April 10-13, 2019; Pasadena, CA; Full Report – Draft

City of Hope’s Levine-Riggs Diabetes Research Symposium took place April 10-April 13 in Pasadena, CA with a strong focus on basic science research and type 1 diabetes. Our full report is below, featuring highlights on topics including fecal transplants, single chain insulin, type 1 autoimmunity, and beta cell plasticity. We’ve included all 13 of our highlights categorized by: (i) Novel therapies and technology in diabetes; (ii) Beta cell function and diabetes pathophysiology; and (iii) Diabetes complications and cancer. Read on!

Novel Therapies and Technology in Diabetes

Very Promising Two-Year TrialNet Data on Low-Dose ATG in New-Onset Type 1 Diabetes: Improved C-Peptide and A1c Reductions

Two-year TrialNet data (n=82, 12-45 year-olds) and six-month mechanistic results continue to suggest that low-dose antithymocyte globulin (ATG) could preserve beta cell function in new-onset type 1 diabetes (<100 days). University of Florida’s Dr. Michael Haller picked up where he left off last June at ADA 2018, where he shared that after one year, participants on placebo experienced a ~0.45 mmol/L decline in C-peptide levels, but this was significantly attenuated to only ~0.1 mmol/L with ATG monotherapy (2.5 mg/kg, p=0.0003) and to ~0.25 mmol/L with ATG in combination with granulocyte colony stimulating factor (6 mg GCSF every two weeks for 12 weeks, p=0.031). The positive results continued into year two with significantly higher C-peptide levels in subjects treated with ATG vs. placebo (p=0.00005) but only a trending increase with ATG/GCSF combination therapy vs. placebo (p=0.032). The improvements in C-peptide were paralleled by A1c reductions in subjects treated with ATG (p=0.011) and ATG/GCSF (p=0.022) vs. placebo. Dr. Haller asserted that there is “no need to use two drugs if we don’t have to” and suggested using ATG as a monotherapy or in combination with other drugs to prevent or delay the onset of type 1 diabetes – wow! We are looking forward to seeing the full results which are currently in press in Diabetes, and Dr. Haller is hoping to move forward with a study examining low-dose ATG for prevention in the near future.

• Low-dose ATG suppresses the immune system by selectively depleting CD4 T-cells and preserving regulatory T-cells (Treg). Thanks to increased funding support, Dr. Haller and his team have been able to conduct mechanistic research using flow cytometric analysis. A first look at six-month results revealed decreases in CD4:CD8 ratio, increased regulatory T-cell (Treg), and increased PD-1+CD4+T cells.

• It looks like TrialNet’s “Brazil Lite” may have found a less aggressive and side-effect-ridden way of achieving the effects of the hotly-debated Brazil Cocktail, which uses high doses of ATG to wipe out immune cells and achieve insulin dependence. While Dr. Jay Skyler called the Brazil Cocktail a “heroic” attempt to fight type 1 diabetes in 2013, others, including JDRF’s Dr. Richard Insel, have stated that the original cocktail would not be safe enough for use in children and adults. While ~60% of patients achieved insulin independence within six months with the Brazil Cocktail and nearly 40% remained completely insulin independent at the 18-month mark, the frequency of severe side-effects was fairly high, including neutropenic fever, alopecia, fever, and one death due to sepsis that may or may not have been related to the therapy. As a result, several approaches to improve the risk/benefit profile of the therapy have been underway. Dr. Haller’s Brazil Lite is the first for which we are seeing promising data. Indeed, Dr. Haller shared that there were self-limited side effects (cytokine release symptoms and serum sickness) that occurred only in the first two weeks following therapy in roughly two-thirds of the subjects. However, there were no differences in the side effect profile as compared to the placebo group after that initial period, suggesting that it is a plausible and tolerable therapy that could be realistically translated to the bedside.

