Glucose Recognition Proteins for Glucose Sensing at Physiological Concentrations and Temperatures

Glucose Recognition Proteins for Glucose Sensing at Physiological Concentrations and Temperatures

Smita Joel , Kendrick B. Turner , and Sylvia Daunert *

Department of Biochemistry and Molecular Biology, Miller School of Medicine, University of Miami, 1011 NW 15th Street, Miami, Florida 33136, United States

ACS Chem. Biol., 2014, 9 (7), pp 1595–1602

DOI: 10.1021/cb500132g

Publication Date (Web): May 19, 2014

Copyright © 2014 American Chemical Society

ABSTRACT: http://pubs.acs.org/doi/abs/10.1021/cb500132g

Motivation:  Current commercially available glucose measurement devices use an electrochemical process that employs enzymes like glucose oxidase and glucose dehydrogenase. Some of the factors that can be improved upon include poor performance in the hypoglycemic range, hematocrit dependence and interference from electrochemically active molecules (low selectivity for glucose) , hypoxemia, and hypotension. Many of these electrochemically active molecules that interfere with glucose measurements are common in medications: acetaminophen, salicylic acid, ibuprofen, ascorbic acid, etc. 

Alternative technologies that have been developed involve optical sensing systems fluorescence and bioluminescence glucose binding proteins (GBP). To optimize the ability of GBP as a sensing protein for glucose, efforts have been made to modify its structure to achieve physiological binding affinities for glucose useful at physiological concentrations between 2 to 20 mM.  Some of these efforts include site mutagenesis, random mutagenesis, and DNA shuffling.

In this paper, the author details the work done by the researchers to employ a different method for improving the binding affinity of GBP.  Unlike other methods focusing on the binding site, the truncation method preserves the first layer of molecular recognition layer allowing better control of the binding affinity without loss or compromise of the recognition and therefore sensing ability.

TESTS ON TRUNCATED PROTEINS

To summarize the structural changes, three truncated versions of GBP (tGRP1, tGRP2, and tGRP3) were constructed retaining the hinge region to preserve the ability for conformational changes and the amino acids responsible for hydrogen bonding to preserve affinity and selectivity for glucose. The binding site consists of 8 amino acids comprising the first layer of binding interactions with sugar ligand stabilized by a second layer of 10 more amino acids. In tGRP1 and tGRP2, the first layer of interaction was retained but portions of the second layer were removed.  In tGRP3, portions were removed resulting in glucose-binding activity.

The binding affinity for glucose and other sugars was characterized for the truncated GRP’s.  Some noted results are:

·         Increasing truncation caused the dissociation constant Kd to increase thus decreasing affinity for glucose.

·         The Kd also increased when more the stabilizing second layer amino acids were removed.

To test whether disrupting the first- and second-layer amino acids would result in loss of selectivity for glucose and galactose, they subjected the tGRP1 and 2 along with native GBP to fluorescence studies of their binding interactions with 6 sugars: maltose, lactose, sucrose, fructose, galactose, and glucose.  Both the tGRPs mimic the binding activity of the native GBP selective primarily for glucose and secondarily for galactose.  No significant response was observed for the other sugars.

The researchers also studied the thermal stability of the truncated proteins using circular dichroism spectroscopy.  The melting point of native GBP is 52.4 C.  Thermal stability is an important characteristic for the use of these proteins for glucose sensing at physiological conditions.  As noted by the authors, “Protein thermal stability is an important consideration when developing protein-based sensors that will be used for extended periods of time at 37 °C, the temperature of the human body. Improved thermal stability should increase the lifetime of the sensor, allowing for long-term, reproducible glucose determination.” Some results:

·         The melting point of the protein decreased as it got smaller upon truncation

·         The thermal stability was significantly affected by the truncation.  The researchers hypothesized that this is due to loss of protein stability which likely caused the high dissociation constant with glucose.

TESTS ON TRUNCATED GRP’S WITH INCORPORATED UNNATURAL FLUORINATED AMINO ACIDS (uGRPs)

It is known that fluorinated amino acids promote chemical, thermal, and conformation stability in proteins. The similarity of the van der Waal’s radii between fluorine and hydrogen makes the substitution amenable “with minimal steric perturbation” in the protein.

Improving thermal stability has the following practical advantages: “increased shelf-lives, longer continuous use capabilities, decrease in refrigeration storing needs, easier packaging/transport of these peptides, and use at ambient, physiological, and extreme environments.”

Some noted results:

·         Glucose binding was maintained with incorporation of unnatural amino acids

·         uGRP with unnatural tryptophan showed lower detection limits compared to uGRP with unnatural leucine

·         increased dissociation constants were observed for all uGRPs

·         Structural studies done using far UV-CD analysis showed that the full-length unnatural tryptophan-incorporated GRP (uGRP-FW) showed no considerable change in secondary structure relative to GBP while indicating secondary structure changes in the full-length unnatural leucine-incorporated GRP (uGRP-FL) compared to GBP. This may explain why the uGRP-FW retained a dissociation constant similar to GBP in the micromolar range while uGRP-FL did not.

·         The same far UV-CD studies indicated structural changes in the uGRP1/2-FW truncated proteins compared to the full length uGRP-FW, with results suggesting a decrease in alpha-helical structures.

·         Secondary structure changes were also indicated for the FL truncated proteins.

·         uGRP-FW and uGRP-FL showed improved thermal stability compared to the full-length GBP and the the truncated proteins.

·         Addition of fluorinated leucines resulted in higher melting points (see table in Supporting Information documents) compared to incorporation of fluorinated tryptophans. Overall, incorporation of FW and FL resulted in higher melting points for the truncated GRP’s.

·         Tests conducted to measure the detection ability of the modified proteins at different physiological conditions indicated improved sensing capacity with ability to detect glucose at physiological millimolar concentrations with the facilitation of a polymerized acrylamide hydrogel on the sensing tip of the optical fiber. The authors conclude, “Thus, the sensors incorporating designer proteins with enhanced thermal stability retain their binding ability toward glucose and can be employed for monitoring glucose at physiological temperatures, from hypothermia to hyperthermia.”

Summary/Conclusion in the authors’ own words:

“In conclusion, herein we present genetically engineered sensing proteins that can function as biosensors with desired properties based upon truncated forms of GBP from E. coli and expansion of the genetic code by incorporation of unnatural amino acids into these proteins. Previous work has shown wildtype GBP to have a binding constant in the micromolar range, which is not ideally suited for the development of a glucose biosensing system at physiologically relevant millimolar concentrations. The apparent binding constant of the truncated proteins is shifted from the micromolar to millimolar range, allowing for glucose determination at physiological ranges.”

And,

“The enhanced thermal stability and altered KD’s of these newly prepared GRPs makes them especially suited for long-term continuous glucose sensing in a variety of platforms and devices as well as for transport and long-term storage.”