Postprandial Chylomicron Output and Transport Through Intestinal Lymphatics Are Not Impaired in Active Crohn’s Disease

The pathophysiology of Crohn's disease (CD) remains obscure. From early descriptions, a defect in lymphatic transport was described,1Kruiningen HJ Van et al.Gut. 2008; 57: 1-4Crossref PubMed Scopus (103) Google Scholar but this issue remains unresolved. A tractable approach for investigating the transport of lymphatic cargo is to quantify postprandial chylomicron appearance in plasma. Chylomicrons, large lipoproteins synthesized by enterocytes during dietary absorption, pass through intestinal and mesenteric lymphatics before arriving in plasma.2Xiao C. et al.Cell Mol Gastroenterol Hepatol. 2019; 7: 487-501Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar Although chylomicron absorption typically occurs in the duodenum and jejunum, variables like enterocyte membrane status can shift absorption to distal parts of the intestine where CD is most commonly involved.3Wang B. et al.Cell Metab. 2016; 23: 492-504Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar Beyond concerns of lymphatic dysfunction, other observations raise the possibility that postprandial chylomicron responses might be abnormal in CD. First, ileal inflammation may cause bile acid malabsorption and thereby impair fat emulsification during absorption.4Uchiyama K. et al.J Immunol Res. 2018; 2018: 7270486Crossref PubMed Scopus (19) Google Scholar Second, down-regulation of genes encoding chylomicron-associated apoproteins apoB, apoA1, and apoC3 has been observed in ileal CD.5Haberman Y. et al.J Clin Invest. 2015; 124: 3617-3633Crossref Scopus (332) Google Scholar Third, enteroendocrine hormones like glucagon-like peptide (GLP) 2 released by specialized enterocytes located in the ileum regulate chylomicron output, increasing mesenteric lymph flow rates in rats.2Xiao C. et al.Cell Mol Gastroenterol Hepatol. 2019; 7: 487-501Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar Here, we quantified postprandial chylomicron transport after feeding a mixed meal containing 13C-triolein to healthy participants compared with those having active CD of the ileum. This study, including 28 participants with ileal CD and 19 healthy control individuals, was conducted as approved by the Human Research Protection Office at Washington University (protocol no. 201712103). Inclusion criteria for participants with CD required the presence of active small bowel inflammatory disease. Patients underwent assessment for active small bowel CD through a combination of ileocolonoscopy and/or computed tomography and/or magnetic resonance enterography. All ileocolonoscopies were scored for disease activity by a single gastroenterologist with fellowship training in inflammatory bowel disease. All enterography scans were conducted at the Mallinckrodt Institute of Radiology, Washington University School of Medicine, and read with clinical reports generated by board-certified and fellowship-trained abdominal radiologists. Time between clinical evaluation and study visit averaged 40 ± 20 (standard deviation) days. Exclusion criteria were previous relevant abdominal resections (eg, bowel resection, gall bladder removal), use of steroids in the past 7 days, or a diagnosis of or treatment for underlying diabetes or other metabolic diseases (eg, gout, thyroid disease). Biologic therapy for CD was not excluded and is reported in Supplementary Table 1, along with other participant characteristics. Control participants, recruited from the community, received a complete metabolic panel screening of fasting plasma to verify healthy status, applying the described exclusion criteria. Other methods used in this study can be found in the Supplementary Methods. We quantified the accumulation of 13C-triolein and associated meal triglycerides within plasma chylomicron fractions over a 6-hour postprandial window. The tracer-to-tracee ratio enrichment of test meal oleate was 2.1% ± 0.5% (standard deviation). Chylomicron-triglyceride oleate derived from the test meal was similar between healthy control participants and CD participants (Figure 1A). The proportion of test meal tracer expired in breath was also similar between groups (Figure 1B), suggesting no obvious malabsorption. Fractional clearance rate of chylomicron-triglyceride from plasma was also not different between groups (Figure 1C). There was a trend (not statistically significant) toward elevated postprandial chylomicron-triglyceride that did not arise from the test meal (Figure 1D). One variable was able to account for this trend: the 8 participants with CD taking adalimumab (all 8) or infliximab (1 participant transitioning from adalimumab) as tumor necrosis factor inhibitors (TNFi) displayed markedly elevated postprandial chylomicron-triglyceride derived from a nonmeal source, compared with control participants or other participants with CD (Figure 1E). Furthermore, the examination of enteroendocrine hormones showed that participants with CD taking TNFi uniquely displayed marked increases in postprandial glucose-dependent insulinotropic peptide (GIP) (Figure 1F and Supplementary Figure 1A), the release of which occurs in response to the secretion of chylomicrons.6Gribble F.M. et al.Nat Rev Endocrinol. 