1st Page

Journal of Animal Science - Animal Nutrition

Evaluation of soluble corn fiber on chemical composition and nitrogen-corrected true metabolizable energy and its effects on in vitro fermentation and in vivo responses in dogs


This article in JAS

  1. Vol. 93 No. 5, p. 2191-2200
    Received: Aug 19, 2014
    Accepted: Mar 12, 2015
    Published: May 8, 2015

    1 Corresponding author(s):

  1. M. R. Panasevich*,
  2. K. R. Kerr*,
  3. M. C. Rossoni Serao*,
  4. M. R. C. de Godoy*,
  5. L. Guérin-Deremaux,
  6. G. L. Lynch,
  7. D. Wils,
  8. S. E. Dowd§,
  9. G. C. Fahey Jr.*,
  10. K. S. Swanson* and
  11. R. N. Dilger 1*
  1. * Department of Animal Sciences, University of Illinois, Urbana 61801
     Roquette Frères, Biology and Nutrition Department, Lestrem, France 62136
     Roquette America, Inc., Geneva, IL 60134
    § MR DNA Molecular Research LP, 503 Clovis Road, Shallowater, TX 79363


Dietary fermentable fiber is known to benefit intestinal health of companion animals. Soluble corn fiber (SCF) was evaluated for its chemical composition, nitrogen-corrected true ME (TMEn) content, in vitro digestion and fermentation characteristics, and in vivo effects on nutrient digestibility, fecal fermentation end products, and modulation of the fecal microbiome of dogs. Soluble corn fiber contained 78% total dietary fiber, all present as soluble dietary fiber; 56% was low molecular weight soluble fiber (did not precipitate in 95% ethanol). The SCF also contained 26% starch and 8% resistant starch and had a TMEn value of 2.6 kcal/g. Soluble corn fiber was first subjected to in vitro hydrolytic–enzymatic digestion to determine extent of digestibility and then fermented using dog fecal inoculum, with fermentative outcomes measured at 0, 3, 6, 9, and 12 h. Hydrolytic–enzymatic digestion of SCF was only 7%. In vitro fermentation showed increased (P < 0.05) concentrations of short-chain fatty acids through 12 h, with acetate, propionate, and butyrate reaching peak concentrations of 1,803, 926, and 112 μmol/g DM, respectively. Fermentability of SCF was higher (P < 0.05) than for cellulose but lower (P < 0.05) than for pectin. In the in vivo experiment, 10 female dogs (6.4 ± 0.2 yr and 22 ± 2.1 kg) received 5 diets with graded concentrations of SCF (0, 0.5, 0.75, 1.0, or 1.25% [as-is basis]) replacing cellulose in a replicated 5 × 5 Latin square design. Dogs were first acclimated to the experimental diets for 10 d followed by 4 d of total fecal collection. Fresh fecal samples were collected to measure fecal pH and fermentation end products and permit a microbiome analysis. For microbiome analysis, extraction of DNA was followed by amplification of the V4 to V6 variable region of the 16S rRNA gene using barcoded primers. Sequences were classified into taxonomic levels using a nucleotide basic local alignment search tool (BLASTn) against a curated GreenGenes database. Few changes in nutrient digestibility or fecal fermentation end products or stool consistency were observed, and no appreciable modulation of the fecal microbiome occurred. In conclusion, SCF was fermentable in vitro, but higher dietary concentrations may be necessary to elicit potential in vivo responses.


Prebiotic fibers in pet foods are becoming increasingly popular due to their favorable effects on gut function and health by increased production of short-chain fatty acids (SCFA) and changes in the intestinal microbiota (Propst et al., 2003). Most prebiotic fibers are rapidly fermentable and, if added at high concentrations to the diet, could result in negative digestive physiologic outcomes such as poor stool consistency and nutrient digestibility. Therefore, it is important to determine appropriate dietary concentrations of prebiotic fibers that modulate the microbiome and increase fermentation characteristics without affecting nutrient digestibility and/or stool consistency.

Common prebiotic fibers often added to pet foods include inulin and oligofructose that promote SCFA production and modulation of the microbiome (Propst et al., 2003). Low digestible carbohydrates are chemically modified starches that increase SCFA production and modify the microbiota in humans and animal models; however, neither the fermentation characteristics nor the altering effects of the fecal microbiome have been studied to any extent in dogs.

Soluble corn fiber (SCF; NUTRIOSE FM; Roquette Frères, Lestrem, France) is a novel low digestible carbohydrate derived from hydrolysis of corn starch by heat and acid. Upon cooling, reformation of mixed β-glycosidic linkages resistant to mammalian enzymatic hydrolysis occurs. Soluble corn fiber is commonly used in the human food industry to aid in colonic health and as a low glycemic food additive (Knapp et al., 2010). Previous research has found it to be fermentable and have positive effects on changing the colonic microbiome of humans and rats; however, there is limited research on its use in dog foods. Therefore, the objectives of this research were to evaluate SCF for nutrient composition, in vitro digestion and fermentability, and in vivo responses (i.e., nutrient digestibility, fermentation end products, and shifts in the intestinal microbiota) in dogs.


