PATHWAY – 27

Project ID: 311876

Final Report Summary – PATHWAY-27 (PIVOTAL ASSESSMENT OF THE EFFECTS OF BIOACTIVES ON HEALTH AND WELLBEING. FROM HUMAN GENOMA TO FOOD INDUSTRY – PATHWAY)

The PATHWAY-27 project started in 2013 and focused on bioactive compounds, i.e. natural components of foods that possess biological activity in addition to their nutritional value. Bioactives naturally occur at very low concentrations in foods, usually far from the optimally effective dose range. To overcome this, the food and drink industry is developing new products containing high concentrations of selected bioactives, so-called bioactive-enriched foods (BEF). In the global scientific scenario, research on bioactive compounds is very active. Notwithstanding, several questions about the effectiveness of bioactives as preventive agents to reduce the risk of disease, as well as the effectiveness of BEF are still waiting for an answer.
Uniquely, PATHWAY-27 did not consider bioactives as discrete molecules, but as ingredients of bioactive enriched food (BEF) that can be part of an everyday diet.
PATHWAY-27 included 8 work packages (WPs), each one devoted to specific activities but all perfectly integrated and inter-connected to achieve the objective of the project (Figure 1).
WP1 was in charge for the coordination and management, and supervised the research activities between WPs, monitored the overall Project achievements and controlled the scientific quality and timing of the Project results. In addition, WP1 managed contacts with EU Commission and among the consortium; supervised the foreground generated by the project activities monitoring that intellectual property was properly managed, ensured an appropriate overall financial and administrative management and reporting, and took care of ethics management. Ethical issues were appropriately managed with the help of the internal Ethical Advisory Board (EAB). All protocols for experiments involving animals and human subjects were approved by the PATHWAY-27 EAB before experiments could start, and this process took place prior to submission of protocols to the relevant local research ethics bodies.
Project coordination meetings were organised at the beginning and end of the project, and every 12 months along the project duration. Several on-line meetings were also organized within different WPs, and an additional “in person” meeting for WP5 partners took place in Paris in July 2015.
The first step of the project was the formulation of the PATHWAY-27 BEF. At the project start 3 foods x 3 matrices were considered, and corresponded to buns, breadsticks, biscuits (bakery), milkshake, milk dessert, pudding (dairy), omelet, egg beverage and pancakes (egg-based). Each food was enriched with 5 bioactive combinations (DHA alone, DHA + AC, DHA + OBG, AC alone and OBG alone). In total, 45 BEF prototypes were developed.
The ingredients containing the different bioactive compounds (AC, OBG and DHA) were produced by the partner SMEs. In particular, ASL produced DHA-enriched egg yolk (2.5% OVO-DHA) and DHA-enriched phospholipids fraction (8% GPL-DHA) from the DHA-enriched egg-yolk; SOF produced OBG-enriched oat bran extracts containing 27-28% OBG; and ABRO produced AC from the skins of red grapes containing 2.5% AC. For the production of BEF, it was important to have the ingredients with high enough concentration of bioactive components as well as ensure the safety of the ingredients. For this purpose, the bioactive content, homogeneity and chemical form were determined for each bioactive formulation by the respective producer SME. The levels of bioactive components were measured and the microbiological safety was analyzed from each production batch. The producer SMEs tested the safety characteristics of bioactive formulations and certified them according to the internal procedures. All ingredients containing the bioactive compounds were in the form of dry powders, so their shelf life was at least one year starting from the production date.
In the 45 prototype BEF, bioactive content was quantified at the end of the manufacturing process, and showed an excellent recovery of DHA in all matrices. Processing led to a decrease in AC concentration in omelet and bakery products, probably due to the cooking step applied during manufacture of these BEF. This loss of AC during manufacture was overcome by increasing the amount of AC in the recipe. Finally, molecular weight of OBG slightly decreased after omelet and bakery products manufacture but the forms present were still bioactive according to the literature.
A pre-screening test to assess the overall sensory quality of the 45 prototypes was organized among partners in order to select only the best two foods in each category. This decreased the number of samples to further analyze from 45 to 30. The 30 selected BEF were in vitro digested to assess the bioaccessibility of the target bioactives. The digestion was done following the oro-gastrointestinal model developed by COST action FA1005, INFOGEST (Minekus et al. 2014). Bioaccessibility (fraction of a compound/nutrient that is released from its matrix and solubilized in the gastrointestinal tract and is potentially available for intestinal absorption) was close to 100% for all the DHA-enriched BEF. All the BEF enriched with AC alone showed higher bioaccessibility than the ones enriched with AC+DHA. OBG bioaccessibility ranged from 57 to 90% for OBG-enriched BEF and from 73 to 119% from DHA+OBG-enriched BEF. Results showed that OBG present after in vitro digestion had a high average molecular weight over the 1000 kDa in most cases. According to the literature OBG molecules higher than 500 kDa should provide similar physiological responses as the native OBG, so we concluded that almost all of the OBG present after in vitro digestion conserved its bioactive properties.
The 30 BEF were extensively studied in order to
• determine the shelf life;
• define the sensory profile;
• understand the consumer acceptance;
• ensure and assess the safety of the samples provided for sensory and consumer tests;
• give inputs for the final selection of BEF to be administered to volunteers in clinical studies;
• set up a questionnaire to be able to measure the preference, overall liking of the developed products during clinical studies.
Comprehensive assessment of food safety risks represented by the experimental samples was carried out. Regarding the microbiological assessment of BEF, challenge test was performed on bakery products only since egg-based products were provided frozen and dairy products as powders. Although preservatives were used to limit bacterial growth during storage of buns at room temperature, this did not avoid the appearance of molds. On the contrary, biscuits, with their low water activity, were highly stable on a microbiological point of view throughout storage.
