Approximately 9.4% of Americans, according to a 2014 NIH survey, currently live with Type 2 diabetes, the equivalent of 29.1 million people¹. Alarmingly, the prevalence of diabetes in the United States has nearly tripled between 1980 and 2011, and the trend continues to climb upwards.² Resulting from this increased prevalence, diabetes has risen to become the seventh leading cause of death of Americans in 2013³. Additionally, improper diabetes management is a risk factor for developing heart disease, the number one leading cause of death in the United States⁴. As diabetes diagnoses continue to rise, the pressure to develop novel treatments to increase the manageability of the condition has increased. Here, we propose our version of a synthetic in vivo Type 2 diabetes treatment.

Type 2 diabetes is often colloquially referred to as “adult-onset diabetes,” because it may be diagnosed later in life. The precise cause of the condition may be a combination of varying genetic predispositions and environmental factors such as diet and exercise habits⁵. Despite varying etiologies, Type 2 diabetes exhibits several definitive clinical symptoms. Similar to Type 1 diabetes, Type 2 diabetes is characterized by the buildup of glucose in the blood stream.

At the onset of the disease, the buildup of glucose in the blood is caused by insulin insensitivity. Under normal physiological conditions, insulin, produced by beta cells in the pancreas, is released into the blood in response to a rise in blood-glucose levels. Once released, insulin works to regulate metabolism by promoting the absorption of glucose from the blood into the surrounding cells. Additionally, the presence of insulin promotes a negative feedback loop that inhibits the synthesis of glucose in the liver by suppressing the production of glucagon⁶. However, insulin insensitivity in individuals with Type 2 diabetes occurs when the pancreas continues to produce ample amounts of insulin, but the hormone receptors do not recognize it properly. In conditions of insulin insensitivity, insulin does not promote the absorption of glucose into cells, leading to incidences of increased blood-glucose levels⁷. When left untreated, insulin insensitivity leads to a condition where the body is no longer able to produce insulin and requires exogenous insulin to maintain blood-glucose homeostasis.

Treatments for Type 2 diabetes vary depending on the severity of the disease. Some newly diagnosed patients may find that lifestyle changes, such as changes in diet and exercise patterns, are enough to manage their condition. Patients may also find that they need the aid of medications to help manage their blood-glucose levels. There is a wide variety of medications, all utilizing different mechanisms of action, available to help Type 2 diabetics maintain their blood-glucose levels. Alpha-glucosidase inhibitors help to slow the digestion of starches to limit blood-glucose level spikes after meals. Thiazolidinediones can be used to increase cell receptor responsiveness to insulin⁸. Eventually as the disease progresses, many patients will become reliant on exogenous insulin.

Current Type 2 diabetes treatments sufficiently treat the disease in a cost-effective and relatively non-invasive way. Nevertheless, currently available treatments still have shortcomings that can be improved upon. A primary concern with insulin therapy in patients with Type 2 diabetes is treatment frequency and timing of treatments. For young and old patients, remembering to monitor blood-glucose levels and deliver insulin when necessary may be difficult, thus placing additional burden on care takers. Delay in insulin treatment can lead to periods of hyperglycemia; chronic hyperglycemia can lead to conditions such as reduced circulation in the extremities, heart disease and stroke⁹.

We propose an insulin independent system that is capable of sensing conditions of hyperglycemia and releasing a pharmacologically-active tripeptide in response. The entire system would be autonomous and can be packaged into a microcapsule that can then be inserted into the body subcutaneously. Because the system can monitor blood-glucose levels in vivo via the manipulation of endogenous EnvZ sensor in E. coli, it can produce therapeutic tripeptides on demand. This design has the capability to reduce treatment burden on caregivers. The proposed treatment design is currently incomplete in regard to its ability to sense blood-glucose levels. Because the EnvZ sensor senses raw osmolarity, any dissolved solutes other than glucose could also induce a response from the sensor. A theoretical glucose sensor based off of a modified lac operon was designed to sense changes in cAMP rather than in osmolarity.

