the conformational and biological analysis of a cyclic anti-obesity peptide from the c-terminal...

E. OgruJ.C. WilsonM. HeffernanW.-J. JiangD.K. ChalmersR. LibinakiF. Ng

Authors' af®liations:

E. Ogru, M. Heffernan, W.-J. Jiang, R. Libinaki

and F. Ng, Department of Biochemistry and

Molecular Biology, Faculty of Medicine, Monash

University, Clayton, Australia.

J.C. Wilson and D.K. Chalmers, Department of

Medicinal Chemistry, Monash University

(Parkville Campus), Australia.

Correspondence to:

Professor Frank Ng

Department of Biochemistry and Molecular

Biology

Faculty of Medicine

Monash University

Clayton 3800

Victoria

Australia

Tel.: 61-3-9905-3777

Fax: 61-3-9905-4699

E-mail: [email protected]

Dates:

Received 15 February 2000

Revised 11 May 2000

Accepted 28 June 2000

To cite this article:

Ogru, E., Wilson, J.C., Heffernan, M., Jiang, W.-J.,

Chalmers, D.K., Libinaki, R. & Ng, F. The conformational

and biological analysis of a cyclic anti-obesity peptide

from the C-terminal domain of human growth hormone.

J. Peptide Res., 2000, 56, 388±397.

Copyright Munksgaard International Publishers Ltd, 2000

ISSN 1397±002X

The conformational andbiological analysis of a cyclic

anti-obesity peptide from theC-terminal domain of humangrowth hormone

Key words: conformation; human growth hormone;

lipogenesis; lipolysis; molecular modeling; NMR; peptide

Abstract: The three-dimensional solution structure of anti-

obesity drug (AOD), a 15-residue, disul®de-bonded, cyclic

peptide, cyclo(6,13)-H2N-Leu-Arg-Ile-Val-Gln-Cys-Arg-Ser-Val-Glu-

Gly-Ser-Cys-Gly-Phe-OH, derived from the C-terminal domain of

the human growth hormone (hGH) (residues 177±191) was

determined using two-dimensional 1H NMR spectroscopy. AOD

stimulates lipolysis and inhibits lipogenesis, in vitro, in rodent,

porcine and human adipose tissues. These biological effects

suggest that AOD is a potential therapeutic candidate for the

treatment of obesity. Conformational studies of AOD were

conducted in aqueous solution and in water/dimethylsulfoxide

mixtures. In general, spectral quality was superior in the water/

dimethylsulfoxide mixtures. The cyclic region of AOD in water/

dimethylsulfoxide adopts type I b-turns at residues Ser8-Val9-

Glu10-Gly11 and Ser12-Cys13-Gly14-Phe15, each preceded by loop-

like structures. Comparison of the conformation of this peptide

with residues 177±191 in the native hGH protein X-ray crystal

structure indicates that the synthetic peptide retains some

structural similarity to the intact protein. This study provides

evidence that the C-terminal region of hGH is a speci®c

functional domain of the multifunctional hGH protein.

Abbreviations: Acm, acetamidomethyl; AOD, anti-obesity drug;

DIC, diisopropylcarbodiimide; DSS, 4,4-dimethyl-4-silapentane-

1-sulfonate; FAB-MS, fast atom bombardment-mass

spectrometer; Fmoc, 9-¯uoreylmethoxycarbonyl; hGH, human

growth hormone; HOBt, 1-hydroxybenzotriazole; MD, molecular

dynamics; TFA, tri¯uoroacetic acid.

Human growth hormone (hGH) is a multifunctional protein

hormone (22±26 kDa). It is synthesized in the pituitary

388

gland and stored in the cells of the anterior pituitary (1). In

mature animals it is released from these cells in response to

insulin-induced hypoglycemia, starvation, amino acid infu-

sion and by neural stimuli, including emotional stress (2).

These responses suggest that growth hormone plays an

integral role in the regulation of the metabolism of adult

animals (3). It is therefore recognized that hGH may play a

vital role in the treatment of obesity. Obesity is an

increasing health problem in af¯uent Western societies

and developing countries that contributes to the prevalence

of serious health disorders such as diabetes and heart disease

(4). The currently available pharmacotherapies for obesity

have adverse side-effects and are not effective in the long-

term (4). Opportunities therefore exist to develop novel

therapeutics to treat this increasingly common and expen-

sive disorder.

The growth hormone molecule comprises several dis-

crete, functionally important, bioactive domains (5). Ng and

colleagues (6,7) reported that hGH fragment 4±15 accent-

uates the decrease in circulating glucose in rats, which is

thought to be due to an increase in the number of insulin

receptors in adipose tissue membrane. Mondon et al. (8)

discovered that a number of N-terminal fragments of hGH

including hGH 1±15, 1±42 and 32±46 are effective in

enhancing insulin-like activities. Human growth hormone

is responsible for biological actions that include the

reduction and redistribution of body fat (3), the regulation

of lipid metabolism and the promotion of growth (3).

Speci®cally, it has been known for some time that hGH

has a signi®cant effect on fat mass reduction by regulating

lipid metabolism (9,10). Chronic exposure of humans to

hGH has been shown to increase plasma free fatty acids, to

inhibit the conversion of glucose into lipid and decrease the

lipid content of adipose tissue (3,10).

Binding of hGH to its receptor is a prerequisite for its

activation (11). The hGH receptor is a member of a group of

receptors that are found on various cell types and are

generally involved in cell growth and differentiation. The

X-ray crystal structure of hGH bound to its receptor reveals

that hGH is comprised of four antiparallel a-helices and that

there are two separate receptor binding sites on hGH to bind

a dimeric form of the receptor (12).

