2. Journal Club Presentation Basis: Feehley T, Nagler CR. Nature. 2014 Oct 9;514(7521):176-7. Nature. 2014 Oct 9;514(7521):181-6.

3. Background: – Non-Caloric Artificial Sweeteners (NAS) – Microbiota Research article data: – Mouse and Human studies Summary Discussion Overview

4. Introduced over a century ago Gained popularity due to: – Reduced costs – Low caloric intake – Perceived health benefits for weight reduction and normalization of blood sugar levels NAS consumption studies: – Some show benefits: little induction of a glycaemic response – Others show associations with weight gain, increased risk of type 2 diabetes – Interpretations complex due to NAS consumption by individuals with existing metabolic syndrome manifestations FDA approved six NAS products for use in the US Background: Non-Caloric Artificial Sweeteners (NAS) Gardner et al. Diab. Care. 35(8):1798-808 (2012). Fitch. J Acad Nutr Dietetics. 112:739–758 (2012). Tordoff et al. Am. J. Clin. Nutr. 51:963–969 (1990). Horwitz, et al. Diab. Care. 11:230–234 (1988). Nettleton et al. Diab. Care. 32:688–694 (2009).

5. Metabolism: – Most NAS pass through GI tract without being digested by the host  directly encounter the intestinal microbiota  central role in regulating multiple physiological processes Background: Non-Caloric Artificial Sweeteners (NAS) Clemente et al. Cell 148, 1258–1270 (2012). http://www.nature.com/nature/journal/v449/n716 4/fig_tab/nature06244_F1.html Stringlike filaments of microbes grow on intestinal cells. Credit: Weizmann Institute of Science https://www.sciencenews.org

6. Background: Microbiota Development Clemente et al. Cell 148, 1258–1270 (2012). Gram negative Gram positive

7. Background: Microbiota/host interactions http://www.quantumrevolution.net/wp- content/uploads/2013/02/microbiome.jpg

8. Question: Can non-caloric artificial sweeteners modulate the composition and/or function of the gut microbiota and thus affect host glucose metabolism? Microbiota and NAS Study Credit: Weizmann Institute of Science Saccharin Sucralose Aspartame

9. Experimental scheme and dosage Commercially available NAS: – 10% solution: Sucrazit (5% saccharin, 95%glucose), Sucralite (5% Sucralose), Sweet’n LowGold (4% Aspartame) – Well below reported toxic doses Pure saccharin: – 0.1mg ml-1 solution – to meet with FDA defined acceptable daily intake (ADI) for saccharin in humans (5mg per kg (body weight)), according to the following calculation: Controls dosage: – 10% solution glucose – 10% solution sucrose 10wk.o.

10. NAS-consuming mice developed glucose intolerance Saccharin *** p < 0.001 Fig. 1 a, b NAS-induced glucose intolerance is mediated through alterations to the commensal microbiota Commercial NAS: – 10% solution: Sucrazit (5% saccharin, 95%glucose), Sucralite (5% Sucralose), Sweet’n LowGold (4% Aspartame) – Well below reported toxic doses Controls dosage: – 10% solution glucose – 10% solution sucrose Antibiotics regimens: Gram-negative targeting regimen A (ciprofloxacin, metronidazole) Gram-positive targeting regimen B (vancomycin)

11. Corroborating the findings in the obesity (HFD) setup: NAS-consuming mice developed glucose intolerance Fig. 1 c d Pure saccharin: – 0.1mg ml-1 solution – to meet the FDA defined acceptable daily intake (ADI) for saccharin in humans (5mg per kg (body weight)) High-Fat Diet (HFD): – 60% kcal from fat * p < 0.03 Also in Swiss-Webster mouse strain NAS-induced glucose intolerance is altered by microbiota

12. Liquids and chow consumption Oxygen consumption Walking distance and energy expenditure Metabolic profiling of normal-chow- or HFD-fed mice showed similar measures between NAS- and control-drinking mice Normal chow HFD Supp. Fig. 3, 4

13. Glucose intolerant NAS-drinking mice display normal insulin levels and tolerance Supp. Fig. 5 b a,c Normal chow HFD

14. Causal role of the microbiota: Faecal transplantation into normal-chow-fed germ-free mice * p < 0.03 Normal chow HFD Metabolic derangements induced by NAS consumption are mediated by the intestinal microbiota Fig 1, Supp. Fig. 2 * p < 0.05 ** p < 0.01

15. Saccharin consuming mice compared to controls: – Considerable dysbiosis in the microbiota of saccharin-consuming mice – Alterations in > 40 operational taxonomic units (OTUs) – Increases in Bacteroides – Decreases in Clostridiales In germ-free recipients of stools from saccharin- consuming donors: – Mirroring of OTUs observed NAS mediate distinct functional alternations to the microbiota Saccharin consumption in various formulations, doses and diets induces dysbiosis with overall similar configurations

