LSTMs and Deep Residual Networks for Carbohydrate and Bolus Recommendations in Type 1 Diabetes Management.

To avoid serious diabetic complications, people with type 1 diabetes must keep their blood glucose levels (BGLs) as close to normal as possible. Insulin dosages and carbohydrate consumption are important considerations in managing BGLs. Since the 1960s, models have been developed to forecast blood glucose levels based on the history of BGLs, insulin dosages, carbohydrate intake, and other physiological and lifestyle factors. Such predictions can be used to alert people of impending unsafe BGLs or to control insulin flow in an artificial pancreas. In past work, we have introduced an LSTM-based approach to blood glucose level prediction aimed at "what-if" scenarios, in which people could enter foods they might eat or insulin amounts they might take and then see the effect on future BGLs. In this work, we invert the "what-if" scenario and introduce a similar architecture based on chaining two LSTMs that can be trained to make either insulin or carbohydrate recommendations aimed at reaching a desired BG level in the future. Leveraging a recent state-of-the-art model for time series forecasting, we then derive a novel architecture for the same recommendation task, in which the two LSTM chain is used as a repeating block inside a deep residual architecture. Experimental evaluations using real patient data from the OhioT1DM dataset show that the new integrated architecture compares favorably with the previous LSTM-based approach, substantially outperforming the baselines. The promising results suggest that this novel approach could potentially be of practical use to people with type 1 diabetes for self-management of BGLs.

Bace2 is a ß cell-enriched protease that regulates pancreatic ß cell function and mass.

Decreased ß cell mass and function are hallmarks of type 2 diabetes. Here we identified, through a siRNA screen, beta site amyloid precursor protein cleaving enzyme 2 (Bace2) as the sheddase of the proproliferative plasma membrane protein Tmem27 in murine and human ß cells. Mice with functionally inactive Bace2 and insulin-resistant mice treated with a newly identified Bace2 inhibitor both display augmented ß cell mass and improved control of glucose homeostasis due to increased insulin levels. These results implicate Bace2 in the control of ß cell maintenance and provide a rational strategy to inhibit this protease for the expansion of functional pancreatic ß cell mass.

An efficient system to generate monoclonal antibodies against membrane-associated proteins by immunisation with antigen-expressing mammalian cells.

BACKGROUND: The generation of monoclonal antibodies specific for protein antigens usually depends on purified recombinant protein for both immunisation and hybridoma screening. Purification of recombinant protein in sufficient yield and purity is a tedious undertaking and can be demanding especially in the case of membrane proteins. Furthermore, antibodies generated against a purified recombinant protein are frequently incapable of binding to the endogenous protein in its native context. RESULTS: We describe a strategy to generate monoclonal antibodies against membrane or membrane-associated proteins that completely bypasses any need for purified recombinant antigen. This approach utilises stably transfected mammalian cells expressing recombinant antigens on their cell surface for immunisation of mice. The transfected cells are also used for measuring seroconversion, hybridoma selection and antibody characterisation. By presenting the antigen in its native conformation for immunisation and hybridoma selection, this procedure promotes the generation of antibodies capable of binding to the endogenous protein. In the present study, we applied this approach successfully for three predicted GPI-anchored proteins of the malaria parasite Plasmodium falciparum. CONCLUSIONS: The described entirely cell-based technology is a fast and efficient approach for obtaining antibodies reactive with endogenous cell-surface proteins in their native conformation.

Guinea pig chymase is leucine-specific: a novel example of functional plasticity in the chymase/granzyme family of serine peptidases.

To explore guinea pigs as models of chymase biology, we cloned and expressed the guinea pig ortholog of human chymase. In contrast to rats and mice, guinea pigs appear to express just one chymase, which belongs to the alpha clade, like primate chymases and mouse mast cell protease-5. The guinea pig enzyme autolyzes at Leu residues in the loop where human chymase autolyzes at Phe. In addition, guinea pig alpha-chymase selects P1 Leu in a combinatorial peptide library and cleaves Ala-Ala-Pro-Leu-4-nitroanilide but has negligible activity toward substrates with P1 Phe and does not cleave angiotensin I. This contrasts with human chymase, which cleaves after Phe or Tyr, prefers P1 Phe in peptidyl 4-nitroanilides, and avidly hydrolyzes angiotensin I at Phe8 to generate bioactive angiotensin II. The guinea pig enzyme also is inactivated more effectively by alpha1-antichymotrypsin, which features P1 Leu in the reactive loop. Unlike mouse, rat, and hamster alpha-chymases, guinea pig chymase lacks elastase-like preference for P1 Val or Ala. Partially humanized A216G guinea pig chymase acquires human-like P1 Phe- and angiotensin-cleaving capacity. Molecular models suggest that the wild type active site is crowded by the Ala216 side chain, which potentially blocks access by bulky P1 aromatic residues. On the other hand, the guinea pig pocket is deeper than in Val-selective chymases, explaining the preference for the longer aliphatic side chain of Leu. These findings are evidence that chymase-like peptidase specificity is sensitive to small changes in structure and provide the first example of a vertebrate Leu-selective peptidase.

Identification of oligomerization and drug-binding domains of the membrane fusion protein EmrA.

Many pathogenic Gram-negative bacteria possess tripartite transporters that catalyze drug extrusion across the inner and outer membranes, thereby conferring resistance. These transporters consist of inner (IMP) and outer (OMP) membrane proteins, which are coupled by a periplasmic membrane fusion (MFP) protein. However, it is not know whether the MFP translocates the drug between the membranes, by acting as a channel, or whether it brings the IMP and OMP together, facilitating drug transfer. The MFP EmrA has an elongated periplasmic domain, which binds transported drugs, and is anchored to the inner membrane by a single alpha-helix, which contains a leucine zipper dimerization domain. Consistent with CD and hydrodynamic analyses, the periplasmic domain is predicted to be composed of a beta-sheet subdomain and an alpha-helical coiled-coil. We propose that EmrA forms a trimer in which the coiled-coils radiate across the periplasm, where they could sequester the OMP TolC. The "free" leucine zipper in the EmrA trimer might stabilize the interaction with the IMP EmrB, which also possesses leucine zipper motifs in the putative N- and C-terminal helices. The beta-sheet subdomain of EmrA would sit at the membrane surface adjacent to the EmrB, from which it receives the transported drug, inducing a conformational change that triggers the interaction with the OMP.