Previous structural studies of fragments derived from

growth hormone have shown that they have secondary

structure elements similar to that of the intact protein (13).

Gooley et al. (14) studied the conformation of fragment

96±133 from bovine growth hormone by NMR and

molecular modeling techniques and found that formation

of an omega loop in the peptide fragment was similar to that

observed in the intact bovine growth hormone. This

fragment was shown to have a stable ordered structure in

aqueous solution. Roongta et al. (15) studied the N-terminal

residues 1±28 of human growth hormone, and found an a-

helix conformation from residues 8 to 24 similar to that

observed in helix-1 of native hGH. This region of the peptide

formed an amphiphilic helix in aqueous solution that

exhibited self-association to form a dimeric species.

We identi®ed the speci®c region of the C-terminal domain

of hGH that is responsible for the lipolytic and antilipogenic

properties of hGH. Moreover, this is the ®rst time that this

speci®c functional region has been identi®ed. This peptide

was identi®ed by examining successive hGH fragments

derived from proteolytic digest and enzymatic cleavage

(6,16). This peptide, anti-obesity drug (AOD), encompassing

residues 177±191 of the C-terminal domain of hGH, is the

minimum length of the hGH sequence that retains these

biological activities. X-ray crystal structure analysis of

intact, native hGH shows that this region of hGH has a

disul®de bond between residues 182 and 189 and that

residues 177±182 form part of an a-helix (12).

The peptide AOD, NH2-Leu-Arg-Ile-Val-Gln-Cys-Arg-Ser-

Val-Glu-Gly-Ser-Cys-Gly-Phe-OH (in cyclic form) has been

synthesized using solid-phase peptide synthesis techniques.

Our investigations show that this synthetic peptide retains

its lipolytic and antilipogenic properties in vitro. A previous

NMR study (17) of a truncated form of this peptide with only

cyclic residues 181±190 of this peptide has been completed

in 70% H2O/30% DMSO-d6 mixtures. Because our biologi-

cal investigations have revealed that residues 177±191 are

the minimum required for the observed lipolytic and

antilipogenic properties, we decided to reinvestigate this

larger peptide fragment with the aim of determining

whether residues 177±181, extraneous to the cyclic loop,

retain any helical structure as found in the native protein. It

was also of interest to us to compare our solution structure

with the X-ray crystal structure of the native hGH protein

and to try to determine whether any speci®c structural

elements are required for the observed biological activity.

Here, we report the three-dimensional structure of AOD

using NMR and molecular dynamics (MD) simulations.

This study may help in the design of new, more potent anti-

obesity analogs of AOD for future studies.

Ogru et al . Conformational analysis of an anti-obesity peptide

J. Peptide Res. 56, 2000 / 388±397 | 389

Experimental Procedures

Synthesis of the peptide

Peptide AOD was synthesized using manual solid-phase

synthesis. 9-Fluoreylmethoxycarbonyl (Fmoc) amino acids

were coupled by diisopropylcarbodiimide (DIC; 1 mmol)/

1-hydroxybenzotriazole (HOBt; 1 mmol) activation. After

synthesis, the peptide was cleaved from the resin and the

side-chain protective groups, except for the acetamido-

methyl (Acm) group of Cys, were removed using Reagent K

(9) by a 2-h treatment. Crude, linear Cys (Acm)182,189 AOD

peptide was puri®ed by reverse-phase high-performance

liquid chromatography (RP-HPLC) using a preparative

octadecyl (C18) column (Supercoil PLC-18 column,

21.23250 nm, 12 mm, 120 AÊ ; Supelco, Bellefonte, PA,

USA) with a H2O±acetonitrile gradient and 0.1% tri¯uor-

oacetic acid (TFA). Fractions corresponding to the major

peak were collected, pooled and freeze dried. The homo-

geneity of the puri®ed peptide was analyzed by RP-HPLC

using an analytical C18 column (Supelco PLC-18 column,

4.63250 mm, 5 mm, 300 A; Supelco) and characterized by

fast atom bombardment-mass spectrometer (FAB-MS) to

give M+H51796. Removal of Acm groups from Cys residues

and simultaneous formation of the disul®de bond between

Cys182 and Cys189 was carried out by iodine oxidation under

acidic conditions (18,19). A 3-mL solution of Cys

(Acm)182,189 AOD peptide (18 mg; 10 mmol) in 80% AcOH

was added to a 6-mL solution of iodine (22.86 mg; 90 mmol)

in 80% AcOH, and the mixture was allowed to react at room

temperature with vigorous mixing. After 2 h, the reaction

was terminated by adding an excess of 10 mm ascorbic acid/

citrate solution (pH55.0).

The mixture which contained crude cyclic AOD peptide

was concentrated by evaporation and puri®ed by RP-HPLC

under conditions identical to those used for the linear

precursor. The molecular mass of AOD peptide was

characterized by FAB-MS, showing the M+H51654. Analogs

were synthesized using an identical procedure.

Animals and tissues

Rodent adipose tissue was from male C57BJ/6L mice,

12 weeks of age. Porcine adipose tissue was taken from

male commercial pigs at 20 weeks of age. Human sub-

cutaneous (abdominal) adipose tissue was taken with

consent from overweight female patients undergoing fat

reduction surgery.