The shelf life tests were carried out first for the targeted shelf life of 3 weeks, which were changed to 84 days in the second schedule to meet the logistic requirements of the pilot studies and to exploit opportunities provided by the changes of the preservation technology.
The complete sensory characterization of the developed products was done by the trained sensory panel. The main aim of this activity was to have clear picture about the main attributes of the products and to be able to identify the key sensory attributes that drive the consumer liking and preference. In addition, an online questionnaire, which contained questions to explore the consumers’ knowledge on health and healthy diet, bioactive enriched foods, or new and not conventional products, was administered to 120 consumers.
Since the PATHWAY-27 pilot study was evaluating the effectiveness of a single product, enriched with all the five bioactive combinations, within a specific matrix (bakery, dairy and egg-based products), results of consumer acceptance and sensory profile of the five enrichments of the same products were averaged, in order to evidence the product in each matrix having the best mean characteristics considering all enrichments. Results of the pre-screening test were also considered.
Based on results of the consumers’ acceptance test, both bakery products were suitable, and no significant difference was observed between the 2 products during sensory analysis. Among egg-based products, omelet received a slightly higher overall scores than pancake. On the contrary, pancake were slightly more popular than the omelet in the pre-screening. Beside the strange appearance of the AC enriched versions, all pancakes were described with typical pancake aromas and flavours. For evaluating the unusual appearance of the pancakes, a group of children (N: 29) were involved in the study. No significant difference between the pancake and omelet were detected in overall liking scores. Sensory evaluation of both products also appeared satisfactory. Regarding the dairy samples, there were not significant differences between the two products. Almost the same positive and negative comments were mentioned regarding both products. Most of the consumers eat these products at least every month, but the majority of the participants reported to drink milkshake every day. So the difference in the frequencies of the consumption gave priority to milkshake.
Given the importance of the product liking in the clinical studies, a questionnaire for the volunteers involved in the pilot study was developed to identify how they like the appearance, aroma, flavour and texture of the samples and how their opinion changes if they consider the health benefits of the products.
The 30 BEF were analyzed by NMR spectroscopy to acquire their molecular profile, and their structural properties were investigated by Confocal Laser Scanning Microscopy.
Comparison of all attributes between the two food products within each matrix was carried out in order to select the most suitable one considering the characteristics coming from the different five enrichments. The selection method applied was the one of the decision sieve. A decision sieve is based on the thresholds of the quality attributes setting the mesh size; BEF can be considered for selection if they positively pass the mandatory requirements above an acceptable quality threshold. In case both candidate BEF in the same matrix fulfil the selection criteria, the secondary attributes are taken into account. Mandatory attributes were safety, chemical stability and nutrient profile. Secondary attributes were bioaccessibility of bioactive compounds, sensory characteristics, ease of preparation, ease of storage, food matrix stability.
For the bakery products, biscuits were selected instead of buns mainly for their microbiological stability during storage. The ease of preparation and storage confirmed the better suitability of biscuits over buns. For the dairy products, milkshake was selected over dairy dessert. The nutrient profile of both products was not satisfactory because of level too high in saturated fatty acids (SFA). However, the recipe of the milkshake was optimized to limit the amount of SFA whereas it was not possible to improve the recipe of the dairy dessert commercially available. The pancake was selected for egg-based products over the omelet since it showed a much better nutrient profile (too much salt and SFA in the omelet). In conclusion, the 3 products selected for the pilot study were biscuits, milkshake and pancake.
The 3 selected products enriched with five different bioactive combinations were produced by two partner SMEs at industrial scale to be used in the three 4-week human pilot studies. DPL produced biscuits (bakery product) and ADX milkshake powders (dairy product) and pancakes (egg-based product).
Milkshake powder and biscuits were dry products meant to be stored at room temperature, but pancakes needed to be stored and delivered frozen at -18 °C to inhibit the growth of spoilage and pathogenic microbes. The safety characteristics of BEF were determined and certified by the corresponding producer SMEs, according to the internal procedure adopted for new products.
Shelf life of BEF was a key issue for both executing the intervention trial and possible marketing of the products. In fact, a longer shelf life allows better management of production and delivery of BEF during the trial, and consumers consider it as an important positive attribute of food products. Regarding BEF, in order to set a longer shelf life not only food safety and sensory characteristics must be considered, but also the stability of the bioactives. Thus, during the pilot study BEF products were exposed to extended shelf-life studies where microbiological and sensory quality of the products was evaluated and the contents of the bioactive compounds analysed. Taking into account the retention of bioactive concentrations as the key criteria, but sensory and microbiological quality of the BEF products as well, 3 months shelf life was considered for BEF pancakes, 8 months for BEF milkshake and 2 months for BEF biscuits.
In addition, to make sure the bioactive contents in the selected BEF were similar in all production batches, so that similar amounts of bioactives were delivered to all volunteers in pilot studies, bioactive concentration was determined in each batch produced for pilot studies.

