SEARCH

SEARCH BY CITATION

Keywords:

  • cholesterol removal;
  • inoculum size;
  • optimization;
  • prebiotic;
  • response surface methodology

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Aims:  To optimize cholesterol removal by Lactobacillus acidophilus ATCC 4962 in the presence of prebiotics, and study the growth and fermentation patterns of the prebiotics.

Methods and Results: Lactobacillus acidophilus ATCC 4962 was screened in the presence of six prebiotics, namely sorbitol, mannitol, maltodextrin, hi-amylose maize, fructo-oligosaccharide (FOS) and inulin in order to determine the best combination for highest level of cholesterol removal. The first-order model showed that the combination of inoculum size, mannitol, FOS and inulin was best for removal of cholesterol. The second-order polynomial regression model estimated the optimum condition of the factors for cholesterol removal by L. acidophilus ATCC 4962 to be 2·64% w/v inoculum size, 4·13% w/v mannitol, 3·29% w/v FOS and 5·81% w/v inulin. Analyses of growth, mean doubling time and short-chain fatty acid (SCFA) production using quadratic models indicated that cholesterol removal and the production of SCFA were growth associated.

Conclusions:  Optimum cholesterol removal was obtained from the fermentation of L. acidophilus ATCC 4962 in the presence of mannitol, FOS and inulin. Cholesterol removal and the production of SCFA appeared to be growth associated and highly influenced by the prebiotics.

Significance and Impact of the Study:  Response surface methodology proved reliable in developing the model, optimizing factors and analysing interaction effects. The results provide better understanding on the interactions between probiotic and prebiotics for the removal of cholesterol.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Interest in the usage of probiotics for human health dated back to 1908 when Metcnikoff suggested that man should consume milk fermented with lactobacilli to prolong life (O'Sullivan et al. 1992). More recently, probiotics have been defined as ‘cultures of live micro-organisms that, applied in animals or humans, benefit the host by improving properties of indigenous microflora’ (Arihara and Itoh 2000). They mainly consist of lactobacilli, streptococci, enterococci, lactococci and bifidobacteria. Over the years, lactobacilli have been associated with the improvement of lactose intolerance, increase in natural resistance to infectious disease in gastrointestinal tract, suppression of cancer, improved digestion and reduction in serum cholesterol level (Gibson and Roberfroid 1995). For hypercholesterolaemic individuals, significant reductions in plasma cholesterol levels are associated with a significant reduction in the risk of heart attacks (Lourens-Hattingh and Viljoen 2001). Various studies reported that lactobacilli could lower total cholesterol and low-density lipoprotein (LDL) cholesterol (Anderson and Gilliland 1999; Sanders 2000).

Prebiotics are defined as nondigestable substances that exert biological effect on humans by selectively stimulating the growth or bioactivity of beneficial micro-organisms either present, or therapeutically introduced to the intestine (Tomasik and Tomasik 2003). Several nonstarchy polysaccharides such as fructo-oligosaccharides, lactulose and β-cyclodextrin have been considered to have prebiotic properties. Recently, polyols such as mannitol, sorbitol and xylitol have been included to the prebiotics group (Klahorst 2000). Prebiotics have been linked with cholesterol-reducing effects. It was previously found that hepatocytes isolated from oligofructose-fed rats had a slightly lower capacity to synthesize triacylglycerol from radiolabelled acetate. This led to the hypothesis that decreased de novo lipogenesis in the liver, through lipogenic enzymes, is the key to reduction of VLDL-triglyceride secretion in rats fed with oligosaccharides (Robertfroid and Delzenne 1998). Administration of oligofructose was postulated to modify lipogenic enzyme gene expression, observed by a 50% reduction of activity of acetyl-CoA carboxylase, malic enzyme and ATP citrate lyase (Delzenne and Kok 2001).

Probiotics and prebiotics simultaneously present in a product are called either synbiotics or eubiotics. Such a combination aids survival of the administered probiotic and facilitates its inoculation into the colon. Additionally, the prebiotic induces growth and increases activity of positive endogenic intestinal flora (Tomasik and Tomasik 2003). Experiments with rats showed that synbiotics protect the organism from carcinogens significantly better than either prebiotics or probiotics individually (Gallaher and Khil 1999). However, there is little information on suitable combinations of probiotics and prebiotics specifically targeting removal or lowering of cholesterol.

Response surface methodology (RSM) is a collection of statistical and mathematical techniques useful for developing, improving and optimizing processes. It also has important applications in design, development and formulation of new products, as well as improvement of existing product designs (Myers and Montgomery 1995). Response surface models may involve main effects and interactions or have quadratic and possibly cubic terms to account for curvature. It has been successfully utilized to optimize compositions of microbiological media (Oh et al. 1995), improving fermentation processes (Lee and Chen 1997) and product development (Gomes and Malcata 1998). Conventional methods (such as one factor at one time) have been applied previously to evaluate the in vitro performance of probiotics and/or prebiotics to remove cholesterol. However, these methods require a large number of experiments to describe the effect of individual factors, were time consuming, and no statistical method was established to distinguish the interaction effects from main effects. Thus, the aim of this study was to optimize cholesterol removal by using L. acidophilus ATCC 4962 in the presence of mannitol, FOS and inulin, through the approach of response surface.