Lipolysis assay

Lipolysis was measured as the release of glycerol into the

incubation medium. Adipose tissue was incubated in

4.0 mL of Krebs-Ringer bicarbonate (KRB) buffer (pH 7.4)

with 4% BSA in the presence of 1 mm of AOD in a 378C

water bath for 60 min. At the conclusion of the incubation

period, tissue samples were removed and discarded. The

amount of glycerol released into the medium was deter-

mined colorimetrically by removing a 10-mL aliquot of the

medium and placing it into a cuvette containing 1.0 mL of

GPO Trinder Reagent A (Sigma-337A). The increase in

absorbance at 540 nm is directly proportional to the glycerol

concentration of the sample. Glycerol output is expressed as

mmol of glycerol released per mg of adipose tissue.

Lipogenesis assay

Lipogenesis was measured as the rate of [14C]-glucose

incorporation into lipid in adipose tissue. Adipose tissue

was sliced into 200-mg segments for use. Tissue segments

were placed in ¯asks containing 2 mL KRB buffer (pH 7.4)

containing 2% BSA and 0.1 mg/mL glucose. Samples were

preincubated for 30 min in a 378C waterbath with constant

shaking (100 r.p.m.) and were gassed with 95% O2/5% CO2

throughout. After preincubation, tissues were removed,

rinsed in buffer, blotted and placed into ¯asks containing

fresh buffer and [14C]-glucose (®nal speci®c activity

0.05 mCi/mmol). Following the 60-min incubation, adipose

tissue samples were rinsed in 0.9% NaCl, blotted, extracted

in glass tubes containing 5 mL of the chloroform/methanol

(2 : 1 v/v) solution and refrigerated (48C) for 18 h. The

resulting tissue extract was placed into 10-mL centrifuge

tubes and tissue was again extracted in 2-mL chloroform/

methanol solution, vortexed and refrigerated for 30 min.

Extracts were pooled and mixed with an equal volume of

methanol/0.1% MgCl2 solution (1 : 1 v/v) for 15 min at 48C.

Tissue extracts were pooled and contents were centrifuged

for 10 min at 6000 g. The upper layer was removed by

vacuum suction and 1 mL of the lower layer was removed,

re-suspended in 8 mL liquid scintillation solution and used

for radioactive determination in a Wallac 1410 liquid

scintillation counter.

Statistics

Values are shown as mean u SEM. Data were analyzed by

Student's t-test when two values were compared. A value of

P,0.05 was considered signi®cant.


390 | J. Peptide Res. 56, 2000 / 388±397

NMR spectroscopy

Samples for 1H NMR spectra were prepared as follows: AOD

(6 mm) in 70% H2O/30% DMSO-d6 (99.96%), pH 3.9; AOD

(5 mm) in 90% H2O/10% 2H2O (99.96%), pH 3.5. Deuter-

ated solvents were obtained from Cambridge Isotope

Laboratories. Spectra were recorded at 278 K on a BruÈ ker

DRX spectrometer operating at 600 MHz with a shielded

gradient unit.

The two-dimensional experiments were recorded in

phase-sensitive mode using time-proportional phase incre-

mentation for quadrature detection in the t1 dimension (20).

The two-dimensional experiments included double quan-

tum ®ltered two-dimensional correlation spectroscopy

(DQF-COSY) (21), exclusive COSY (E-COSY) (22) and

TOCSY using a MLEV-17 spin-lock sequence (23) with a

mixing time of 120 ms, and NOESY (24) with mixing times

of 250 and 400 ms. The temperature calibration of the probe

was achieved by comparison to ethylene glycol chemical

shifts. All chemical shifts (p.p.m.) were referenced exter-

nally to the methyl resonance of 4,4-dimethyl-4-silapen-

tane-1-sulfonate (DSS, 0 p.p.m.).

In E-COSY experiments, solvent suppression was

achieved using selective low-power irradiation of the

water resonance during a relaxation delay of 1.8 s. Solvent

suppression for NOESY and TOCSY experiments was

achieved using modi®ed WATERGATE sequence (25) in

which two gradient pulses of 1 ms were applied on either

side of a binomial 3-9-19 pulse. Spectra were routinely

acquired over 6024 Hz with 4096 complex data points in the

F2 and 512 increments in the F1 dimension, with 32 scans

per increment (64 for NOESY). Slowly exchanging NH

protons were detected by acquiring a series of one-dimen-

sional and TOSCY spectra of the fully protonated peptide

immediately following dissolution in 2H2O.

The 3JHa-Hb coupling constants were measured from E-

COSY spectra and 3JNH-Ha coupling constants were mea-

sured either directly from one-dimensional spectra or a

DQF-COSY spectra, strip Fourier transformed to 8K31K.

Spectra were processed on a Silicon Graphics Indigo work-

station using xwin-nmr, version 2.5 (Bruker) software.

The t1 dimension was zero-®lled to 2048 real data points,

and 908 phase-shifted sine-bell window functions were

applied prior to Fourier transformation. Spectra were

analyzed using xeasy version 1.3 (26).

Structural restraints

Distance restraints were derived from a 400-ms NOESY

spectrum at 278 K. Automatic integration of peak volumes

was achieved using the protocols of xeasy (26). Peak

volumes were calibrated using an average of the volumes

of two well-resolved geminal CbH cross-peaks. Distance

constraints were calculated using volumes proportional to

r±6. Appropriate pseudo-atom corrections were applied to

nonstereospeci®cally assigned methylene and methyl pro-

tons (27). A 0.5-AÊ correction was added to the upper limits of

distances involving methyl protons. In addition, a 1.0-AÊ

correction was then added to backbone distance constraints

to allow for conformational averaging, errors in volume

integration and the effects of spin diffusion (which are likely

to be more signi®cant at the low temperature that the

spectra were acquired) (28). The ®nal constraint set

consisted of 91 intraresidue, 73 sequential, 27 medium

range (1,| i-j |#4) and eight long-range NOEs in 70% H2O/

30% DMSO-d6 (99.96%), and 84 intraresidue, 57 sequential,

14 medium range (1,| i-j |#4) and four long-range NOEs in

the 90% H2O/10% 2H2O solvent system. Backbone dihedral

restraints were calculated from 3JNH-Ha coupling constants

with w restrained to ±60u40 for 3JNH-Ha ,6 Hz. Speci®c

residues being Cys6, Ser8, Val9 in 70% H2O/30% DMSO-d6

(99.96%), and residues Arg7, Val9 and Glu10 in 90% H2O/

10% 2H2O.