Overall, PATHWAY-27 developed a strategy to formulate and select BEF to be used in intervention studies, evaluating food characteristics (acceptability, bioactive accessibility and retention, etc) that are seldom considered although they could interfere with results obtained in human trials.

The three selected products (milkshakes, biscuits, pancakes) enriched with five different bioactive combinations were used in pilot intervention studies. The purpose of pilot studies was to select, for each of the three food matrices, the enrichment delivering the greatest reduction in serum triglycerides (TG) or increase in HDL cholesterol (HDL-c). Three pilot studies were conducted in three clinical research centers (Max Rubner-Institute, Karlsruhe, Germany; University of Leeds, United Kingdom; and Human Nutrition Research Centre of Auvergne, France), each participating center focusing on a specific food matrix. Each pilot study was randomized, double blind, parallel dietary intervention study without a placebo.
In each pilot, participants were divided in five groups, each receiving BEFs enriched with DHA, BG, or AC alone or DHA+BG, or DHA+AC based on bakery, dairy or egg products (Figure 2). Participants were advised to consume the allocated BEF daily for a period of four weeks. The BEF should be consumed in replacement of similar foods and not in addition to the usual diet. Besides BEF consumption, participants were instructed to consume food rich in DHA, OBG and AC only in limited amounts. Therefore they received a list of respective foods.
At baseline and after 4 weeks of intervention, fasting blood samples were collected for further analysis. Additionally, blood pressure and anthropometric data have been determined. A schematic view of the study protocol is given in figure 3.
To standardize and harmonize procedures among the study centers, a set of standard operation procedures (SOPs) were developed for the pilots. Each clinical centre obtained the approval from local Ethics Commitees and additional regulatory bodies before the start of the trial.
Pilots were designed as exploratory trials aiming to identify the best matrix × enrichment combination that would result in the strongest effect on the designated endpoints. Each pilot explored 5 different BEF within the same matrix, and evaluated the effects compared to baseline.
Pilot studies have also the meaning to solve possible problems arising before/during the intervention.
Considering that MetS is defined as “a group of conditions that put you at risk for heart disease and diabetes” more than a disease itself, criteria for MetS diagnosis (Table 1) were used as inclusion criteria in the study. Subjects presenting two to four of the criteria, at least one of them being alteration of fasting triglycerides or HDL cholesterol were considered eligible to participate. Additionally, major exclusion criteria were selected (among others, regular drug therapy with impact on serum lipids, diabetes, celiac disease, history of allergy or intolerance to any components used in BEFs; recent history of cancer or cancer treatment; familial dyslipidemia; intestinal malabsorption; pregnancy or lactation; etc).
In total, 167 men and women at risk for MetS participated in the pilot study.
The number of volunteers in each intervention subgroup and the duration of the intervention itself did not allow identifying statistically significant changes in the primary endpoints (blood triglicerides -TG and HDL-cholesterol – HDL-c), but some trends were clearly evidenced, allowing the selection of the most promising enrichment within each matrix.

