Fermentation characteristics of resistant starch, arabinoxylan, and β-glucan and their effects on the gut microbial ecology of pigs: A review.

Dietary fibers (DF) contain an abundant amount of energy, although the mammalian genome does not encode most of the enzymes required to degrade them. However, a mutual dependence is developed between the host and symbiotic microbes, which has the potential to extract the energy present in these DF. Dietary fibers escape digestion in the foregut and are fermented in the hindgut, producing short-chain fatty acids (SCFA) that alter the microbial ecology in the gastrointestinal tract (GIT) of pigs. Most of the carbohydrates are fermented in the proximal part, allowing protein fermentation in the distal part, resulting in colonic diseases. The structures of resistant starch (RS), arabinoxylan (AX), and β-glucan (βG) are complex; hence, makes their way into the hindgut where these are fermented and provide energy substrates for the colonic epithelial cells. Different microbes have different preferences of binding to different substrates. The RS, AX and βG act as a unique substrate for the microbes and modify the relative composition of the gut microbial community. The granule dimension and surface area of each substrate are different, which influences the penetration capacity of microbes. Arabinose and xylan are 2 different hemicelluloses, but arabinose is substituted on the xylan backbone and occurs in the form of AX. Fermentation of xylan produces butyrate primarily in the small intestine, whereas arabinose produces butyrate in the large intestine. Types of RS and forms of βG also exert beneficial effects by producing different metabolites and modulating the intestinal microbiota. Therefore, it is important to have information of different types of RS, AX and βG and their roles in microbial modulation to get the optimum benefits of fiber fermentation in the gut. This review provides relevant information on the similarities and differences that exist in the way RS, AX, and βG are fermented, and their positive and negative effects on SCFA production and gut microbial ecology of pigs. These insights will help nutritionists to develop dietary strategies that can modulate specific SCFA production and promote beneficial microbiota in the GIT of swine.

Introduction

Diet induces a change in the microbial ecology and fermentation end products in the gut, which in turn, influences the nutritional, physiological, and immunological functions of pigs (Brestoff and Artis, 2013, Jha et al., 2019). Cereal grains and different agro-industrial coproducts represent major portions of the pig diet which contains a considerable amount of fermentable carbohydrates like resistant starch (RS) and non-starch polysaccharides (NSP) such as AX and βG (Tiwari and Jha, 2016). Most parts of the RS and NSP are not digested in the small intestine and passes to the large intestine where microbes ferment these substrates and produce short-chain fatty acids (SCFA), which in turn influence microbial ecology and overall gut health of pigs (Pieper et al., 2009). Saccharolytic fermentation predominantly takes place in the proximal colon as most microbes prefer to utilize carbohydrates over proteins (Giuberti et al., 2015). Whereas, proteolytic fermentation takes place in the distal colon producing branched-chain fatty acids and potentially harmful metabolites like ammonia (from deamination of amino acids and hydrolysis of urea), indoles, and phenols (from carboxylation of amino acids). These harmful metabolites cause several colonic diseases in the distal colon as the availability of fermentable carbohydrates in the distal colon is minimal (Jha and Berrocoso, 2016). Fermentation of carbohydrates mainly produces SCFA (acetate, propionate, and butyrate) and lactate as major metabolic end products. Production of SCFA is dependent upon the fermentation substrate available and microbial ecology in the gut (Jha and Berrocoso, 2015). Hence, the composition of SCFA produced in the gut can be manipulated by changing the substrate that reaches the colon (Bach Knudsen, 2015). Production of lactic acid bacteria as a result of carbohydrate fermentation is considered beneficial whereas protein fermentation represents a potential risk factor for disruption of the intestinal ecosystem (Jha et al., 2019). Hence, it is of utmost importance to make a dietary strategy that can increase SCFA production constantly throughout the colon by promoting beneficial gut microbiome without compromising growth performance and health of an animal. Due to the difference in fermentation characteristics of various fibrous feed ingredients, nutritionists require a thorough understanding on the inclusion of RS and NSP in the diet of pigs to get the optimum benefits of fiber fermentation in the gut. This review has attempted to critically analyze the role of different types of RS and NSP like AX and βG in swine nutrition. More specifically, this review is focused on describing the structural variation of RS, AX, and βG in different feed ingredients and the way they affect the physiology of digestion, fermentation, and modulation of microbial ecology in the gastrointestinal tract (GIT) of pigs (Table 1).

Table 1

Effects of resistant starch (RS), arabinoxylan (AX) and β-glucan (βG) on microbial community in the gastrointestinal tract of swine.