Structure calculations

Initial structures were calculated using the torsion angle

dynamics program dyana version 1.5 (29). Several rounds of

structure calculation were performed to resolve ambiguous

NOEs and violated distance constraints. This process was

repeated until all the distance and angle restraints produced

a set of structures that had no NOE distance violations

.0.3 AÊ or dihedral angle violations .38. Once the ®nal set of

reliable distance restraints was obtained these data were

used as input for calculation of the three-dimensional

structures using a simulated annealing and energy mini-

mization protocol performed using x-plor 3.1 (30). A

simulated annealing protocol (31) was used to generate a

set of 200 structures, starting from template structures with

randomized w and y angles and extended side-chains.

Disul®de bonds were included as pseudo-NOE restraints.

The simulated annealing protocol consisted of 20 ps of high

temperature molecular dynamics (1000 K) with a low

weighting on the repel force constant and NOE restraints

followed by for a further 10 ps with an increased force


J. Peptide Res. 56, 2000 / 388±397 | 391

constant on the experimental NOE restraints. The disul®de

bonds were then formally included and the dihedral force

constant increased prior to cooling the system to 0 K and

increasing the repel force constant over 15 ps of dynamics.

The NOE restraints were checked for violations, and

ambiguous cross-peaks were resolved on the basis of

interproton distances in the initial family of structures. A

further 200 structures were calculated with the inclusion of

w dihedral angle restraints derived from spin±spin coupling

constants. Re®nement of these structures was achieved

using the conjugate gradient Powell algorithm with 2000

cycles of energy minimization (32) and a re®ned force®eld

based on the program charmm (33).

No hydrogen-bonding restraints were used and all peptide

bonds were de®ned as trans. Structures were analyzed using

promotif (34) and procheck-nmr (35) and displayed using

insightii (Biosym Technologies, San Diego, CA).

Results and Discussion

Biological data

Human growth hormone (hGH) has diverse physiological

actions, particularly concerned with lipid and carbohydrate

metabolism (3). According to Gertner (3) and Ho et al. (10),

hGH elicits a lipolytic and antilipogenic effect that is

mediated by reduced peripheral uptake and utilization of

glucose through reducing the responsiveness of the adipose

tissue to insulin. This study of the growth hormone peptide

fragment supports the concept that growth hormone is

multifunctional and may act at effectors distant from the

hGH receptor upon fragmentation of the molecule into

smaller peptides (6).

hGH is capable of inducing triglyceride breakdown

(lipolysis) and inhibiting lipid accretion (antilipogenesis)

in adipose tissues (2). Evidence suggests that AOD, a

15-residue cyclic peptide derived from residues 177±191 of

the C-terminal domain of hGH, is the lipolytic/antilipo-

genic domain of the intact hGH molecule (9,10). This study

investigated the in vitro ability of AOD to induce metabolic

changes with respect to lipogenesis (Fig. 1) and lipolysis

(Fig. 2) in rodent, swine and human adipose tissues.

Figure 1 shows the in vitro antilipogenic effect of AOD at

1 mm in rodent, swine and human adipose tissue. This effect

was determined by measuring the rate of [14C]-glucose

incorporation into triglyceride in isolated adipose tissue

using the method established by Jungus (36). These results

show a signi®cant reduction in lipogenic activity with AOD

treatment in all three adipose tissue types. For example,

Fig. 1 shows a 22% decrease in [14C]-glucose incorporation

in rodent, followed by a 30% decrease in porcine and a 40%

decrease in human adipose tissue. Lipolytic activity was also

measured (Fig. 2), with the greatest increase in human

adipose tissue, with a 272% increase in glycerol release,

followed by rodent with 141% and porcine with 124% upon

AOD treatment. The data shown indicate that AOD has a

higher ef®cacy in human adipose tissue compared with

porcine and rodent models. This may be due to the fact that

AOD originates from the human sequence of hGH.

Interestingly, it has been observed that AOD elicits its

lipolytic effect within minutes of incubation, whereas the

response induced using intact hGH is much slower, < 4 h.

The overall lipolytic effect of AOD and hGH is similar,

however, it has been observed that long-term administration

Figure 1. The antilipogenic effect of anti-obesity drug on rodent,

porcine and human adipose tissue. Results are the mean value of ®ve

tissues u SEM. *P,0.05.

µmol

gly

cero

l rel

ease

/g t

issu

e/h

Figure 2. The lipolytic effect of anti-obesity drug on rodent, porcine

and human adipose tissue. Results are the mean of ®ve tissues

u SEM. *P,0.05.


392 | J. Peptide Res. 56, 2000 / 388±397

of hGH results serious side-effects such as insulin-resistance

(37).

