Importance and differences of fatty acids in the human fecal metabolome

Several fatty acids were already detected to be significantly changed between the baseline, HRS or LRS diet (Maier et al. 2017) through the correlation studies or the pathways classification analysis.

However, in this chapter, the overall fatty acid profile impacted through the different diets will be shown (Figure 2.4-18). In general, twenty-eight fatty acids, both, saturated and unsaturated were found to be significantly changed between fecal samples of the baseline diet and both RS samples.

Figure 2.4-18: Fatty acids significantly changed through diet.

Fold change values for fatty acids significantly increased or decreased in HRS (red) or LRS diet (green), analyzed in (-) FT-ICR-MS mode. Left: Log2 fold change values of fatty acids significantly increased or decreased in HRS (red) or LRS (green) diet compared to baseline diet. Right: Log2 fold change values of fatty acids significantly increased or decreased in the HRS diet compared to the LRS diet. Further details are given in Table 6.1-16. From Maier, T. V.; Lucio, M.; Lee, L. H.; VerBerkmoes, N. C.; Brislawn, C. J.; Bernhardt, J.; Lamendella, R.; McDermott, J. E.; Bergeron, N.; Heinzmann, S. S.; Morton, J. T.; González, A.; Ackermann, G.; Knight, R.; Riedel, K.; Krauss, R. M.; Schmitt-Kopplin, P.; Jansson, J. K.: Impact of Dietary Resistant Starch on the Human Gut Microbiome, Metaproteome, and Metabolome. mBio vol. 8 no. 5 e01343-17 (2017).Illustration and data depiction modified from (Maier et al. 2017). Copyright (2017) Maier et al., Information about the creator and respective contributions, as well as the original material are available: http://mbio.asm.org/content/8/5/e01343-17.full with the original title:

Classes of fatty acyls grouped according to their relative abundance following a specific diet category. Licence notice: https://creativecommons.org/licenses/by/4.0/.

The saturated fatty acids, heptadecanoic acid (C17:0), octadecanoic acid (C18:0), nonadecanoic acid (C19:0), icosanoic acid (C20:0), docosanoic acid (C22:0), tricosanoic acid (C23:0), tetracosanoic acid (C24:0), pentacosanoic acid (C25:0) and hexacosanoic acid (C26:0) were significantly increased to a similar extend by through dietary starch intake and showed a log2 fold change > 1 for the HRS compared to the baseline diet, and a log2 fold change > 0.53 for the comparison of the LRS diet and the baseline diet.

-6.00 -4.00 -2.00 0.00 2.00 4.00 Log2 fold change

Log2 FC HRS vs LRS

On the contrary, the fecal samples of the participants consuming the HRS diet were additionally characterized by increased levels of unsaturated fatty acids, namely icosapentanoic acid (C20:5), icosenoic acid (C20:1), heneicosenoic acid (C21:1), docosatrienoic acid (C22:3), docosadienoic acid (C22:2), docosenoic acid (C22:1), tricosenoic acid (C23:1), tetracosenoic acid (C24:1), pentacosatrienoic acid (C25:3), hexacosatrienoic acid (C26:3), hexacosadienoic acid (C26:2) and nonacosatrienoic acid (C29:3) with log2 fold changes > 0.7. Thereof, nonacosatrienoic acid (C29:3) and pentacosatrienoic acid (C25:3) showed decreased levels in the LRS diet compared to the baseline diet.

Comparing the fatty acid profiles between the HRS and LRS diet, almost all of the above mentioned fatty acids, except C18:0, C22:0, C24:0, C26:0 and C20:1, were strongly increased in the HRS diet (log2 fold change from 0.8 – 3.2 HRS compared to LRS). Interestingly, pentadecanoic acid (C15:0) was increased in both RS diets. However, the increase observed in the LRS diet was slightly higher, compared to the HRS diet.

Additionally, octadecenoic acid (C18:1, Figure 2.4-19 C) and octadecadienoic acid (C18:2, Figure 2.4-19 B) were two of the most abundant metabolites (mean intensity >1x10¹¹) and significantly altered between the baseline diet and the two RS diets (Maier et al. 2017). Further, palmitoleic acid (C16:1, Figure 2.4-19 A) was increased in the baseline diet.

Figure 2.4-19: Unsaturated fatty acids increased in baseline diet.

Boxplots of 3 unsaturated fatty acids significantly increased in the baseline diet (blue) compared to the HRS diet (red) and the LRS (green), analyzed in (-) FT-ICR-MS mode. p-values were calculated through the Kruskal-Nemenyi Test. Further details are listed in Table 6.1-16.

In contrary, Sun et al. found octadecadienoic acid and octadecenoic acid to be enriched in samples of pigs consuming a raw potato starch diet, though in cecal samples of the pigs (Sun et al. 2016). They hypothesized that through raw potato starch, the significantly higher concentrations of unsaturated fatty acids in cecal samples may appear by a lower absorption of those fatty acids in the cecum (Sun et al.