Bacteria and media preparation

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Lactobacillus acidophilus ATCC 4962 is a human-derived strain that was obtained from the Australia Starter Culture Collection Center (ATCC) (Werribee, Australia). All stock cultures were stored in 40% glycerol at −80°C, and transferred successively three times in sterile de Mann, Rogosa, Sharpe (MRS) broth using 1% inoculum and 20 h incubation at 37°C prior to use. The culture was then centrifuged at 4°C for 15 min at 2714 g (Sorvall RT7, Newtown, CT, USA). The supernatant was discarded while the pellet was washed twice with sterile distilled water, resuspended by vortexing in 50 ml of 0·1 m phosphate buffer (pH 6·8), and recentrifuged at 2714 g at 4°C for 15 min. After discarding the supernantant, 50 ml of 0·1 m phosphate buffer (pH 6·8) containing 2·0% (w/v) of food grade cryoprotectant UnipectinTM RS 150 (Savannah Bio Systems, Balwyn East, Australia) was added to the pellet. The mixture was vortexed, poured into large Petri dishes and freeze-dried (Dynavac FD300, Airvac Engineering Pty. Ltd, Rowville, Australia) at −88°C for 40 h for primary freezing and 8 h for secondary freezing. After freeze-drying, the hygroscopic cultures were transferred into sterile sealed bags and stored at −18°C until used. Six types of commercially available prebiotics were used, namely sorbitol (Sigma Chemical Co., St Louis, MO, USA), mannitol (Sigma), maltodextrin (Grain Processing Corp., Muscatine, IA, USA), hi-amylose maize (Starch Australasia Ltd, Lane Cove, NSW, Australia), inulin (Orafti Pty. Ltd, Tienen, Belgium) and FOS (Orafti). FOS used was Raftilose P95 that contained 5% of glucose, fructose and sucrose. It contained oligofructose with degree of polymerization (DP) ranging from 2 to 7, with an average DP of 4. Inulin used was Raftiline ST with a purity of 92%, an average DP of 10. Hi-amylose maize contained >70% amylose, and 32·5% total dietary fibre.

All prebiotics were used at concentrations as per the experimental design (Table 1). Prebiotics were prepared in phosphate buffer (0·1 m, pH 6·0) containing ammonium citrate (2·0 g l−1), sodium acetate (5·0 g l−1), magnesium sulphate (0·1 g l−1), manganese sulphate (0·05 g l−1), dipotassium phosphate (2·0 g l−1) and Tween-80 (1·0 ml l−1). Freeze-dried cells of L. acidophilus ATCC 4962 were inoculated at appropriate levels as described in the experimental design.

Table 1.  Treatment combinations and response for screening experiments
Standard orderFactors*Response: cholesterol assimilated (μg ml−1)
Inoculum size (% w/v)Sorbitol (% w/v)Mannitol (% w/v)FOS (% w/v)Hi-maize (% w/v)Inulin (% w/v)Maltodextrin (% w/v)
  1. *Inoculum size: 0·10–0·30% w/v; sorbitol: 0·50–1·50% w/v; mannitol: 0·50–1·50% w/v; maltodextrin: 0·50–1·50% w/v; hi-amylose maize: 0·50–1·50% w/v; FOS: 0·50–1·50% w/v; inulin: 0·50–1·50% w/v.

1−1−1−1−1−11131·36
31−1−1−1−1−1−133·13
5−11−1−1−1−1−125·52
711−1−1−11136·09
9−1−11−1−1−1127·71
111−11−1−11−139·17
13−111−1−11−132·53
15111−1−1−1136·15
17−1−1−11−1−1−127·50
191−1−11−11139·01
21−11−11−11131·51
2311−11−1−1−134·90
25−1−111−11−134·58
271−111−1−1136·15
29−1111−1−1130·64
311111−11−139·58
33−1−1−1−111−128·70
351−1−1−11−1134·22
37−11−1−11−1126·30
3911−1−111−136·20
41−1−11−11−1−128·49
431−11−111138·54
45−111−111131·09
47111−11−1−134·01
49−1−1−111−1125·25
511−1−1111−138·23
53−11−1111−130·16
5511−111−1135·73
57−1−11111133·59
591−1111−1−136·82
61−11111−1−131·61
63111111140·52
65000000032·81
66000000031·98
67000000033·02
68000000031·88
69000000033·96

Cholesterol removal

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Freshly prepared media containing prebiotics were added with water-soluble filter-sterilized cholesterol (polyoxyethanyl-cholesteryl sebacate), at a final concentration of 70–100 μg ml−1, inoculated with appropriate levels of freeze-dried L. acidophilus ATCC 4962 (Table 1), and incubated anaerobically at 37°C for 48 h. After the incubation period, cells were centrifuged and the remaining cholesterol concentration in the spent broth was determined using the o-phtaldealdehyde colorimetric method as described previously (Rudel and Morris 1973).

Growth of L. acidophilus ATCC 4962 in the presence of prebiotics

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

The growth was determined using the plate count method. Bacilli generally divide in one plane, and can produce chains of cells due to the failure to separate completely. Thus, at the end of the fermentation time, fermentation broth containing probiotic cultures sonicated for 5 s to disrupt clumps of lactobacilli (Bermudez et al. 2001) before serial dilutions were performed. Subsequent serial dilution blanks were vortexed for 30 s individually. One millilitre sample was taken after the incubation period, and 10-fold serial dilutions were made using peptone water diluent. MRS agar was used for plating and the plates were incubated anaerobically at 37°C for 24 h in an anaerobic jar (Becton Dickinson Microbiology Systems®, Sparks, MD, USA) with a Gas Generating Kit® (Oxoid). Growth was calculated as log 10 colony-forming units (CFU ml−1) and expressed as percentage difference between initial growth values obtained at time = 0 and at the end of the incubation period.