Dietary fiber	Type of dietary fiber	Microbial proliferation	Site	Reference
RS	RS2	Increased growth of Bifidobacterium, Ruminococcus bromii, Eubacterium rectale	Colon	Martinez et al. (2010)
		Increased growth of Bacteroidetes and Actinobacteria	Colon	Young et al. (2012)
		No effect on microbial proliferation	Feces	Martinez et al. (2010)
	RS3	Increased growth of R. bromii	Colon	Haenen et al. (2013)
		Increased growth of E. rectale	Colon	Martinez et al. (2010)
		Increased bacterial group belonging to Clostridium cluster IV, IX, XV, XVI, XVII	Colon	Haenen et al. (2013)
		Decreased bacterial group belonging to Clostridium cluster XIVa	colon	Haenen et al. (2013)
	RS3 (pattern A)	Increased growth of Atopobium	Colon	Lesmes et al. (2008)
	RS3 (pattern B)	Increased growth of Bifidobacterium	Colon	Lesmes et al. (2008)
	RS4	Increased growth of Bacteroidetes and Actinobacteria	Colon	Martinez et al. (2010)
		Increased growth of Bifidobacterium and ParaBacteroides distasonis	Colon	Martinez et al. (2010)
		No effect on microbial proliferation	Caecum	Metzler-Zebeli et al. (2015)
AX	AX (from wheat)	Increased Lactobacilli, no effect on Escherichia coli	Colon	Van Laere et al., 2000, Mirande et al., 2010
	AX	Arabinoxylan not fermented by Lactobacilli, Enterococci, E. coli, Clostridium perfringens	In vitro study	Crittenden et al. (2002)
		Oligosaccharide with more than 5 units of xylan increased proliferation of Bacteroides	Colon	Mirande et al. (2010)
	AX	Oligosaccharide with less than 5 units of xylan increased proliferation of Bifidobacteria	Colon	Moura et al. (2007)
	AX	Among oligosaccharides xylotriose and xylotetraose was best utilized by Bifidobacteria	Colon	Ejby et al. (2013)
	AX	B. breve and B. infantis are not able to degrade AX, whereas B. longum and B. ovatus could only partially degrade AX and B. adolescentis and B. vulgatus could completely degrade AX	In vitro study	Van Laere et al. (2000)
βG	Mixed linked βG	Reduced Lactobacilli, Enterobacteria and Streptococci	Colon	Pieper et al. (2008)
	βG	Improved growth of overall microbes	Stomach	Leterme et al. (2000)
	βG (from barley)	Increased proliferation of Lactobacilli and Bifidobacteria when compared with wheat based diet	Colon	Garry et al. (2007)
	βG (from wheat)	Increased proliferation of Bifidobacteria and decreased the concentration of Clostridium species	Colon	Pieper et al. (2008)
	βG (from hulled barley)	Increased proliferation of Lactobacilli	Colon	Pieper et al. (2008)
	βG (from hulless barley)	Decreased proliferation of Lactobacilli	Colon
	βG (from oat as well as microbial beta glucan)	Increased growth of Bifidobacteria	Colon	Mårtensson et al. (2005)
	βG (from barley)	Did not increase growth of Bifidobacteria (B. infantis, B. adolescentis, B. longum)	In vitro study	Crittenden et al. (2002)
	βG (from barley)	Increase growth of Bifidobacteria (B. infantis, B. adolescentis, B. longum)	In vitro study	Zhao and Cheung (2011)
	Purified βG	Increased Lactobacilli	Ileum	Pieper et al. (2008)
	βG (from cereal)	Decreased proliferation of Lactobacilli	Ileum
	βG (from cereal)	Increase proportion of Lactobacilli and decreased amount of coliforms	Feces	Pieper et al. (2008)

Structural difference of resistant starch, arabinoxylan, and β-glucan

Resistant starch is a homopolysaccharide of glucose, i.e., a linear molecule of α-1-4-D-glucan, which is resistant to digestion by endogenous enzymes of pigs. There are 5 types of RS based on their physicochemical properties. RS1: this group contains starches which are physically inaccessible, i.e., starches locate inside the fiber-protein matrix, e.g., coarsely ground or whole kernel grains), and RS1 does not break down with normal cooking. RS2: this group contains the granular type or the non-gelatinized native starch granules (e.g., raw potato, green banana, and cornstarch) and can be reduced by thermal treatment. RS3: this group contains heat-stable starches that are produced by gelatinization and retrogradation (slow recrystallization), e.g., cooked and cooled starchy foods. RS4: this group is produced by chemical modification (etherification, esterification and cross-linking), and RS4 is resistant to hydrolysis by host enzymes as well as by bacterial amylase (Birt et al., 2013).

Arabinoxylan is a heteropolysaccharide of D-xylose units joined by β-linkage and substituted by arabinose randomly along the chain, which also allows random substitution of the acetyl group and D-glucuronic acid. Arabinose residue further gets feruloylated, i.e. forms dimer or trimer with ferulic acid, hence forming a heterogenous intermolecular complex. This matrix makes it difficult for the enzymatic degradation and leads to potential encapsulation of nutrients (Pedersen et al., 2014, Tiwari and Jha, 2017). The arabinose substitution on the xylan backbone determines the digestibility and fermentability of AX (Tiwari and Jha, 2016, Tiwari et al., 2018). Arabinose substitution in wheat and rye differs significantly. In wheat, one third of arabinose are linked to the singly substituted xylan backbone, and rest two thirds are linked to the doubly substituted xylan backbone. Whereas in rye, two thirds of arabinose are linked to the singly substituted xylan backbone, and only one third are linked to the doubly substituted (Höije et al., 2008). The structure of AX differs depending upon the botanical origin as well as the specific part of the grain. There is lower arabinose substitution in the aleurone layer and higher in pericarp or testa (Bach Knudsen et al., 2017). Pericarp and testa in grain are the places where almost all lignin is located. Hence, AX from the aleurone layer is readily fermented whereas those from pericarp and testa are slowly fermented. The cell wall in the endosperm layer contains a lower amount of AX than those in bran rich fractions. Aleurone contains a higher amount of insoluble polysaccharides than remaining of other endosperm layer (Bach Knudsen et al., 2017). Aleurone and pericarp also consist of larger amount of ferulic acids than is found in any other starchy endosperm layer (Barron et al., 2007). Aleurone layer in wheat contains a large amount of AX (Bach Knudsen et al., 2017) whereas the aleurone layer in oat bran contains a higher concentration of βG (Wood, 2010). The AX present in the endosperm layer of wheat and rye are less branched with arabinose-to-xylose (A:X) ratio ranging from 0.50 to 0.70 and from 0.48 to 0.55, respectively, whereas that of rice (0.80) and sorghum (0.87) are heavily branched and contains more arabinose, galactose and glucuronic acid substituents (Zhang et al., 2015).