2019; 15: 226-237Crossref PubMed Scopus (231) Google Scholar Lesser to no changes in GLP-1)and GLP-2 were observed (Supplementary Figure 1B and C). Immunoblots identified apoB48, not apoB100, protein in the fractions, confirming that the traced lipids were from chylomicrons (Figure 1G). The total apoB48 in the chylomicron fractions was similar after TNFi treatment (Supplementary Figure 1D), such that the triglyceride-to-apo48 ratio was increased (Figure 1H), suggesting enlarged chylomicrons in participants taking TNFi (Figure 1I). Electron microscopy confirmed elevated size of plasma chylomicron remnants (Figure 1I). We addressed whether the lymphatic-dependent transport of chylomicrons was decreased in individuals with active CD, given the possibility that lymphatic defects may occur in CD. However, chylomicron transport was not impaired. Rather, a detailed analysis of chylomicron-triglyceride appearance in the postprandial state indicated normal lipid absorption and delivery to plasma in CD. It remains possible and perhaps likely that, because chylomicrons are mainly carried by lymphatics in the duodenum and jejunum,3Wang B. et al.Cell Metab. 2016; 23: 492-504Abstract Full Text Full Text PDF PubMed Scopus (73) Google Scholar upstream of where disease is focused, ileal lymphatic transport defects exist and were not captured in this study. Methods to trace ileal lymphatic cargo remain to be defined but will be of interest for future studies. This study also does not support the idea that fat malabsorption typifies CD, at least in the inflammatory phenotype of CD without fistulizing complications or prior resection of bowel segments or gall bladder. Although we did not set out to study the impact of specific therapeutics on the postprandial response, we observed that TNFi promoted an early output of enlarged chylomicrons loaded with lipids not derived from the test meal. These lipids may arise from fats from prior meals retained in an epithelial storage pool.2Xiao C. et al.Cell Mol Gastroenterol Hepatol. 2019; 7: 487-501Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar The enhanced output of chylomicrons was strongly associated with increased secretion of the enteroendocrine hormone GIP, secreted downstream of chylomicron release.6Gribble F.M. et al.Nat Rev Endocrinol. 2019; 15: 226-237Crossref PubMed Scopus (231) Google Scholar GIP has been linked to weight gain and increased bone mass.7Møller C.L. et al.J Clin Endocrinol Metab. 2016; 101: 485-493Crossref PubMed Scopus (40) Google Scholar Although GIP has scarcely been investigated in association with CD or CD therapy, it is known that treatment of CD with TNFi increases fat mass8Parmentier-Decrucq E et al. 2009;15:1476-1484.Google Scholar and bone mass. The mechanisms underlying how TNFi might affect lipid loading on chylomicrons and whether a rise in GIP might account for some effects linked with TNFi should be addressed in future prospective studies. It will also be important in future studies to determine if the effect of TNFi in the postprandial response is observed only in participants with CD or in other patient populations taking TNFi. The authors are indebted to those who had important roles in advising on methods including statistics, providing technical assistance, recruiting patients, and/or providing overall advice and resources, including Washington University colleagues Nicole K. H. Yiew, Shashi B. Kumar, George P. Christophi, Adewole L. Okunade, Ling Chen, Sewuese E. Akuse, Adam J. Bittel, W. Todd Cade, and Nicholas O. Davidson. We thank Shaji Chacko (Baylor University) and the Baylor Children's Nutrition Research Center core facility for mass spectroscopy analysis of breath samples. The authors thank Ross Kossina and Gregory Strout for assistance with electron microscopy, performed at the Washington University Center for Cellular Imaging, which is supported in part by the Children's Discovery Institute of Washington University and St Louis Children's Hospital (CDI-CORE-2015-505 and CDI-CORE-2019-813) and the Foundation for Barnes-Jewish Hospital (3770 and 4642). The authors extend additional thanks to the nursing and kitchen staff of the Clinical Research and Translational Unit for their expert assistance. The authors are grateful for the advice and insight generously shared