Mean doubling time

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Mean doubling time was calculated as described previously (Shin et al. 2000). The specific growth rate (μ) of the cultures was obtained using the following equation:

  • image

where X2 and X1 are the cell density at time t2 and t1 respectively. Mean doubling time (Td) was calculated as Td =  ln 2/μ, and expressed in minutes.

Short-chain fatty acids determination

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

The fermentation of prebiotics was determined by measuring short-chain fatty acids (SCFA) as the end products of fermentation using high-performance liquid chromatography (HPLC, Varian Australia Pty. Ltd, Mulgrave, Australia). At the end of the incubation period, fermentation broths containing L. acidophilus ATCC 4962 and the prebiotics used were centrifuged at 2714 g at 4°C for 15 min, and the supernatant was prepared for HPLC analysis using the method as described previously (Dubey and Mistry 1996). Briefly, 5 ml of supernatant was added to 100 μl of 15·5 n HNO3 and 5 ml of 0·009 n H2SO4. The mixture was vortexed for 10 s and recentrifuged at 14 000 g for 10 min. The supernatant was filtered (0·20 μm) and stored at 4°C until analysed. SCFA was expressed as the total acetic, butyric and propionic acids.

Experimental design and statistical analyses

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Screening experiments to select prebiotics were performed with seven independent factors namely, inoculum size of L. acidophilus ATCC 4962 (X1), sorbitol (X2), mannitol (X3), maltodextrin (X4), hi-amylose maize (X5), inulin (X6) and FOS (X7), using a two-level partial factorial design 27−2 resulting in 64 experimental runs (including duplicates) and five middle point runs. The units and the coded levels of the independent factors are shown in Table 1. First-order empirical equation was used to exclude insignificant factors and to generate steepest ascent. Optimization was performed using a rotatable central composite design (CCD) with an alpha value of ±2·00 for four factors. The treatment combinations of CCD were allocated in two blocks, with the first block representing the first day of the experiment and contained all factorial runs accompanied by four centre runs. The second block, representing the second day of the experiment, contained all axial runs accompanied by two centre runs. These modelling and statistical analyses were performed using the Design Expert version 5·07 software (Stat-Ease Corp., Minneapolis, MN, USA). All data presented are mean values of triplicate experiments, n = 3.

Screening of factors and steepest ascent

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

The results from the two-level partial factorial design are shown in Table 1, while analysis of variance (anova) for the evaluation of the first-order model is shown in Table 2. anova showed that the model used was suitable, lack-of-fit test was insignificant with only 9·58% total variation that was not explained by the model. Removal of cholesterol was significantly influenced by inoculum size of L. acidophilus ATCC 4962 (X1), mannitol (X3), FOS (X6) and inulin (X7), while the other prebiotics were found to have insignificant influence and were not included in the anova. Thus, further optimization processes will only involve these four factors. A first-order equation (coded term) was generated from this first-degree order model, for response of cholesterol removal (Y), with the significant factors now redefined as inoculum size (X1), mannitol (X2), FOS (X3) and inulin (X4):

  • image
Table 2.  Analysis of variance and coefficient estimates for the evaluation of the first-order model
Source of variationSum of squaresd.f.*Mean squareF-valueP-value
Model†1115·3714278·84148·73<0·0001
Curvature1·4111·410·750·3890
Residual118·11631·87  
Lack-of-fit49·73271·840·970·5269
Pure error68·38361·90  
Correlation total1234·9068   
FactorCoefficient estimated.f.s.e.t-ValueP-value
  1. *d.f., degree of freedom.

  2. R2 = 0·9042.

  3. ‡Significant at alpha = 0·05.

Inoculum size (X1)3·5010·1720·430·0001‡
Mannitol (X3)1·1710·176·830·0001‡
FOS (X6)0·8310·174·850·0001‡
Inulin (X7)1·7710·1710·360·0001‡

From the equation and coefficient estimate, inoculum size (X1) produced greatest effect and was used as the fundamental scale in the next step, steepest ascent. In this study, the steepest ascent design was based on the increase of 0·50 (% w/v) concentrations for X1. This produced five design units (0·50/0·10 = 5). Thus, movement for X2 was 1·67 design units [(1·17/3·50)(5) = 1·67], for X3 was 1·19 design units [(0·83/3·50)(5) = 1·19] and for X4 was 2·53 design units [(1·17/3·50)(5) = 2·53]. The following steepest ascent coordinates were generated as shown in Table 3. Steepest ascent coordinates showed that removal of cholesterol decreased after the fifth step, with highest value of 50·938 μg ml−1, from the combination of inoculum size (2·20% w/v), mannitol (4·36% w/v), FOS (3·40% w/v) and inulin (6·08% w/v). This combination was used as the middle point for optimization experiments.

Table 3.  Coordination path of steepest ascent for all chosen factors in coded and natural levels
StepCoded factors*Natural factors†Cholesterol removed (μg ml−1)
ξ1ξ3ξ6ξ7X1X2X3X4
  1. *ξ1: inoculum size (% w/v), ξ3: mannitol (% w/v), ξ6: FOS (% w/v); ξ7: inulin (% w/v).