Descriptive statistics data on TG revealed a moderate nominal reduction after 4 weeks of intervention for all matrix groups, more evident for dairy. Considering all matrices, the most prominent effect was for DHA+OBG enrichment. The data indicates that DHA+BG showed for dairy the greatest reduction in blood TG, AC and DHA+AC showed the greatest reduction in blood TG for egg-based and DHA+AC for bakery products. Among matrices, egg-based products had the most prominent effect on HDL-c while DHA+AC was the most effective enrichment.
To further evaluate the impact of BEF intervention we included the following secondary endpoint into the statistical analysis: systolic and diastolic blood pressure, waist circumference and fasting blood glucose. Additionally, the anthropometric measurements of height, body weight and calculated BMI have been considered. Although no significant effects of BEF intervention on any single secondary endpoint were identified, some interesting trends were observed, particularly on blood pressure.
Sensory acceptance of BEFs was also evaluated, and it evidenced that the volunteers were open to eat such unusual products. The products received relatively high scores for all attributes. Volunteers stated in all three pilot studies that they would eat these products, if proof is provided that the consumption of the products improves their health, or helps to avoid medication. Sensory characteristics and acceptance among enrichments within the same matrix were almost similar, so they did not interfere with the selection.
Results obtained in pilot studies were integrated to select the most promising enrichment in each food matrix. Taking the promising effects of DHA+BG on blood TG into account and considering also the positive and stable effects on HDL-cholesterol and blood pressure, the DHA+OBG enrichment was selected for the dairy product (milkshake). For the egg-based product (pancake), considering the positive effect on blood TG and HDL-c, and on blood pressure the enrichment with DHA+AC was selected. The same enrichment (DHA+AC) was selected for the bakery product (biscuits) considering the promising effect in blood TG and the increase of HDL-c (Figure 4).
The three selected BEF were then used in the PATHWAY-27 large intervention study (LIS), a multicentre, randomized, placebo-controlled, parallel-arm dietary intervention study carried out in four European recruiting centres (RC): Human Nutrition Research Centre of Auvergne – CRNH, France; Max Rubner-Institute – MRI, Karlsruhe, Germany; University of Leeds – ULE, United Kingdom; University of Bologna – UNIBO, Italy.
The primary objective of the LIS was the evaluation of the relationship between TG and HDL-c blood levels and the intake of BEF selected in pilot studies in subjects at risk for or affected by MetS. Secondary objectives were to evaluate the effect of the consumption of BEF on other markers of MetS (secondary end-points) and on body composition using total-body Dual Energy X-Ray Absorptiometry (DEXA) in a subgroup of volunteers. In addition, blood, urine, stool and white adipose tissue (WAT) samples were used for omics analyses.
According to the study design, participants were randomly assigned to one of four groups to receive either:
1. Dairy BEF + egg placebo + bakery placebo;
2. Egg BEF + dairy placebo + bakery placebo;
3. Bakery BEF + dairy placebo + egg placebo;
4. Dairy, egg and bakery placebo.
Volunteers were asked to participate to a sub-study, including additional analysis. Participation was on a voluntary basis.
Participants were required to consume all three of the allocated products each day for 12 weeks. Fasting blood and urine samples were collected from all participants at baseline (T0), midpoint (T6 –urine were collected in subgroup participants only) and endpoint (T12). Additionally, blood pressure, body weight (and height) and waist-circumference were collected at baseline, midpoint and endpoint from all participants. At baseline, a dietary assessment was performed for all participants.
As soon as the RCs received the authorization from the respective Ethics Committee, then they actively started contacting and screening the volunteers. Inclusion and exclusion criteria were the same as in the pilot studies.
Although a high number of volunteers was screened during the first 2 months of the trial, the recruitment rate was lower than foreseen, therefore the recruitment phase was extended till the end of July 2017. The LIS concluded in October 2017.
Considering food production, the pilot studies were valuable source of experience for the LIS, both for the manufacturing SMEs and the research centers. The difficulties in BEF distribution and storage encountered during the pilot studies were solved, allowing successful production, delivery and storage of the required amount of BEF during the LIS. Therefore, to conduct pilot studies represented a good strategy to reduce the cost of the large intervention not only because they allowed selecting the most promising BEF, but also solving problems related to food production and delivery to recruiting centres.
Microbiological quality of BEF pancakes and milkshakes was analysed right after production, and the food was sent to research centers only after the analysis results were ready and the batch thus declared as safe. Biscuits have very low water activity (<0.6) and the data obtained from a microbiological challenge test was enough to assure microbiological safety and stability.
For the LIS, 25 000 active and 73 000 placebo pancake portions were produced by ADX. For milkshakes the corresponding numbers were 27 000 and 70 000 (produced by ADX), and for biscuits 19 000 and 60 000 (produced by DPL). To ensure research centers an un-stoppable supply of fresh BEFs and placebos, the food productions were arranged based on pre-planned schedules where different shelf lives of the products were taken into account and the number production batches minimized.
Reproducibility of tested food is one of the basic requirements of EFSA for the substantiation of the health claims. In order to establish the cause and effect relationship between consumption of the food constituent and the claimed effect, a standardised concentration of the constituent must be verified by checking the variation of the initial data between different batches. Thus during the LIS, for each lot produced, the concentration of bioactives in the BEF was analysed. The results from the analyses showed that bioactive concentration was quite constant in all production batches of the LIS. The foodomics approach was applied to assess the effect of the incorporation of a functional ingredient on the reproducibility and/or stability of three different food matrices. Foodomics has the advantage of capturing a holistic view of the molecular complexity of the food extracts, and allowed a simplified appreciation, granted by 1H-NMR spectroscopy, of the molecular homogeneity among the production batches of the 3 different food matrices before storage and the subsequent stability performances of several products (enrichment, storage time) produced and consumed during the LIS.
Due to the difficulties in recruitments and the LIS extension, food production lasted longer than foreseen. To minimise food waste coming from low recruitment and the SMEs’ efforts, an excel-based planning protocol was developed and taken in use. It was based on a continuous flux of information from RC to producing SMEs and on re-labeling of placebo servings that were not used earlier, but still had long enough shelf-life. To ensure the blindness in the study, re-labelling was performed by contact persons (one in each RC), otherwise not involved in the study.
The SMEs were able to continue the productions also after the original ending date of the study and provide BEFs and placebos for the research centres for the whole duration of the LIS, altogether 1.5 years.

