Transcript Levy APS - Indiana University Bloomington

Optimization of the C-terminal Sequence in Glucagon to Maximize Receptor Affinity
J. Levy, V. Gelfanov, and R. DiMarchi
Department of Chemistry Indiana University, Bloomington Indiana 47405 U.S.A.
Table 1. Bioactivity of Glucagon Alanine Analogs
Experimental Design & Results
Abstract
Synthesis of Glucagon Analogs: Each analog was synthesized on a solid support using Boc-based in
situ neutralization chemistry as described by Kent, et al. (3) on either a highly-modified Applied
Biosystems 430A peptide synthesizer or a CSBio Model CS336 peptide synthesizer. Peptides were then
cleaved from the support using HF/p-cresol, 95:5, for 1hr at 0°C. Following HF removal and ether
precipitation, the peptides were extracted into 10% HOAc and lyophilized. Each peptide was then purified
using RP-HPLC in 0.1% TFA using a linear gradient of CH3CN, on a Waters Associates preparative HPLC
system. Fractions containing the desired peptide were pooled and lyophilized. The identity and purity of
each analog was confirmed by analytical HPLC and MS analyses.
Glucagon is a linear peptide of 29 amino acids that is highly conserved among vertebrates. The
C-terminal -helical region is central to receptor recognition. The depth of structural analysis in this
section of the hormone pales in comparison to those at the N-terminal region, where the presence of the
first amino acid and the nature of the ninth are vital to receptor signaling (1). Since native glucagon
constitutes a life-saving medicine with notable biophysical deficiencies, the individual contribution of each
C-terminal residue to glucagon’s pharmaceutical properties should be known. We have completed an
alanine scan of the C-terminal region of glucagon in the range of amino acid residues 20-28. This portion
of the molecule is highly homologous to the related hormone GLP-1.
Glucagon
GLP-1
Biological Activity (cAMP induction): The ability of each glucagon analog to induce cAMP was
measured in a firefly luciferase-based reporter assay. HEK293 cells co-transfected with either glucagonor GLP-1 receptor and luciferase gene linked to cAMP responsive element were employed for bioassay.
The cells were serum deprived by culturing 16h in DMEM (Invitrogen, Carlsbad, CA) supplemented with
0.25% Bovine Growth Serum (HyClone, Logan, UT) and then incubated with serial dilutions of either
glucagon, GLP-1 or novel Glucagon analogs for 5 h at 37oC, 5% CO2 in 96 well poly-D-Lysine-coated
“Biocoat” plates (BD Biosciences, San Jose, CA). At the end of the incubation 100 microliters of LucLite
luminescence substrate reagent (Perkin Elmer, Wellesley, MA) were added to each well. The plate was
shaken briefly, incubated 10 min in the dark and light output was measured on MicroBeta-1450 liquid
scintillation counter (Perkin-Elmer, Wellesley, MA). Inhibitory 50% (IC50) and effective 50% concentrations
(EC50) were calculated by using Origin software (OriginLab, Northampton, MA).
QDFVQWLMN
KEFIAWLVK
Similar alanine substitution in GLP-1 demonstrated that four of these same nine amino acids and, in
particular, the single Phe were vital for full potency (2). We have studied the glucagon bioactivity of each
glucagon analog, as well as selectivity with the GLP-1 receptor. Our results demonstrate that glucagon
behavior is highly dependent on the nature of amino acid sequence in this region of the hormone.
Additional modifications with non-alanine based amino acids are in progress.
Glucagon Receptor
Peptide
EC50, nM
n*
EC50, nM
n
Glucagon
0.160.11
3
6.482.40
3
A-20
0.220.12
3
4.962.54
3
A-21
0.230.09
2
42.4313.12
3
A-22
97.6535.35
4
6150.31293.79
3
A-23
18.9610.49
4
533.819.97
3
A-24
0.090.03
2
10.785.92
3
A-25
2.221.07
3
133.6470.21
3
A-26
6.440.19
2
312.03118.56
3
A-27
0.660.19
2
72.0337.63
3
A-28
0.210.04
2
31.6010.04
3
2908.50357.09
2
0.050.02
5
GLP-1
* Number of experiments
Q20
K20
V23
1
His
5
Ser
Gln
Gly
Thr
Vydac C-4, 4.6 x 15cm, 1.5 ml/min, 220nm
0.1%TFA in Acetonitrile Gradient
Phe
Thr
Ser
20
Arg
Arg
Ser
Tyr
Leu
Theoretical Mass = 3453.78
Phe
Val
Gln
Trp
Leu
L26
10
Tyr
Lys
Ser
Met
Asn
Thr --COOH
E21
W25
F22
Figure 3
GLP-1
Results & Discussion
Figure 1. Representative Glucagon Alanine Analog (Ala-29)
Chromatographic & Mass Spectral Analysis
Introduction
Glucagon is a linear peptide hormone of 29 amino acids of central importance in physiology. For more
than half a century it has been used as a critical care medicine in the treatment of life-threatening insulininduced hypoglycemia. The biophysical properties of natural sequence glucagon are not conducive to
formulation in a patient-friendly formulation. The hormone is poorly soluble at physiologic pH and prone to
physical aggregation to insoluble fibrils. Consequently, glucagon is commercially supplied as a lyophilized
powder to be solubilized in dilute aqueous HCl immediately prior to administration. To a patient that is
semi-conscious or unconscious this represents an obstacle to proper administration and could constitute a
fatal flaw.