  2. X1: inoculum size (% w/v), X2: mannitol (% w/v), X3: FOS (% w/v); X4: inulin (% w/v).

1: Base0 00 00·201·001·001·0016·478
Δ51·671·192·53(5)(0·1) = 0·5(1·67)(0·50) = 0·84(1·19)(0·50) = 0·60(2·53)(0·50) = 1·27 
2: Base + Δ51·671·192·530·701·841·602·2736·563
3: Base + 2Δ103·342·385·061·202·682·203·5444·375
4: Base + 3Δ155·013·577·591·703·522·804·8150·781
5: Base + 4Δ206·684·7610·122·204·363·406·0850·938
6: Base + 5Δ258·355·9512·652·705·204·007·3548·813
7: Base + 6Δ3010·027·1415·183·206·044·608·6247·497

Optimization of cholesterol removal

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Optimization was performed using CCD with fixed middle point of inoculum size (2·20% w/v), mannitol (4·30% w/v), FOS (3·40% w/v) and inulin (6·00% w/v). Design matrix for CCD and responses are shown in Table 4, while the adequacy and fitness were evaluated by anova and regression coefficients (Table 5). anova results indicated that the quadratic regression to produce the second-order model was significant. Lack-of-fit test was insignificant and a good coefficient regression was obtained. Inoculum size, mannitol, FOS and inulin significantly influenced cholesterol removal.

Table 4.  Combination matrix of the central composite design using coded levels for the response of cholesterol removal
Standard runBlock*FactorsCholesterol removal (μg ml−1)†
Inoculum size (X1)Mannitol (X2)FOS (X3)Inulin (X4)
  1. *1, first day of experiment; 2, second day of experiment.

  2. †All factorial and axial points are mean values of duplicates.

 11−1−1−1−130·367
 211−1−1−146·304
 31−11−1−129·586
 4111−1−141·461
 51−1−11−126·461
 611−11−142·086
 71−111−131·929
 81111−147·086
 91−1−1−1128·023
1011−1−1140·367
111−11−1123·648
12111−1139·117
131−1−11118·023
1411−11138·179
151−111124·273
161111134·351
171000053·179
181000063·648
191000056·304
201000060·054
212−200015·211
222200033·414
2320−20032·164
242020023·804
25200−2034·586
262002024·976
272000−225·523
282000235·836
292000060·836
302000050·523
Table 5.  Analysis of variance of the second-order model* and coefficient estimates for the response Y0 and factors X1, X2, X3 and X4
SourceSum of squaresd.f.Mean squareF-valueP-value
Model†4302·4214307·3210·780·0001
Residual399·171428·51  
Lack-of-fit284·111028·410·990·5541
Pure error115·07428·77  
Total4870·6029   
Factor‡Coefficient estimated.f.s.e. t-Value P-value
  1. *Y0 = 56·58 + 6·38X1 − 0·63X2 − 1·49X3 − 1·19X4 − 7·34X12−6·42X22 − 5·97X32− 5·75X42 − 0·72X1X2 + 0·34X1X3− 0·034X1X4 + 1·51X2X3− 0·50X2X4− 1·01X3X4.

  2. R2 = 0·9540.

  3. X1: inoculum size (% w/v), X2: mannitol (% w/v), X3: FOS (% w/v), X4: inulin (% w/v).

  4. §Significant at alpha = 0·05.

Interceptc = 56·5812·21  
X1c1 = 6·3811·095·850·0001§
X2c2 = −0·6311·09−0·580·5735
X3c3 = −1·4911·09−1·360·1938
X4c4 = −1·1911·09−1·100·2915
inline imagec11 = −7·3411·02−7·200·0001§
inline imagec22 = −6·4211·02−6·300·0001§
inline imagec33 = −5·9711·02−5·860·0001§
inline imagec44 = −5·7511·02−5·640·0001§
X1X2c12 = −0·7211·33−0·540·5993
X1X3c13 = 0·3411·330·2500·8044
X1X4c14 = −0·03411·33−0·0260·9799
X2X3c23 = 1·5111·331·130·2774
X2X4c24 = −0·5011·33−0·380·7120
X3X4c34 = −1·0111·33−0·760·4615

The effect of each factors were further assessed using perturbation plots, to show how the response changes as each factor moves from the chosen reference point, with all other factors held constant at reference values (Oh et al. 1995). In this study, as one particular chosen factor was assessed, the other factors were held constant at the optimum point. Figure 1 shows the perturbation plot of the factors used in this study. Although all factor showed significant quadratic effect, the curve with the most prominent change was the perturbation curve of inoculum size, compared with the other factors that were fixed at their maximum levels. Thus, we believe that inoculum size was the most significant factor that contributed to the removal of cholesterol with the most obvious quadratic effect. Although the P-values of both FOS and inulin showed similar levels of significance, it could be clearly seen from the perturbation plot that the response curve of inulin was less prominent than that of FOS.

image

Figure 1. Perturbation plot of inoculum size (A), mannitol (B), FOS (C) and inulin (D)

Download figure to PowerPoint

The best explanatory equation to fit the second-order model and subsequently produce the response surface was expressed as:

  • image

where cc23 are regression coefficients and X1, X2, X3, X4 are the coded independent factors. Here, the second-order regression model involved four factors, thus producing four linear, four quadratic and six interaction terms. Response surface was generated (Fig. 2) based on the second-order equation:

  • image
image

Figure 2. Response surface for cholesterol removal (μg ml−1) from the effects of (a) FOS and mannitol and (b) inoculum size and inulin. Factors that were not included in the axes were fixed at their respective optimum levels

Download figure to PowerPoint

An optimum point was produced with optimum cholesterol removal obtained at 58·142 μg ml−1. The combination that produced the optimum point was (X1, X2, X3, X4) =(0·437, −0·082, −0·115, −0·092). The original levels that correlated with those coded values were found to be inoculum size at 2·64% w/v, mannitol at 4·14% w/v, FOS at 3·28% w/v and inulin at 5·82% w/v.