Chemical Analyses

Soluble corn fiber (NUTRIOSE FM; Roquette Frères), experimental diets, and fecal samples were analyzed for DM, OM, and ash according to standardized procedures (AOAC, 2006; methods 934.01 and 942.05). Crude protein was calculated from LECO (models FP2000 and TruMac; LECO Corp., St. Joseph, MI) total nitrogen values (AOAC, 2006; method 992.15). Total starch concentration of SCF was determined according to the AOAC, 2006; method 979.10). Total lipid content (acid-hydrolyzed fat) of each substrate was determined according to the methods of the American Association of Cereal Chemists (1983) and Budde (1952). Total dietary fiber and high and low molecular weight soluble fiber of SCF were determined by AOAC (2005) method 2001.03. Briefly, high molecular weight soluble fiber was determined as the portion that precipitated in 95% ethanol and low molecular weight soluble fiber that did not precipitate in 95% ethanol was determined by HPLC. Experimental diets were analyzed for total dietary fiber, insoluble dietary fiber, and soluble dietary fiber concentrations according to Prosky et al. (1992). Free glucose and digestible starch concentrations were determined according to Muir and O’Dea (1993). Resistant starch was determined by subtracting digestible starch and free glucose from total starch concentration. Gross energy was measured using an oxygen bomb calorimeter (model 1261; Parr Instruments, Moline, IL). Free monosaccharide and oligosaccharide concentrations were determined according to Smiricky et al. (2002).

Fecal SCFA and branched-chain fatty acid (BCFA) concentrations were determined by gas chromatography according to Erwin et al. (1961) using a gas chromatograph (model 5890A series II; Hewlett-Packard, Palo Alto, CA) and a glass column (180 cm by 4 mm i.d.) packed with 10% SP-1200/1% H3PO4 on 80/100+ mesh Chromosorb WAW (Supelco Inc., Bellefonte, PA). Nitrogen was the carrier with a flow rate of 75 mL/min. Oven, detector, and injector temperatures were 125, 175, and 180°C, respectively. Fecal ammonia concentrations were determined according to the method of Chaney and Marbach (1962). Fecal phenol and indole concentrations were determined using gas chromatography according to the methods described by Flickinger et al. (2003). Biogenic amine concentrations were quantified using HPLC according to methods described by Flickinger et al. (2003).

In Vitro Hydrolytic Digestion/Fermentation Simulation

The in vitro hydrolytic digestion/fermentation study was conducted according to Panasevich et al. (2013) with some modifications. Briefly, approximately 500 mg of SCF was weighed in triplicate and incubated with 12.5 mL phosphate buffer and 5 mL of a pepsin/hydrochloric acid solution at 39°C to simulate gastric digestion. After 6 h, the pH was adjusted to 6.8 and 5 mL pancreatin solution (Sigma-Aldrich Co., St. Louis, MO) was added to each tube. Incubation continued at 39°C for 18 h to simulate small intestinal digestion (Boisen and Eggum, 1991). The set of samples prepared for enzymatic digestion then was assayed for released free sugars to correct for free glucose entering the in vitro fermentation.

In vitro fermentation was performed using a modification of the method of Bourquin et al. (1993). Following the in vitro digestion procedures described above, samples were hydrated overnight in 26 mL of anaerobic media. Fecal samples from 3 dogs were collected within 10 min of defecation and maintained at 39°C to prepare fresh inoculum. Before collection of feces, dogs had been maintained on a commercially available food for 1 mo (Iams Weight Control; Procter & Gamble Pet Care, Cincinnati, OH). The fecal inoculum was prepared by blending 10 g of each fecal sample with 90 mL anaerobic diluting solution for 15 sec in a Waring blender (Fisher Scientific Inc., Pittsburgh, PA) under a stream of CO2. The resulting solution was filtered through 4 layers of cheesecloth and sealed in 125-mL serum bottles pending the in vitro experiment.

Samples, blanks, and standards were inoculated with 4 mL of diluted feces. Solka-Floc (International Fiber Corp., North Tonawanda, NY) and high-methoxy pectin (TIC Gums Inc., Belcamp, MD) were used as negative and positive fermentation controls, respectively. Tubes were incubated at 39°C with periodic mixing. A subset of tubes was removed from the incubator at 0, 3, 6, 9, and 12 h after inoculation and processed immediately for analyses. A 2-mL subsample of the fluid was removed and acidified for SCFA and BCFA analyses. Concentrations of SCFA and BCFA were determined by gas chromatography as previously described.

In Vivo Studies

Rooster Study: True Metabolizable Energy.

