Location: Diet, Microbiome and Immunity Research
2024 Annual Report
Objectives
Objective 1 Define associations between diet and gut microbiota composition and function.
Sub-objective 1A (Phenotyping Study): Examine the association between dietary features (e.g. fiber intake), gut microbial composition (bacterial taxa) and gut microbial functional capacity.
Sub-objective 1B (Phenotyping Study): Use ex vivo culture models to examine the difference between high and low fiber groups in gut microbial functional capacity and colonization resistance to a food-borne pathogen.
Sub-objective 1C (Longitudinal Study): Determine which bacterial taxa are consistently present over time and which bacterial taxa vary and correlate with dietary patterns for each subject.
Sub-objective 1D (Intervention Study): Examine the specific effects of an inulin intervention on short term changes in composition and functional capacity of the gut microbial community.
Objective 2 – Assess the association of diet and microbiota with gut health.
Sub-Objective 2A (Phenotyping Study): Determine how intake of dietary fiber is associated with markers of gut health in a cross-sectional study.
Sub-Objective 2B (Phenotyping Study): Determine whether dietary fiber intake and gut microbiome functional capacity are correlated with markers of gut health.
Sub-Objective 2C (Longitudinal Study): Determine whether a long-term habitual low fiber diet is associated with markers of chronic gut inflammation relative to high fiber-consuming controls in a longitudinal study.
Sub-Objective 2D (Intervention Study): Determine if consumption of dietary inulin reduces gut inflammation and impairs intestinal permeability when perturbed by an oral typhoid fever vaccine.
Objective 3 – Determine if dietary patterns that promote gut health also promote systemic immune health.
Sub-Objective 3A (Phenotyping Study): Determine if dietary features or nutritional status, gut microbial composition or functional capacity, and gut inflammation markers are associated with markers of systemic inflammation, specific immune cell types, or their level of activation.
Sub-Objective 3B (Longitudinal Study): Determine if the associations identified in 3A are also seen in the baseline samples from the Longitudinal Study and determine if these associations are constant across time.
Sub-Objective 3C (Intervention Study): Determine if consumption of 12 g/d inulin for 10 wk (for 4 wk before, 1 wk during and 1 wk after administration of the Vivotif® vaccine) will increase the vaccine-specific ALS IgG and IgA responses (primary endpoints), the plasma antibody, and stool IgA and T-cell responses (secondary endpoints) to the vaccine, relative to 12 g/d maltodextrin.
Objective 4: Investigate the immunological properties of peanuts, other nuts and alternative proteins.
Objective 5 - Investigate whether peanut consumption affects immune cell function and inflammation in healthy adults, including those at risk for immune dysfunction due to underlying conditions such as intestinal dysbiosis, obesity or chronic stress.
Approach
Our central hypothesis is that immunological health is a function of both dietary intake and the functional capability of gut microbes to respond to that diet. We will use three human studies to examine our central hypothesis: a cross-sectional Phenotyping Study, a Longitudinal Study, and a Fiber Intervention Study. The Western Human Nutrition Research Center (WHNRC) Nutritional Phenotyping Study is a cross-sectional study of healthy adults balanced by sex, age and body mass index with the recruitment phase to be completed in 2019. We will use stool samples from this project in ex vivo culture models—stool fermentations, pathogen challenge, and intestinal cell response—to address how the microbial environment interacts with substrate and how it affects physiology. The WHNRC Longitudinal Study is an observational cohort of middle-aged non-obese human participants selected at baseline to have adequate or low fiber intake. This cohort will be followed for up to 20 years, subject to renewal, with baseline and year 1 occurring in the current project cycle. Primary outcomes are measures of gastrointestinal and systemic inflammation. The WHNRC Fiber Intervention Study is a randomized controlled trial designed to test whether dietary inulin improves response to an oral vaccine that includes a live attenuated enteric pathogen.