Alanine scan results

Figure 3 shows the in vitro effect of AOD peptide and a

series of analogs on the rate of lipogenesis in isolated rat

adipose tissue. This in vitro lipogenesis assay measures the

rate of glucose incorporation into lipid (pmol/mg tissue/

min). The analogs are identical to AOD except that each has

had an amino acid residue successively replaced with

alanine. Peptides with the cysteine residues replaced by

alanine show dramatically reduced antilipogenic activity.

These data suggest that formation of the disul®de is

important for biological activity. Moreover, the results of

lipogenesis assays using linear AOD have shown that there

is no antilipogenic or lipolytic activity (data not shown).

These results suggest that maximal lipolytic/antilipogenic

functions of the peptide are related to the cyclic conforma-

tion of the molecule.

Replacement of either of the charged residues Arg7 and

Glu10 results in a dramatic loss of antilipogenic activity. It

was therefore of interest to try and determine whether the

observed biological effects could be rationalized in terms of

the three-dimensional structure of AOD.

1H NMR resonance assignments

1H NMR spectra of AOD were acquired in two solvent

systems; aqueous solution (90% H2O/10% 2H2O) and 70%

H2O/30% DMSO-d6. Assignment of all proton resonances in

both solvent systems was accomplished using standard

NMR procedures (38). The amide region of the TOCSY

spectrum of AOD showed good chemical shift dispersion in

both solvents and individual amino acid spin systems could

readily be identi®ed. Each spin system was assigned

unambiguously to a speci®c amino acid residue in the

AOD sequence from the NOESY spectrum. These data are

presented in Table 1.

Although a signi®cant number of NOE connectivities was

observed in both solvent systems, the spectra acquired in

70% H2O/30% DMSO-d6 proved to be of better quality than

those acquired in aqueous solution because these spectra

exhibited less overlap of signals and more intense NOE

cross-peaks.

Figure 4 shows the ®ngerprint region of the 250 ms

NOESY spectrum of AOD acquired in 70% H2O/30%

DMSO-d6 at 278 K and 600 MHz. This spectrum shows a

complete cycle of aH-NH sequential connectivities for the

entire peptide sequence.

Secondary structure

Figure 5 presents a summary of the 1H-1H NOE connectiv-

ities for AOD in 70% H2O/30% DMSO-d6 (A) and 90% H2O/

10% 2H2O (B). It is evident from Fig. 5A that in 30% DMSO-

d6/70% H2O solution AOD has a large number of dNN(i,i+2)

and daN(i,i+2) NOE connectivities from residues Cys6 to

Cys13, with an uninterrupted span of such NOEs from

residues Glu10 to Phe15. Residues Cys6 to Phe15 encompass

the cyclic region of the AOD which is constrained by a

disul®de bond between Cys6 and Cys13. The dNN(i,i+2) and

daN(i,i+2) NOE connectivities are indicative of turn or loop-

like structures (17) throughout this region of the peptide.

The lack of NOE data in the N-terminal region of the peptide

from residues Leu1 to Gln5 (Fig. 5A,B) suggests that this

region of the peptide is devoid of secondary structural

elements and is predominantly conformationally averaged

in both solvent systems. The motion of this region also leads

to some averaging of the disul®de bond as the dbNH NOEs of

Cys6 were overlapped and there was a general paucity of

NOE data in this region. Consequently, it was not possible

to stereospeci®cally assign the b-methylene side-chain

protons of Cys6. In a previous study by Jois et al. (17) of

the truncated form of AOD stereospeci®c assignment of the

side-chain protons of Cys residues was possible on the basis

of dbNG and dab NOEs intensities and 3Jab coupling constants

leading to well-de®ned structure in this region. The

motional characteristics of the elongated tail of AOD

peptide no doubt in¯uence the conformation of the disul®de

Figure 3. In vitro effect of anti-obesity drug and related analogs with

mono-alanine substitutions on the rate of lipogenesis in isolated

rodent adipose tissue. Results are mean u SEM of 30 (in control or

anti-obesity drug) or six (in analogs) determinations. *P,0.05.


J. Peptide Res. 56, 2000 / 388±397 | 393

bridge which manifests itself as conformational averaging in

this region.

In aqueous solution, 90% H2O/10% 2H2O, there were

signi®cantly fewer dNN(i,i+2) connectivities observed

between residues Cys6 and Phe15, indicating that the

secondary structure is not as well de®ned as the

70% H2O/30% DMSO-d6 spectra.

Three-dimensional structure

Most information about secondary structures was obtained

from NOE data in the 70% H2O/30% DMSO-d6 solvent

mixture. Comparison of the chemical shifts of proton

resonances (Table 1) and NOE patterns (Fig. 5A,B) indicated

that the backbone conformation of the peptide is conserved

in different solvent systems. A set of 200 structures was

calculated using 158 NOE distance restraints and three

dihedral angle restraints for 70% H2O/30% DMSO-d6.

Figure 6 shows the superimposition of the backbone of the

20 lowest energy structures which demonstrates that,

p.p.

m.

p.p.m.

9.0 8.8 8.6 8.4 8.2 8.0 7.8 7.6 7.4

3.40

3.50

3.60

3.70

3.80

3.90

4.00

4.10

4.20

4.30

4.40

4.50

4.60

Figure 4. Fingerprint region of the 250 ms NOESY spectrum of anti-

obesity drug acquired in 70% H2O/30% DMSO-d6 at 278 K and

600 MHz.

Table 1. 1H Chemical shifts of AOD (p.p.m.)