GLP-1 Receptor
mediated cAMP Induction
Glucagon Receptor
mediated cAMP Induction
5000
A
B
30000
Glucagon
A20
A21
A22
A23
A-24
A-25
A26
A-27
A-28
GLP-1
4000
3000
CPS
We explored the structural modification of native glucagon with the intent of enhancing the physical
properties with minimal change in pharmacology. As a first step in optimizing the pharmaceutical
properties of glucagon we have completed an alanine scan in the C-terminal region of the hormone
(residues 20-28) to identify those residues that are central to biological function. Alanine scaning is a
proven approach to initial segregation of those amino acids that contribute structurally to receptor
signaling through peptide backbone conformational effect versus direct side chain interaction. In a similar
study with the highly homologous peptide GLP-1 (2) it was demonstrated that in vitro bioactivity was
extremely sensitive to substitution with alanine at the amino acids comparable to residues 22 and 23 in
this region of the peptide. Our study explores changes in potency at the glucagon receptor and specificity
for in vitro action at the GLP-1 receptor.
W25
Glucagon
25
Gln
Asp
Asp
F22
Asp
15
Ala
D21
A24
HPLC Purified Glucagon(Ala-29)
2000
1000
25000
20000
15000
10000
5000
0
0
1E-3 0.01
0.1
1
10
100
[Peptide], nM
1000 10000
1E-4 1E-3 0.01 0.1
1
10
100 1000 10000
[Peptide], nM
Figure 2. Glucagon (A) and GLP-1 (B) receptor-mediated cAMP induction by glucagon analogs
CPS
NH2 --
I23
Q24
L26
Structure of Native Glucagon
GLP-1 Receptor
The synthesis and purification of the nine glucagon analogs was relatively straightforward and each
peptide was obtained in total yields in excess of 20% based on the weight of the starting amino acid resin.
None of the peptides proved any more problematic than native glucagon in the physical handling and
formulation for bioassay. A single example in the synthesis of this set of alanine-substituted glucagon
analogs is shown in Figure 1. The representative chromatographic and mass spectral analyses
demonstrate the integrity and purity of the analogs studied in this report.
The results of the bioassay at each of the two receptors are shown in Figure 2 and Table 1. It is
immediately obvious that glucagon has a number of amino acids in this C-terminal region of the peptide
where the bioactivity at both receptors is extremely sensitive to substitution with alanine. Consistent with
previously reported alanine scanning studies with GLP-1, residues 22 and 23 were extremely sensitive to
alanine substitution (2). Residues 25 and 26 were also significantly reduced in bioactivity with alanine
substitution but to a more modest degree. Most notably, the amino acids that border these four residues
were relatively insensitive to substitution and support the report that this region of glucagon is prone to
alpha-helix formation (4). The directional changes in bioactivity at the two receptors with each substitution
studied was consistent and differed in magnitude only at positions 21 and 28 where in both instances the
GLP-1 activity appeared to decrease to an appreciably larger extent.
Figure 3 provides a helical wheel representation of this C-terminal region and illustrates the amphiphatic
structure and high homology within these two peptide hormones, particularly in the biologically sensitive
hydrophobic residues (colored in red). Of particular note on the hydrophilic side of this helix (colored in
yellow) is the change of the glutamines at residues 20 and 24 in glucagon with lysine-20 and alanine-24.
We believe that these two differences are the structural basis for the physical properties that render
glucagon more prone to physical aggregation and formation of high molecular weight fibrils (5), since
glutamine is much more supportive of secondary beta-structure than either alanine or lysine.
These results in concert with prior reports (6) of structure-activity in this region of glucagon form the basis
for design of more potent and physically stable glucagon agonists.
References
1. Unson, C. G. and Merrifield, R. B. (1994) PNAS 91, 454-458.
2. Adelhorst, K. et al. (1993) J. of Biol. Chem 269, 6275-6278.
3. Schnolzer, M., Kent, S.B.H. et al. (1992) Int. J. Peptide Protein Res. 40, 180-193.
4. Ying, J., Hruby, V. J., et al. (2003) Biochemistry 42, 2825-2835
5. Onoue, S. et al. (2004) Pharm. Res. 21, 1274-1283.
6. Hruby, V.J., Ahn, J-M.,Trivedi, D. (2001) Curr. Med. Chem.–Imm., Endoc. & Metab Agents 1, 199-215.