To address the hypothesis that dietary fiber consumption is associated with altered gut microbiome composition and function, stool samples from the studies will be sequenced for DNA content. Stool samples from the Phenotyping Study will additionally be assessed for fermentation capability, and pathogen resistance. To address the hypothesis that dietary fiber consumption is associated with altered gastrointestinal health, stool samples from the studies will be assessed for markers of inflammation and tested in an in vitro culture model of intestinal epithelial cells. In the intervention trial, intestinal permeability will be measured by quantifying the permeability of non-metabolizable sugar molecules. To address the hypothesis that dietary fiber consumption is associated with altered systemic immunity, blood samples from the studies will be assessed for measures of innate and adaptive immunity. These include plasma markers and complete blood count (CBC) in all trials as well as flow cytometry and ex vivo cytokine production by PBMC in the Phenotyping Study and measurement of vaccine-specific lymphocyte and antibody responses in the Intervention Study. Both gastrointestinal and systemic response will also be analyzed with gut microbiota as a mediator to determine whether these responses are microbiota-dependent.
The most challenging aspect of all of these studies is the recruitment and retention of human participants, particularly for the Longitudinal Study. If we are unable to recruit enough participants, we may pursue new partnerships (e.g. UC Davis alumni association) or open a second study site (e.g. Sacramento). If we are unable to retain enough participants, we could consider the subset of outcomes that can be assessed remotely or backfill by recruiting more participants.
Progress Report
This is the final report for project 2032-51530-026-000D, Impact of Diet on Intestinal Microbiota, Gut Health and Immune Function, which has been replaced by new project, 2022-21000-021-000D, Effect of Diet on Gut Microbiome, Gastrointestinal Health, and Immune Function. For additional information, see the new project report.
In support of Sub-objective 1A, microbiome data were generated from fecal samples from participants in the USDA Nutritional Phenotyping Study, including both 16S rRNA (n=365) and shotgun metagenomes (n=330). Diet-microbiome interactions in healthy adults were analyzed, resulting in numerous manuscripts which detailed the associations of (a) dietary fiber diversity with microbiome diversity, (b) increased dietary fiber diversity with reduced antimicrobial resistance, and (c) dietary monosaccharide composition with changes in microbial taxonomy. To determine the contribution of microbial taxa to outcomes of interest, ARS researchers in Davis, California, developed software, called TaxaHFE, that reduces the feature set based on known taxonomic relationships of microbes and their information content relative to the outcome. Application of TaxaHFE to data from six different studies demonstrates superior performance and feature set reduction.
For Sub-objective 1B, ARS researchers in Davis, California, conducted fermentations using participant stools to determine microbial function and its relation to diet. After pilot studies showed that a 13-day sequence was required to complete one set of continuous flow fermentations with 4 carbohydrates (mucin, pectin, arabinoxylan, and resistant starch type 3) for each participant stool, the decision was made to perform fermentations with 10 participant stools total. Five stools were selected from subjects who habitually consumed low amounts of dietary fiber and five from subjects with diets high in dietary fiber. The fermentations were completed. An initial analysis of the base use data collected during the continuous flow fermentations as well as microbial 16SrRNA amplicon data shows that in general the most acid was produced during fermentation of resistant starch type 3, and the least during fermentation of mucin. Pectin fermentations resulted in increases in relative abundance of the Lachnospiraceae family regardless of habitual diet. Fermentations for which there are SCFA estimates available showed a positive correlation between Lachnospiraceae.NK4A13.group and butyrate (high fiber consumer) or acetate (low fiber consumer) production.
Sub-objective 1C was cancelled, due to Maximum Telework. As one of the replacement studies, the Mixed Dish Estimation study aims to validate a novel tool for recording the contents of mixed dishes in dietary records. The Mixed Dish Estimation study is in progress (target n=160, complete n=82).
For Sub-objective 1D, the Fiber Intervention Study (n=60) was delayed due to Maximum Telework; 35 participants have been enrolled, 22 completed the study, eight dropped out, five currently active.