Thermalin’s Novel Single-Chain Insulin Designed to be Stable in Extreme Temperatures Outperforms Glargine, Lispro, and Lantus in Thermal Stress Assays – Effective After 57 Days of Heat, Agitation

Indiana University School of Medicine’s Dr. Michael Weiss presented the latest on his novel single-chain insulin (SCI) formulation, which is very stable in extreme temperatures, shows little tendency to form hexamers, and is rapid-acting. We first heard about his efforts to develop single-chain insulin at the DTM 2009! Dr. Weiss, co-founder of Thermalin, aspires to recapitulate all insulin products with ultra-stability in order to obviate the need for the costly (and ineffective?) global cold delivery chain. Thermal stress assays were conducted in rats to assess residual insulin activity following prolonged agitation at elevated temperatures. The SCI-a (rapid-acting) and SCI-b (biphasic) molecules showed no reduction in pharmacodynamic activity after 57 days of stress exposure, while glargine and lispro were completely inactive after 45 days. Dr. Weiss hypothesizes that the SCI variants could have lasted even longer than 57 days! In a separate, unpublished study, that SCI-c (basal) had no change in activity after 100 days of agitation while Lantus was completely inactivated after 10 days. We eagerly await SCIs’ entry into clinical testing. As a reminder, Thermalin is partnered with Sanofi on two next-gen insulins, though the identity of those candidates hasn’t been publicly disclosed , to our knowledge.

• A single refrigerator is shared by 1 in 12 rural villages in Bangladesh according to Dr. Weiss. Consequently, there is a huge unmet need to simplify insulin for global distribution, especially given the bona fide pandemic of diabetes mellitus in developing countries. SCIs may play a role in addressing this global need!

• SCI contain a shortened 24-Angstrom connecting peptide (“C domain”). This reduces the propensity to form amyloid-associated cross-beta sheets, the dominant form of degradation and inactivation at high temperatures, while maintaining high receptor affinity for the insulin receptor and native biological activity. Dr. Weiss said the SCIs are indistinguishable from human insulin in terms of degradation and are cleared within 10-15 minutes.

Amsterdam Group’s Puzzling Fecal Transplant Results for T1D Beta Cell Preservation: Autologous More Effective than Allogenic?

University of Amsterdam’s Dr. Max Nieuwdorp presented exciting unpublished data suggesting fecal microbiota transplantation (FMT) may have beneficial effects on residual beta cell function in type 1 diabetes patients. Dr. Nieuwdorp’s group previously showed that populating the intestines of insulin-resistant subjects with microbiota from lean donors via allogenic FMT had beneficial effects on glucose metabolism, mediated by an increase in production of butyrate (a short chain fatty acid). The group then conducted randomized controlled trial, where individuals with recently-diagnosed type 1 diabetes received three fecal transplants (at 0, 2, and 4 months) from allogenic (healthy; n=10) or self (autologous; n=10) sources. At one year, beta cell function (measured by AUC C-peptide response) had not deteriorated in the autologous group; while C-peptide declined in the allogenic group, it did so at a slower rate than expected. We wonder how the curves will look out to two years and beyond. Dr. Nieuwdorp suggested that the transplants may “stabilize” endogenous insulin production and extend the honeymoon period. It’s certainly odd that the autologous group saw such significant benefit given that there was theoretically no change in their microbiota; in fact, the data could also tell the story of a placebo effect from the transplant procedure, which is countered by deleterious introduction of a healthy donor’s “bugs.” [It’s also possible that the sample size was too small for definitive conclusions to be drawn.] We look forward to seeing further analysis of these curious results. Even if his hypothesis is validated, Dr. Nieuwdorp emphasized that FMT is “no panacea,” but has the potential to be a “therapeutic intervention from a different angle.”

• Dr. Nieuwdorp alluded to other microbiome-related interventions for beta cell functioning: (i) Oral butyrate supplementation had no effect on residual beta cell function in both type 1 and type 2 diabetes (despite the fact that it appeared to be a mediating variable in the aforementioned study); and (ii) Treatment with the live E. halli L2-7 strain has been shown to be safe and to improve insulin sensitivity in diabetes mice models. In that study, there was a positive correlation between change in fecal E. halli levels and change in peripheral insulin sensitivity in all animals after four weeks of treatment (p=0.055).