Most of the βG contain pure glucose as their sugar component except for seaweed (laminarin) which also contains mannose (Zhao and Cheung, 2011). Most commonly used βG from cereals in general, such as barley, consists of both β(1-3) and β(1-4) linkage on the main chain in the ratio of 1:3 (Lambo et al., 2005) and requires β-glucanases with both β(1-3) and β(1-4) cleavage activity to degrade them completely (Hughes et al., 2008). Fungal βG (e.g., mushroom) have the most complex structure with β(1-3) linked glucose on the main chain and varying ratio of β(1-6) and β(1-4) on the side chain (Wong et al., 2005). The βG in algae have β(1-3) linked linear glucan backbone with β(1-6) linked glucose on the side chain in the ratio of 3:1 (Read et al., 1996). The βG from bacteria (e.g., curdlan) are linear, unbranched, and highly insoluble whereas, βG from seaweeds (e.g., laminarin) are highly branched and soluble (Zhao and Cheung, 2011). The βG from oat and barley are similar in structure, but the ratio of β(1-3) and β(1-4) varies (Wood and Beer, 2002). Concentration of βG varies among the most commonly used cereals in the diets of pigs, i.e. highest in oat (29 to 63 g/kg) and barley (36 to 99 g/kg), intermediate in wheat and rye (7 to 17 g/kg), and lowest in corn (1 g/kg) (Bach Knudsen et al., 2017). Most brans contain a higher amount of insoluble fiber than cereal grains except oat bran which is more soluble as it contains a higher amount of βG (Wood, 2010). The βG from oat has a high molecular weight and is more insoluble than barley βG as they contain a higher proportion of β(1-4) linkage (0.7) and lower proportion of β(1-3) linkage (0.3; Duss and Nyberg, 2004). Hence, βG from barley is a more readily fermentable substrate for microbes in the GIT of pigs because of their higher solubility.

Normal microbial community in the gastrointestinal tract of pigs

The GIT of pigs consists of a complex and diverse group of microbes. Bacteria comprise the majority of the microbial population in the GIT of pigs, which consists of over 50 genera and more than 500 species of bacteria (Jensen and Jorgensen, 1994). Almost 90% of bacteria in the GIT of the pig are Gram-positive, and rest are Gram-negative (Gaskins, 2001). Species diversity of the majority of bacteria is high in the large intestine due high retention time of the digesta, and their population is in the range of 1010 to 1011 per gram of intestinal content (Jensen and Jorgensen, 1994). Due to the acidic environment in the stomach and proximal small intestine, the number of bacteria is lower (103 to 105 per gram of intestinal content) than that in the distal part. The pig gut is sterile at the time of birth. However, colonization of microbes in the GIT of pigs begins as soon as they are born. Streptococcus spp. and Escherichia coli create an anaerobic environment and start the colonization which paves the way for other microbes like Bacteroides and Bifidobacteria, to start forming a colony (Konstantinov et al., 2006). Before weaning of piglets, Lactobacillus predominates the whole small intestine (Petri et al., 2010). The predominant strains of Streptococci in the small intestine before weaning ferments lactose, whereas the predominant strains of Streptococci after weaning do not ferment lactose (Fouhse et al., 2016).

On the other hand, predominant anaerobes in the colon before weaning are Bacteroides spp. However, their number starts increasing only after weaning. Increased growth rate after weaning has been associated with an increase in the proliferation of Provetella as they are responsible for an increase in the concentration of immunoglobulin A (IgA) (Mach et al., 2015). Immunoglobulin A restricts pathogenic microbes from getting entry to the epithelial cells and prevents their colonization (Inman et al., 2010).

Resistant starch fermentation and its effect on microbial population

A diverse range of starches that are included in RS are fermented in the large intestine instead of being enzymatically digested in the small intestine of pigs (Bird et al., 2007, Jha et al., 2010a, Jha and Leterme, 2012). Starch digestion in pigs is more desirable than its fermentation because digestion products of starch are a better source of energy, whereas fermentation products of starch (SCFA) are less energy efficient (Giuberti et al., 2015). The SCFA can provide up to about 15% of the maintenance energy requirement of the growing pigs and 30% in gestating sows (Varel and Yen, 1997). However, an increase in the concentration of SCFA, more specifically of butyrate, improves the gut mucosal health as well as the immune system of pigs (Jha et al., 2019). Diet rich in RS increases butyrate production in the proximal part of the colon in pigs and modulates the microbial composition (Haenen et al., 2013). Microbes in the GIT has been well known to play a vital role in the development of the immune system and preventing the host from infection (Jha et al., 2019). The RS induced change in the gut microbiota is influenced by several factors like the initial composition of gut microbes (Walker et al., 2010), type of RS (Martinez et al., 2010), crystalline polymorphisms of same type of RS (Lesmes et al., 2008), and the location of the intestine where they are fermented. High amylose-containing starches are effectively degraded by Bifidobacteria (Macfarlane and Macfarlane, 2003).

When RS1 was supplied to pigs through a long-term intake of raw potato starch, it improved the integrity of mucosa by increasing butyrate production and reducing damage to colonocytes (Nofrarías et al., 2007). Increasing dietary amylose or RS content in the diet of pigs can cause a potential physiological change, leading to a proliferation of commensal microbes in the GIT of pigs, which in turn, increase SCFA production (Giuberti et al., 2015). Effect of RS2 on the microbial population is not consistent. However, it helps in the proliferation of Eubacterium and Ruminococcus spp. The granular type RS2 which are extracted and available in pure form are found to increase butyrate production (Weaver et al., 1992). Butyrate is involved in changing the composition of intestinal microbiota by increasing the proportion of Bifidobacterium spp., Ruminococcus bromii and Eubacterium rectale (Martinez et al., 2010). In rats, supplementation of RS2 increased the proliferation of Bacteroidetes and Actinobacteria in colonic digesta (Young et al., 2012). However, in pigs, a significant increase in those 2 genera (Bacteroidetes and Actinobacteria) was reported in RS4, not in RS2 (Martinez et al., 2010). Supplementation of RS2 did not affect the microbial population in the feces (Martinez et al., 2010) or in the cecum and colon (Sun et al., 2016). However, the significant proliferation of R. bromii and E. rectale was seen in RS2 at the species level (Martinez et al., 2010). Raw or boiled RS2 was effectively degraded by R. bromii than by Bacteroides thetaiotaomicron (Ze et al., 2012). Increase in proliferation of Ruminococcus species is associated with improved butyrate production and gut health (Pryde et al., 2002).