90% H2O/10% 2H2O 70% H2O/30% DMSO-d6

NH a-H b-H Others NH a-H b-H Others

Leu1 3.87 1.41 d-CH2 0.64 Leu1 3.73 1.51

Arg2 8.39 4.10 1.54 c-CH2 1.08

d-CH3 1.38

Arg2 8.53 4.13 1.59 c-CH2 0.907

c-CH3 1.20

d-CH2 0.58

Ile3 8.71 4.32 1.73 c-CH2 1.62

d-CH3 0.84

Ile3 8.25 3.88 1.34

Val4 8.30 3.99 1.93 c-CH2 0.82 Val4 8.18 3.80 1.73 c-CH2 2.07

Gln5 8.48 4.24 1.85, 1.94 c-CH2 2.23 Gln5 8.37 4.06 1.69, 1.78 c-CH2 2.07

Cys6 8.53 4.55 2.98, 3.04 Cys6 8.43 4.37 2.83, 2.90

Arg7 8.68 4.32 1.48, 1.57 c-CH2 1.69

c-CH3 1.81

d-CH2 3.03

Arg7 8.55 4.15 1.29, 1.36 c-CH2 1.46

c-CH3 1.62

Ser8 8.01 4.39 3.70, 3.77 Ser8 7.87 4.19 3.52, 3.61

Val9 8.25 4.03 2.05 c-CH2 0.86 Val9 8.08 3.84 1.89 c-CH2 0.67

Glu10 8.39 4.18 1.90, 2.08 c-CH2 2.35 Glu10 8.25 4.01 1.74, 1.88 c-CH2 2.18

Gly11 8.06 4.02 3.70 Gly11 8.06 3.84

3.51

Ser12 8.05 4.39 3.74 Ser12 7.90 4.24 3.58

Cys13 8.60 4.59 2.94, 3.09 Cys13 8.45 4.43 2.76, 2.89

Gly14 8.36 3.78 Gly14 8.20 3.61

Phe15 7.83 4.41 2.89, 3.11 2, 6H 6.90

3, 5H 7.30

Phe15 7.60 4.24 2.90, 2.71 2, 6H 6.80

3, 5H 6.70


394 | J. Peptide Res. 56, 2000 / 388±397

despite its small size, the peptide leads to a well-de®ned

three-dimensional structure in the 70% H2O/30% DMSO-

d6 solvent mixture.

The cyclic region of AOD adopts a compact, tightly folded

structure. Examination of the ®nal set of structures using

promotif (34) revealed that the majority of the conformers

exhibited type I b-turns at residues Ser8-Val9-Glu10-Gly11

and Ser12-Cys13-Gly14-Phe15. Type IV b-turns were observed

at Arg7-Ser8-Val9-Glu10 and Glu10-Gly11-Ser12-Cys13 but, as

these turns precede type I b-turns, it is probably more

accurate to describe these structures as a loop-turn

conformation.

The quality of the structures in aqueous solution is, in

general, poorer than those seen in 70% H2O/30% DMSO-d6

solution, which is consistent with the reduced number of

distance constraints obtained from spectra acquired in this

solvent. In aqueous solution there was not one single family

of structures that adequately described the adopted fold in

solution. In fact, analysis of the three-dimensional struc-

tures generated revealed at least six families of conformers.

There were, however, structural features that were retained

throughout these families of conformers, for example, the

cyclic region of the peptide which remained tightly folded

with b-turn characteristics. Moreover, in four of the six

families of conformers a b-turn was retained at residues Ser8

to Gly11, and of these, two additionally had the b-turn from

residues Ser12 to Phe15, as seen for the 70% H2O/

30% DMSO-d6 structure.

It was of interest to examine the structural features of

AOD and try to rationalize the alanine scan results which

had insinuated that the charged residues Arg7 and Glu10 are

essential for biological activity. Figure 6 shows that Glu10 is

exposed on one side of the peptide, creating a negative

potential; and that Arg7 is exposed on the other side of the

peptide, creating a positive potential. This observation may

be important when considering potential receptor interac-

tions of AOD. Competition binding studies in the presence

(A)

(B)

Figure 5. Summary of 1H-1H NOE connectivities for anti-obesity drug

peptide in (A) 70% H2O/30% DMSO-d6 and (B) 90% H2O/10% 2H2O

at 278 K. The intensity of NOE cross-peaks is indicated by the

thickness of the lines, grouped into strong, medium, and weak.

(3JNH-Ha,6 Hz .).

Figure 6. Superimposition of the backbone

heavy atoms of the 20 ®nal structures of

AOD, superimposed over residues Cys6 to

Cys13. Residues colored blue were shown to

be most important for biological activity as

revealed by the alanine scan results.


J. Peptide Res. 56, 2000 / 388±397 | 395

of hGH or AOD have demonstrated that AOD does not

interact with the hGH receptor. However, to date, the

receptor that AOD binds to is unknown, but investigations

of its binding properties are underway.

Structure statistics

Geometric and energetic statistics that de®ne these 20

lowest energy structures (70% H2O/30% DMSO-d6) are

given in Table 2. In general, the structures exhibit no

signi®cant deviation from ideal covalent geometry and ®t

the experimental restraints with minimal violations. There

were no NOE violations .0.3 AÊ , and no dihedral angle

restraints violations .38. Superimposition of the family of

structures in Fig. 6 shows that most regions are well de®ned

but there is some residual disorder over residues Leu1 to

Gln5. The mean pairwise RMSDs over the whole molecule

were 0.30 AÊ for the backbone atoms and 0.55 AÊ for all heavy

atoms (Table 2).