A nitrogen-corrected true ME (TMEn) coefficient was determined using conventional single comb white leghorn roosters (n = 4) according to Kim et al. (2010). Briefly, roosters were deprived of feed for 24 h and then crop intubated with approximately 15 g of SCF and 15 g of corn with a known GE and nitrogen value (Sibbald et al., 1980). Roosters were crop intubated and excreta (urine plus feces) were collected for 48 h on plastic trays placed under each cage. Excreta samples were subsequently lyophilized, weighed, and ground to pass a 60-mesh screen and analyzed for GE content as described for samples above. Endogenous corrections for energy were made using roosters that had been food deprived for 48 h. The TMEn values, corrected for endogenous energy losses, were calculated using the following equation: TMEn (kcal/g) = [energy intake (kcal) – energy excreted by fed birds (kcal) + energy excreted by fasted birds (kcal)]/feed intake (g).

Dog Study: Animals and Diets.

Ten female dogs with hound bloodlines (6.4 ± 0.2 yr and 22 ± 2.1 kg) were used. Dogs were housed in individual kennels (2.4 by 1.2 m) in 2 temperature-controlled rooms with a 16:8 h light:dark cycle. A replicated 5 × 5 Latin square design experiment was conducted with 5 diets and 10 dogs in 2 different rooms for five 14-d periods. The first 10 d of each period served as an adaptation phase followed by 4 d of total fecal collection. Five diets containing SCF were formulated to contain approximately 32% CP and 18% crude fat (DM basis; Table 1). Each diet contained graded concentrations of SCF (0, 0.5, 0.75, 1.0, or 1.25% [as-is basis]) that replaced cellulose (Solka-Floc; International Fiber Corp.) in the diet. Low-ash poultry byproduct meal, poultry fat, brewer’s rice, ground corn, and vitamin and mineral premixes constituted the remainder of the dry, extruded, kibble diets. All diets were formulated to exceed NRC (2006) recommended allowances for an adult large breed dog. Diets were mixed and extruded at the Kansas State University Bioprocessing and Industrial Value-Added Program facility (Manhattan, KS) under the supervision of Pet Food and Ingredient Technology, Inc. (Topeka, KS). Dogs were offered 155 g of diet twice daily (0800 and 1700 h) to meet the required energy needs based on the estimated ME content of the diet. Food refusals were recorded daily and fresh water was provided to the dogs ad libitum. Chromic oxide was added as a digestion marker but was not needed because of excellent stool quality and ease of fecal collection from the pen floor.

View Full Table | Close Full ViewTable 1.

Chemical composition of soluble corn fiber

Item Concentration
DM, % 96.5
DM basis
OM, % 100.0
CP, % 0.0
Acid-hydrolyzed fat, % 0.5
Total dietary fiber, % 78.3
Insoluble dietary fiber 0.0
Soluble dietary fiber 78.3
HMWSF1 22.8
LMWSF2 55.5
Starch, %
    Digestible 17.7
    Resistant 7.8
    Total 25.5
Free sugars, mg/g
    Arabinose 0.7
    Galactose 0.3
    Glucose 24.9
    Sucrose 0.7
    Mannose 0.1
    Fructose 1.8
    Total 28.5
    GE, kcal/g 4.1
    TMEn,3 kcal/g 2.6
1HMWSF = high molecular weight soluble fiber; defined as the portion that precipitated in 95% ethanol.
2LMWSF = low molecular weight soluble fiber; defined as the portion that did not precipitate in 95% ethanol.
3TMEn = nitrogen-corrected true ME.

Sample Handling and Processing

Total feces excreted during the collection phase of each period were taken from the pen floor, weighed, and frozen at –20°C until analysis. All fecal samples during the collection period were subjected to a consistency score according to the following scale: 1 = hard, dry pellets and small hard mass; 2 = hard, formed, dry stool that remains firm and soft; 3 = soft, formed, and moist stool that retains shape; 4 = soft, unformed stool that assumes shape of container; and 5 = watery liquid that can be poured.

Fecal samples were dried at 55°C in a forced-air oven and ground in a Wiley mill (model 4; Thomas Scientific, Swedesboro, NJ) through a 2-mm screen. On d 11 of each period, fresh fecal samples were collected within 15 min of defecation. An aliquot of fresh feces was immediately transferred to sterile cryogenic vials (Nalgene, Rochester, NY) and snap-frozen in liquid nitrogen. Once frozen, vials were stored at –80°C until used for DNA extraction for microbial analysis. Aliquots for analysis of phenols, indoles, and biogenic amines were frozen at –20°C immediately after collection. One aliquot was collected and placed in approximately 2 mL of 2 N hydrochloric acid for ammonia, SCFA, and BCFA analyses. Additional aliquots were used for pH measurement and fresh fecal DM determination.

Microbiome Analysis

Fecal DNA Extraction and 454 Pyrosequencing.