In support of Sub-objective 2A, associations between diet and gastrointestinal health were analyzed in healthy adults, resulting in the following key findings: (a) dietary intake, esp. saturated fat, and the stress hormones cortisol and norepinephrine predict stool consistency, (b) low fecal pH, a proxy for fermentable fiber intake, was associated with better bone density, (c) diversity of dietary non-glucose monosaccharides was negatively associated with GI inflammation, (d) dietary fiber intake was negatively associated with GI inflammation, and (e) lower diet quality was associated with subclinical GI inflammation even in healthy adults. Polyphenol intake was quantified from dietary data; total dietary polyphenols were associated with lower GI inflammation and olive-source polyphenols were associated with lower plasma lipopolysaccharide binding protein. Experiments with Caco-2 cells exposed to fecal waters from healthy participants (n=317) are complete, exceeding our target number; analysis is on-going.
In support of Sub-objective 2B, all shotgun metagenomes from the USDA Nutritional Phenotyping Study were mapped to microbial genes in the Carbohydrate-Active Enyzme (CAZyme) Database and to other databases (KEGG, MetaCyc, etc.) A new metric, Muc2Plant, was defined as the ratio of mucin-unique CAZymes to plant unique CAZymes. Muc2Plant was positively associated with GI inflammation, in support of the hypothesis that mucin degradation promotes or co-occurs with GI inflammation. However, dietary fiber intake was not associated with Muc2Plant. SCFAs were measured from all fecal and plasma samples available (n=x fecal, n=y plasma) from healthy adults. Fecal pH correlated with fecal SCFA abundance. A higher quality diet was associated with a compositional increase in fecal butyrate, relative to acetate and propionate. SCFAs were associated with markers of subclinical GI inflammation and inversely related to plasma lipopolysaccharide binding protein. Diet and microbiome features were far more predictive of fecal SCFA abundances compared to plasma SCFA abundances. The top diet and microbiome predictors of fecal butyrate included potatoes and the thiamine biosynthesis pathway, respectively.
For Sub-objective 2C: See Sub-objective 1C, which relies on the same clinical trial. For Sub-objective 2D: See Sub-objective 1D, which relies on the same clinical trial.
In support of Sub-objecvtive 3A, data from the USDA Nutritional Phenotyping Study was used to examine the association of diet with systemic inflammation and markers of innate and adaptive immunity. Higher diet quality was associated with lower total lymphocyte concentrations and percentages of Natural Killer (NK) T-lymphocytes, lower activation of innate immune cells (monocytes and neutrophils) and altered levels of cytokine production by immune cells. Additionally, associations were examined between the gut microbiota and 17 immune factors made up of a total of 79 individual markers of systemic immune activation. Results showed the families S24-7, Rikenellaceae, Pseudomonoadaceae and one uncharacterized family were associated with four different immune factors describing systemic inflammation, T lymphocyte activation, monocyte activation and matrix metalloproteinase activity, indicating that intestinal bacteria may directly affect systemic immunity in healthy adults. The post-prandial response to consumption of the high-fat challenge meal in the USDA Nutritional Phenotyping Study was also examined. Relative to fasting conditions, the circulating monocyte pool after the meal was significantly larger and comprised of fewer “classical” monocytes which are important for replenishing immune cells in tissues, but more “patrolling” monocytes which are important for maintaining vascular integrity. 3B: The Longitudinal Study was cancelled due to Maximum Telework. As one of the short-term replacement studies, the WHNRC postprandial monocyte study aims to delineate the function of postprandial non-classical monocytes following consumption of a mixed macronutrient challenge meal. The monocyte trial is almost complete (target n=25, n=20 complete, n=5 active).