The retrograded type RS3 occurs in 3 different patterns. The A type pattern is generally observed in cereals and low amylose starches as they have an open structure and are highly digestible. The B type pattern is seen in potatoes and high amylose starches as they have a close-packed structure and are resistant to digestion by amylase. The C type pattern is found in legume starch and is highly resistant to enzymatic action (Biliaderis, 1991). Although Bifidobacterium spp. and Clostridium butyricum are known to have amylase activity that can utilize high amylopectin and soluble starch, but they were able to utilize RS3 retrograded in B type pattern containing a high amount of amylose (Birt et al., 2013). On the other hand, RS3 retrograded in A type pattern resulted in a proliferation of Atopobium spp. (Lesmes et al., 2008). The RS3 is involved in increasing the proportion of R. bromii (Haenen et al., 2013) and E. rectale (Martinez et al., 2010, Ze et al., 2012). Supplementation of RS3 increased the proliferation of bacterial group belonging to Clostridium cluster IV, IX, XV, XVI, XVII, whereas those belonging to Clostridium cluster XIVa were decreased (Haenen et al., 2013). This difference in the bacterial proliferation of Clostridium cluster can be because of the dependence of 2 important butyrate-producing bacteria (Roseburia and E. rectale) present in Clostridium cluster XIVa on residual carbohydrate as well as pH level (Louis et al., 2007). The pattern in which RS3 is retrograded affects the way they interact with microbes. The RS3 with low amylose content increases the proliferation of Atopobium, whereas those with high amylose content help in the proliferation of amylolytic microbes like Bifidobacterium spp. and C. butyricum. Chemically modified RS4 was developed by modifying the target starch molecules that helps to control reduction in starch digestibility. For examples, chemical modification like acetylation or butyralation of high amylose starch promotes better growth when challenged with necrotic enteritis. Chemical modification helped to increase acidity in the gut and suppress the growth of pathogenic organism. The RS4 augments the proportion of Bifidobacterium adolescentis and Parabacteroides distasonis (Martinez et al., 2010). However, cecal microbial diversity and abundance of phylum were not impacted by RS4 (Metzler-Zebeli et al., 2015).

Different microbes have a preference of binding to the different substrate. Each type of RS act as a unique substrate for the microbes and modifies the relative composition of the gut microbial community. The granule dimension and surface area of each RS are different, which influences the penetration capacity of microbes into the starch granules. Hence, microbial proliferation in the GIT of pigs is highly influenced by the type of RS being offered to the pigs. Different types of RS favor different types of microbes. The RS2 favors Ruminococcus and Eubacterium; RS3 helps to increase the proliferation of all the Clostridium clusters except the one that possesses the butyrogenic bacteria like Roseburia; and RS4 favors production of Bifidobacteria. Changes and variation in the microbial population in the colon of pigs are not just because dietary composition (use of different types of RS as substrate) but are also associated with host, microbes and their interaction (Jha et al., 2019).

Arabinoxylan and β-glucan fermentation and its effect on microbial population

The major NSP in common cereals fed to pigs are AX, cellulose and mixed linked βG (Bach Knudsen et al., 2017). The structure of cell wall is complex, and their composition and properties vary depending upon the location of tissues. Cell wall is thick and hydrophobic and consists of xylans, cellulose and a significant amount of lignin. On the other hand, endosperm (aleurone layer) is thin and hydrophilic and consists mainly of AX and βG (Izydorczyk and Dexter, 2008). The composition and structure of AX and βG are different in various cereals and coproducts. For example, the solubilities of AX and βG in corn and wheat distillers dried grain with solubles (DDGS) are different, so is the A:X ratio or the degree of substitution of arabinose on xylan backbone (Pedersen et al., 2014). Hence, AX and βG act differently when they are in extracted forms or when they exist as a part of grain matrix. The viscous property of these 2 polysaccharides is related to their molecular structure and molecular weight. Beta-glucan is more viscous than AX as the molecular weight of βG (2.1 × 106 to 2.3 × 106) is higher than that of AX (0.07 × 106 to 0.6 × 106) (Saulnier et al., 2007, Wood, 2010). The viscous property created by AX and βG in the intestinal lumen increases the digesta retention time and delays the process of digestion and absorption (Bach Knudsen, 2015, Tiwari et al., 2018). However, this viscous property of βG can be beneficial in the sense that it helps in the removal of cholesterol from the body pool by reducing the reabsorption of bile acids from the small intestine (Tiihonen et al., 2015).

Bacteroides possess the most expanded glycolytic gene that can degrade xylan (Zhang et al., 2014) by producing extracellular endoxylanase (glycoside hydrolase family 10) (Mirande et al., 2010, Ejby et al., 2013). The AX extracted from wheat is not used by E. coli, but increases the growth of Lactobacilli (Van Laere et al., 2000). However, mixed-linked βG in the diet significantly reduced the number of lactobacilli, enterobacteria, and streptococci (Pieper et al., 2008). The utilization of oligosaccharides of AX by gut microbes is limited to Bifidobacteria, and few Firmicutes like Lactobacillus brevis (Moura et al., 2007) as Bacteroides can only degrade those oligomers or polymers which have more than 5 units of xylan or which are larger than xylo-pentose (Mirande et al., 2010). Among the lower xylo-oligosaccharides, xylotriose and xylotetraose are utilized much efficiently than xylobiose by Bifidobacteria (Ejby et al., 2013). Bifidobacteria comprises about 25% of the total cultivable gut microflora (Pastell et al., 2009) and favors xylo-oligomers as a substrate over hexose sugars, but the selectivity of xylo-oligosaccharides is affected by arabinose substitution on the xylan backbone. A substrate with high arabinose substitution is favored by Bifidobacterium when compared with Bacteroides spp., but substrate with lower arabinose substitution results in overall better fermentation (Amrein et al., 2003). Lower arabinose substitution (A:X = 0.34) significantly increased levels of cecal Bifidobacteria after 2-week intake of 2.5 g/kg of AX in chickens but had no effect on Enterobacteriaceae and Lactobacilli (Courtin et al., 2008). Hence, lowering the arabinose substitution will improve the overall fermentation, and AX will be utilized by Bacteroides. However, a higher degree of substitution favor Bifidobacteria proliferation over Bacteroides. Hence, Bacteroides are the primary AX degraders which degrade the larger polymers (hexose or more) of AX into smaller fragment mainly in the proximal colon. Arabinoxylan utilization by Bifidobacteria in the distal part depends upon how efficiently the primary AX degraders worked. Bacteroides are a better degrader of xylan whereas Bifidobacteria utilizes arabinose more efficiently. Hence, metabolic syntrophy among these dominant commensal microbes is maintained by their preference to different chain length as well as arabinose substitution on the xylan backbone, subsequently establishing microbial ecology.