Bacterial DNA was extracted according to McInnes and Cutting (2010) using the PowerSoil Kit (MO BIO Laboratories, Carlsbad, CA). Extracted DNA concentrations were quantified using a Qubit 2.0 Fluorometer (Life Technologies, Carlsbad, CA) and diluted to 5 ng/mL. Quality of DNA was assessed by electrophoresis using precast agarose gels (E-Gel EX Gel 1%; Invitrogen, Grand Island, NY). Amplification of a 600-bp sequence of the V4 to V6 variable region of the 16S rRNA gene was done using barcoded primers (Cephas et al., 2011). Amplicons from PCR then were further purified using AMPure XP beads (Beckman Coulter Inc., Indianapolis, IN). Amplicons were combined in equimolar ratios to create a DNA pool that was used for pyrosequencing. Quality of DNA from amplicon pools was assessed before pyrosequencing using a 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA). Pyrosequencing was performed at the Roy J. Carver Biotechnology Center at the University of Illinois using a 454 Genome Sequencer and FLX titanium reagents (Roche Applied Science, Indianapolis, IN).


High-quality (quality value > 25) sequence data derived from the sequencing process was processed using a proprietary analysis pipeline and as previously described (Dowd et al., 2008a,b, 2011; Edgar, 2010; Capone et al., 2011; Eren et al., 2011; Swanson et al., 2011). Briefly, sequences were depleted of barcodes and primers, short sequences (<200 bp), sequences with ambiguous base calls, and sequences with homopolymer runs exceeding 6 bp. Sequences then were denoised and chimeras were removed. Operational taxonomic units were defined after removal of singleton sequences and clustering at 3% divergence (97% similarity). Then, operational taxonomic units were taxonomically classified using a nucleotide basic local alignment search tool (BLASTn) against a curated GreenGenes database (; accessed January 2012; DeSantis et al., 2006) and compiled into each taxonomic level into both “counts” and “percentage” files. Only genera and species that represented greater than 0.01% of the total sequences were reported.

Statistical Analysis

Data were analyzed as a completely randomized design using the Mixed procedure of SAS (version 9.2; SAS Inst., Inc., Cary, NC). The UNIVARIATE procedure was used to assure equal variance and normal distribution and to identify outliers. Any observation that was more than 3 SD away from the mean was considered an outlier. Data were transformed by log or square root if the normality assumption was not met. The in vitro experimental data were analyzed using mean separation with a Tukey’s adjustment to determine differences among substrates. For the in vivo dog experiment, diet was considered a fixed effect, whereas random effects included animal and period. Linear and quadratic effects were tested using orthogonal polynomial contrasts. Differences among dietary treatments were determined using the LSD method. A probability of P < 0.05 was accepted as being statistically significant. Additionally, sequence percentages were compared using single degree of freedom orthogonal contrasts to test linear and quadratic effects of providing graded concentrations of dietary SCF, and all SCF treatments (0.5 to 1.25%) were compared to the 0% SCF control using a single degree of freedom contrast. Principal component analysis was used to assess shifts in variability between diets and Chao 1 and rarefaction curves were used to assess microbial diversity and species richness.


Substrate Chemical Analysis

Soluble corn fiber was devoid of CP and ash and had very low concentrations of acid-hydrolyzed fat (Table 1). It contained a high amount of total dietary fiber (78.3%) that was completely soluble. Starch concentration was 25.5% with a notable proportion of resistant starch 7.8%. These values sum to 103.8% because a portion of the resistant starch was included in the total dietary fiber value. More low molecular weight soluble fiber (55.5%) was present in SCF compared with high molecular weight soluble fiber (22.8%). The concentration of total free sugars was very low (28.5 mg/g DM), with glucose serving as the predominant free sugar (24.9 mg/g DM).

In Vitro Hydrolytic Digestion/Fermentation

Concentrations of SCFA produced over time from cellulose, SCF, and pectin are shown in Fig. 1. During the in vitro hydrolytic–enzymatic digestion, SCF was only 7% digestible (data not shown), leaving 93% as indigestible material for subsequent in vitro fermentation. Once hydrolytic–enzymatic digestion was complete, the fermentation experiment was corrected for release of free sugars.

Figure 1.
Figure 1.

In vitro experiment: concentrations of acetate (A), propionate (B), butyrate (C), and total short-chain fatty acids (SCFA; D) and pH values during a 12-h in vitro fermentation. Standard error bars are presented for each mean value.*Significant (P < 0.05) time by treatment interaction. #Significant difference (P < 0.05) between pectin and soluble corn fiber within each time point.


Over the 12-h in vitro fermentation, a numerical decrease in pH due to concomitant increases (P < 0.05) in acetate, propionate, and butyrate concentrations with SCF was noted. Concentrations of acetate, propionate, and total SCFA were greater (P < 0.05) for SCF at each time point compared with cellulose. Soluble corn fiber elicited higher (P < 0.05) butyrate concentrations at 6, 9, and 12 h compared with cellulose. In comparison with pectin, SCF produced lower (P < 0.05) acetate, propionate, butyrate, and total SCFA concentrations throughout the 12 h fermentation, which translated into less (P < 0.05) of a decrease in pH over time.