The most promising BEF were selected in pilot studies and used in the large intervention trial. All problems related to the production of BEF to be used in human intervention trials were successfully solved. Overall, PATHWAY-27 activities paved the road for the industrial exploitation of PATHWAY-27 BEF.

During the LIS, 5401 putative volunteers were contacted by the 4 clinical centres, of whom 1361 accepted to carry out the screening visit and/or blood sampling for lipid profile and exclusion criteria: 420 resulted eligible and 325 accepted to enter the study and were randomized. Among them, 109 accepted to carry out additional analysis included in the sub-study. During the trial, 89 subjects dropped out while 236 volunteers (82 in the sub-study) completed the 12-week trial and were considered in the Intention to Treat (ITT) statistical analysis.
Subjects with available TG and HDL-c value at all 3 main time points and at least 70% compliance to consumption for each administered food were included in the per protocol (PP) statistical analysis. Four subjects with TG>400 mg/dl (possibly indicating familial hypertriglyceridemia) were excluded from the analysis. In total, 195 volunteers were included in the PP analysis. Restrictive inclusion/exclusion criteria allowed us to select a homogeneous cohort of high metabolic-risk subjects who potentially could have benefited from BEF consumption.

In volunteers with low HDL-c (≤50 mg/dl for females, ≤40 mg/dl for males) at baseline, all BEF significantly increase HDL-c. HDL-c is the well-behaved “good” cholesterol. High HDL levels reduce the risk for heart disease but low levels increase the risk. HDL cholesterol scavenges and removes LDL-c (“bad” cholesterol) by transporting it to the liver where it can be reprocessed. The positive effect of BEF on HDL-c was already achieved after 6 weeks of consumption, and it lasted over the whole period of treatment. Since HDL-c is biologically stable, the HDL-c improvement may be long standing with potential clinical benefits.
The HDL-c increasing effect was more evident using BEF enriched DHA+AC (pancake and biscuits). The nutritional evaluation of volunteers evidenced that consumption of these BEF did not induce a significant increase in AC intake while DHA intake was significantly higher than at baseline. So, we could speculate that DHA is the more active compound, AC having a facilitating activities.
The group receiving DHA+OBG experienced a clear reduction of LDL-c level (mean changes T0-T12 = -6.844; % differences of T0= -3.69%). It is worth noting that DHA+OBG treatment was the only one reducing LDL-c, other treatments having an increasing effect. A modest increase of LDL-c level after DHA administration has been already reported in clinical trials, although the clinical relevance of such finding is uncertain (Mozaffarian D, 2012). Anyway, the association of OBG to DHA could have antagonized the LDL-c increasing effect. A decrease in LDL-c level was also observed in the placebo group. This could reflect changes in the dietary habits of volunteers, who possibly turned to a healthier diet during the trial.
Consumption of all food enriched with DHA was associated to significant reduction of systolic blood pressure (SBP) compared to placebo, both at T6 and T12. The effect appeared mainly ascribable to administration of DHA+OBG and DHA+AC embedded in pancakes. Consumption of DHA+AC enriched biscuits had no effect on SBP.
Importantly, the observed metabolic effect seems to be independent by changes in BMI and waist-circumference.