For Sub-objective 3C: See Sub-objective 1D, which relies on the same clinical trial. In a subordinate project on glycans as food biomarkers (0000065329), ARS researchers in Davis, California, collaborated with colleagues at University of California, Davis to create Davis Food Glycopedia version 2 (DFG2), which contains detailed glycan content of 250 high-priority foods. ARS researchers mapped DFG2 to NHANES to determine that dietary glycan features are more predictive of insulin resistance in adults than are existing dietary features (e.g. total dietary fiber). This project supports Objective 1 to understand diet and microbial function. In a subordinate project on dairy consumption, lactase persistence genotypes, and gut microbiome (58-2032-2-01100), the objective was to determine how lactase persistence genotypes and dairy consumption interact to impact the gut microbiome. Increased abundance of families Lactobacillaceae and Lachnospriaceae was found in the gut microbiomes of lactase non-persistent (LNP) individuals consuming >12.4 g lactose/day. Independent of genotype, dairy consumption was not associated with increased GI inflammation. This project supports Objectives 1 and 2.
In a subordinate project on gut microbiomes of 1000 mother-infant dyads (2032-51530-026-10T), ARS researchers in Davis, California, extracted DNA from approximately 3000 fecal samples. Analysis of infant gut microbiomes in the context of milk cells and infant morbidity is ongoing. This project supports Objective 1 to understand diet and microbial function. In a subordinate project on honey (2032-51530-026-09T), the primary aim is to assess the effects of minor components of honey on the composition and function of the small intestine microbial community. In vitro gastric digestion of honey or sugar control together with ETEC showed that honey inhibited ETEC growth under gastric conditions relative to sugar control. Fermentation of the gastric digests with a small intestine microbial community showed that honey does not appear to inhibit ETEC growth outside of the gastric environment. In a subordinate project, as part of the ARS Dairy Grand Challenge, relationships among milk components – lactose, oligosaccharides and fatty acids– milk microbiota, and somatic cell count, were examined in lactating dairy cows consuming different diets. Increased fiber from forages was associated with decreased Gammaproteobacteria, a class containing opportunistic pathogens. Fatty acids within the milk were negatively correlated with specific opportunistic pathogens to a greater degree than milk oligosaccharides (complex sugars). These results suggest that dietary fiber from forages and the resulting changes in milk fatty acids may contribute to reduced pathogen load.
Accomplishments
1. New software enables health predictions from diet or microbiome data. Microbiome data are organized in a taxonomic tree that encodes the parent-child relationships of different microbes. Diet data can similarly be organized into a tree of relationships between types of food. Prior to this work, analyses of diet or microbiome data had to be done at a fixed level, without leveraging the relationships between foods or between microbes. ARS researchers in Davis, California, developed a method of hierarchical feature engineering called TaxaHFE, which dynamically collapses tree-based data based on taxonomic information together with information gain, maximizing the information contained at various taxonomic levels while reducing redundancy. The TaxaHFE software improves predictions of health outcomes from diet and/or microbiome data while simultaneously increasing the ability to interpret the results. This new technology enables nutrition scientists to discover diet and/or microbiome relationships with health.
2. Benchmark dataset enables evaluation of image-based dietary assessment. Photo-based dietary assessment methods are becoming more feasible as artificial intelligence methods improve. However, advancement of these methods to the level usable in nutrition studies has been hindered by the lack of a dataset against which to benchmark algorithm performance. ARS researchers in Davis, California, conducted the Surveying Nutrient Assessment with Photographs of Meals (SNAPMe) Study (ClinicalTrials ID: NCT05008653) to develop a benchmark dataset of food photographs paired with traditional food records. The SNAPMe database contains 3,311 unique food photos linked with 275 food records from 95 participants who photographed all foods consumed and recorded food records in parallel for up to three study days each. Publicly available algorithms for ingredient prediction from food photos performed poorly on the SNAPMe benchmark, especially for single-ingredient foods and beverages. The SNAPMe database will provide nutrition and computer science researchers with a benchmark to test future algorithms for the improvement of photo-based dietary assessment.