In Vivo Experiments

Rooster Study: Nitrogen-Corrected True Metabolizable Energy.

The TMEn value was determined to be 2.6 kcal/g (Table 1).

Dog Study.

Table 2 presents the ingredient composition of the basal diet fed to dogs, and Table 3 presents the analyzed chemical composition of all experimental diets. All diets had similar DM, OM, CP, acid-hydrolyzed fat, total dietary fiber, and GE concentrations. During the feeding study, food intake was similar among treatments at 310 g/d (as-is basis; 288 g DM/d), and dogs consumed all of their food allotment (data not shown). Fecal output, apparent total tract nutrient digestibility, and fecal consistency scores are all presented in Supplementary Table S1 (see online version of the article at Briefly, fecal output and nutrient digestibility were not affected by increasing concentrations of dietary SCF. Fecal consistency was excellent across all diets, and no differences due to dietary treatment were observed.

View Full Table | Close Full ViewTable 2.

Ingredient composition of the basal diet fed to dogs1

Ingredient Percent, as fed
Brewer’s rice 46.55
Low-ash poultry byproduct meal 25.50
Ground yellow corn 12.00
Poultry fat 8.00
Soluble corn fiber2 Variable
Cellulose3 6.00
Salt 0.70
Potassium chloride 0.56
Chromic oxide 0.20
Mineral mix4 0.18
Vitamin mix5 0.18
Choline chloride, 50% 0.13
1Soluble corn fiber was added at 0, 0.5, 0.75, 1.0, or 1.25% of the diet at the expense of cellulose and brewer’s rice to main isofibrous and isonitrogenous diets.
2NUTRIOSE FM (Roquette Frères, Lestrem, France).
3Solka-Floc (International Fiber Corp., North Tonawanda, NY).
4Provided per kilogram of diet: 66.00 mg Mn (as MnSO4), 120 mg Fe (as FeSO4), 18 mg Cu (as CuSO4), 1.20 mg Co (as CoSO4), 240 mg Zn (as ZnSO4), 1.8 mg I (as KI), and 0.24 mg Se (as Na2SeO3).
5Provided per kilogram of diet: 5.28 mg vitamin A, 0.04 mg vitamin D3, 120 mg vitamin E, 0.88 mg vitamin K, 4.40 mg thiamine, 5.72 mg riboflavin, 22.00 mg pantothenic acid, 39.60 mg niacin, 3.52 mg pyridoxine, 0.13 mg biotin, 0.44 mg folic acid, and 0.11 mg vitamin B12.

View Full Table | Close Full ViewTable 3.

Chemical composition of experimental diets fed to dogs

Soluble corn fiber, %
Item 0 0.5 0.75 1.0 1.25
DM, % 92.9 92.7 93.4 92.9 93.2
% DM basis
OM 94.5 94.6 94.5 94.2 94.1
CP 23.4 23.8 24.2 24.3 24.9
Acid-hydrolyzed fat 14.1 13.8 13.7 14.0 14.0
Total dietary fiber 7.34 7.38 7.39 7.43 7.44
GE, kcal/g 4.99 4.98 4.98 5.00 4.99

Table 4 presents fecal SCFA, BCFA, and ammonia concentrations as well as fecal pH values for dogs. Fecal concentrations of acetate, propionate, and total SCFA were lowest (P < 0.05) when dogs were fed the 0.75% SCF diet. When compared with the 0.75% SCF diet, dogs fed 1.25% SCF had higher (P < 0.05) fecal acetate, propionate, and total SCFA concentrations. Fecal butyrate concentrations were not affected by treatment. Fecal BCFA and ammonia concentrations were low and showed no significant differences due to treatment. Other markers of protein fermentation, including phenols, indoles, and biogenic amines, were measured; however, these compounds were present at low concentrations and were not affected by dietary SCF concentration (data not shown).

View Full Table | Close Full ViewTable 4.

Fecal short-chain fatty acid (SCFA), branched-chain fatty acid (BCFA), and ammonia concentrations and pH values for dogs fed diets containing graded soluble corn fiber concentrations

Soluble corn fiber, %
Item 0 0.5 0.75 1 1.25 SEM P-value Linear Quadratic
pH 6.67 6.45 6.47 6.62 6.23 0.13 0.11 0.09 0.73
SCFA, μmol/g DM
    Acetate 268.7ab 313.6ab 254.7a 292.2ab 317.2b 17.1 0.02 0.15 0.58
    Propionate 96.7ab 116.5b 91.8a 111.9ab 118.9b 7.5 0.01 0.09 0.64
    Butyrate 53.8 57.8 44.8 48.9 53.8 6.1 0.19 0.98 0.79
    Total 419.2ab 487.9b 391.3a 453.1ab 490.0b 27.6 0.02 0.21 0.57
BCFA, μmol/g DM
    Isobutyrate 8.34 9.00 7.72 8.23 8.14 0.71 0.51 0.66 0.86
    Isovalerate 12.93 14.63 12.44 13.27 13.16 1.23 0.10 0.22 0.77
    Valerate 0.87 0.94 0.84 0.82 1.04 0.15 0.63 0.65 0.62
    Total 22.13 24.96 21.00 22.32 22.34 2.01 0.53 0.85 0.79
Ammonia, μmol/g DM 176.1 194.9 162.4 178.8 177.9 13.94 0.22 0.80 0.91
a,bMean values within a row with unlike superscript letters differ (P < 0.05).