All these predictions by the regression model were further ascertained by a validation experiment. We compared the cholesterol removal patterns over a 24-h period using four different media: the optimum medium (inoculum size: 2·60% w/v; mannitol: 4·10% w/v; FOS: 3·30% w/v; inulin: 5·80% w/v), the centre-point medium (inoculum size: 2·20% w/v; mannitol: 4·30% w/v; FOS: 3·40% w/v; inulin: 6·00% w/v), the high-point medium (inoculum size: 3·20% w/v; mannitol: 6·30% w/v; FOS: 4·40% w/v; inulin: 8·00% w/v) and the low-point medium (inoculum size: 1·20% w/v; mannitol: 2·30% w/v; FOS: 2·40% w/v; inulin: 4·00% w/v). The cholesterol removal curves are shown in Fig. 3. Although the exact cholesterol removal quantities were different from the predictions, the patterns were in tandem with predictions by the model. Highest cholesterol was removed from the optimum medium, and lower from the centre-point medium. Least cholesterol was removed from both high-point and low-point media, as supported by the response surface of cholesterol removal (Fig. 2).

image

Figure 3. Cholesterol removal by Lactobacillus acidophilus ATCC 4962 in the optimum (bsl00001), centre-point (•), high-point (bsl00066) and low-point (bsl00063) media, for the validation experiments. Factors combination for optimum medium were: inoculum size 2·60% w/v, mannitol 4·10% w/v, FOS 3·30% w/v and inulin 5·80% w/v. Centre-point medium were: inoculum size 2·20% w/v, mannitol 4·30% w/v, FOS 3·400% w/v and inulin 6·00% w/v. High-point medium were: inoculum size 3·20% w/v, mannitol 6·30% w/v, FOS 4·40% w/v and inulin 8·00% w/v, and low-point medium were inoculum size 1·20% w/v, mannitol 2·30% w/v, FOS 2·40% w/v and inulin 4·00% w/v. Error bars represent s.e.m., n = 3

Download figure to PowerPoint

Growth, mean doubling time and production of SCFA

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

We further studied patterns of growth, mean doubling time and production of SCFA from the fermentation of prebiotics, at the experimental regions used to obtain optimum removal of cholesterol. The response obtained using the CCD is shown in Table 6. The statistical analyses with coefficient estimates and the significance of each response model are presented in Table 7.

Table 6.  Combination matrix of the central composite design using coded levels for the factors and five responses
Standard runBlock*Factors†Responses‡
X1X2X3X4Y1Y2Y3
  1. *1, first day of experiment; 2, second day of experiment.

  2. X1 = inoculum size, X2 = mannitol, X3 = FOS, X3 = inulin.

  3. Y1 = growth (%), Y2 = mean doubling time (min), Y3 = short-chain fatty acid (mmol l−1).

 11−1−1−1−139·629288·6776·308
 211−1−1−135·996290·79713·064
 31−11−1−138·381288·3038·220
 4111−1−133·925290·64916·503
 51−1−11−128·365284·4065·992
 611−11−135·774288·43516·711
 71−111−130·550286·9898·915
 81111−136·249290·79115·324
 91−1−1−1128·398287·9015·131
1011−1−1132·935288·41824·531
111−11−1123·948285·81311·966
12111−1132·318288·53017·959
131−1−11120·730286·9117·239
1411−11132·278288·57935·922
151−111124·742286·8407·448
161111131·398291·75062·947
171000038·706290·24367·026
181000048·981291·17553·419
191000038·739290·37246·826
201000042·216291·50567·139
212−200019·677284·73436·543
222200031·106292·09145·701
2320−20024·825292·16931·714
242020024·734293·19522·015
25200−2032·519291·31023·119
262002027·326290·10222·252
272000−246·054290·71616·866
282000231·942289·10815·285
292000045·946290·79144·787
302000038·688291·46572·814
Table 7.  Regression coefficients of the second-order equation* for the five responses†
CoefficientY1Y2Y3
  1. *Y = c + c1X1 + c2X2 + c3X3 + c11X12 + c22X22 + c33X32c12X1X2 + c13X1X3 + c23X2X3.