3. Intestinal bacteria are associated with both higher and lower immune activation in healthy adults. Chronic inflammation may develop in otherwise healthy adults for several reasons, including activation of the immune system by bacteria that are normal constituents of the intestinal microbiome. Identifying such bacteria may help provide dietary advice to minimize microbiome-induced immune activation and thus prevent the development of chronic inflammation. ARS scientists from Davis, California, studied 355 healthy adults, measuring 79 markers of innate and adaptive immune activation in blood samples and characterizing the intestinal microbiome using stool samples. Some bacteria were associated with higher innate and adaptive immune activation suggesting these bacteria should be further examined for roles in development of chronic inflammation. Other bacteria which ferment dietary fiber to produce inflammation-dampening molecules were associated with lower innate and adaptive immune activation. These bacteria, which are normal residents of the gut microbiome, should be examined further for health benefits in humans, and to determine if fiber or other dietary components can increase their abundance.
4. Higher lactose intake by lactose intolerant adults associated with lactose-metabolizing gut bacteria. People who avoid dairy products due to lactose intolerance may be missing key nutrients. ARS scientists in Davis, California, examined the amount of lactose consumed, consumer genetics, and consumer gut bacteria in a healthy multiethnic U.S. adult cohort. Individuals who were genetically lactose intolerant tended to consume less lactose per day than individuals who were not genetically lactose intolerant. However, lactose intolerant individuals who consumed more than 12.4 g/d of lactose per day, had significantly more families of gut bacteria that are capable of metabolizing lactose. These data suggest that the gut bacteria of lactose intolerant individuals who consume high amounts of lactose may be altered to accommodate increased lactose in the gut.
Review Publications
Benatzy, Y., Palmer, M.A., Lutjohann, D., Ohno, R., Kampschulte, N., Schebb, N., Fuhrmann, D.C., Snodgrass, R.G., Brune, B. 2024. ALOX15B controls macrophage cholesterol homeostasis via lipid peroxidation, ERK1/2 and SREBP2. Redox Biology. 72. Article 103149. https://doi.org/10.1016/j.redox.2024.103149.
Beckner, A., Arnold, C.D., Bragg, M.G., Caswell, B.L., Chen, Z., Cox, K., DeBolt, M.C., George, M., Maleta, K., Stewart, C.P., Oakes, L.M., Prado, E.L. 2023. Examining infants' visual paired comparison performance in the US and rural Malawi. Developmental Science. 27(5). Article e13439. https://doi.org/10.1111/desc.13439.
Kable, M.E., Chin, E., Huang, L., Stephensen, C.B., Lemay, D.G. 2023. Association of estimated daily lactose consumption, lactase persistence genotype (rs4988235), and gut microbiota in healthy adults in the United States. Journal of Nutrition. 153(8):2163-2173. https://doi.org/10.1016/j.tjnut.2023.06.025.
Bouzid, Y.Y., Chin, E.L., Spearman, S.S., Alkan, Z., Stephensen, C.B., Lemay, D.G. 2023. No associations between dairy intake and markers of gastrointestinal inflammation in a healthy adult cohort. Nutrients. 15(16). Article 3504. https://doi.org/10.3390/nu15163504.
Riazati, N., Kable, M.E., Stephensen, C.B. 2023. Association of intestinal bacteria with immune activation in a cohort of healthy adults. Microbiology Spectrum. 11(6). Article e0102723. https://doi.org/10.1128/spectrum.01027-23.
Larke, J.A., Chin, E.L., Bouzid, Y.Y., Nguyen, T.T., Vainberg, Y., Lee, D., Pirsiavash, H., Smilowitz, J.T., Lemay, D.G. 2023. Surveying nutrient assessment with photographs of meals (SNAPMe): A benchmark dataset of food photos for dietary assessment. Nutrients. 15(23). Article 4972. https://doi.org/10.3390/nu15234972.
Oliver, A., Kay, M., Lemay, D.G. 2023. TaxaHFE: A machine learning approach to collapse microbiome datasets using taxonomic structure. Bioinformatics. 3(1). Article vbad165. https://doi.org/10.1093/bioadv/vbad165.