Pyrosequencing of 16S rRNA gene-barcoded amplicons resulted in a total of 769,200 sequences, with an average of 15,384 reads (range: 8,776 to 32,349) per sample. Samples had an average read length of 506 bp. According to Chao 1 values and rarefaction curves (data not shown), microbial diversity and species richness were similar among dietary treatments. Principal component analysis revealed no separation among dietary treatments (data not shown).

Regardless of dietary treatment, Firmicutes (24.78 to 92.69% of all sequences) was the predominant bacterial phylum in dog feces followed by Fusobacteria (0.11 to 52.21% of all sequences) and Tenericutes (2.58 to 54.04% of all sequences; Table 5). Actinobacteria (0 to 9.68% of all sequences), Bacteroidetes (0 to 3.22% of all sequences), and Proteobacteria (0 to 8.41% of all sequences) were also present. No statistically significant changes were noted among treatments, but there was a numeric increase in the proportions of Firmicutes and a numeric decrease in Fusobacteria with SCF supplementation.

View Full Table | Close Full ViewTable 5.

Predominant bacterial phyla and genera expressed as a percentage of total sequences in feces of dogs fed diets containing graded soluble corn fiber concentrations1

Soluble corn fiber, %
Phylum Family Genus 0 0.5 0.75 1 1.25 SEM P-value
Actinobacteria 0.26 0.34 0.33 0.17 0.36 0.21 0.74
Bifidobacteriaceae Bifidobacterium 0.24 0.32 0.31 0.15 0.34 0.21 0.73
Bacteroidetes 1.00 0.56 0.91 0.68 0.71 0.21 0.44
Bacteroidaceae Bacteroides 0.49 0.28 0.31 0.47 0.38 0.08 0.08
Prevotellaceae Prevotella 0.34 0.26 0.46 0.41 0.31 0.19 0.91
Firmicutes 56.94 56.44 59.06 59.26 64.89 5.30 0.40
Acidaminococcaceae Acidaminococcus 0.22 0.22 0.27 0.32 0.18 0.05 0.35
Clostridiaceae Clostridium 24.73 21.81 25.17 22.34 26.54 3.87 0.54
Eubacteriaceae Eubacterium 1.41 2.07 2.08 1.63 3.08 0.83 0.21
Lachnospiraceae Blautia 10.97 12.80 14.09 12.71 15.15 2.23 0.34
Dorea 0.24 0.08 0.08 0.25 0.04 0.08 0.08
Lachnospira 0.02a 0.92ab 1.52b 1.57b 1.51b 0.42 0.01
Roseburia2 0.69 0.56 0.54 0.50 0.35 0.14 0.23
Lactobacillaceae Lactobacillus 3.38 4.92 2.75 4.20 5.88 2.86 0.49
Paenibacillaceae Paenibacillus 0.19 0.25 0.37 0.18 0.39 0.14 0.47
Peptococcaceae Delsulfotomaculum 0.75 1.00 1.68 0.75 1.42 0.59 0.39
Ruminococcaceae Fecalibacterium 2.79 3.64 2.43 3.15 5.50 0.97 0.20
Oscillospira 0.79 0.82 0.62 1.11 0.66 0.30 0.35
Ruminococcus2 5.64 5.06 3.96 5.00 4.12 1.04 0.13
Turcibacteraceae Turicibacter 0.54 0.59 1.23 0.98 1.39 0.48 0.55
Veillonellaceae Megamonas3 1.27 0.69 1.13 0.90 1.80 0.32 0.06
Phascolarctobacterium 4.90 4.38 4.86 6.97 4.05 1.05 0.15
Fusobacteria 28.37 27.05 23.69 28.69 17.70 4.52 0.20
Fusobacteriaceae Fusobacterium 28.37 27.05 23.69 28.69 17.70 4.52 0.20
Proteobacteria 2.67 1.92 1.86 2.72 2.61 0.68 0.13
Succinivibrionaceae Succinivibrio 0.16 0.09 0.12 0.08 0.25 0.08 0.20
Anaerobiospirillum 0.22 0.38 0.14 0.57 0.17 0.18 0.22
Alicaligenaceae Sutterella 1.99 1.31 1.45 1.76 1.12 0.47 0.28
Tenericutes 10.75 12.75 14.14 10.78 14.06 4.10 0.32
Erysipelotrichaceae Allobaculum 7.71 9.20 10.32 4.74 8.13 3.96 0.12
Bulleidia 0.32 0.22 0.35 0.18 0.26 0.11 0.62
Catenibacterium2 0.14 0.26 0.16 0.24 1.33 0.39 0.08
Coprobacillus2,4 0.33b 0.17a 0.19ab 0.25ab 0.14a 0.05 0.01
a,bMean values in the same row with unlike superscript letters differ (P < 0.05).
1Genera included have least squares means of 0.01 or higher.
2Linear effect (P < 0.05).
3Quadratic effect (P < 0.05).
4Difference between 0% soluble corn fiber vs. all other soluble corn fiber diets (P < 0.05).