Samples from volunteers enrolled in the LIS were analysed with state-of-the-art omics techniques.
An intensive literature search led to the selection of 39 single nucleotide polymorphisms (SNPs) to be investigated in all recruited volunteers to highlight the possible correlation of the genetic profile to the risk of MetS. Among selected SNPs, one (ADIPOQ_ rs17300539) was found associated with the risk of MetS in all national sub-cohorts, and the presence of the polymorph allele A (in homo- and heterozygosis) increased the risk.
An NMR-based metabolomics approach applied to serum and urine evidenced that samples collected at the beginning and the end of the study could be discriminate with high accuracy, so indicating a clear modification of the metabolome (i.e. the complete set of small-molecule chemicals found within a biological sample) due to treatment. From the combined analysis of the serum and urine samples in the different BEF groups, it appeared that the enrichment with DHA+AC was more effective than DHA+OBG in introducing effects in the metabolomic profiles of the volunteers, confirming the clinical observation on HDL-c. In particular, in serum we highlighted significant changes in the lipoprotein profiles of the volunteers receiving DHA+AC embedded in the bakery product, but not in volunteers receiving the same bioactives embedded in the egg-based product.
Investigations on the DNA methylome (i.e. the set of DNA methylation modifications that controls the activity of genes in a particular cell) in human white adipose tissue (WAT) and peripheral blood mononuclear cells (PBMCs) of a subgroup of volunteers (n=35) also revealed significant changes. However, we observed little overlap between both tissues, indicating that PBMCs could not be considered as a surrogate for changes observed in WAT. Differentially methylated sites partly overlapped with stable and dynamic enhancer regions of adipocyte differentiation.
mRNA expression changes between the beginning and the end of the intervention of genes associated with differential methylation clustered the 33 subjects with available RNA-seq data in two relatively stable clusters. Most of the placebo subjects were located in one cluster, while the second cluster represented most subjects of the dairy-based BEF group as well as several subjects of the egg-based BEF group, potentially indicating differences in response to intervention depending on the food matrix. Anyway, due to the small number of subjects and multiple co-variables (age, gender, nationalities, BMI and other clinical parameters, smoking, etc.), interpretation of the results needs to be done with caution.
The effect of BEF intervention on unbalances in gut microbiota and intestinal metabolites that impact on human physiology was monitored in a subgroup of volunteers (n=82). At the bacterial species level, some components of gut microbial population were apparently positively associated to changes in the lifestyle of volunteers due to participation in the study and consumption of any BEF. Other shifts were more related to the effect of one or two BEF. Interestingly, increase in some microbial groups seemed to exclude the increase of pathogenic members of the group, suggesting an ecological barrier function induced by BEF consumption.
Single omics datasets were merged into an integrated multi-omics collection to test correlations between pairs of variables and perform multivariate analysis. The most important finding obtained was a weak tendency to separate subjects in the multi-omics space consequent to the consumption of BEF, associated to two genetic variants of ADIPOQ and FTO genes, together with LDL-c, HDL-c and total cholesterol, as well as a couple of microbial species and the lipids region of the NMR spectra.