  2. Y1 = growth (%), Y2 = mean doubling time (min), Y5 = short-chain fatty acid (mmol l−1).

  3. ‡Significant at alpha = 0·05.

c41·97291·2160·03
c12·46‡1·53‡6·67‡
c2−0·120·320·62
c3−1·49‡−0·282·30
c4−3·35‡−0·313·29
c11−3·90‡−0·97‡−6·08‡
c22−4·05‡0·095−9·65‡
c33−2·77‡−0·40‡−10·69‡
c44−0·50−0·60‡−12·34‡
c12−0·220·340·66
c131·66‡0·42‡3·80
c141·63‡−0·164·84
c230·890·66‡1·45
c24−0·08−0·211·29
c340·530·70‡3·20
R20·91730·93770·8448
P-value0·00010·00010·0016

The response surface of growth (Y1) is shown in Fig. 4, and was generated based on the following coded factor equation:

  • image
image

Figure 4. Response surface for growth (%) from the effects of (a) FOS and mannitol and (b) inoculum size and inulin. Factors that were not included in the axes were fixed at their respective optimum levels

Download figure to PowerPoint

The response surface clearly indicated that an optimum point (45·21%) was produced with X1, X2, X3 and X4 at 2·23, 4·21, 3·04 and 4·00% w/v respectively. Growth increased with increasing inoculum size level from 1·20 to 2·23% w/v. Further increase in concentrations of inoculum size beyond 1·69% w/v generated a decrease in growth. Similarly, increasing concentrations of mannitol and FOS from 2·30 to 4·21% w/v and 2·40 to 3·04% w/v, respectively, increased growth, but further increase in the prebiotics concentration generated a decrease in growth. Inulin produced highest growth at its lowest concentration of 4·00% w/v, and produced lowest growth at its highest concentration of 8·00% w/v. It appeared that growth of L. acidophilus ATCC 4962 was influenced by inulin in a linear manner, while inoculum size, mannitol and FOS showed significant quadratic effects. Other than main quadratic effects, interactions between inoculum size and FOS, and inoculum size and inulin produced strongest influence towards growth, while the other interactions were insignificant.

In this study, patterns of mean doubling time (Y2) were studied using the response surface (Fig. 5) that was generated from the equation:

  • image
image

Figure 5. Response surface for mean doubling time (min) from the effects of (a) inoculum size and FOS and (b) FOS and inulin. Factors that were not included in the axes were fixed at their respective optimum levels

Download figure to PowerPoint

Inoculum size, FOS and inulin showed significant quadratic effect, while mannitol did not (Table 7). FOS mainly contributed to the interaction effects, with only interaction terms involving FOS showed significant influence on mean doubling time. All these significant interaction terms also showed positive regression coefficients, indicating that either a decrease or increase in both factors will contribute to an increase in mean doubling times.

The SCFA (Y3) was obtained as a total of individual fatty acids, namely acetic, butyric and propionic acids. A response surface (Fig. 6) was generated from the second-order equation:

  • image
image

Figure 6. Response surface for the production of short-chain fatty acid (SCFA, mmol l−1) from the effects of (a) FOS and mannitol and (b) inoculum size and inulin. Factors that were not included in the axes were fixed at their respective optimum levels

Download figure to PowerPoint

All factors produced significant quadratic effects on production of SCFA. Response surfaces produced showed that the production of SCFA appeared to be growth associated.

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References

Various factors normally affect the response surfaces that are produced. Thus, screening experiments are needed to segregate important main effects from less important ones (Montgomery 1996). In this study, first degree order equation was generated and significance of factors was tested using screening experiments. A complete replication of the seven factors using a 2x factorial design would need 128 experimental runs. However, only seven degree of freedoms would be needed to estimate main effects, and 21 degree of freedoms would estimate two-factor interaction effects, while the remaining 99 degree of freedoms would estimate error and/or three or higher factor interaction effects (Cox and Reid 2000). Thus, a partial two-level factorial design (27−2) was applied in this study. Partial factorial designs are capable of identifying important factors using less number of experimental runs without loss of information on main factor effects and their interactions (Li et al. 2002). Following the screening of significant factors, design points were subjected to steepest ascent before subsequent optimization steps. Steepest ascent or steepest descent involved the generation of mathematical movements along an ascending or descending path until no improvement occurred (Montgomery 1996).

A significant quadratic regression, insignificant lack-of-fit and a small total variation (4·60%) that was not explained by the model, suggested that the model accurately represented data in the experimental region. This also indicated that second-order terms were sufficient and higher-order terms were not necessary (Oh et al. 1995). It must also be noted that the t value of the quadratic term of inoculum size (X12) was higher than others (Table 5), indicating that the quadratic effect of inoculum size had the strongest effect on cholesterol removal, which was also confirmed using the perturbation plot. Validation experiments showed that the predicted value was 58·142 μg ml−1 while the actual experimental result was 52·941 μg ml−1. However, it must be noted that the conditions for both were slightly different. The predicted value was obtained at the predicted 2·64% w/v inoculum size, 4·14% w/v mannitol, 3·28% w/v FOS and 5·82% w/v inulin, while the actual experiments were conducted with 2·60% w/v inoculum size, 4·10% w/v mannitol, 3·30% w/v FOS and 5·80% w/v inulin. Under such dissimilarity, the difference between the prediction and actual data was only 8·95%. The obvious difference of cholesterol removal between the optimum, high-point, low-point and centre-point media proved the validity of the model and the reproducibility of the prediction.