Supplementation of βG has been found to selectively increase the proliferation of Lactobacilli, Bifidobacterium as well as other butyrate-producing bacteria. Most of the bacterial groups has been found to degrade βG except Enterobacteriaceae (Beckmann et al., 2006). Beta-glucan can promote the proliferation of bacteria in the stomach of weaned pigs. Beta-glucans are viscous, and they increase the retention time of digesta in the GIT, thereby extra time for the bacteria to proliferate (Leterme et al., 2000, Jha et al., 2010a). Also, βG produced by Lactobacilli increases their acid tolerant capacity by 15 times (Stack et al., 2010). This shows the positive and protective effect that βG can have on the beneficial microbes in the GIT. Beta-glucan in the barley-based diet increased the proliferation of Lactobacilli and Bifidobacteria when compared to a wheat-based diet in the colon (Garry et al., 2007). The reason behind this might be the presence of a higher concentration of βG in barley than in wheat (Jha et al., 2010a). On the other hand, βG in the wheat-based diet increased Bifidobacteria and decreased the concentration of Clostridium species (Pieper et al., 2008). Beta-glucan in hulled barley promoted the growth of Lactobacilli whereas βG in hulless barley decreased the growth of Lactobacilli and promoted the growth of xylan-degrading bacteria (Pieper et al., 2008). The isolates of βG, as well as microbial βG from oat-based products, stimulated the growth of Bifidobacteria spp. (Mårtensson et al., 2005). However, in vitro model using βG from barley as carbon substrate did not help in the proliferation of 3 Bifidobacterium spp (B. infantis, B. adloescentis, and B. longum) (Crittenden et al., 2002). Whereas all 3 Bifidobacterium spp. proliferated from 1 to 2.3 log10 cfu using purified βG substrates (Zhao and Cheung, 2011). Purified βG increased the ileal Lactobacilli content whereas βG in cereal matrix decreased the proportion of Lactobacilli (Pieper et al., 2008). This indicates that purified βG as a better substrate for the proliferation of beneficial bacteria like Lactobacilli and Bifidobacteria than βG found in grain matrix. It has been found that the Bifidobacteria and Lactobacilli can utilize βG isolated from the cereals. Beta-glucan and RS have been found to show a similar effect on fecal microbes. Both of them increase the proportion of Lactobacilli whereas decrease the number of coliforms (Pieper et al., 2008). Hence, it is not just the content of mixed linked βG that influences the microbial population in GIT of pigs but also their physical forms (purified or grain matrix).

Microbial utilization of short chain fatty acids

The end products of RS, AX, and βG fermentation are SCFA like acetate, propionate and butyrate and various gasses like hydrogen, carbon dioxide and methane (Englyst et al., 1992). The most abundant end product of fermentation in the proximal GIT is acetate which accounts for more than 90% of total SCFA produced, and the concentration of propionate and butyrate is very minimal. However, the condition changes in the distal part where the concentration of SCFA increases with a ratio of approximately 60% acetate, 25% propionate, and 15% butyrate. Most of the SCFA (more than 90%) absorption occurs in the anionic dissociated form, as they are weak acids (Velázquez et al., 1997). The SCFA are absorbed from the apical membrane by 3 different processes: passive diffusion in lipid soluble form (dissociated form) (Velázquez et al., 1997), the anion exchange between bicarbonate and SCFA (Kawamata et al., 2007), and by the help of transporters like monocarboxylate transporter 1 (MCT1) and sodium-coupled monocarboxylate transporter 1 (SMCT1). The MCT1 is coupled to transmembrane hydrogen gradient, and SMCT1 mediates SCFA absorption by enterocytes (Cuff et al., 2005). Different microbes utilize deoxy-sugars (like rhamnose, fucose) or lactate and produce 1,2 propanediol (Saxena et al., 2010). Microbes like Salmonella enterica serovar Typhimurium, metabolize this 1,2 propanediol to propionate or propanol (Bobik et al., 1999). For example, Roseburia inulinivorans has been found to produce propanediol from fucose, which is converted to propionate in human GIT. However, the same bacteria when grown on glucose produces butyrate instead of propionate (Scott et al., 2006). Similarly, butyrate is produced by the Megasphaera elsdenii (Clostridium cluster IX) when they are grown on glucose, but the same bacteria produce propionate when grown on lactate (Hino and Kuroda, 1993). Both Firmicutes and Bacteroidetes enter the succinate pathway via methylmalonyl-CoA (decarboxylation of methylmalonyl-CoA produces propionyl-CoA). Firmicutes produces propionate from organic acids (Flint et al., 2012), whereas Bacteroides utilize peptides and polysaccharides for the production of propionates (Watanabe et al., 2012). Phascolarctobacterium succinatutens can grow only on succinates (Watanabe et al., 2012), whereas Viellonela parvula uses lactate as the main substrate; however, they can get additional energy from succinates (Janssen, 1992). Hence, the production of SCFA varies depending upon the microbes and the substrate available for fermentation.

Fermentation of resistant starch, arabinoxylan, and β-glucan in the small intestine

Purified and isolated βG are readily and easily fermented in the proximal GIT or small intestine of pigs, whereas the βG which occur in grain matrix are fermented in the distal parts of the GIT (Pieper et al., 2008, Jha et al., 2010a). Presence of higher amount of βG would result in increased production of SCFA, which ultimately would reduce the pH of the small intestine. A major group of bacteria with several health benefits is Lactobacilli, and their proliferation increase in the small intestine when barley containing a higher amount of βG is fed. Lactobacilli is acid tolerant and can survive in the acidic environment caused due to higher fermentation. Bach Knudsen and Canibe (2000) fed oat bran containing a higher amount of βG to cannulated pigs and found a higher concentration of lactic acid in the small intestine. Weiss et al. (2016) found a higher ratio of Lactobacilli to Enterobacteriaceae when pigs were fed barley-based diet (2.3) compared to wheat-based diets (1.3). This increased ratio suggests improved resistance against pathogenic microbes in the small intestine with the use of barley in the diet. Lactobacilli have shown to outcompete other bacterial groups when it comes to colonization and nutrient availability in GIT (Lawley and Walker, 2013). It might be the result of this competitive effect of Lactobacilli that there was a significant reduction in the number of other microbes like Bacteroides, Clostridium, Roseburia when pigs were fed a barley-based diet (Weiss et al., 2016). It can be concluded that βG present either in oat- or barley-based diet increases the proliferation of Lactobacilli as well as the concentration of the lactic acid in the small intestine. This decreases the pH and leads to decrease in a number of pathogenic bacteria like E. coli or other members of Enterobacteriaceae which are sensitive to the acidic environment.