To evaluate food digestion and bioactive availability, considered as a possible variable of BEF effectiveness, in vitro and in vivo studies were performed. In vivo, experiments were performed in a pig model since it is considered as the most physiologically-relevant model to mimic the human digestive system. Each pig was fed the three BEF selected after the pilot study and used in the large intervention study. Effluents and plasma were collected continuously during 7h after the meal and the resulting digested material was analyzed. The kinetics of DHA concentration in plasma appeared to be affected by the type of BEF ingested. After pancake ingestion, the peak of DHA concentration appeared earlier but it was of lower magnitude than for milkshake and biscuits. DHA concentration peaked at later time when biscuits were administered, but it decreased slower than using pancake and milkshake.
Within the PATHWAY-27 project, in vitro studies were also performed to understand the protective role and mechanism of action of the three bioactive compounds (DHA, OBG, and AC) used to enrich food tested in clinical studies. Two main cell model systems were used, adipocytes and hepatocytes, due to their pivotal role in development of the MetS.
Despite the wide use of cell cultures for the determination of bioactive effectiveness and mechanism of action, many studies reporting a protective effect of bioactives have been performed supplementing cells with putative active compounds in the form they are present in the food. In vivo, foods are digested and some bioactives are extensively metabolized, so that the effective molecules are very different from parent compounds. In PATHWAY-27, we took it into account, and we used for cell supplementation protocatechuic acid (PCA), the major metabolite of most anthocyanins (Vitaglione et al. 2007; Czank et al. 2013), and propionic acid (PRO), one of the short fatty acid produced in the colon upon OBG fermentation by the microbiota. PRO produced in the colon is absorbed, and around 90% is metabolized by the liver (Al-Lahham et al. 2010). Since DHA is absorbed by intestinal cells and it is delivered to peripheral cells in its parent form, cells were supplemented with DHA.
An extensive literature search was performed to identify the physiological range of concentrations of the studied bioactives. In addition, since cell cultures represent a close system, and their direct and continuous exposure to bioactives could alter the cell response, inducing cytotoxicity due to the lack of the continuous detoxification and clearance of compounds occurring in the whole body, the possible cytotoxicity of supplemented compounds was assessed prior to further experiments.
A systemic analysis of the action of bioactive compounds, alone and in combination, on lipogenesis, lipolysis, adipokine secretion (i.e. inflammatory status) and adipogenesis of human in vitro differentiated adipocytes was made. As model, primary human pre-adipocytes from subjects between 26-65 years of age with a body mass index (BMI) ranging from 21-28 kg/m2 (mean 25) were isolated from subcutaneous white adipose tissue (WAT) obtained as a waste product from cosmetic liposuction. No cytotoxic effect was observed up to 6 days of treatment with bioactives at the tested concentrations (DHA 0.5 µM; PRO 100 µM and PCA 13 µM) . As well, no effects of the selected compounds on lipid accumulation nor on gene expression of selected marker genes (PPARG, CEBPA, PLIN1, LIPE, UCP1 and ADIPOQ) were observed after 6 days of treatment. To investigate the effect of the bioactives on the function of adipose cells, a panel of the most important pro- and anti-inflammatory adipokines, were measured, as well as the effect on lipolysis and insulin-stimulated lipogenesis. All bioactives decreased basal lipolytic activity and secretion of inflammatory markers in human fat cells, which might explain some of the reported health benefits induced by bioactive compound-containing diets in humans, including reduced insulin resistance, hypertension and dyslipidemia.
In preliminary studies on hepatocytes, possible bioactive cytotoxicity was assessed by means of different methods and the highest concentrations of bioactives to be used in further experiments were set as follow: DHA 50 μM; PRO 70 μM; PCA 20 μM; DHA 50 μM + PRO 70μM; DHA 50 μM + PCA 20 μM (Di Nunzio et al, 2017). Then it was evidenced that DHA supplemented alone or in combination with PRO or PCA is significantly taken up by cells in a time-dependent manner, and modified cell fatty acid composition. A significant decrease in intracellular cholesterol content was detected in DHA supplemented cells compared to controls. This effect was not due to an increased cholesterol secretion but mainly to the reduced expression of genes involved in cholesterol metabolism. Results showed that in normal condition (no insulin resistance) 24h supplementation with the bioactives increase glycogen storage and alter the expression of genes of the insulin pathway (AKT, GYS, GLUT2).
The overall effect of bioactive supplementation on hepatocyte lipidome and metabolome was investigated using NMR combined with traditional techniques. DHA supplementation, alone or in combination with PCA or PRO, strongly altered the cell lipid profile, and cells receiving DHA (alone or in combination) were perfectly discriminated from other cells. This “DHA signature” was identified not only in the cell lipidome, but also in the metabolome (Ghini et al, 2017).
We also compared metabolic changes occurring in vitro in human hepatocytes to changes observed in serum and urine of volunteers enrolled in the intervention trial. Overall, we provide evidence that the NMR-based metabolomics approach can predict modifications in the metabolome consequent to the treatment with bioactives in vitro or administration of BEF in vivo.
At molecular level, the analysis of the hepatocytes transcriptome showed a significant modulation of genes involved in cholesterol and fatty acid synthesis following DHA supplementation. PRO appeared to potentiate the effects of DHA whereas there was little effect of PCA co-supplementation. Supplementation with PCA or PRO alone had almost no effect on the transcriptome.
To study the epigenetic effect of bioactives in hepatocytes, two approaches were used: acute treatment (24h) and chronic treatment (4 weeks). In the short treatment, bioactives induced chromatin changes on a global and on a genome wide level. The changes were mainly induced by DHA, alone or in combination. The observed chromatin changes potentially regulate genes that were identified as differential expressed by RNA-seq. All genes are referred to mechanisms which are involved in the pathogenesis of metabolic syndrome, more precisely fatty acid metabolism and chronic inflammatory processes, and were downregulated after the short-term treatment with DHA (alone and in combination).
After the long-term exposure, we observed a clear separation between DHA-treated and not-treated cells at the transcriptomic level, with additional distinction of DHA and the combination treatments. In general, fatty acid-related metabolic processes were downregulated by DHA. Methylation change was mostly very small, so we suppose that DNA methylation is not a major mechanism that contributes to the regulation of expression of the considered genes in the selected cell model (HepG2 cells), but eventually reinforces a gene regulatory effect.
Other molecular studies were performed in the human pre-adipocyte cell system Simpson-Golabi-Behmel syndrome (SGBS) cells. In these cells, an extremely useful new strategy for administration of fatty acids to human adipocytes was developed. This strategy is based on differentiation under serum free conditions followed by adaptation of adipocytes to serum prior to administration of DHA. As in hepatocytes, DHA appeared by far the bioactive with the greatest impact on the transcriptome and promoted subtle but consistent activation of a subset of known adipocyte genes and suppression of proinflammatory genes, which is suggested to collectively promote fatty acid storage and turnover. Bioinformatic analyses of the data indicated that these effects of DHA in adipocytes are primarily mediated by activation of PPARγ. Depending on the subgroup of genes, PRO and PCA displayed positive or negative cross-talk with DHA. This gene group dependent cross-talk between the different bioactives in human adipocytes is novel and uncovering the mechanisms for this cross-talk will be important for the interpretation of the physiological results.
Long term treatment of SGBS preadipocytes with DHA and combinations induced strong changes in DNA methylation, mainly hypomethylation, while PCA and PRO had little effects on this epigenetic mark. Differentially methylated CpG sites identified both in vitro and in vivo (WAT from LIS volunteers) partly overlapped with stable and dynamic enhancer regions of adipocyte differentiation.