2016). This might be an explanation, why in this study lower concentration levels of octadecadienoic

Intensity x 10¹¹

acid and octadecenoic acid were detected in the fecal samples after RS consumption. By contrast, already in 1974, Eyssen and Parmentier investigated the impact of starch diets, compared to lactose diets on the intestinal microflora of germfree and conventional rats (Eyssen and Parmentier 1974). They were interested in the fecal fatty acid profiles, which were analyzed by gas-liquid chromatography and revealed fatty acids, such as octadecanoic acid (C18:0) to be increased in conventional rats fed the starch diet and octadecenoic acid (C18:1) and octadecadienoic acid (C18:2) to be decreased in the conventional rats fed the starch diet. This was in agreement with our results on the profile of the above mentioned fatty acids in the human fecal samples through dietary starch intake. Remarkably, three fatty acids, which were decanoic acid (C10:0, Figure 2.4-20 A), dodecanoic acid (C12:0, Figure 2.4-20 B) and tetradecanoic acid (C14:0, Figure 2.4-20 C) showed increased levels in the fecal samples of participants receiving the LRS diet. In order to undoubtedly identify the significant fatty acids enriched solely in samples of participants consuming the LRS diet, this result was confirmed by applying a lipidomics-MS/MS approach in (-) ESI mode (method description in chapter 2.3.1.6) including respective standards (extracted ion chromatograms and MS/MS spectra are illustrated in Figure 5.1-1) (Maier et al. 2017).

Figure 2.4-20: Saturated fatty acids significantly increased in samples of LRS diet.

Boxplots of 3 saturated fatty acids significantly increased in the LRS diet (green) compared to the baseline diet (blue) and HRS diet (red), analyzed in (-) FT-ICR-MS mode (top) and (-) ToF-MS (bottom) experiments. p-values were calculated through the post hoc Kruskal-Nemenyi Test. Further details are listed in Table 6.1-16. From Maier, T. V.; Lucio, M.; Lee, L. H.; VerBerkmoes, N. C.; Brislawn, C. J.; Bernhardt, J.; Lamendella, R.; McDermott, J. E.;

Bergeron, N.; Heinzmann, S. S.; Morton, J. T.; González, A.; Ackermann, G.; Knight, R.; Riedel, K.; Krauss, R. M.;

Schmitt-Kopplin, P.; Jansson, J. K.: Impact of Dietary Resistant Starch on the Human Gut Microbiome, Metaproteome, and Metabolome. mBio vol. 8 no. 5 e01343-17 (2017).Illustration of Figure B and C was modified from (Maier et al. 2017). Copyright (2017) Maier et al., Information about the creator and respective contributions, as well as the original material are available: http://mbio.asm.org/content/8/5/e01343-17.full with the original title:

Identification of decanoic (C12:0) and tetradecanoic acid (C14:0). Licence notice:

https://creativecommons.org/licenses/by/4.0/.

Intensity x 10³ Intensity x 10⁵

0

In order to find out where those differences originate from, the composition of both diets was considered.

Certainly, diets were matched for protein, fat and total carbohydrate, and differed only with respect to the source of starch used (which was either high or low in resistant starch). It is therefore unlikely that the differences in decanoic acid, dodecanoic acid and tetradecanoic acid between the HRS and LRS diet occurred from differences in fat content between the diets. For this reason, those fatty acids appeared to be affected by the digestion of cornstarch low in resistant starch.

Further, several dicarboxylic acids, such as dodecenedioic acid, octadecanedioic acid, eicosanedioic acid and tricosanedioic acid were detected to be significantly increased in the HRS diet, compared to both, the baseline and the LRS diet (Figure 2.4-21).

Figure 2.4-21: Dicarboxylic acids significantly increased in the HRS diet.

Boxplots of 4 dicarboxylic acids significantly increased in the HRS diet (red) compared to the baseline diet (blue) and the LRS diet (green), analyzed in (-) FT-ICR-MS mode. p-values were calculated with the Kruskal-Nemenyi Test. Further details are listed in Table 6.1-17.

Through the conversion of monocarboxylic acids to their dicarboxylic acids in rat liver in vitro, Jin et al.

wanted to investigate unsaturated dicarboxylic acids and their excretion in urine (Jin and Tserng 1990).

They hypothesized dicarboxylic acids, especially dodecenedioic acid to be a metabolic precursor of octenedioic acids. In turn, octenedioic acids are derived from oleic and linoleic acids through several metabolic oxidation processes (Jin and Tserng 1990). However, they also concluded that the origin of dicarboxylic acids can be caused by multiple metabolic processes (Jin and Tserng 1990). In our study, increased levels of octadecadienoic acid (linoleic acid) and octadecenoic acid (oleic acid) were observed in the baseline diet. Accordingly, higher levels of dicarboxylic acids were detected through HRS consumption. Considering the results from Jin et al. in the 90’s, our result lead to assume that through RS consumption the metabolic conversion to several dicarboxylic acids is affected.

Intensity x 10⁸