We investigated plasma biomarkers of inflammation, immune activation, kynurenines, and mucosal injury in relation to ETEC proliferation during human experimental ETEC infection. CRP, neopterin, and serum amyloid A levels increased substantially and reached maximum levels 3 days after dose ingestion, while no clear changes were seen in these levels for any of the LP volunteers. Plasma calprotectin levels decreased from baseline in the LP group. Among the kynurenines, KTR increased until day 3 and Pic decreased in the SP group. B6 vitamers, cofactors of kynurenines, PLP, and PL were lower in the SP group on day 3. The mucosal injury marker Reg3a increased in circulation until day 2, particularly in the SP group.

CRP levels increased in the SP group on day 3 after dose ingestion, indicating a mild systemic inflammatory condition induced by ETEC. A rise in CRP usually reaches the max level within 48 h after microbe invasion and tissue damage, and this resonates well with the peak point for CRP since the median time to first symptom debut was 25.5 h for volunteers with the diarrheal symptom in the SP group. CRP creates a pro-inflammatory environment by activating complement, phagocytosis, nitrogen oxide (NO), and the production of pro-inflammatory cytokines, and eliciting adaptive, humoral immune reactions [43].

SAA is another major acute-phase reactant, and this pro-inflammatory protein exists in several isoforms [44]. In our study, the levels of total SAA and most isoforms increased until day 3 in the SP group. SAA, especially SAA1, contributes to innate immunity by opsonization, especially against Gram-negative bacteria, including Escherichia coli and Pseudomonas aeruginosa, via phagocytosis by IL-10- and TNF-α-stimulated macrophages [45]. Therefore, it can be argued that SAA plays an active role in protective innate immunity during ETEC infection. In addition, SAA contributes by inducing cytokines, activating cellular surface receptors such as Toll-like receptors 2 and 4, CD36 and stimulating the inflammasome cascade [46].

The truncation pattern of SAA1.1 was altered during the infection phase, where relatively lower levels of SAA1.1 were found as truncated variant on day 2 and 3, and normalizing by day 7. This finding of cessation of SAA1.1 truncation during the early infection phase indicates a potential role of SAA1.1 in early immunity; however, their exact roles and truncation are yet to be studied. Truncations are mainly the result of alternative splicing and cleavage and may also be associated with changes in the expression of truncated forms, protease activities, or its potential function in opsonization.

In the present study, plasma neopterin levels were increased on day 3 in the SP group. A similar increase has been previously reported in individuals infected with Shigella [47], an invasive pathogen that normally causes more severe infections than ETEC. In contrast, previous studies found lower fecal neopterin levels in infants positive for ETEC and other gut pathogens in infants in LMIC [8, 48]. This discrepancy can be explained by neopterin functioning as an immunoattractant in circulation rather than secretion into the gastrointestinal tract or biological differences in infants in LMIC.

Calprotectin levels showed a slow decrease until day 3 in the LP group. The S100A8-S100A9 complex functions as an antimicrobial by chelating Zn, Fe, and Mn ions, which are essential for bacterial growth, as observed in an in vivo experiment with E. coli [49]. S100A9 itself is also known to enhance phagocytosis by neutrophil granulocytes, as reported for E. coli in vivo [50]. In Bangladesh, higher fecal calprotectin levels have been observed in ETEC-positive infants [8]. In patients with inflammatory bowel diseases (IBD) with high fecal calprotectin, fecal and plasma calprotectin usually showed positive associations [51].

KTR showed a sharp increase until day 3 in the SP group. Increased KTR has been highlighted as a candidate for prognostic biomarker for the severity of environmental enteric dysfunction (EED), tested in infants in low-income countries [52]. It is worth noting that in pigs experimentally infected with ETEC, supplementary tryptophan did not have any effect on CRP-level after infection, diarrheal disease, or ETEC shedding, but increased serum serotonin levels, which in turn increased post-weaning performance [53].

An increase in KTR is a result of increased IDO activity, as IDO transforms tryptophan to kynurenine [19], and increased IDO activity can be both beneficial and detrimental for host immunity [54]. IDO protects the host by two mechanisms: first, by degrading tryptophan, a crucial amino acid for the survival of many microorganisms, including E. coli [55], and second, by producing kynurenine, which has a bactericidal function. However, IDO activity limits host immunity by inducing the proliferation of regulatory T (Treg) cells, blocking the conversion of Tregs into T effector cells, and increasing T cell apoptosis [20, 54]. Interestingly, uropathogenic E. coli (UPEC) is known to induce IDO in the host environment to utilize these immunosuppressive effects for survival and colonization in the host [55].

Picolinic acid (Pic) levels decreased in both groups until day 3, with a more pronounced decrease in the SP group. Pic has been reported to exert antiviral and antimycobacterial effects during HIV and Mycobacterium avium infections, respectively [56, 57]. Similar to discussions on the KTR, Pic has been shown to have both beneficial and non-beneficial immunological properties during infection. Pic can activate macrophages [57, 58], but can also inhibit the proliferation of CD4+ T cells [59].