Overcoming Eroom’s Law with Pancreas-on-a-Chip for Faster Transition to Phase 1 Trials and Translation to Humans

Dr. Aline Zbinden (University of Tübingen) described her “pancreas-on-a-chip” technology as a promising alternative to animal models which fail to resemble human physiology (look no further than the fact that diabetes has been “cured” in lab animals literally hundreds of times). Dr. Zbinden’s in vitro model combines human genetic background and physiological relevance by trapping human conditionally immortalized EndoC-beta-H3 pseudo-islets into a perfusable microfluidic platform. Taking advantage of the optical accessibility of the chip, Dr. Zbinden’s team utilized Raman micro-spectroscopy as a label-free, non-destructive, real-time technique to monitor the pseudo-islet functionality, insulin secretion, mitochondrial activation, and lipid composition in situ. To understand the role of white adipose tissue (WAT) as a contributor to diabetes, they have also developed and validated WAT-on-a-chip. The ultimate goal is to create multi-organ chips with “lego-style-plug-and-play” to culture organs separately with different conditions and connect them together to understand the interaction between different organs. Getting candidates to phase 1 faster won’t completely solve Eroom’s law – the idea that drug development is becoming more sluggish, despite vastly improved technology – but it could certainly help. This model also raises the possibility of creating “pancreata-on-chips” with genetic makeups tailored to every individual with diabetes, so that therapies could be assessed for effectiveness before the person is ever exposed.

• Rodent models provide significant insights, but translation into humans is difficult. Glucose regulation mechanisms are highly species-specific and there is an enormous need for development of novel approaches to study diabetes from a molecular to an organ level. The pancreas-on-a-chip is promising for the study of insulin kinetics, toxicity screening, disease modeling and personalized medicine.

Dr. Colin Dayan’s Three Mindset Changes to Accelerate Immunotherapy Development; Ambitious Goal to Eradicate Type 1 Diabetes by 2041

In a lively and much-anticipated session, Cardiff University’s Dr. Colin Dayan stressed the potential of immunotherapies to transform diabetes care and identified three main obstacles limiting their development. At a time when less than 30% of people with type 1 diabetes achieve target levels of glycemic control, Dr. Dayan highlighted immunotherapy as a promising complimentary treatment. After all, as he put it: “the best insulin is your own insulin.” Dr. Dayan referenced a 2013 study (n=342) showing that Danish children three to six years after type 1 diabetes onset who retained just 1%-2% of their beta cell function (meal-stimulated C-peptide >0.2 nmol/L) were five times less likely to experience severe hypoglycemia than those who did not retain any. We wonder how common it is to achieve this small degree of function across a broader population. In the study, 8% of participants had residual beta function (RBF) >0.2 nmol/L, although those with C-peptide between 0.04 nmol/L and 0.2 nmol/L were also at a significantly lower risk of severe hypoglycemia and were present at a larger proportion (19%). Not surprisingly, older age at diabetes onset increased the likelihood of stimulated RBF >0.04 nmol/L and longer diabetes duration increased the probability of having RBF <0.04 nmol/L. In a more recent study, individuals who showed no change in C-peptide over one year exhibited no change in time-in-range, whereas those who saw a decline in C-peptide exhibited a decrease in time-in-range.

• Dr. Dayan pointed out that many immunotherapies are in development, including five promising interventions that have been shown to slow beta cell loss in controlled trials. In order to bring these treatments over the line, Dr. Dayan proposed three common mindsets/attitudes that must be remediated.

  • The “curse of insulin.” Although a life-saving discovery, the existence of an adequate treatment for type 1 diabetes has led healthcare providers and pharmaceutical companies to become complacent in their search for other therapies. Dr. Dayan called on both parties to further prioritize the development of immunotherapies.