Clostridium clusters are generally involved in the production of butyrate (Duncan et al., 2004, Louis et al., 2007). Faecalibacterium and Roseburia, which are 2 bacterial genera with the ability to produce butyrate, are associated with Clostridium cluster IV and XIVa, respectively (Louis et al., 2007). The growth of these butyrate-producing bacteria was higher in pigs fed wheat-based diet compared to barley-based diet (Weiss et al., 2016). This might be because of the presence of a higher amount of fructans in wheat (15 g/kg) compared to barley (6 g/kg). Weiss et al. (2016) recommended that it is not the AX content, but the amount of fructans that causes an increase in butyrate production. The amount of soluble AX in wheat was similar to that of barley whereas barley has a higher amount of insoluble AX. Barley also possesses a higher amount of βG when compared to wheat. However, it was not the barley but wheat that increased butyrate production. Hence, it can be ruled out that an increase in butyrate concentration in the small intestine is due to the presence of a higher amount of βG in the diet. In the case of oat bran, Bach Knudsen et al. (1993) claimed that AX but not βG is responsible for the enhanced butyrate production. Ivarsson et al. (2014) found a strong positive correlation between butyrate production and the intake of xylose, but there was no correlation between butyrate production and AX intake. This can be an indication of xylan acting as the substrate for the production of butyrate in the small intestine, not the arabinose. Production of butyrate in the hindgut was found in a higher proportion when pigs were fed a higher amount of AX (Högberg et al., 2006), which is in accordance to the finding of Ivarsson et al. (2014) who showed a relation between butyrate production and arabinose degradation in the colon. Also, a positive correlation between intake of arabinose and production of propionate was observed in the ileum (Ivarsson et al., 2014), indicating that fermentation of arabinose results in an increased proportion of propionate in the small intestine. Prevotella spp. in the colon is responsible for the production of acetate. However, a positive correlation between Prevotella spp. and butyrate production in the small intestine was observed (Ivarsson et al., 2014), indicating that Prevotella spp. degrade xylans responsible for butyrate production in the small intestine. Thus, it can be concluded that fermentation of xylan results in the increased production of butyrate in the small intestine, and the bacteria responsible for this increased butyrate production is Prevotella spp. The fermentation of arabinose increases the production of propionate in the small intestine and butyrate in the large intestine. Also, the role of fructans and xylan in increased production of butyrate in the small intestine is much higher than that of βG.

Besides producing butyrate, AX also plays an important role in maintaining the integrity of gut by increasing proliferation of goblet cell as well as secretion of IgA. Goblet cells in the GIT produce mucin. Mucin production is increased by Lactobacillus (Che et al., 2014), as well as other species which can help to improve the gut barrier as pathogenic microbes cannot penetrate through the dense mucous layer (Mendis and Simsek, 2015). The AX from wheat bran has been found to increase the number of goblet cells (Dock-Nascimento et al., 2007) which secretes not only mucin but also protein barrier factors (Bergstrom et al., 2008) hence protects intestinal epithelial cells. Arabinoxylan from wheat bran increases the concentration of IgA (Chen et al., 2015), which protects mucosal epithelia by preventing pathogenic microbes from getting attached to epithelial cells.

Fermentation of resistant starch, arabinoxylan, and β-glucan in the large intestine

Microbial fermentation in the hindgut depends on the amount of RS, AX, and βG available for the microbes, i.e., substrate and microbial interaction (Choct, 1997). A significant portion of the soluble AX and βG are fermented in the proximal colon as the fragment length of soluble AX is smaller whereas the fermentation of larger fragments or insoluble AX and βG takes place in the distal colon (Choct, 1997, Tiwari et al., 2018). Fermentation of soluble AX or βG improves gut health comparatively better than insoluble ones (Wellock et al., 2007). However, RS delays the fermentation of other fiber fractions like mixed linked βG or soluble AX in the large intestine of pigs by shifting the microbial metabolism towards utilization of starch (Jonathan et al., 2012).

Source of RS affects the production of SCFA and changes the molar ratio of production of acetate, propionate, and butyrate, but most of RS are butyrogenic (Giuberti et al., 2015). Degradation of RS2 and RS3 is the highest in the proximal colon, and so is the production of lactic acid and SCFA by these RS. However, degradation decreases as there is a progressive decrease in the flow of digesta towards the distal colon leading to change in fermentation metabolite and bacterial profile (Jha and Berrocoso, 2015). However, RS4 has a modification in the structure of starch due to cross-linking, transglycosylation or esterification which prevents hydrolysis of starch both by host enzymes and by bacterial amylases (Birt et al., 2013). Production of butyrate due to fermentation of RS is 2 times higher than that produced due to the fermentation of NSP (Birt et al., 2013). The molar ratio of butyrate production is affected by the source as well as the amount of RS available for the microbes for fermentation, which ultimately influence the proliferation of butyrate-producing bacteria (Pieper et al., 2008). Energy provided by butyrate is vital to maintain the gut ecosystem as well as the health of pigs. In the absence of energy (butyrate), fermentation shifts towards amino acids, i.e., carbon skeleton from deamination of amino acids is used as energy source, and ammonia is absorbed and disposed of as urea (Jha et al., 2019). However, in the presence of energy, ammonia is removed as microbial biomass (Bach Knudsen et al., 1993), i.e., the resident microbes in the large intestine retain more nitrogen for their growth. The RS gets depolymerized quicker than AX and βG. Thus, RS are rapidly fermented in the proximal part whereas AX and βG are fermented slowly in the distal part of the large intestine. In other words, fermentation starts only after the substrate (RS, AX, or βG) gets depolymerized by microbial hydrolytic enzymes. Faster the rate of depolymerization of a substrate, faster the carbohydrates will be available for fermentation by the bacteria.

