Fleur Jansen

The browning process that gives cooked foods their appealing aromas, flavors and textures is known as the Maillard reaction. This non-enzymatic reaction between proteins and sugars, also referred to as protein glycation, leads to the formation of covalent adducts on lysine and arginine residues; a heterogenous group of compounds collectively called advanced glycation end products (AGEs). Following the discovery of glycated hemoglobin (HbA1c) as a diagnostic marker for diabetes in the 1960s, endogenous protein glycation was shown to be relevant to human disease, which subsequently raised concerns about the safety of dietary AGEs. These concerns were reinforced when high consumption of processed foods in the Western diet became associated with an increased risk of the same non-communicable diseases.

Dietary AGEs enter the body through the digestive tract. After digestion, they are mainly absorbed as free AGEs or dipeptides, and current literature on the health effects of high AGE intake therefore predominantly focuses on the systemic effects of these bioavailable forms. However, only protein-bound AGEs are able to bind to the receptor for AGEs (RAGE), which is involved in inflammatory signaling. In this context, the intestine would be the primary target organ, as it might be the only tissue to encounter undigested glycated dietary proteins. However, their impact at the intestinal level remains largely unknown. Yet, intestinal health is known to play a key role in general wellbeing, so this interaction might be of influence to systemic health outcomes. Therefore, the aim of this thesis was to increase our understanding of how intact dietary glycated proteins affect intestinal health at the epithelium and immune level, using in vitro models.

Main findings
The currently available in vitro studies on the inflammatory effects of glycated dietary proteins on the intestinal epithelium face multiple methodological limitations and should therefore be interpreted with caution (Chapter 2). Key concerns include the limited chemical characterization of the glycated samples, including insufficient endotoxin quantification, and the absence of appropriate positive controls for RAGE activation. The subsequent chapters of this thesis aimed to address these uncertainties. In Chapter 3, both monomers and aggregates of undigested glycated proteins (beta-lactoglobulin) were shown to cross a conventional air-liquid interface monolayer, mimicking the ileal villous epithelium, most likely via paracellular transport. This finding challenges the prevailing assumption that undigested glycated proteins are not bioavailable, leaving open the possibility that they might interact with cell types beyond the epithelium, for example intestinal immune cells. Since protein-bound AGEs are thought to induce inflammation through interaction with RAGE, these glycated proteins were subsequently tested on primary immune cells in Chapter 4. Some notable observations emerged. First, RAGE appeared predominantly expressed intracellularly across multiple cell models, while RAGE was only substantially expressed on the cell surface of specific in vitro models, for example in M-CSF differentiated M0 macrophages. Second, LPS scavenging drastically reduced the inflammatory responses attributed to glycated dietary proteins, despite endotoxin levels being considered neglectable based on common quantification methods. In contrast, a RAGE antagonist did not attenuate these responses, suggesting that LPS may act as a confounding factor in such experiments. Furthermore, stimulation with known RAGE agonists (e.g. HMGB1 or Aβ-42) did not induce an inflammatory response, raising the question whether LPS might function as a necessary co-factor in RAGE-mediated processes. This hypothesis was explored in Chapter 5, but the preliminary results obtained were mixed. RAGE agonists did not enhance the inflammatory potency of LPS in the primary cell model used previously, but occasional amplification of NF-kB activity was observed in NF-kB-reporter cells. This result was however not consistently reproducible and was not accompanied by increased pro-inflammatory cytokines production, as would be expected, suggesting that alternative downstream outcome may be involved.

Discussion and conclusion
Based on the main findings of this thesis, it is proposed that a combination of stimuli is required to induce RAGE-dependent inflammatory signalling, suggesting that glycated dietary proteins alone are unlikely to induce inflammation in healthy individuals. When placed in the broader context of dietary AGE consumption and its implications for human health, glycation can also contribute to lysine deficiency and some absorbed free AGEs (particularly HMGB1) are suggested to exert nephrotoxic effects. However, as discussed in Chapter 6, the consequences of a high-GI diet and endogenous AGE formation appear to be far more pronounced. From this perspective, the societal relevance of further dietary AGE research, particularly in healthy populations, seems limited.