  • The “expectation of a cure.” Dr. Dayan criticized the notion that new diabetes therapies should focus only on one-off treatments curing the disease. While a cure is obviously the end goal, he noted that for other autoimmune diseases (multiple sclerosis, rheumatoid arthritis, Crohn’s disease, ulcerative colitis) non-curative immunotherapies have significantly enhanced disease management. The ability of immune-based therapies to offer even modest improvements in glycemic control should not be overlooked.

  • The fear of side effects. While intolerable side effects have halted novel therapy development in certain areas, Dr. Dayan emphasized that selective immunosuppression and boosted immune regulation are associated with low or very low risk of side effects. In addition, immunotherapy may call for the reevaluation of accepted risk-benefit ratios. Given sufficient therapeutic efficacy, providers and patients may be more willing to tolerate certain side effects, he posited; we are not sure that is the case. While many patients may be willing to ensure various “side effects,” that is different in our view from safety factors.

• In an ambitious call for action, Dr. Dayan ended his talk with a goal to eradicate type 1 diabetes by 2041. His timeline predicted the licensing of the first beta cell preserving therapy by 2021, the development of beta cell regenerative therapies by 2025, and the possibility of antigen-specific therapies by 2031. Afterwards, he foresees focus shifting from delaying disease progression to preventing the onset of type 1 in individuals at risk. “Immunotherapy is a completely tractable problem,” he concluded, “we just have to be focused on it.”

Dr. Jose Florez Provides Ten Reasons GWAS Remains Relevant in Diabetes Research, Dr. Stephen Rich Calls for Application of GWAS Data to Identify Novel Drug Targets

In a stimulating debate, Drs. Jose Florez (Massachusetts General Hospital, Boston, MA) and Stephen Rich (University of Virginia School of Medicine, Charlottesville, VA) discussed the extent to which diabetes research should prioritize genome-wide association studies (GWAS). Both speakers agreed that over the last ten years GWAS has revolutionized our understanding of the genetic factors implicated in type 1 and type 2 diabetes. However, while Dr. Florez outlined ten key components of GWAS that continue to make this approach relevant in diabetes research today, Dr. Rich emphasized that the next phase of research should focus on translating GWAS findings into actionable information, such as the identification of drug targets. See below for a more detailed summary of these speakers’ positions.

• Dr. Florez offered the “GWAS Decalogue” in which he provided ten reasons for why GWAS remains a valuable tool. Dr. Florez asserted that human GWAS data can: (i) highlight fundamental biology; (ii) lead to the discovery of novel pathways/identification of new molecular targets; (iii) clarify pathogenesis; (iv) provide casual inference/identify off-target effects; (v) support prior epidemiological observations; (vi) temper expectations around prediction; (vii) explain population differences; (viii) resolve diagnostic dilemmas; (ix) stratify the population; and (x) elucidate disease heterogeneity. Continued use of GWAS, Dr. Florez explained, combined with other approaches such as systematic fine-mapping, can enhance models of disease prediction and therapeutic decision-making.

• Dr. Rich acknowledged that, while GWAS has identified hundreds of single nucleotide polymorphisms (SNPs) associated with diabetes, the challenge lies in using that information to identify functional targets for diabetes therapy. Paving the way forward, Dr. Rich suggested using a classification system to prioritize SNPs of interest and implementing methods such as RNA-seq (gene expression), ATAC-seq (open chromatin), and promoter-focused whole-genome Capture-C (chromatin interaction) be used to identify the target genes of the associated SNPs. With this approach, GWAS could meaningfully enhance the use of precision medicine in diabetes care.

Dr. Leif Groop: Identification of Type 2 Subpopulations Present Opportunities for Personalized Care, Enhanced Reimbursement

Recognizing the diversity of the type 2 population, Dr. Leif Groop (Institute for Molecular Medicine Finland) urged the adoption of a more nuanced classification system as a step toward precision medicine in diabetes care.

