 2014; Wolff, 2014).

According to Calamia et al. (2010, 2014), differences in the mechanisms of action could explain why the combination of chondroitin and glucosamine is more effective than When given individually. These compounds produced different patterns of protein modification when tested alone or in combination, both intracellularly (Calamia et al., 2010) and extracellularly (Calamia et al., 2014). However, in most cases, the synergistic effect was demonstrated when cells were exposed to both compounds, which supports combined supplementation for the treatment of locomotor disorders.

It should be noted that a previous study by Martins et al. (2020a) found that the inclusion of sulfates in the diet of broilers reduced the incidence of locomotor problems, and, additionally, glucosamine sulfate alone increased weight gain at 42 d of age, which may justify its economic viability. However, further studies are needed to assess the economic feasibility of using sulfates in the diet of broilers.

In conclusion, chondroitin and glucosamine sulfates can be used in broiler diets in order to favor the development of the structure of the locomotor system (bones and joints), thus preventing locomotion problems.