Fusobacterium (0.11 to 52.21% of all sequences), Clostridium (7.31 to 53.29% of all sequences), Blautia (4.38 to 34.04% of all sequences), and Allobaculum (0.27 to 53.58% of all sequences) were the predominant genera in dog feces (Table 5). Fecal Lachnospira increased (P < 0.05) with increasing concentrations of dietary SCF. The proportions of Roseburia and Ruminococcus decreased (P < 0.05) linearly with increasing concentrations of SCF. Within the Tenericutes phylum, the proportion of Catenibacterium increased linearly (P < 0.05) with increasing SCF concentrations. Fecal Coprobacillus exhibited a linear decrease (P < 0.05) and, overall, dogs fed diets containing SCF had lower (P < 0.05) Coprobacillus compared with dogs fed the 0% SCF diet. Bacterial populations at the species level are presented in Supplementary Table S1 (see online version of the article at


Functional food ingredients that are becoming increasingly popular include low digestible carbohydrates that induce prebiotic effects. Prebiotics are defined as nondigestible food ingredients that, when consumed in sufficient amounts, selectively stimulate the growth, activity, or both of one or a limited number of microbial genera or species in the gut microbiota that ultimately benefits health of the host (Tremaroli and Backhed, 2012). Common prebiotic fibers present in companion animal and human foods include fructooligosaccharides, inulin, and resistant starch (Tomasik and Tomasik, 2003). Low digestible carbohydrates often are low molecular weight and resist mammalian hydrolytic/enzymatic digestion and will enter the large bowel to be fermented by microbes to produce SCFA and lower digesta pH (Mussatto and Mancilha, 2007). They are similar to dietary fiber in that they have the ability to provide health-promoting effects on the host (Knapp et al., 2010).

Soluble corn fiber contained 78% total dietary fiber, all of which was soluble fiber, with 55% in a low molecular weight form. Soluble corn fiber is a purified fiber source having no or very low concentrations of ash, CP, acid hydrolyzed fat, and free sugars and a moderate amount of both digestible and type 4 resistant starch. The SCF used in this study was a soluble fiber dextrin derived from corn starch that is considered a low digestible carbohydrate due to its high proportion of low molecular weight soluble fiber. Normally, corn starch is made up of α-1,4 and α-1,6 glycosidic bonds that are easily degraded by mammalian pancreatic amylase. The dextrinization process uses heat and acid to hydrolyze the α-glycosidic bonds. Upon cooling, the reformation of both digestible glycosidic bonds (α-1,4 and α-1,6) as well as nondigestible glycosidic bonds (β-1,4, β-1,6, β-1,3, and β-1,2) make up the short-chain oligosaccharides that then can enter the large bowel for fermentation by the resident microbiota (Knapp et al., 2010). Previous studies have shown that SCF is a good candidate for low caloric and low glycemic dog diets (de Godoy et al., 2013), but no information was available regarding prebiotic potential of SCF fed to dogs.

The in vitro hydrolytic–enzymatic digestion experiment suggested that SCF was only 7% digestible, leaving 93% of the substrate available for fermentation. The SCF was highly fermentable throughout the entire 12 h fermentation, exhibiting increases in SCFA concentrations at each time point. Wheat dextrin soluble fiber (NUTRIOSE FB06; Roquette Frères, Lestrem, France) was tested for in vitro fermentation properties using human fecal inoculum (Hobden et al., 2013). In that study, acetate, propionate, and butyrate were reported to increase in the distal portion of the gut model compared with the proximal portion, indicating that the wheat dextrin soluble fiber substrate was potentially fermentable throughout the gastrointestinal tract. Furthermore, wheat dextrin soluble fiber modulated the microbiota in the model, with increases in butyrate-producing taxa (Hobden et al., 2013). Knapp et al. (2010) determined that a variety of soluble fiber dextrins, including those derived from corn starch, may be potential substrates for hindgut fermentation due to their ability to resist in vitro hydrolytic–enzymatic digestion. This was further supported by low glycemic responses in dogs (Knapp et al., 2010).