• Dr. Groop noted that, as the prevalence of type 2 diabetes has increased, so has its heterogeneity. Rather than diagnosing diabetes by measuring only 1 metabolite, glucose, he proposed a model in which six clinical variables are used to divide the type 2 population into five clusters. This widely-publicized model, which comes from Dr. Groop’s paper that Dr. Jim Gavin recently highlighted at EASD, divides the type 2 population into those with: (i) Severe autoimmune diabetes (SAID); (ii) Severe insulin-deficient diabetes (SIDD); (iii) Severe insulin-resistant diabetes (SIRD); (iv) Mild obesity-related diabetes (MOD); and (v) mild age-related diabetes (MARD). Each cluster has different genetic markers, as well as a unique disease prognosis. By recognizing that certain clusters are at a greater risk of complications such as kidney disease or retinopathy, providers can offer more personalized care to their patients beginning at diagnosis. We’re not sure what to think about this – would busy PCPs even have time to understand the types? If they did, we believe the personalization would really help patients early on in disease progression.

• Since the model’s proposal in 2018, efforts are already underway to learn more about the classifications. These include replicating the sub-classification system in other ethnic cohorts (in China and in India), using genome-wide association studies (GWAS) to identify additional variants associated with each subtype, and further studying the response to therapies such as metformin, SUs, and TZDs in each group. In the future, Dr. Groop suggested it might even be possible to identify a prediabetes phenotype.

• As a final note, UMass’ Dr. David Harlan mentioned that simplistic type 1/type 2 delineation can be “particularly toxic” in the US, where diagnosis dictates coverage. Should the new classification system grain traction – at least in medical communities – we wonder how long it would take for payers to begin tweaking their reimbursement policies.

Beta Cell Function & Diabetes Pathophysiology

Rachmiel Levine Award Winner Dr. Steven Kahn on NIH-Funded RISE Study; Type 2 Diabetes Pathophysiology Differs in Children vs. Adults

This year’s winner of the prestigious Rachmiel Levine Award, VA Puget Sound Health Care System and University of Washington’s Dr. Steven Kahn, presented data from the NIH-sponsored RISE study (n=91) showing that children (10-19 year-olds) with impaired glucose tolerance (prediabetes) or recently-diagnosed type 2 diabetes have lower insulin sensitivity, hyper-responsive beta cells, and reduced insulin clearance compared to adults with the same status. We assume the adults used for the comparator group were in part from the RISE Adult Medication Study (n=255), which is ongoing and will read out at ADA in June. Specifically, insulin sensitivity was 46% lower in children compared to adults, and children had 2.3-fold greater acute and 1.3-fold greater steady-state C-peptide than the adults. Insulin response to glucose was three times higher and steady-state insulin secretion was found to be over twice as high as compared to adults. These observations suggest that the pathogenesis of type 2 diabetes may differ between adults and youth, which is concerning given the increasing prevalence of pediatric type 2 diabetes. These results add to RISE data presented at ADA 2018, where UNC’s Dr. John Buse commented on the concerning youth data and noted that more insight on the pathophysiological differences between youth and adult will come when RISE Adult reads out. At the time, Dr. Buse’s biggest takeaway from RISE Peds was the need to address obesity in youth as a public health crisis – we agree and we’d like to see more investment in the built environment and in access to healthy food.

• The RISE Consortium is testing interventions to preserve or improve beta cell function in prediabetes or early type 2 diabetes. In the pediatric study, participants were randomized to one of two therapy regimens: (i) metformin for 12 months; or (ii) three months of insulin glargine followed by nine months of metformin. Neither treatment halted progression of beta cell deterioration and in both groups clamp-measured beta cell function fell from a low baseline. Despite the negative results from this study overall, we learned at last year’s ADA that there was a small but significant decrease in BMI with metformin therapy at three months (from 37 kg/m² to 36 kg/m², p<0.05) and A1c declined to 5.6% (baseline: 5.7%) with insulin glargine after three months (p<0.05 – we wonder if this could be significant even just in terms of the accuracy of the A1c test alone – speculation on our part), and also dropped significantly with both medication regimens at six months (p<0.05 vs. baseline). At the time, Dr. Buse positioned it as a positive that the impact on BMI and A1c was as expected with metformin and with insulin glargine.