Comparison of AOD and hGH(177±191)

To compare the conformation of the AOD in solution with

the native protein, we overlaid the conformation of the

peptide with the sequence comprising residues 177±191

from the X-ray crystal structure of the intact hGH protein

(39). The synthetic peptide AOD maintains structural

similarity with the analogous region of intact protein. In

the native protein there is a type II b-turn at Val9-Glu10-

Gly11-Ser12. However, a type I b-turn is observed at residues

Ser8-Val9-Glu10-Gly11 and conformational averaging is

observed between residues Leu1 to Gln5. The conformation

of the synthetic peptide in solution is somewhat similar to

that of intact protein; this may be due to the physical

restriction of the disul®de bond on the conformation of the

peptide.

This study further supports the concept that the intact

hGH serves as a precursor for small peptides with speci®c

metabolic actions (5,6,10). This concept is substantiated by

earlier ®ndings by Ng et al. (6,10) that the N-terminus and

the C-terminus of hGH are responsible for the insulin-like

and the insulin-antagonistic actions of the molecule,

respectively. This present study con®rmed that the C-

terminal region of hGH is the lipolytic/antilipogenic

domain of the multifunctional hGH molecule (9,40). The

outcomes of this study may facilitate a novel approach to

generate new therapeutics for the treatment of human

obesity.

Acknowledgment: Stuart Thomson is thanked for acquiring the

mass spectral data. This work is supported by Monash University

Research fund. EO is a recipient of an industry postgraduate

stipend from Metabolic Pharmaceuticals Limited, Australia.

Table 2. Geometric and energeticstatistics for the 20 energy-minimizedstructures of AOD in 70% H20/30%DMSO-d6

RMS deviations from idealized geometry

Bonds (AÊ ) 0.009u0.00019

Angles (deg) 2.034u0.065

Impropers (deg) 0.154u0.0012

Energies (kcal/mol)

ENOEa 2.015u0.53

Ecdiha 0.034u0.01

EL-Jb ±51.37u2.64

Mean pairwise RMS (AÊ )

Backbone heavy atoms 0.30u0.012

All heavy atoms 0.55u0.11

Values in the table are given as meanuSD.None of the structures had distance violations.0.3 AÊ or diherdal angle violations .38. a.Force constants for the calculation of square-well potentials for the NOE and dihedral anglerestraints were 50 kcal/mol AÊ and 200 kcal/mol/rad2, respectively. b. The Lennard±Jonesvan der Waal's energy was calculated with theCHARMM empirical energy function.

References1. Goodman, V. & Hirsch, J. (1968) Multiple

effects of growth hormone on lipolysis.

J. Clin. Endocrinol. 2, 551±555.

2. Pullin, A., Block, K. & Anker, H.S. (1981)

Effects of growth hormone on adipose tissue.

Biochemistry 183, 441±445.

3. Gertner, J.M. (1993) Effects of growth

hormone on body fat in adults. Horm. Res. 40,

10±15.

4. Elks, M.L. (1996) Appetite suppressants as

adjuncts in the treatment of obesity. J. Fam.

Pract. 42, 287±292.

5. Ng, F.M. (1990) Human growth hormone

fragments. In New Antidiabetic Drugs

(Bailey, C.J. & Flatt, P.R., eds). Smith-Gordon,

London; Nishimura, Niigata-Shi, Japan,

pp. 197±209.

6. Ng, F.M. (1987) The insulin-like actions of

human growth hormone and its fragments.

Asia Paci®c Comm. Biochem. 1, 163±169.

7. Ng, F.M., Bornstein, J., Pullin, C.E., Bromley,

J.O. & Macauley, L.S. (1970) The minimal

amino acid sequence of the insulin-

potentiating fragments of human growth

hormone. Diabetes 11, 216±222.


396 | J. Peptide Res. 56, 2000 / 388±397

8. Mondon, C.E., Reaven, G.M., Ling, N., Lewis,

U.J. & Frigeri, L.G. (1988) Amino-terminal

peptide of growth hormone enhances insulin

action in normal rats. Endocrinology 123,

827±833.

9. Effendi, W. & Ng, F.M. (1993) Effect of an

anti-lipogenic fragment of human growth

hormone on glucose transport in rat

adipocytes. Biochem. Mol. Bio. Int. 31, 543±

552.

10. Ho, K.K., O'Sullivan, A.J. & Hoffman, D.M.

(1996) Metabolic actions of growth hormone

in man. Endocr J. 43 (Suppl.), S57±S63.

11. Behncken, S.N. & Waters, M.J. (1999)

Molecular recognition events involved in the

activation of growth hormone receptor by

growth hormone. J. Mol. Recog. 12, 355±362.

12. De Vos, A.M., Ultsch, M. & Kossiakoff, A.A.

(1992) Human growth hormone and

extracellular domain of its receptor: crystal

structure of the complex. Science 255, 306±

312.

13. Surowy, T.K., Bartholomew, R.M. &

VanderLaan, W.P. (1984) Antigenic sites of

human growth hormone and related

molecules detected by monoclonal antibodies

to human growth hormone. Mol. Immunol.

21, 345±352.

14. Gooley, P.R., Plaisted, S.M., Brems, D.N. &

MacKenzie, N.E. (1988) Determination of

local conformational stability in fragment

96±133 of bovine growth hormone by high-

resolution 1H NMR spectroscopy.


15. Roongta, V., Powers, R., Jones, C., Beakage,

M.J., Shields, J.E. & Gorenstein, D.G. (1989)

Solution conformation of a synthetic

fragment of human pituitary growth

hormone. Two-dimensional NMR of an

a-helical dimer. Biochemistry 28, 355±360.