in conversations with Drs Robert Hirten and Jean-Frederic Colombel (Mount Sinai) and Elizabeth Parks (University of Missouri). Li-Hao Huang, PhD (Data curation: Equal; Formal analysis: Equal; Investigation: Lead; Methodology: Equal; Validation: Lead; Visualization: Equal; Writing – original draft: Equal; Writing – review & editing: Supporting); Parakkal Deepak, MD (Conceptualization: Supporting; Investigation: Supporting; Methodology: Supporting; Writing – review & editing: Supporting); Matthew A. Ciorba, MD (Conceptualization: Supporting; Methodology: Supporting; Supervision: Supporting; Writing – review & editing: Supporting); Bettina Mittendorfer, PhD (Conceptualization: Supporting; Data curation: Supporting; Methodology: Supporting; Resources: Supporting; Writing – review & editing: Supporting); Bruce W. Patterson, PhD (Conceptualization: Supporting; Formal analysis: Lead; Methodology: Lead; Supervision: Supporting; Validation: Equal; Visualization: Supporting; Writing – review & editing: Supporting); Gwendalyn J. Randolph, PhD (Conceptualization: Lead; Data curation: Equal; Formal analysis: Equal; Funding acquisition: Lead; Investigation: Supporting; Methodology: Supporting; Project administration: Lead; Resources: Lead; Supervision: Lead; Validation: Lead; Visualization: Equal; Writing – original draft: Lead; Writing – review & editing: Supporting). The day before the study visit, participants consumed standardized test meals containing 2165 calories (46% energy from carbohydrates, 40% from fat, 14% from protein). For the study, fasted participants reported to Washington University's Clinical and Translational Research Unit at 7:00 AM. An intravenous catheter was inserted into a vein of the hand or forearm, which was then was placed in a thermostatically controlled box at 55°C for 15 minutes before blood draws. A second catheter was inserted into a vein on the contralateral arm to administer the glycerol tracer. After collection of baseline blood and breath in the fasted state, participants consumed a liquid mixed test meal consisting of 13.5 g Sol Carb, 516 g Boost Plus, 3.2 g canola oil, 0.2 g lecithin, and 5 mg/kg of glycerol tri(olein-1,2,3,7,8-13C5) (no. 772941; Sigma-Aldrich, St Louis, MO) and 24 g of water. Four equal aliquots of this meal were consumed in a time-controlled fashion over a 16-minute period, with the study time course clock starting at initial meal consumption. At 30 minutes, a 75 μmol/kg bolus infusion of [1,1,2,3,3-2H5]glycerol (no. DLM-1229-MPT-PK; Cambridge Isotope Laboratories, Tewksbury, MA) was administered intravenously in the second venous catheter to assess the plasma chylomicron fractional clearance rate.1Kruiningen HJ Van et al.Gut. 2008; 57: 1-4Crossref PubMed Scopus (103) Google Scholar,2Xiao C. et al.Cell Mol Gastroenterol Hepatol. 2019; 7: 487-501Abstract Full Text Full Text PDF PubMed Scopus (41) Google Scholar Blood samples (10 mL, EDTA tubes) were collected immediately before the meal was consumed (t = 0, baseline) and at 15, 30, 45, 60, 90, 120, 150, 180, 240, 300, and 360 minutes after meal initiation while participants reclined in a hospital bed. Plasma was immediately obtained by centrifugation at 300g. Breath samples to determine 13C expiration were collected before consuming the meal (baseline) and then hourly. Baseline blood was used to monitor disease activity through the analysis of plasma C-reactive protein (Washington University Core Laboratory, St Louis, MO). The first fecal sample after the study visit began was collected on the day of the study or the day after, with the collected sample used to monitor levels of fecal calprotectin (no. EK-CAL, no. B-CALEX-C50-U; Bühlmann Diagnostics Corp, Amherst, New Hampshire). Immediately at the end of the 6-hour study, 1 mL of each plasma sample was transferred into ultracentrifuge tubes (no. 344057; Beckman Coulter Life Sciences, Indianapolis, IN), overlaid with a saline solution (density = 1.006 g/mL), and spun for 35 minutes at 47,096g, 16°C in a SW 55Ti rotor (Beckman Instruments, Inc, Pasadena, CA). After centrifugation, the top layer, enriched in apo48-containing chylomicrons, was collected by using a 3-mL syringe fitted with a 20-gauge needle, and the volume recovered (∼1.5 mL) was recorded. Lipids from these chylomicron-enriched fractions were extracted by using a 0.5:1:3 ratio of the chylomicron fraction to chloroform and methanol (volume/volume/volume). Triglycerides from the chylomicron fractions were purified by thin-layer chromatography, and fatty acid methyl esters and heptafluorobutyryl glycerol were prepared for gas chromatography/mass spectrometry analysis1Patterson B.W. et al.J Lipid Res. 2011; 43: 223-233Abstract Full Text Full Text PDF Google Scholar for measurement of 13C5-oleate and 2H5-glycerol