Previous studies that have determined TMEn values of various SCF substrates that were obtained by different methods of starch hydrolysis reported values as low as 1.7 kcal (Knapp et al., 2010) and as high as 3.0 kcal (de Godoy et al., 2014). Our TMEn value of 2.6 kcal/g is accurate because the SCF used in this study was treated with hydrochloric acid and was consistent with the TMEn value obtained previously using the same processing method (2.4 kcal/g; de Godoy et al., 2014). The variation in TMEn values of different SCF substrates has been attributed to differences in processing methods and molecular structures of the carbohydrates (de Godoy et al., 2014).

Nutrient digestibility was not affected by SCF inclusion in this study. There were no significant changes in fecal SCFA concentrations, fecal pH, or fecal consistency relative to the 0% SCF diet. Similarly, we observed no changes in markers of protein fermentation as evidenced by a lack of change in fecal BCFA, phenolic and indolic compounds, and biogenic amines with increasing SCF supplementation. Higher dietary concentrations of SCF than those used here may be necessary to affect these outcomes. Dogs may also have a lower ability to ferment the SCF substrate compared to humans, perhaps due to differences in abundance of select microbial taxa (i.e., Firmicutes, Fusobacteria, Bacteroidetes, Proteobacteria, and Actinobacteria), volume of the large bowel, and anatomical/physiological differences, such as sacculations and transit time.

Previous in vivo studies in rodent and human models investigating dietary SCF determined this substrate to be highly fermentable (defined by SCFA concentration), modulatory of the microbiome, and overall favorable to indices of gut health. The objectives of these studies, however, were to determine the efficacy of dietary SCF as a fermentable fiber source. Therefore, the concentrations of SCF added to the diet were high to elicit these responses. Normally, prebiotic fibers are added at low concentrations in the diet to increase SCFA production and stimulate the growth of potentially beneficial bacteria. The objective of this current study was to assess the prebiotic potential of SCF, which entails adding graded concentrations of SCF of only up to 1.25% of the diet. The previous in vivo studies in humans and rodents have shown the potential health benefits of SCF as a fiber source at higher dietary concentrations. Dietary SCF concentrations of 5% or higher in rats has been found to improve cecal and colonic fermentation characteristics as well as indices of gut health (i.e., increased crypt depth, goblet cell numbers, and acidic mucins; Guerin-Deremaux et al., 2010; Knapp et al., 2013). Similarly, in humans, consumption of SCF at 20 g/d resulted in a decrease in colonic pH, suggesting increased fermentative activity (Lefranc-Millot et al., 2012).

In recent years, the development of novel high-throughput sequencing techniques has led to a more comprehensive understanding of the microbial populations present in the colon. Specifically, the effect of diet on the microbial populations, as well as their functional capacity to metabolize nutrients, can be measured using these techniques. Very limited research using these techniques has been conducted with dietary SCF. In humans, consumption of 21 g/d of SCF elicited increases in select butyrate-producing taxa (i.e., Fecalibacterium spp. and Faecalibacterium prausnitzii) as well as increases in lactobacilli (Hooda et al., 2012).

In our study, there were no significant differences in diversity of gut bacteria among diets and there was no clear clustering by diet as indicated by principal component analysis. The predominant bacterial phyla present in feces of dogs fed all diets in this study were Firmicutes and Fusobacteria. Previously published data showed that a normal dog fecal microbiome was variable, with studies showing 14 to 48% Firmicutes and 7 to 40% Fusobacteria (Middelbos et al., 2010; Suchodolski et al., 2008; Swanson et al., 2011). At the microbial genus level, there was only slight modulation of the fecal microbiome with increasing dietary SCF. Dogs fed the 1% SCF diet showed increases in Lachnospira, which is a part of the butyrate-producing superfamily Lachnospiraceae (Marounek and Dušková, 1999). However, this did not translate into increases in fecal butyrate concentrations.

Previous studies in humans have suggested that SCF and other fibers similar in chemical composition (e.g., wheat dextrin soluble fiber) are fermentable in vitro and in vivo and can beneficially modulate the microbiome, but these effects were observed only at dietary concentrations well above the highest concentration used in the present study (Pasman et al., 2006; Lefranc-Millot et al., 2012; Hobden et al., 2013). The integration of both in vitro fermentation and in vivo dog data suggests that SCF elicits some modulation of the microbiome; however, the doses provided were insufficient to induce a robust response.

Soluble corn fiber added at concentrations similar to proven prebiotics did not elicit the same effects on SCFA concentration, pH decline, or shifts in the microbial populations in dogs. Prebiotic fibers such as fructooligosaccharides and galactooligosaccharides are effective at all of the dietary concentrations tested in this experiment, putting this particular novel fiber at a disadvantage because of the higher concentrations that ostensibly would be required to elicit an effect. Overall, establishing an effective dose of SCF to elicit effects of increased SCFA concentrations, modulation of the microbiome, and other indices of gut health is needed.





Be the first to comment.

Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.