Using cell experiments and AI to predict the outcome of in-vivo studies

Using-cell-experiments-and-AI-to-predict-the-outcome-of-in-vivo-studies.jpg
© Natali Mis | Getty (Getty Images)

Research innovators at the nutraceutical research firm Institut-Kurz have combined cell experiments with brand new software to efficiently predict the outcome of in-vivo studies.

The new simulation program, named ADME (Administration, Distribution, Metabolization, Excretion), is claimed to remove the need for unnecessary in-vivo research by accurately quantifying molecules of interest in food and following their journey through the body.

"Processes in the gastric environment are difficult to measure in vivo but the gastric environment and its effect on the molecules of interest can be determined quite well in-vitro," Dr Helmut Weidlich, the firm's managing director told NutraIngredients. "As can the effect of the chemical and microbiological environment in the different sections of the small intestine, the transfer of molecules through the intestinal wall of intestinal membranes, and the various chemical reactions in the cells."

In fact, he asserted some processes which are not accessible to in-vivo measurement, such as the passage of molecules of interest from the small intestine into the blood flowing to the portal vein, can be measured well in-vitro too.

“In-vivo, only the content of interesting molecules in the blood of the arms and legs can be measured, which is only the composition of the blood after passing through the liver at least once," he explained.

“However, the liver passage of blood can also be readily mimicked in-vitro and thus the effect of the liver on the molecules of interest to us in each case.”

Quantifying bodily processes

The effect of food ingredients on the human body is complex. In addition to the mechanical transport processes, trillions of chemical reactions take place per second in the human body, Dr Weidlich explained.

“For this reason, medicine and nutritional science have so far been able to elucidate individual bio-molecular processes in the body, but have not been able to reconstruct the entire interaction of all cells and the molecules contained inside and outside the cells.”

The gold standard to prove a health claim is true is a multi-center, randomized, placebo-controlled intervention study lasting 30 years. But hardly any manufacturer can afford or wait for that.

Therefore, it is accepted scientific practice to look for changes in certain effect-specific biomarkers that make it credible that a desired health effect is triggered by the consumption of, for example, a particular dietary supplement. For example, the reduction in the usual excretion of F2 isoprostanes after the consumption of vitamin C is accepted as evidence of the antioxidant effect of vitamin C. This approach is widely accepted by EFSA.

A single individual biomarker - when measured in subjects in vivo - is considered sufficient for the complex responses in the human body.

"F2 isoprostanes (IsoPs) are detected in urine in order to deduce whether there has been less or more oxidative attack on cell walls, thereby suggesting whether a food ingredient classified as antioxidant actually helps the body to protect itself against oxidative radicals or not," Weidlich explained.

“Therefore, even the health effect in the context of in-vivo studies is usually not given credence on the basis of holistic overall effects, but on the basis of quantification of characteristic individual substances in blood or urine.”

He noted that health effects, for which no biomarkers can be defined as endpoints, are doubtful and difficult to prove - namely only through observational studies lasting decades. 

However, the molecular pathway by which individual substances are formed in the affected body cells and transported can be described very well on the basis of in-vitro cell experiments in combination with simulative computer programs, he asserted.

“Because we usually quantify only single characteristic biomarkers as outcome in in-vivo proband studies, the development of exactly these biomarkers can also be predicted sufficiently well with in vitro cell experiments combined with a good whole-body simulation computational program.”

Combining in-vitro knowledge and AI

With a wealth of knowledge and experience providing test labs across Europe since the early 70's, Institut-Kurx developed this new ADME tool with support from the German Federal Ministry of Economy through its innovation program ZIM (Zentrale Innovation Mittelstand).

The firm’s clients have so far utilised this programme to predict the outcome of in-vivo-studies on: isoprostanes as markers for antioxidative effects of food supplements; advanced glycation end products (AGES) after consumption of foods suspected of reducing the build-up of such AGES; and bio-availability of biotin after administration of different forms of food supplements.

Those clients are now preparing dossiers for submission to EFSA (European Food Safety Authority).