• According to the CDC’s 2017 Diabetes Report Card, unadjusted incidence of type 2 diabetes grew a staggering 7.1% annually, from nine cases/100,000 youths per year in 2002-2003 to 12.5 cases in 2011-2012 (p<0.001). Unfortunately, obesity prevalence in children remains very high. In 2017, the CDC’s data brief on the NHANES 2015-2016 survey found childhood obesity prevalence to be 19%, up from 17% in 2013-2014 (nearly an 8% increase although not technically statistically significant – but surely the chance from for example 2010 would’ve been statistically statistically). This is all the more concerning given the sobering RISE results, which stress the limited pharmacotherapy options to address the underlying pathophysiology of type 2 diabetes in youth.

Dr. Andrew Stewart on The Combined Inhibition of DYRK1A and TGFß to Induce Robust Proliferation in Human Beta Cells

Icahn School of Medicine’s Dr. Andrew Stewart presented data showing that the combined action of harmine and TGFß-inhibitors results in robust beta cell proliferation. Picking up right where he left off from his talk at AACE 2018, Dr. Stewart reminded the audience how his team has previously shown the small molecule harmine to block DYRK1A, which in its un-inhibited state essentially “takes the breaks off beta cell proliferation.” The problem, Dr. Stewart explained, is that the resulting beta cell replication via harmine occurs at an impractically low rate relative to the little beta mass remaining in patients with diabetes. Separately, TFGß triggers a pathway that has been characterized to activate cell cycle inhibitors in insulinomas – rare, benign tumors that overproduce insulin and are critical to Dr. Stewart’s work (let Dr. Andrew Stewart know if you have insulinomas to donate). By inhibiting TGFß in beta cells, the cell cycle is no longer inhibited, potentially allowing for beta cell proliferation. Dr. Stewart hypothesized that by combining the inhibition of both DYRK1A and TGFß, perhaps the resulting beta cell proliferation would be stronger. Indeed, in a paper recently published in Cell Metabolism, Dr. Stewart demonstrated that by combining harmine with a TGFß-inhibitor, beta cell proliferation rates were on average 5%-8%, reaching as high as 15%-18%. Dr. Stewart considers these to be “crazy rates,” especially compared to the ~2% replication rate observed with harmine alone. To put this into perspective, in a period of just one week, the synergistic action of inhibiting DYRK1A and TGFß restored beta cell mass to practically normal levels following a 60% pancreatectomy. The increase in proliferation was shown to hold up not only in rodents but also in human beta cells transplanted into severe combined immunodeficient (SCID) mice – an approximation of an in vivo human experiment. These results are incredibly exciting and we are eager to see his team continue this important work

• Dr. Stewart estimates that for people with type 2 diabetes, it could take roughly six months to restore beta cell mass with the combined inhibitor, whereas for those with type 1 diabetes, it will be “still a tough slog, but maybe enough.” Of course, there is still the issue of managing autoimmunity in type 1 diabetes. Although, Dr. Stewart pointed out, autoimmunity is not present in type 2 diabetes, and since there are many more people with type 2 diabetes as compared to type 1 diabetes, it may make sense to first test the principle in type 2. Additionally, due to harmine’s off-target hallucinogenic effects, it will be necessary to target the therapy to beta cells. To this end, Dr. Stewart’s collaborators have predicted next-generation version of harmine, developing over 500 versions of harmine and other DYRK1A inhibitors that might be targeted to beta cells.