16. Functional characterization of saccharin-modulated microbiota To compare relative species abundance: – Shotgun metagenomic sequencing of faecal samples – genetic analysis to examine environmental samples abundant in microscopic species Fig. 2a, Supp. Fig. 7a, 7b Saccharin induced the largest changes in microbial relative species abundance High Low

17. Functional characterization of saccharin-modulated microbiota Genetic pathways abundance: – Mapped metagenomic reads to a gut microbial gene catalogue, grouping genes into KEGG (Kyoto Encyclopedia of Genes and Genomes) – Found changes in pathway abundance to be inversely correlated between commercial saccharin- and glucose-consuming mice -> Saccharin greatly affects microbiota function: - among over-represented pathways is increased glycan degradation: glycans are fermented to form various compounds including short chain fatty acids (SCFAs) – obesity association Fig. 2b, c, d  Glycan degradation pathways

18. Higher glycan degradation is attributed to five bacterial taxa Gram-negative and positive species Consistent with the sharp increase of the species in the 16S rRNA analysis (marker of bacterial abundance) Consequence of higher glycan degradation – elevated acetate and SCFAs propionate Other pathways enriched: – Starch and sucrose metabolism – fructose and mannose metabolism – glycerolipid and fatty acid biosynthesis Fig. 2e,f,g

19. Saccharin modulates the composition and function of the microbiome causing dysbiosis ** p < 0.01 Fig. 3, Supp. Fig 8  In microbiomes of diabetic mice Does saccharin directly affect the microbiota?

20. Does the human microbiome function similarly after NAS consumption? Population study: Non-randomized 381 non-diabetic individuals: 44% males and 56% females; age 43.3 ± 13.2 High-NAS consumers (40) and non-consumers (236) Examined the relationship between long-term NAS consumption (based on a validated food frequency questionnaire) and various clinical parameters Clinical parameters found to be increased in NAS consumers compared to non-consumers: Weight and waist-to-hip ratio Haemoglobin (HbA1C%) – indicates glucose [c] over the previous 3 months Glucose tolerance test (GTT, measures impaired glucose tolerance) Serum alanine aminotransferase (ALT, measures hepatic damage that is likely to be secondary, in this context, to non-alcoholic fatty liver disease) Human microbiome functioning

21. Acute saccharin consumption impairs glycaemic control in humans by inducing dysbiosis Non-consumers (236) and high-NAS consumers (40): Randomly characterized 16S rRNA (172) Found positive correlations between multiple taxonomic entities and NAS consumption: – Enterobacteriaceae family (Pearson r=0.36, FDR corrected P<10 -6 ) – Deltaproteobacteria class (Pearson r=0.33, FDR corrected P<10 -5 ) – Actinobacteria phylum (Pearson r=0.27, FDR corrected P<0.0003) Did not detect statistically significant correlations between OTU abundances and BMI  correlations are not due to distinct BMI Fig. 4 ** ** p < 0.002

22. Acute saccharin consumption impairs glycaemic control in humans by inducing dysbiosis Initial assessment of NAS consumption/ blood glucose causation: 7 healthy volunteers (NAS non- consumers): 5 males, 2 females, 28 – 36 y.o. 7 day consumption of commercial saccharin (5 mg per kg (body weight)) as 3 divided daily doses equivalent to 120 mg Continuous monitoring by glucose measurements Fig. 4 Responders Non-responders

23. Acute saccharin consumption impairs glycaemic control in humans by inducing dysbiosis Fig. 4 Responders developed poorer glycemic response 5-7 d after treatment Microbiome configuration (16s rRNA analysis) from responders clustered differently from non- responders Microbiome composition changed in NAS responders

24. Acute saccharin consumption impairs glycaemic control in humans by inducing dysbiosis Fig. 4 Transfer of transplanted sample from responder induced glucose intolerance in recipient germ-free mice Germ-free mice transplanted with ‘responders’ microbiome replicated some of the donor saccharin- induced dysbiosis * p < 0.004 Orders: Bacteroidales Lactobacillales, Clostridiales

25. Summary NAS-consuming mice developed glucose intolerance NAS-regulated glucose intolerance is mediated by gut microbiota NAS modulate microbiota to induce glucose intolerance NAS-altered gut microbiota is functionally altered Acute NAS consumption may impair glycaemic control in humans by inducing dysbiosis

26. Discussion Several of the bacterial taxa that were altered by NAS consumption – previously associated with type 2 diabetes: – Increased Bacteroides, lowered Clostridiales Enrichment for glycan degradation pathway – link to enhanced energy harvest and thus regulation of multiple processes in the organism Comparing current report to other reports is complex, due to diverse ways of microbiota analysis Human response to NAS may be personalized Personalized nutrition – personalized medical outcome

27. Thank you!