The natural Envz inner membrane protein is an osmoregulator in Escherichia coli. The Envz protein controls the production of two membrane proteins, Outer membrane porin protein C (OmpC) and outer membrane porin protein F (OmpF), by regulating the phosphorylation state of a transcriptional regulatory protein (OmpR). At low osmolarity, the Envz’s reactions with the OmpR proteins are slower. As a result, small amounts of the OmpR-P proteins are produced. The activity of the Envz protein rises with increasing osmolarity¹¹ (Figure 1).

After the OmpR is phosphorylated to OmpR-P, it will bind to the promoters that regulate OmpC and OmpF genes. There are three domain to each promoter and two OmpR-P binding sites per domain. The OmpR-P binds to the promoter regions in a galloping motion. Two OmpR-P needs to bind to one domain, forming a dimer, before the promoter is turned on (Figure 2.) There is a hierarchy to the way OmpR-P binding to the promoters. The hierarchy for the binding of the promoters are as followed: F1, C1>F2, F3>C2>C3, F4¹⁰. As more domains of the promoter region are occupied, the expression of the protein downstream will increase. The promoter upstream of the OmpF protein has a high affinity for OmpR-P. Thus OmpF is made at a low ambient osmolarity. OmpF protein’s channel is smaller compared to the channels of its cousin protein, OmpC. As a result large molecules cannot move out of the cell by filtrated diffusion when the osmolarity is low. At a high osmolarity, there is higher Envz activity resulting in more OmpR-P formation. With the OmpF promoter sites occupied, the free OmpR-P will begin to bind to the OmpC promoter, turning on its expression. OmpC is a porin protein with large channels. At high ambient osmolarity, a large pore allows for the free movement of nutrients into the cell¹⁰.

Our osmolarity sensor utilizes the natural Envz-OmpR sensor in E. Coli. We added our protein of interest, a tripeptide, downstream on the OmpC promoter. When there is an increase in glucose level in the blood stream, the Envz sensor will phosphorylate OmpR to OmpR-P. Then OmpR-P will bind to the OmpC promoter turning on the expression of our tripeptides.

Canagliflozin, also known as Invokana, is a drug developed by Mitsubishi Tanabe Pharma as an insulin independent treatment for type II diabetes. The drug targets the SGLT2 protein in the proximal tubule of the kidney. SGLT2 is responsible for the reabsorption of 90% of glucose in the kidney. By directly inhibiting the actions of SGLT2, Invokana is able to reduce the amount of glucose reabsorbed into the body and excrete the glucose via urination. The activity of the drug reached its optimal peak one to two hours after intake. According to test trials done by the developer, side effects such as urinary tract infection and hypoglycemia were not detected. Although consuming the drug did increase the risk of dehydration¹².

Our team looked for a biological drug capable of doing the same job as Invokana. QSP is a tripeptide produced by humans naturally that has a similar function to Invokana. The human RSC1A1 gene codes for the tripeptide. Although QSP has the same effects as Invokana, its working mechanism is significantly different. Invokana works by directly inhibiting the STLG2 membrane proteins and thereby reducing the reabsorption of glucose. However, QSP works as a post-transcriptional inhibitor of STLG2 at the trans-Golgi network (TGN). The tripeptides reduces the expression of STLG2 by 40%-50% at the plasma membrane¹³.

The model we used for tripeptide production is the sausage-protease model. Since the tripeptides are coded for by only nine base pairs, it is hard for transcriptional factors to transcribe it. In addition, the start codon in the beginning of the sequence disrupts the function of the protein. Thus we came up with the sausage model of tripeptides. The sequence for the tripeptides will be repeated several times so when the gene is transcribed there is three copies of tripeptides. At the end of the tripeptide link there is a HLY secretion tag. We also engineered another cell that contains a protease attached to the outer membrane of E. coli via an ompA membrane protein. The tripeptides will be secreted out of the cell into the extracellular environment where it will meet up with the protease to receive its final modifications. The protease will cut the sausage tripeptides into individual units before it enters the bloodstream.