Scientific Achievement Award Lecture on The Role of Succinate Dehydrogenase in Beta Cell Dysfunction and Diabetes Pathogenesis

Scientific achievement award winner Dr. Sooyeon Lee (Stanford University, Palo Alto, CA) described a pathway involving succinate dehydrogenase, which may contribute to beta cell dysfunction and diabetes pathogenesis. She explained that succinate dehydrogenase, a critical enzyme in the citric acid cycle and electron transport chain, is known to link mitochondrial dysfunction and diabetes pathogenesis. Succinate dehydrogenase expression has been shown to be reduced in human patients with diabetes, and loss of succinate dehydrogenase leads to progressive diabetes in mice. In these knockout mice, islets failed to utilize alternative carbon sources, and both beta cell replication and response to glucose were significantly reduced. Mitochondrial respiration was also impaired, resulting in decreased ATP production. In explaining these results, Dr. Lee noted that in other cell types, loss of succinate dehydrogenase characteristically increases succinate and succinylation, impacting protein function. Sure enough, Dr. Lee found increased succinate in the knockout mice and a strong increase in succinylation. Interestingly, protein succinylation has been observed in islets of old-age mice, as well as islets derived from young prediabetic mice. For these reasons, Dr. Lee believes succinylation may be important in diabetes pathogenesis. The obvious question here, Dr. Lee posed, is whether the phenotype can be rescued. While this discovery is still quite early stage, the more we understand the myriad pathways that contribute to diabetes, the closer we can get to a cure.

Transcription Factor NKX2.2 is Essential for Beta Cell Identity; Beta Cell Plasticity a Blessing and a Curse; What Can Beta Cell Phenotypic Heterogeneity Tell Us About Disease-Modifying Strategies?

Dr. Lori Sussel (Barbara Davis Center for Diabetes, Aurora, CO) shared research demonstrating the importance of transcriptional networks in maintaining beta cell identity. In both type 1 and type 2 diabetes, loss of beta cell function is a hallmark of disease progression. While this loss of function has traditionally been attributed to beta cell death, Dr. Sussel and her team believe it could be a result of beta cell plasticity. Beta cells have a remarkable ability to dedifferentiate and transdifferentiate, and transcriptional networks are needed to promote beta cell function while repressing non-beta cell programs. To make her point, Dr. Sussel focused on one transcription factor in particular, NKX2.2. While prior work has shown that NKX2.2 is necessary for beta cell formation in both mice and human fetuses, Dr. Sussel hypothesized that this factor was also essential for beta cell maturation and maintenance. As expected, she found that, in young mice, disruption of NKX2.2 led maturing beta cells to transdifferentiate and acquire other endocrine cell identities. In adult mice, deletion of NKX2.2 also led beta cells to lose their identity, but to only adopt the characteristics of pancreatic delta (somatostatin-producing) cells. This finding was replicated in human islets, where inactivation of NKx2.2 again led to the upregulation of somatostatin. At the end of her talk, Dr. Sussel suggested that beta cell plasticity is both a blessing and a curse: On the one hand, it offers the possibility of reprograming beta cells and restoring their natural function (or inducing transdifferentiation from other pancreatic lineages to beta cells); on the other, it could lead to phenotypic instability. In the future, understanding the cues that influence pancreatic cell identity will also be essential for ensuring the success of islet transplants, said Dr. Sussel. In related news, the UCSF Hebrok lab recently reported that they had successfully produced highly-functional “enriched beta clusters” via a novel induction procedure.

• Throughout her talk, Dr. Sussel also pointed out that beta cell heterogeneity can offer new insights into developing effective therapies. She noted that, when the NKX2.2 transcription factor was knocked out in the beta cells of young mice, only 50% of beta cells underwent reprogramming. Among these, four different beta subpopulations were identified: (i) single-hormone insulin expressing cells; (ii) beta cells that coexpress insulin and other islet hormones; (iii) cells that only express other islet endocrine hormones; and (iv) cells that do not express any hormones. Further studies, she said, should focus on characterizing these subpopulations, each of which could play a different role in diabetes progression and suggest a different strategy for intervention.

