1. In silico modelling of the physiology of the
digestive system, digestion and absorption
Autor: George van Aken
2. Underappreciated aspect in digestion modelling:
Interaction of food with the body is both ways and dynamic
The food’s perspective:
The food is selected, masticated, digested,
absorbed and processed
The body’s perspective:
The body receives mechanical, nutrient
and pharmacological signals;
affected by
time of day, mood, stress, activity, …
The body adapts:
release of digestive fluids,
residence times,
absorptive capacity,
post-absorptive processing,
appetite
3. 3
Demonstrate how bio feedback regulation of digestion can be modelled in silico
Modelling published physiological studies:
• Digestive fluid secretion
• Gastrointestinal transit
• Gut hormone release by receptor cells
• Absorption rates in the small intestine
Modelling material science knowledge on in vitro behaviour of
food matrials:
• Enzyme kinetics of digestion
• Buffer capacities of proteins, mucin
• Gelling behaviour of proteins
• In silico (or in vitro)
4. 4
Main physiological mechanisms for gastrointestinal control of nutrient absorption
• Digestive fluid release
• saliva, gastric fluid, gastric mucin, bile pancreatic secretion and intestinal secretion
• Adjusted to needs
• Gastric emptying rate
• Limiting the delivery of calories to the small intestine
• Avoiding harmfull acity in the small intestine
• Intestinal absorption rate
• Limiting the absorption rate of calories by the small intestine
• Seems to avoid blocking of protein absorption by carbohydrates and fat
• Satiety
• Limits the amount of calories ingested, by nervous and hormonal signals by the small intestine
Receptor cells are present in the absorptive cell layer covering the inside of the small
intestine, reacting on the detection of food material and absorbable nutrients in the
small intestine
6. About the model
• 1 single model that has been extended and refined over the years
• Mechanisms and parameters are derived from physiological and in
vitro studies
• Some parameters have been derived by fitting to in vivo studies and
then retained by adding them to the model
• The modelling results shown are not fits to the data but results from
the program as is with parameters derived from previous studies
• Complication encountered: Variations between individuals in a study
are larger than the deviations between measured and simulated
averages
• LIKE TO DISCUSS HOW TO DEVELOP IT FURTHER:
• With whom?
• What modelling platform? 6
7. Modelling gastric motility
7
fundus
antrum
corpus
duodenum
Solid
proteins
Free proteins
Fed state:
3 peristaltic cycles/min;
0 < C < 1
0< C < 1/3:
X% fundus empties
into corpus
* based on
measurements by
Tack 1998 and
Janssen 2011
pepsin
Fractional emptying F: 0<F<1:
F is a function of intestinal and blood
hormone stressors:
pH, osmotic value, nutrient detection,
blood CCK, GIP, PYY
8. Modelling gastric fluid secretion
8
fundus
antrum
corpus
water 96%
mucus 2%
HCl 0.1 M
pepsin 0.5%
duodenum
Solid
proteins
Buffering free
proteins pH
pH
pH
pH
Basis rate
0.1 ml/min
Stimulated* mole H+/min =
Reduce stimulated
secretion by a factor:
̶̶
*Based on fit of measurements
by Konturec (1971)
+
= 1
(simulates the effects of
gastrin and secretin)
9. Physiological limit to intestinal caloric absorption rate
measured in pig
9
Governing equation
derived and applied
in the model
(Weber, Ehrlein, Digestive diseases and science, 1998, 43(6), 1141-1153)
Caloric rate of absorption increases with infusion rate,
independent of the relative ratios carbohydrate, protein and fat
10. Modelling results
- Liquid beverages: gastric emptying, fullness and hunger
- Leucin blood levels after protein solutions
- Cheese gastric digestion and effect of fullness and hunger
- Glycemic effects of carbohydrates
10
11. Liquid beverages: gastric
emptying, fullness and hunger
Gastric emptying of beverages
Mixed nutrient beverages varying in calories and viscosity
In vivo experiments: Guido Camps, Thesis, Wageningen University (2017)
11
Gastric volume
measured by MRI-
imaging
12. Gastric emptying of beverages
12
In vivo experiments: Guido Camps, Thesis, Wageningen University (2017)
How do caloric content and viscosity affect gastric volume during emptying?