From Table 5, it must be noted that the coefficient estimates of the interaction terms of (X2, X4) and (X3, X4) had negative signs (X24 = −0·50, X34 = −1·01). These negative signs may imply that for an increase of the response, the coded levels of (X2, X4) and (X3, X4) must have different signs, either one must be higher than zero and the other lower than zero (Oh et al. 1995). However, it must be noted that the optimum was achieved at (X2 = −0·082, X4 = −0·092) and (X3 = −0·115, X4 = −0·092), which would produce a positive sign instead. This may be due to other terms that may dominate this particular interaction term (Oh et al. 1995). Considering that the lack-of-fit test was insignificant, other higher terms would not have contributed to this, thus, we postulate that the linear term might have played a role.

The response surface of growth showed similar patterns with the response surface of removal of cholesterol, indicating a strong correlation between removal of cholesterol and growth. Previous studies also showed that cholesterol assimilation by strains of L. acidophilus during refrigerated storage of nonfermented milk was associated with bacterial growth and their viability, and was growth dependent (Piston and Gilliland 1994; Pereira and Gibson 2002). This has led us to postulate that cholesterol removal in vitro was growth associated. Significant interaction terms of inoculum size with FOS and inulin showed that these two prebiotics strongly encouraged growth of L. acidophilus ATCC 4962. Comparing these two, a higher coefficient of regression for X1X3 than X1X4 indicated that FOS was more preferred than inulin. Studies using bifidobacteria showed that the bifidogenic effects of inulin and FOS are independent of chain lengths or GFn type. FOS of the GF2 and GF3 moiety were also found to be more rapidly consumed compared with GF4 (Kaplan and Hutkins 2000). All these may have contributed to the preference of L. acidophilus ATCC 4962 on FOS than on inulin, and the fact that linear decrease in concentration of inulin contributed to an increase in growth.

Mean doubling time was used as a measure of the effectiveness of a specific carbon source in modulating bacterial growth rate (Bruno et al. 2002). Of all factors, FOS contributed significantly in the interaction patterns of mean doubling time, and higher growth rates (lower mean doubling time) were obtained at lower concentration of FOS (Fig. 5). It was previously reported that both the uptake and hydrolysis of FOS are induced by higher oligosaccharides but repressed by products of their hydrolysis (Kaplan and Hutkins 2003). In this experiment, it appeared that at higher concentration of FOS, more product of hydrolysis were produced and repressed bacterial growth rate, producing a higher mean doubling time. It must also be noted that the interaction between FOS and inulin produced lower mean doubling times when one factor was at lower levels and the other at higher levels. This indicated that when FOS was at its lower level, L. acidophilus ATCC 4962 utilized a higher level of inulin for higher growth rate and vice versa. It appeared that although L. acidophilus ATCC 4962 preferred FOS over inulin, but under conditions of substrate limitation, inulin was beneficially utilized for the modulation of growth rate.

The major products of metabolism of prebiotics are SCFA, carbon dioxide and hydrogen, and bacterial cell mass (Cummings et al. 2001). Although much work has been done on SCFA production and the significance of the individual acids, no particular pattern of SCFA production from prebiotic fermentation has emerged as yet. Hence, in this study, we analysed the SCFA production from fermentation of mannitol, FOS and inulin by L. acidophilus ATCC 4962. Production of SCFA appeared to be growth associated and correlated with the patterns of cholesterol removal. Although all factors significantly affected the production of SCFA, mannitol exhibited the strongest effect (Table 7). While the utilization of FOS and inulin has been widely reported, the utilization of mannitol to produce high concentration of SCFA was less studied and was also found to be strain dependent. Lactic acid bacteria that produced NADH oxidase would have the alternative NADH-H+-oxidizing mechanism, resulting in higher ability to grow on substrates more chemically reduced than glucose, such as mannitol (Stanton et al. 1999). This may contribute to the better growth of L. acidophilus ATCC 4962 in the presence of mannitol and subsequently produced higher amount of SCFA and higher cholesterol removal. Previous study showed that strains of L. acidophilus that utilized mannitol also exhibited capability of cholesterol uptake (Gupta et al. 1996).