A higher degree of substitution of arabinose or more branched AX is slowly fermented as endoxylanase enzyme produced by bacteria acts on xylan backbone and has to pass through the branch of arabinose before it can reach to xylan (Tiwari et al., 2018). This is the opposite in the case of RS. The RS which is heavily branched or the one that contains a higher amount of amylopectin provides a larger surface area for the enzymes to act on, hence are broken down in smaller fragments (monomers and dimers) and are rapidly fermented (Giuberti et al., 2015). Degradation of the more linear polymer of RS or the RS that contains a large amount of amylose and low amount of amylopectin gets slowly fermented as their degradation yields larger fragments (larger oligomers) which cannot be directly used by bacteria and further needs to be broken down to smaller fragments. Furthermore, a higher degree of substitution with arabinose impose the risk of forming dimers or trimers with ferulic acids which makes the AX structure more complex and more difficult to break through, hence delays the fermentation (Tiwari et al., 2018). Ferulic acid is the most abundant or predominant phenolic acids present in most cereals as well as in wheat and rye brans which are esterified to AX. Physico-chemical properties are affected by the crosslinking of these diferulates with lignin. Branching increases surface area for the enzyme in solution to act on starch granules in RS (solid substrate), hence surface area accessible to enzymes is an important parameter (Tester et al., 2006). Lower branching or high amylose starch forms smaller surface area and more intramolecular bonds and delays starch degradation. However, not just the branching but the type of surface of starch in RS also affects fermentation. The RS present in tubers have a larger smooth surface, hence are more resistant to enzymatic hydrolysis than the RS present in cereals which have a granular surface (Lehmann and Robin, 2007) and a more open structure (Englyst et al., 1992, Regmi et al., 2011). Cereals high in amylose favor proliferation of C. butyricum, whereas cereals high in amylopectin favor Clostridium ramosum and Bacteroides (Pieper et al., 2009, Bindelle et al., 2011). Barley high in amylose content increased butyrate production both in vivo (Bird et al., 2007) and in vitro (Jha et al., 2011a). Thus, linear RS and branched AX are resistant to degradation and fermentation, whereas linear AX and branched RS are easily and rapidly degraded. According to Jha et al. (2011b), lower half-time of fermentation (T1/2) is an indicator of fermentation taking place throughout the colon whereas higher T1/2 means fermentation is taking place mainly in the distal part of the colon. Hence, it is very important to figure out the ingredients having lower T1/2 values as they would minimize protein fermentation since microbes prefer carbohydrates over protein. Reducing protein fermentation would prevent the release of toxic compounds as well as prevents the proliferation of protein fermenting pathogenic microbes (Williams et al., 2005, Jha and Berrocoso, 2016). Usually, there is a shortage of fermentable carbohydrates in the distal part of the colon. Hence including more linear RS and more branched AX in the diet would help to prevent protein fermentation as they would get slowly depolymerized and therefore lately fermented. Distal fermentation is more important for a healthy colon and distal fermentation of RS is more desirable as that would contribute to the higher uptake of butyrate by the colonocytes.

Arabinoxylan and βG are not completely fermented in the colon of pigs. The βG and RS are fermented to a greater extent than AX and hence are capable of modulating the physicochemical properties of digesta. In vitro studies have shown that various microbes such as Lactobacilli, Enterococci, E. coli, Clostridium perfringens are not able to ferment AX (Crittenden et al., 2002). Viscous property of βG increases the retention time of digesta in small intestine and are fermented in the small intestine. Whereas, RS (from rapeseed meal) reduces the retention time of digesta and are not fermented (Vries et al., 2016). The solubility of AX also affects SCFA production as insoluble AX are less fermentable compared to soluble ones because insoluble AX contains about 100 folds more of ferulic acid as compared to soluble ones (Bunzel et al., 2001). Besides SCFA production, soluble AX also influences gut health by increasing fecal bulk, reduction in transit time, lowering pH in the intestinal lumen as well as bile acid profiles (Tungland and Meyer, 2002). Soluble AX and βG are responsible for changing the viscosity of luminal digesta (Zijlstra et al., 1999, Tiwari et al., 2018). Increase in viscosity acts as a physical barrier between nutrients and enterocytes absorption which results in immune stimulation, villus cell loss, increase proliferation of cells in crypts making it deeper and atrophy in chronic cases. Utilization of energy in supporting immune response diverts the energy utilization in promoting the growth of pigs (Jha et al., 2019). Different Bifidobacterium spp. produce SCFA differently when βG is used as the substrate. The higher amount of SCFA was produced by B. infantis than B. longum, and the ratio of acetate, propionate, and butyrate produced by B. adolescentis was 8:1:1 (Zhao and Cheung, 2011). Amount of insoluble βG is higher in oat than in barley. Thus, the presence of a more significant amount of soluble βG in barley-based diet fed to pigs would produce a higher concentration of total SCFA as well as a higher molar proportion of propionic and butyric acids in the cecum and colon. Arabinoxylan in corn resulted in a higher production of SCFA when compared with AX in wheat or rice bran when human feces was used as microbial inoculum in an in vitro study (Rose et al., 2010). This might be because AX in corn is comparatively less branched than rice bran; hence, are easily degraded and produces a higher amount of SCFA.