13. Leucin blood levels after protein
solutions
• Leucin promotes muscle protein growth, relevant for top sporters and the eldely
• Faster and higher peaks of blood Leucin are thought to be advantageous
13
14. Modelling protein digestion and absorption
Included in the model:
• pH dependence of pepsin activity
• Adjust the dynamic release of pepsin in such a way that an activity in
gastric juice of 2000 U/ml is obtained for the fed state at the pH
optimum
Based on: Minekus et al, A standardized static in vitro digestion method suitable for food – an international
consensus, Food and Function (2014) 5, 113-1124)
• Curves of pH versus gastric acid addition for mucin and alimentary
proteins (-casein, BSA and -lg)
• Luminal CCK Releasing Protein, competing with alimentary
digestable protein and peptides for digestion by pancreatic
proteases.
It’s detection leads to CCK secretion by I-cells, exciting pancreatic fluid an bile release and slowing gastric
emptying and reducing gastric tone
• Formation of a protein gel under gastric conditions for -casein
14
• -casein: Pertzoff and Carpenter, 1932.
• BSA: Curvale, 2009.
• -lg: Tanford and Nozaki, J. Biological Chem.,
1959.
• Gastric mucin: Li, et al., 15th International
Conference on Miniaturized Systems for
Chemistry and Life Sciences, 2011.
15. Example: Appearance of Leucine in the blood plasma
15
Experimental work from publications
by Dangin 2001, 2002 and 2003
Appearance of exogenous
Leucine
Tracer Leucine appearance in blood.
• Tracer Leucine added separately
• Tracer Leucin incorporated in the protein
P L
L
L
L
P
L
L
whey
casein
Whey,
casein
Gelling protein
(casein)
Non gelling protein
(whey)
Preferred:
Fast peaked release and
no leucine off taste
16. Extrapolation, predicted effect of coadministration of glucose or fat
16
Same as Figure 2, but now with a
co-administration of 20 g glucose.
Same as Figure 2, but now with a
co-administration of 20 g fat.
20 g whey protein supplement
17. Cheese gastric digestion and
effect of fullness and hunger
- Cheese clumps remain in the stomach for longer
17
18. Effect of solids in the stomach
18
Control meal:
Yoghurt with emulsified fat
Active meal:
Yoghurt with grated cheese
Similar nutrient composition
and energy content
In vivo experiments in collaboration with IFR: Mackie,
A.R., Rafiee, H., Malcolm, P., Salt, L., van Aken, G.A., Specific
structuring of food emulsions leads to increased satiation and
hunger suppression. Am. J. Physiol. Gastrointestinal and Liver
physiology (2013), 304, G1038-G1043.
Included in the model: cheese particles
erode away at their surfaces by pepsic
digestion.
This is slower than digestion of the protein
in solution
19. Experimental data versus simulation
-8,00
-6,00
-4,00
-2,00
0,00
2,00
4,00
6,00
8,00
-50,00 0,00 50,00 100,00 150,00 200,00 250,00
Time after meal (minutes)
Change in Hunger
Active
Control
-4,00
-2,00
0,00
2,00
4,00
6,00
8,00
10,00
-50,00 0,00 50,00 100,00 150,00 200,00 250,00
Time after meal (minutes)
Change in Fullness
Active
Control
Hunger calculated
from simulated
nutrients in SI
Fullness calculated
from simulated
gastric tone
simulated
simulated simulated
(a.u.)
simulated
21. SI absorption glucose measured in vivo
• The Role of Sodium in Intestinal Glucose Absorption in Man, Ward A.
Olsen and Franz J. Ingelfinger, J Clin Invest. 1968 May; 47(5): 1133-1142
• Absorption of glucose, sodium, and water by the human jejunum studied
by intestinal perfusion with a proximal occluding balloon and at variable
flow rates, R. Modigliani, J. J. Bernier, Gut, 1971, 12, 184-193
Appears to follow Menten kinetics,
Kmax and Ki measured for oligosacharides, sugars, petides, amino acids
22. Only fast monomer absorption of
glucose, galactose and fructose
Transporter proteins across the apical cell membrane of absorptive enterocytes
• Glucose and galactose co-absorbed with Na+ and water by SGLT-1
(high capacity)
Sodium-dependent GLucose coTransporter 1 co-transports 2 Na+ and 260 H2O per glucose.
About 3 liter of water every 100 grams of glucose absorbed!