Among vitamin B6 forms, PLP, the prevailing bioavailable form in plasma, decreased over time and was negatively correlated with ETEC proliferation. These findings imply exhaustion or reduced uptake of vitamin B6. Thus, these results imply that vitamin B6 could be a beneficial supplement for ETEC infection. Vitamin B6 has antibacterial functions, and its deficiency can negatively affect antibody production and IL-2 and T cell proliferation [60]. Vitamin B6 is a cofactor in the kynurenine pathway, which is associated with ETEC infection [20]. It is also worth noticing that pyridoxine is damaged in autoclaved formula for infants [61], stressing the value of breastfeeding and/or proper administration of vitamin B6 in other forms to young children residing in endemic areas.

The PAr index is known to increase in a chronic inflammatory environment, especially in cardiovascular diseases [18, 62], but no changes were observed in the present study. This may indicate that vitamin B6 catabolism is not significantly affected during acute, temporary, or mild infection.

Reg3a levels increased to day 3 in the SP group. In porcine diarrheal infection with Lactobacillus casei, murine Reg3a promotes intestinal cell proliferation and reduces the bacterial load [63], thus indicating the possibility that Reg3a also plays a role in intestinal host immunity in ETEC infection. In addition, Gazi et al. found higher levels of fecal Reg1b, the same family as Reg3a, in enteroaggregative E. coli (EAEC)-positive infants but not in ETEC-positive infants [8]. Reg1b is a newly suggested gut-specific marker related to cell proliferation and regeneration and has been suggested as a prognostic marker for stunting among infants in Bangladesh and Peru [64]. This suggests that plasma Reg3a may also be a biomarker of intestinal injury. Elevated systemic inflammatory markers and Reg3a, may also signify a compromised gut barrier function due to inflammation, commonly referred to as 'increased intestinal permeability' or 'leaky gut'.

Reg3a levels positively correlated with calprotectin levels (total S100A8/A9) on days 0 and 3. Interestingly, plasma Reg3a levels increased over time, but calprotectin levels tended to decrease. Further studies on the relationship between these two markers are warranted.

Although not significant, a quick increase in iFABP on day 1 in both groups can indicate mucosal damage after ETEC infection, returning to the baseline level on day 3. A previous human challenge study with ETEC strain H10407 found an increase in plasma iFABP peaking on day 3 in six volunteers with moderate to severe diarrhea and in six volunteers without symptoms [2]. This may be due to the ETEC strain used, as H10407 is regarded as an atypical, virulent ETEC strain with three toxin types and normally causes more severe diarrheal disease than the two strains used in the present study. Instead of iFABP, we identified Reg3a and calprotectin as potential plasma markers that showed changes over time.

Plasma CRP, SAAt, and PAr levels on day 3 showed positive correlations with stool ETEC DNA levels, indicating that higher levels of intestinal ETEC proliferation are associated with increased levels of systemic inflammation.

Our findings support the presence of induced systemic inflammation and mucosal injury during non-invasive ETEC infection, notably marked by an increase in CRP and SAAt with a decreased truncation pattern in SAA1.1 and Reg3a during the early infection phase in the SP group. Inflammatory markers were correlated with mucosal injury markers and maximum ETEC proliferation levels. The plasma concentrations of some kynurenines, namely KTR and Pic, and their correlations with inflammation markers indicate their roles in the inflammatory response during ETEC infection. Plasma PLP is a known cofactor and decreased over the follow-up period in the SP group, inversely correlating with the maximum ETEC proliferation level, which emphasizes the importance of the kynurenine pathway. Reg3a, a mucosal injury marker, increased in the SP group, indicating its potential protective role for the gut mucosa. These findings add to our understanding of how ETEC infection can impact host inflammatory and immune responses and may have implications for long-term negative health outcomes following ETEC infection.

In this study, we only measured biomarkers in plasma, but incorporating measurements of some of these markers in stool samples would have offered more gastrointestinal-specific validation of our findings. In contrast to other studies that used only one strain for experimental infection, we used two strains, TW11681 and TW10722, which might vary somewhat in their ability to elicit mucosal injury and inflammation. However, we did not see a clear difference between the strains' ability to cause inflammation in our data, and the number of volunteers was too small to directly compare the two ETEC strains. We chose to use the maximum proliferation levels for the formation of groups in this study, rather than inoculum dose or diarrheal disease, since gut lumen levels of bacteria vastly outnumber the initial dose when ETEC colonizes.

It is possible that some biomarker actually peaked on days 4, 5, or 6 during the experimental infection phase, when samples were not collected. Also, some of the stools that were passed by the volunteers could not be analyzed as detailed in Vedøy et al. [24, 27], with a possibility that we did not capture the absolute maximum ETEC stool concentrations. This may be the reason for poor correlations between plasma biomarkers and ETEC proliferation.

This study was performed on healthy young Norwegian adults, and our results may therefore not be directly comparable to young children living in ETEC endemic areas, where ST-ETEC has been shown to cause severe, sometimes fatal, diarrhea and longer-term stunting [4]. There are several reasons contributing to this higher disease burden and sequels in young children, such as frequent enteric co-pathogens, malnutrition and small bodies more quickly becoming dehydrated [65, 66]. Although data are still scarce on ETEC induced systemic inflammation in children, they may play an important role for disease severity in young children.