tracer-to-tracee (TTR) ratios (MSD 5973; Hewlett-Packard, Palo Alto, CA). Test meal triglycerides were processed similarly to measure 13C5-oleate TTR in the test meal. Chylomicron triglyceride concentration was measured using the Wako L-type TG M test (Fujifilm Wako Diagnostics, Mountain View, CA). Additional plasma (after removal of 1 mL for chylomicron recovery) was frozen at –80°C for later drug, plasma hormone, or metabolite analysis. Whole plasma was used in enzyme-linked immunosorbent assays to quantify GLP-1 (no. EZGLP1T-36BK), GLP-2 (no. EZGLP2-37BK), and GIP (no. EZHGIP-54K) from EMD-Millipore (Massachusetts, MA). Enzyme-linked immunosorbent assay for apoB48 was from Fujifilm Wako Chemicals (no. 637-10641). C-reactive protein, total triglycerides, total cholesterol, plant sterol campesterol, cholesterol precursor lathosterol, low-density lipoprotein, and high-density lipoprotein were measured at the Core Laboratory for Clinical Studies at Washington University or by liquid chromatography–mass spectrometry in the Diabetes Research Lipid Core. Human apoB100 and apoB48 were detected in chylomicron fractions by immunoblot, as previously described.3Blanc V. et al.J Lipid Res. 2012; 53: 2643-2655Abstract Full Text Full Text PDF PubMed Scopus (12) Google Scholar Adalimumab or infliximab drug levels were quantified at the Mayo Clinic (Rochester, MN). Fraction of test meal 13C tracer expired in CO2 was calculated by integrating the product of CO2 production rate (indirect calorimetry) × 13CO2 enrichment over the 6-hour meal period by the moles of 13C in the meal (calculated by kilogram of body weight × 5 mg/kg of tracer). An index for the fractional clearance rate (pools/hour) of chylomicron-triglyceride (TG) was calculated as the negative of the initial monoexponential slope of TG 2H5-glycerol TTR vs time beginning 1 hour after the bolused glycerol tracer (1.5 hour after the start of the test meal). The fraction (F) of chylomicron-TG oleate derived from test meal triolein was calculated by dividing the TTR of chylomicron TG 13C5-oleate by the TTR of test meal TG 13C5-oleate; the fraction of chylomicron from nontest meal oleate was calculated as 1 – F. The absolute concentration of 13C-oleate in chylomicron-TG (μmol/L) was calculated as the product of chylomicron-TG concentration × 3 mol fatty acid/mol TG × 35% oleate fatty acid content (assumed) × 13C5-oleate TTR. Sample processing, negative staining, and scanning were performed on chylomicron fractions through the Washington University School of Medicine Center for Cellular Imaging using a JEOL (Tokyo, Japan) JEM-1400Plus, 120-kV transmission electron microscope. Images were exported for analysis in Fiji ImageJ software (NIH, Bethesda, MD). Approximately 10 pictures of each sample were taken, and the diameters of more than 100 particles in each sample were quantified. Prism 8 software (GraphPad, San Diego, CA) was used for calculations and graph generation. Variables of age, sex, and body mass index were evaluated between control participants and participants with CD by using Fisher's exact test and chi-square analysis, with no statistically significant variables observed. Shapiro-Wilks tests were carried out on data sets where comparisons were of interest and generally found to exhibit a nonnormal distribution, even after log transformation. Thus, data were compared using nonparametric tests: Mann-Whitney U test for 2-column data comparisons, Kruskal-Wallis test for 3 or more columns of data (in the table), or Friedman tests for time course data. Data are plotted as mean ± standard error of the mean in the figures. Multiple comparisons tests, using aggregated data for an experimental group across the time course, were applied by using Dunn's post hoc test. ∗P < .05, ∗∗P < .01, ∗∗∗P < .001, ∗∗∗∗P < .0001. Download .xlsx (.01 MB) Help with xlsx files Supplementary Table Download .pdf (.35 MB) Help with pdf files Supplementary Figure Lymphatic-dependent Transport of Chylomicrons in Inflammatory Bowel DiseaseGastroenterologyVol. 160Issue 6PreviewIntestinal lymphatic obstruction, remodeling, expansion, and impaired contraction, which characterize inflammatory bowel diseases including Crohn's disease, may impair lymphatic pumping and lead to lymphangiogenesis and immune dysregulation, supporting lymphangitis as a cause or consequence of Crohn's disease. We read with interest the article by Huang et al.1 regarding postprandial chylomicron output and transport through intestinal lymphatics in patients with Crohn's disease. In this study, the authors investigated whether the lymphatic-dependent transport of chylomicrons was decreased in Crohn's disease, and the results indicated that chylomicron transport was not impaired in patients with active Crohn's disease of the ileum. Full-Text PDF