In conclusion, cholesterol removal was optimized after selecting a combination of inoculum size and prebiotic, with the predicted optimum removal of 58·142 μg ml−1 obtained at 2·64% w/v inoculum size, 4·14% w/v mannitol, 3·28% w/v FOS and 5·82% w/v inulin. Validation experiment showed that RSM was reliable in developing a model, optimization of factors, and analysis of interaction effects. Analysis of growth, mean doubling time and production of SCFA showed that cholesterol removal and the production of SCFA was growth associated.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Bacteria and media preparation
  6. Cholesterol removal
  7. Growth of L. acidophilus ATCC 4962 in the presence of prebiotics
  8. Mean doubling time
  9. Short-chain fatty acids determination
  10. Experimental design and statistical analyses
  11. Results
  12. Screening of factors and steepest ascent
  13. Optimization of cholesterol removal
  14. Growth, mean doubling time and production of SCFA
  15. Discussion
  16. References
  • Anderson, J.W. and Gilliland, S.E. (1999) Effect of fermented milk (yoghurt) containing Lactobacillus acidophilus L1 on serum cholesterol in hypercholesterolemic humans. Journal of the American College of Nutrition 18, 4350.
  • Arihara, K. and Itoh, M. (2000) UV-induced Lactobacillus gasseri mutants resisting sodium chloride and sodium nitrite for meat fermentation. International Journal of Food Microbiology 5, 227230.
  • Bermudez, L., Inderlied, C.B., Kolonoski, P., Wu, M., Aralar, P. and Young, S.L. (2001) Telithromycin is active against Mycobacterium avium in mice despite lacking significant activity in standard in vitro and macrophage assays and is associated with low frequency of resistant during treatment. Antimicrobial Agents and Chemotherapy 45, 22102214.
  • Bruno, F.A., Lankaputhra, W.E.V. and Shah, N.P. (2002) Growth, viability and activity of Bifidobacterium spp. in skim milk containing prebiotics. Journal of Food Science 67, 27402744.
  • Cox, D.R. and Reid, N. (2000) The Theory of the Design of Experiments. London: Chapman and Hall.
  • Cummings, J.H., Macfarlane, G.T. and Englyst, H.N. (2001) Prebiotic digestion and fermentation. American Journal of Clinical Nutrition 73, 415S420S.
  • Delzenne, N.M. and Kok, N. (2001) Effects of fructans-type prebiotics on lipid metabolism. American Journal of Clinical Nutrition 73, 456S458S.
  • Dubey, U.K. and Mistry, V.V. (1996) Effect of bifidogenic factors on growth characteristics of bifidobacteria in infant formulas. Journal of Dairy Science 79, 11561163.
  • Gallaher, D.D. and Khil, J. (1999) The effect of synbiotics on colon carcinogenesis in rats. Journal of Nutrition 129 (Suppl.), 1483S1487S.
  • Gibson, G.R. and Roberfroid, M.B. (1995) Dietary modulation of the human colonic microbiota: introducing the concept of prebiotics. Journal of Nutrition 125, 14011412.
  • Gomes, A.M.P. and Malcata, F.X. (1998) Development of probiotic cheese manufactured from goat milk: response surface analysis via technological manipulation. Journal of Dairy Science 81, 14921507.
  • Gupta, P.K., Mital, B.K. and Garg, S.K. (1996) Characterization of Lactobacillus acidophilus strains for use as dietary adjunct. International Journal of Food Microbiology 29, 105109.
  • Kaplan, H. and Hutkins, W. (2000) Fermentation of fructooligosaccharides by lactic acid bacteria and bifidobacteria. Applied and Environmental Microbiology 66, 26822684.
  • Kaplan, H. and Hutkins, W. (2003) Metabolism of fructooligosaccharides by Lactoacillus paracasei 1195. Applied and Environmental Microbiology 69, 22172222.
  • Klahorst, S.J. (2000). Food Product Design. Northbrook: Weeks Publishing.
  • Lee, S.L. and Chen, W.C. (1997) Optimization of medium composition for the production of glucosyltranferase by Aspergillus niger with response surface methodology. Enzyme and Microbial Technology 21, 436440.
  • Li, C., Bai, J.H., Cai, Z.L. and Ouyang, F. (2002) Optimization of a cultural medium for bacteriocin production by Lactococcus lactis using response surface methodology. Journal of Biotechnology 93, 2734.
  • Lourens-Hattingh, A. and Viljoen, B.C. (2001) Yogurt as probiotic carrier food. International Dairy Journal 11, 117.
  • Montgomery, D.C. (1996) Design and Analysis of Experiments. New York: John Wiley and Sons.
  • Myers, R.H. and Montgomery, D.C. (1995) Response Surface Methodology: Process and Product Optimization Using Design Experiments. New York: Wiley-Interscience Publication.
  • O'Sullivan, M.G., Thornton, G., O'Sullivan, G.C. and Collins, J.K. (1992) Probiotic bacteria: myth or reality? Trends in Food Science and Technology 3, 309314.
  • Oh, S., Rheem, S., Sim, J., Kim, S. and Baek, Y. (1995) Optimizing conditions for the growth of Lactobacillus casei YIT 9018 in tryptone-yeat extract-glucose medium by using response surface methodology. Applied and Environmental Microbiology 61, 38093814.
  • Pereira, D.I.A. and Gibson, G.R. (2002) Cholesterol assimilation by lactic acid bacteria and bifidobacteria isolated from the human gut. Applied and Environmental Microbiology 68, 46894693.
  • Piston, R.L. and Gilliland, S.E. (1994) Influence of frozen and subsequent refrigerated storage in milk on ability of L. acidophilus to assimilate cholesterol. Cultured Dairy Products Journal 29, 919.
  • Robertfroid, M.B. and Delzenne, N. (1998) Dietary fructans. Annual Review of Nutrition 18, 117143.
  • Rudel, L.L. and Morris, M.D. (1973) Determination of cholesterol using o-phtaldealdehyde. Journal of Lipid Research 14, 364366.
  • Sanders, M.E. (2000) Considerations for use of probiotic bacteria to modulate human health. Journal of Nutrition 130, 384S390S.
  • Shin, H.S., Lee, J.H., Pestka, J.J. and Ustunol, Z. (2000) Growth and viability of commercial Bifidobacterium spp. in skim milk containing oligosaccharides and inulin. Journal of Food Science 65, 884887.
  • Stanton, T.B., Everett, L.R., Kennedy, M.J., Jensen, N.S. and Bosworth, B.T. (1999) Isolation, oxygen sensitivity, and virulence of NADH oxidase mutants of the anaerobic spirochete Brachyspira (Serpulina) hyodysenteriae, etiologic agent of swine dysentery. Applied and Environmental Microbiology 65, 50285034.
  • Tomasik, P.J. and Tomasik, P. (2003) Probiotics and prebiotics. Cereal Chemistry 80, 113117.