The Lipid Status in Patients with Ulcerative Colitis: Sphingolipids are Disease-Dependent Regulated

Abstract

The factors that contribute to the development of ulcerative colitis (UC), are still not fully identified. Disruption of the colon barrier is one of the first events leading to invasion of bacteria and activation of the immune system. The colon barrier is strongly influenced by sphingolipids. Sphingolipids impact cell-cell contacts and function as second messengers. We collected blood and colon tissue samples from UC patients and healthy controls and investigated the sphingolipids and other lipids by LC-MS/MS or LC-QTOFMS. The expression of enzymes of the sphingolipid pathway were determined by RT-PCR and immunohistochemistry. In inflamed colon tissue, the de novo-synthesis of sphingolipids is reduced, whereas lactosylceramides are increased. Reduction of dihydroceramides was due to posttranslational inhibition rather than altered serine palmitoyl transferase or ceramide synthase expression in inflamed colon tissue. Furthermore, in human plasma from UC-patients, several sphinglipids change significantly in comparison to healthy controls. Beside sphingolipids free fatty acids, lysophosphatidylcholines and triglycerides changed significantly in the blood of colitis patients dependent on the disease severity. Our data indicate that detraction of the sphingolipid de novo synthesis in colon tissue might be an important trigger for UC. Several lipids changed significantly in the blood, which might be used as biomarkers for disease control; however, diet-related variabilities need to be considered.

Keywords: DHA; EPA; LC–MS/MS; S1P; ceramide; patient; sphingolipid; ulcerative colitis.

Figure A1

In healthy colon tissue, tight junctions stabilize the epithelial layer. The functionality of the tight junctions is dependent on proper membrane lipid composition, which is warranted by, e.g., vesicular delivery of sphingolipids that are built de novo in the endoplasmic reticulum (ER). In colon tissue from ulcerative colitis patients, de novo sphingolipid synthesis is reduced, leading to the disruption of membrane integrity. Furthermore, LacCers, which can function as pattern-recognition receptors, are upregulated, leading to an enhanced binding of pathogens and thereby, activation of immune cells.

Figure 1

Patient data. (A) Time of disease duration in our ulcerative colitis (UC) patient cohort. (B) Overview of medical treatments of UC patients included in this study.

Figure 2

Sphingolipids in colon tissue. Colon samples from inflamed and control tissue were collected from UC patients undergoing diagnostic colonoscopy and sphingolipid concentrations where analyzed by LC–MS/MS. n = 27. Statistical differences were analyzed by a paired T-test, α = 0.05, (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 3

qRT-PCR and immunohistochemical staining of CerS and SPT in colon tissue. (A) mRNA levels of Sptlc2, Sptlc3 (SPT), CerS1-6 in inflamed and control tissue were detected by quantitative RT-PCR. CT levels were normalized to GAPDH and subsequently, related to the level of control (set to 1), using 2^(-ΔΔCT) method. Data are means ± SEM; n = 10–15. Statistical differences were analyzed by GraphPad Prism using unpaired T-test, p ≤ 0.05. (B) Immunohistochemistry of CerS3 in human colon tissue. CerS3 was stained with an anti-CerS3 antibody and a secondary antibody coupled to Cy3 and tissue was co-stained with DAPI. Original magnification: 63x and 20x; white arrow indicates cytoplasmic CerS3 staining in cell localized in the lamina propria.

Figure 4

Sphingolipids in human plasma. Sphingolipids in human plasma was determined by LC–MS/MS. n_control = 148; n_remission = 102, n_moderate = 43; n_acute = 24, Statistical analysis was performed with R, significant differences between groups were assessed using two-tailed ANOVA with Tukey multiple comparison, α = 0.05. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 5

Sphingolipids in human plasma in relation to medical treatment. n_none = 6; n_{anti TNF-α} = 21; n_{aminosalicylate} = 50. Statistical analysis was done with R, significant differences between groups were assessed using two-tailed ANOVA with Tukey multiple comparison, α = 0.05. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 6

qRT-PCR of CerS in white blood cells from colitis patients and controls. mRNA levels of CerS1-6 in white blood cells (WBCs) were detected by quantitative RT-PCR. CT levels were normalized to GAPDH and subsequently related to the level of the control cohort (set to 1), using 2^(-ΔΔCT) method. Data are means ± SEM; n = 9–15. Statistical differences were analyzed by one-way ANOVA with α Bonferroni post-hoc test. α = 0.05. (* p < 0.05, ** p < 0.01, *** p < 0.001).

Figure 7

Lipids in human plasma. (A) Volcano blot of lipids in human plasma determined by LC-QTOFMS. Shown are the significances versus fold-changes of all lipids measured in plasma of UC patients who are in remission or suffer from mild disease. n_remission = 109, n_moderate = 33. Statistical analysis was performed with R, significant differences between groups were assessed using two-tailed ANOVA with Tukey multiple comparison, α = 0.05. (B) Relative amounts of EPA and DHA in plasma of UC patients with different disease stages. n_remission = 109, n_moderate = 33; n_acute = 19. Statistical analysis was performed with GraphPad, significant differences between groups were assessed using two-tailed ANOVA with Tukey multiple comparison, α = 0.05. (* p < 0.05, ** p < 0.01, *** p < 0.001).