Diabetes Complications & Cancer

Inflammatory Resolution Agonist Lipoxin Attenuates Human Carotid Plaque Inflammatory Response Ex Vivo, Mitigates Renal Dysfunction in Diabetic Mouse Model; Synthetic Development Underway

University of College Dublin’s Dr. Catherine Godson proposed that considering inflammatory resolution agonists rather than anti-inflammatory interventions may be more effective in preventing diabetes complications. She explained that inflammation does not stop independently in the body but is actively resolved via lipid mediators that clear lipocytes from the site of inflammation and switch molecules from pro-inflammatory to pro-resolving. These pathways have been shown to occur in humans, and inflammatory resolution can be accelerated with a cocktail of pro-resolving mediators. Given that there is a growing acceptance that inflammation plays a major role in diabetes complications, Dr. Godson asked wither lipoxins, a category of lipid mediators, might target diabetes complications. Indeed, using a mouse model of diabetes complications, Dr. Godson’s team showed that treatment with lipoxins was associated with a decrease in urinary albumin excretion at 10 weeks, with attenuation at 20 weeks. Not only were lipoxins shown to attenuate renal dysfunction in a diabetic mouse, it also was shown to attenuate renal dysfunction in mice with already established diabetes kidney disease. Additionally, lipoxins were shown to reduce diabetes-associated atherosclerosis, markedly reducing plaque size. Importantly, Dr. Godson noted that transcriptome profiling shows the mouse model is “a reasonably good model” for human disease. To explore the potential for translation in humans, Dr. Godson’s team tested lipoxins in human carotid plaques ex vivo. Excitingly, the lipoxins were shown to attenuate the inflammatory response of the plaque. She believes lipoxins act to switch macrophage expression into a more resolving phenotype. Given these strong results, work is underway to develop synthetic lipoxin mimetics. Investigators are currently characterizing the effect of these synthetic molecules in different screens and looking to see how they behave at specific sites. We’re intrigued by this approach of turning the problem on its head, harnessing the body’s innate pathways to resolve inflammation, rather than trying to introduce a novel anti-inflammatory agent

AMPK Phosphorylation of Acetyl-CoA Decreases Proliferation of Human Liver Cancer Cells; Implications for Relationship between Type 2 diabetes Treatments and Cancer

Dr. Gregory Steinberg (McMaster University, Ontario, Canada) shared research elucidating the mechanism through which AMP-activated protein kinase (AMPK) protects against liver cancer. Dr. Steinberg’s work builds on prior knowledge that the cellular energy sensor AMPK inhibits acetyl-CoA carboxylase (ACC), whose actions play a role in de novo lipogenesis (DNL). DNL is a contributing factor in the development of non-alcoholic fatty liver disease (NAFLD), and in many cancers including hepatocellular carcinoma (HCC). In a novel study, Dr. Steinberg sought to investigate whether AMPK-mediated phosphorylation of ACC is important for inhibiting tumor genesis and proliferation. As expected, in mice and human liver cells, it was found that blocking AMPK’s activity on ACC increased both liver DNL and cancer cell proliferation. Corroborating these results, a novel, liver-specific ACC inhibitor (ND-654) suppressed liver DNL and the development of HCC. Taken together, these studies provide convincing evidence for the role of AMPK-mediated ACC phosphorylation in the development of hepatocellular carcinoma. Not only might these results lead to future novel drug targets, they also point to a potential link between cancer and NAFLD. Better understanding this relationship may help in the development of therapies for both diseases.

• Given the inhibitory effect of AMPK on liver cancer, Dr. Steinberg emphasized that certain medications already used to treat type 2 diabetes may be protective against other types of cancer as well. In prior work, Dr. Steinberg has shown that metformin indirectly activates AMPK and synergizes with salicylate to inhibit de novo lipogenesis, reducing the survival of prostate and lung cancer cells ex vivo. The SGLT-2 inhibitor canagliflozin also indirectly activates AMPK, lowering liver lipid content and cancer cell proliferation. These data are exciting, and we can’t wait to learn more about the therapies that affect pathways at the interface of diabetes and cancer.

--by Caterina Florissi, Meghna Ray, Brian Levine, Maeve Serino, and Kelly Close