Butyrate production by resistant starch, arabinoxylan, and β-glucan in the large intestine

Butyrate-producing bacteria are widely distributed across the different clusters of Clostridium. The butyrogenic bacteria are gram-negative, anaerobic Firmicutes having low mol% of guanine-cytosine content. However, the bulk of potent butyrate-producing bacteria (Faecalibacterium prausitzii, E. rectale, and Roseburia spp.) belongs to Clostridium cluster IV and XIVa (Louis and Flint, 2009). In humans, out of 3 butyrate-producing bacteria, Faecalibacterium is present in the largest amount which comprises almost 5% to15% of the total microbial population (Eckburg et al., 2005), whereas the other 2 butyrate producers (Eubacterium and Roseburia) comprises 5% to 10% of the total microbial species (Aminov et al., 2006). These butyrate-producing bacteria are lactate utilizing bacteria which produces acetyl CoA from lactate and condensation of acetyl CoA with subsequent reduction to butyryl CoA results in the formation of butyrate (Pryde et al., 2002). In the absence of acetate, 75% of the supplied glucose is converted to lactate; however, the presence of acetate results in the production of butyrate (Diez-Gonzalez et al., 1999). Both the proportion of butyrate-producing bacteria as well as the concentration of butyrate were higher at pH 5.5 whereas Bacteroidetes dominated at pH 6.5. This indicates that mild acidic pH allows butyrate-producing bacteria to grow well and be able to compete against gram-negative xylan degrading bacteria (Bacteroides spp). The R. inulinivorans has been found to produce propanediol from fucose, which gets converted to propionate in human GIT. However, the same bacteria when grown on glucose produces butyrate instead of propionate (Scott et al., 2006). Similarly, butyrate is produced by the M. elsdenii when they are grown on glucose, but the same bacteria produce propionate when grown on lactate (Hino and Kuroda, 1993).

Both RS and AX produce butyrate differently. Most of the studies done with pigs either in vitro (Weaver et al., 1992) or in vivo (Marsono et al., 1993) or with humans (Topping et al., 1993) claim RS to be superior to AX, βG, or any other NSP in butyrate production (Bird et al., 2000). This is because the amount of butyrate produced by RS is 20 to 28 mmol% whereas NSP fermentation results in 10 to 15 mmol% of butyrate (Brouns et al., 2002). Though RS have been suggested to produce more butyrate than NSP, Ingerslev et al. (2014) and Nielsen et al. (2014) found the opposite. They observed AX derived from whole grain rye to be superior to RS2 in butyrate and acetate production in pigs. Also, a positive correlation was found between digested AX and net butyrate absorption in catheterized pigs fed a cereal-based diet (Bach Knudsen and Lærke, 2010). Arabinoxylan from rye flakes stimulates butyrate-producing bacteria thereby amount of butyrate more efficiently than RS from raw potato as well as high amylose corn starch (RS2) in the proximal and mid colon (Bach Knudsen and Lærke, 2010). However, AX derived from wheat did not affect colonic or cecal butyrate concentration (Belobrajdic et al., 2012). Among the different types of RS, RS3 is considered as the most powerful butyrogenic substrate (Brouns et al., 2002). Length of the 1,4-α-D-glucan chain and the degree of polymerization of glucose affect the butyrogenic properties of RS3. However, 20 to 25 units of glucose polymerization on the chain of RS3 produce a higher amount of butyrate (Jacobasch et al., 2006). Fermentation of AX concentrate derived from wheat was rapid with the decrease of pH only in the cecum whereas fermentation of AX from whole grain matrix was slower with the decrease in pH both in the proximal colon and cecum (Bach Knudsen, 2015). Hence, it can be concluded that parent grain (rye, wheat, or any other cereals) from which AX is derived as well as pH in the intestinal lumen affects butyrate production. Arabinoxylan is a relatively better butyrate producer when compared to RS2. However, RS3 is the most potent butyrogenic substrate not only among different types of RS but also better than AX, βG or any other NSP.

Increase in production of butyrate not only results from the increased proliferation of butyrate-producing bacteria but can also be as a result of increased acetate produced by Prevotella (Ivarsson et al., 2014) and lactate produced by Bifidobacterium. This is because about 90% of butyrate is derived from acetate (Duncan et al., 2004) and in ruminants (sheep), 60% of the butyrate has been found to be synthesized directly from extracellular acetate (Leng and Leonard, 1965). Acetate and lactate are produced as a result of fermentation by Prevotella and Bifidobacterium, and those acetates can be consumed by butyrate-producing bacteria in the gut to produce butyrate (Belenguer et al., 2006, Rios-Covian et al., 2015). Hence, it is not just RS, AX, βG or any other substrate passing through the colon that affects proliferation of butyrate-producing bacteria and subsequently the butyrate production but also the metabolites like lactate or acetate produced by other microbes like Prevotella and Bifidobacterium contribute as precursors of butyrate production. The in vitro study with different barley and oat cultivars confirmed that βG increased the molar ratio of butyrate (Pieper et al., 2009, Jha et al., 2010b). However, feeding high level of βG increased production of lactate and propionate in the colon with no effect on the production of butyrate (Pieper et al., 2012). Despite lactate and propionate being precursors of butyrate production through the cross-feeding mechanism, it has no effect in increasing the production of butyrate.

Conclusion

Structural variation, degree of polymerization, and branching of RS, AX, βG, and types of RS being offered to interact with the digestive process throughout the GIT, leading to change in the fermentation characteristics and modulation of the microbial community. Degradation of RS on the proximal or distal part of GIT depends on the type of RS. The solubility of AX and βG also affects SCFA production as insoluble AX and βG are less fermentable compared to soluble ones. Branching in RS increases the surface area for the enzymes to act on. However, branching in AX decreases area for xylanase to act on the xylan backbone. Though arabinose and xylan occur in the form of AX, arabinose has been found as a butyrogenic substrate in the large intestine. However, in the small intestine, the role of xylan as a butyrogenic substrate is more pronounced. Most of the bacterial group except Enterobacteriaceae can degrade βG, and a larger polymer of AX is degraded by Bacteroides while smaller oligomers of AX is degraded by Bifidobacteria. It is not just the content of RS, AX or βG but also their physical forms (purified or grain matrix) influence the microbial population in the GIT of pigs. The pH reduction in the hindgut as a result of fermentation suppresses the growth of pathogenic organism, whereas beneficial microbes flourish. Though several studies have started to take into consideration the levels of RS, AX, and βG and its fraction, further information is needed to identify an appropriate source and the amount of RS, AX, and βG that can improve gut health while maintaining or improving the performance of pigs.

Conflict of interest

None.