• Fructose and Glucose by GLUT5
(lower capacity)
GLUcose Transporter type 5
22
23. Modelling brush border transporters and digestive enzymes
GLUCOSE absorption rate form in
vivo intubation measurements,
SUCROSE, ISOMALTULOSE and
STARCH as extension on glucose
absorption model by including
modelled amylase and brush
border digestion; includes
calculation of the concentrations
at the brush border
24. Glucose homeostasis
Periferal
blood
glucose
Toliċ at al. J. Theor. Biol. (2000)
Liver
Muscles,
adipose
Brain, nerves
Liver, gut
insulin
−
Delay time ~
36 min
+
++
+
Glycemic peak becomes high if the rate of
glucose absorption is high: fast increase of
carbohydrates that are quickly absorbed as
glucose by the gut.
based on mathematical fits of many in vivo and in vitro physiological studies
+
difference
Empirical
25. Glycemic effects of 25 g glucose or sucrose in 250 ml water
B.M. Lee and T.M.S. Wolever, European Journal of
Clinical Nutrition (1998) 52, 924-928
Blood glucose peaks:
• Higher for
glucose than for
sucrose
• For both at about
30 min
Blood insulin peaks:
• Higher for
glucose than for
sucrose
• For both at about
30 min
glucose
Sucrose
reference
bread
fructose
26. Sucrose and isomaltulose
75 g of either sucrose or isomaltulose + 400 g
water in healthy pre-diabetics
Van Can et al. 2009, Br. J. Nutrition, 102, 1408–1413
Isomaltulose
Sucrose
Vmax for brush border
digestion ~70% smaller than
for sucrose
27. L-Arabinose uncompetitively inhibits sucrase
R. Nolles and A. Benschop, L-Arabinose: A Novel Food Ingredient For A Healthier
Living, Biotech, Biomaterials and Biomedical: TechConnect Briefs 2016, 5-8
L-arabinose
The small discrepancy may suggest that the
absorption of arabinose is modelled slightly too fast
28. Ideas on how to bring this further?
- Academic collaborations?
- In silico modelling platforms?
- Direct inclusion into in vitro modelling?
- Commercial applications?
29. Proteolysis simulation
symplified scheme for the digestion of protein “1” as used in the model
1Pep
1
1PepPan
pepsin
pancreatin
1AA
> 1 kDa
> 1 kDa
< 1 kDa
1Pan
> 1 kDa
< 1 kDa
< 1 kDa
DH
DH
DH
Leu
EAA
NEA
pancreatin 1Peptides
< 1 kD
BB peptidases
30. Modelling brush border transporters and digestive
enzymes
Brush border enzyme kinetics is usually not rate limiting for the pure
carbohydrates
However:
• many real foods contain inbitors
• various intolerances related to inadequacies (lactose, sucrose intolerance)
31. Steady state enzyme inhibitor equations
•Michaelis-Menten equation for competitive, uncompetitive and non-
competitive enzyme inhibition
• Kinetic inhibitory constants have been measured in vitro for various substances
• Applicable for digestion modelling if the concentrations at the absorptive surface is
known.
For example sucrase is inhibited competitively by maltose, isomaltulose and acompetitively by
arabinose and xylose.
32. Diffusive and convective
transport, absorption and
production/removal by brush
border enzymes
Small intestinal lumen
Stirred by perstalsis
Unstirred mucus layer
Absorptive enterocites
Portal blood
cb
cl
Diffusion + liquid drag
from water absorption
and secretion
Absorption
Transport to blood
Receptor
cells
Mucus coating thickness
Factorial surface increase F
Circulation
Brush
border
enzymes
1 cm2
cb calculated analytically:
33. Calculation cycle
Calculate
concentrations
at the brush
border
Calculate
digestion by
brush border
enzymes into
absorbable
species
Calculate
absorption rate
Correct
absorption rates
to maximum
caloric
absorption rate
33
Calculate
brush border
neural and
hormone
signals
Calculate
digestive
fluids release
Calculate
gastric
tension,
motility and
emptying
Calculate
adjusted gut
transit rate
Calculate
digestion
Calculate
entry to
portal blood
Calculate
insulin
release and
degradation
Calculate
blood glucose
Input from in
vitro
digestion
studies
About 1
cycle/second
34. Gastric emptying of beverages
34 Together to the next level
In vivo experiments: Guido Camps, Thesis, Wageningen University (2017)
Possibly nutrients detected further down the SI enhance
CCK secretion, further reducing gastric tone
Low calory drinks give less
hunger reductionLow calory drinks give
relatively more fullness
Various references for physiology of fullness and hunger, e.g.;
Little, Gastroenterology 2007, 133, 1124-1131
Welch et al, Gut (1988), 29, 306-311