The integrated analysis of all results indicated that DHA, alone or in combination, is the main modulator at molecular and cellular level although the effects in adipocytes and hepatocytes are different, as well as the cross-talk between DHA and PRO or PCA. Overall, studies performed in PATHWAY-27 suggest that beneficial effects of DHA in relation to the MetS is mediated in part by the ability to reduce cholesterol synthesis in hepatocytes and adipocytes, and to promote lipid storage and adipokine secretion and to reduce the inflammatory tone in adipocytes.

One of the main aim pf PATHWAY-27 was the preparation, publication and implementation of guidance documents that could inform and assist the food industry sector, especially SMEs, to produce bioactive-enriched foods (BEF) with supportive health claims according to the EU legislation, thereby supporting implementation of the legislation claim framework and enhancing cooperation between complementary scientific disciplines, public-private partnership in the food sector, stakeholder interactions, and – ultimately – the health of consumers.
Two set of guidelines were produced:
➢ Guidelines for the scientific community
This set of Guidelines is addressed to scientists from both academia and the food industry. The guidance helps to understand and apply the relevant steps of the health claim substantiation process. These steps include:
1. Thoroughly reviewing the published evidence concerning any putative beneficial physiological effect(s) of the food or food constituent of interest (e.g. a bioactive compound); and
2. Correctly designing, conducting, interpreting, and reporting any necessary human dietary interventions. The guidance focuses on randomised controlled trials as they are the most rigorous type of interventions to investigate the claimed health effect of a specific food. In this regard, the critical aspects of randomisation, blinding and control are discussed, with hands-on examples from the PATHWAY-27 project of issues that can be encountered when dealing with bioactive compounds and how to solve them. Furthermore, it addresses sample size and data analysis, emphasising the need to involve an experienced (bio)statistician from the outset. Study duration and (non-)compliance are also discussed. As recruitment of study participants can often be a major bottleneck, various recruitment strategies are described. This guidance document also refers to successfully submitted dossiers as well as failed applications and other publically available relevant resources which will help with the appropriate scientific substantiation of health claims.

➢ Guidelines for the food industry
The first step of the work was to conduct a survey amongst relevant stakeholders to collect and identify industry/SME needs and difficulties in establishing and submitting health-claims for food products enriched with health-promoting bioactives. The results of the survey were used to ensure that both industry and scientific guidelines address the issues mentioned by the industry/SMEs (Hegyi et al, 2015).
The proposed integrated Guidelines for food industry/SMEs thus offer a structured product development approach addressing all aspects that SMEs and their suppliers of material, knowledge and related services should consider when designing products with health claims in Europe (Figure 5):
• Description of the targeted cause-effect relationship to obtain the claimed effect;
• Composition of the product including the characterization of the bioactive constituent;
• Measures to ensure the requested minimum quantity of the active constituent within a batch, between different bathes, lots and through the whole shelf-life of the product;
• Selection and verification of health claims to prepare scientific substantiation;
• Food safety assessments such as determination of shelf-life;
• Sensory analysis methods for product acceptability testing;
• Manufacturing ability;
• Financial feasibility;
• Marketing strategy and market launch of products with health claims.

Finalisation of both Scientific and Industry Guidelines was based on interaction with key stakeholders. The Guidance Paper Workshops held in Brussels in September 2017 allowed reviewing the Guidelines before publication (Figure 6).

During this two-day event, workshops in a combination of presentations and working groups sessions were organized to give the opportunity to the end users representatives (industry/SMEs and academic representative, EFSA staff, etc.) of the guidelines to provide feedback on how to improve and strengthen the usefulness and readiness of both documents. The workshop on the Scientific Guidelines was attended by 64 stakeholders and the one on the Industry Guidelines was attended by 50 stakeholders. The feedback received during these events was critical to finalize these Guidelines.
The final version of the PATHWAY-27 guidelines is downloadable free of charge on the PATHWAY-27 website.
An additional set of information supporting the implementation of the European legislation frameworks on health claims was prepared and uploaded on PATHWAY-27 website. The information provided refers to the current Regulation on nutrition and health claim made on foods, a series of workshops organized by EFSA and related Guidance documents that brings key information on how to address health claims related to different health functions (immune system, cognition, etc.) but also reference to peer-reviewed articles showing the importance of how the selection of relevant makers to show an action of the bioactive compound(s) is key when applying for a claim in Europe.
An interactive diagram was produced to ease the understanding of the regulatory framework and to facilitate the finding of the relevant information (Figure 7).

PATHWAY-27 provided two sets of guidelines based on PATHWAY-27 experience. The PATHWAY-27 guidelines are downloadable free of charge on the PATHWAY-27 website.