16. Bornstein, J., Ng, F.M., Heng, D. & Wong,

K.P. (1983) Metabolic actions of pituitary

growth hormone ± inhibition of acetyl CoA

carboxylase by human growth hormone and a

carboxyl terminal part sequence acting

through a second messenger. Acta

Endocrinol. 103, 479±486.

17. Jois, D.S., Matthew, W.C., Soma, C. &

Siahaan, T. (1997) Conformational analysis of

cyclo(2,9)Ac-QCRSVEGSCG-OH from the

C-terminal loop of the human growth

hormone. J. Peptide Res. 49, 15±22.

18. King, D.S., Fields, C.G. & Fields, G.B. (1990)

A cleavage method which minimizes side

reactions following Fmoc solid phase peptide

synthesis. Int. J. Peptide Protein Res. 36, 255±

266.

19. Kamber, B., Hartmann, A., Eisler, K., Riniker,

B., Rink, H., Sieber, P. & Rittel, W. (1980) The

synthesis of cysteine peptides by iodine

oxidation of S-trityl-cysteine and S-

acetamidomethyl-cysteine peptides. Helv.

Chim. Acta 63, 899±915.

20. Marion, D. & Wuthrich, K. (1983)

Application of phase sensitive two-

dimensional correlated spectroscopy (COSY)

for measurements of 1H±1H spin±spin

coupling constants in proteins. Biochem.

Biophys. Res. Commun. 113, 967±997.

21. Rance, M.N., Sorenson, O.W., Bodenhausen,

G., Wagner, G., Ernst, R.R. & Wuthrich, K.

(1983) Improved spectra resolution in cosy1H NMR spectra of proteins via double

quantum ®ltering. Biochem. Biophys. Res.

Commun. 117, 479±485.

22. Greisinger, C., Sorenson, O.W. & Ernst, R.R.

(1987) Practical aspects of the E.COSY

technique, measurement of scalar spin±spin

coupling constants in peptides. J. Magn.

Reson. 75, 474±492.

23. Bax, A. & Davis, D.G. (1985) MLEV-17 based

two-dimensional homo-nuclear

magnetization transfer spectroscopy. J. Magn.

Reson. 65, 355±360.

24. Kumar, A., Ernst, R.R. & Wuthrich, K. (1980)

Studies of J-connectivities and selective1H±1H Overhauser effect in H2O solutions of

macromolecules by two-dimensional NMR

experiments. Biochem. Biophys. Res.

Commun. 96, 1156±1163.

25. Piotto, M., Saudek, V. & Sklenar, V. (1992)

Gradient-tailored excitation for single-

quantum NMR spectroscopy of aqueous

solutions. J. Biol. NMR 2, 661±665.

26. Bartels, C. & Xia, T. (1994) X-EASY manual ±

ETH automated spectroscopy for X Windows

Systems. Institute of Molecular Biology and

Biophysics, Honggerberg, Zurich.

27. Hind, M.G. & Norton, R.S. (1997) NMR

spectroscopy of peptides and proteins,

practical considerations. Mol Biotechnol. 7,

315±331.

28. Pallagy, P.K., Alewood, D., Alewood, P.F. &

Norton, S.R. (1997) Solution structure of

robustoxin, the lethal neurotoxin from the

funnel-web spider Atrax robustus. FEBS Lett.

419, 191±196.

29. Guntert, P., Mumenthaler, C. & Wuthrich, K.

(1997) Torsion angle dynamics for NMR

structure calculation with the new program

dyana. J. Mol. Biol. 273, 283±298.

30. Brunger, A.T. (1992) X-PLOR, Version 3.1. A

system for X-ray crystallography and NMR.

Yale University, New Haven, CT.

31. Nilges, M., Grorenborn, A.M., Brunger, A.T ,

Clore, G.M. (1998) Determination of three

dimensional structures of proteins by

simulated annealing with interproton

distance restraints. Application to crambin,

potato carcarboxypeptidase inhibitor and

barley serine proteinase inhibitor 2. Protein

Eng. 2, 27±38.

32. Clore, G.M., Brunger, A.T., Karplus, M. &

Grorenborn, A.M. (1986) Application of

molecular dynamics with interproton

distance restraints to three-dimensional

protein structure determination. A model

study of crambin. J. Mol. Biol. 191, 523±551.

33. Brooks, B.R., Bruccoleri, R.E., Olafson, B.D.,

States, D.J., Swaminathan, S. & Karplus, M.

(1983) charmm: a program for

macromolecular energy minimization and

dynamics calculations. J. Comput. Chem. 4,

187±217.

34. Hutchinson, E.G. & Thornton, J.M. (1996)

promotif ± a program to identify and analyze

structural motifs in proteins. Protein Sci. 5,

212±220.

35. Laskowski, R.A., Rullmann, J.A.C.,

MacArthur, M.W., Kaptein, R. & Thornton,

J.M. (1996) aqua and procheck-nmr:

programs for checking the quality of protein

structures solved by NMR. J. Biolmol. NMR

8, 477±486.

36. Jungus, R.L. (1968) Fatty acid synthesis in

adipose tissues incubated in titrated water.


37. Grundy, S.M. (1999) Hypertriglyceridemia,

insulin resistance, and the metabolic

syndrome. Am. J. Cardiol. 222, 421±426.

38. Wuthrich, K. (1986). NMR of Proteins and

Nucleic Acids. Wiley, New York.

39. Ultsch, M.H., Somers, W., Kossiakoff, A.A. &

De Vos, A.M. (1994) The crystal structure of

af®nity-matured human growth hormone at

2 AÊ resolution. J. Mol. Biol. 236, 286±299.

40. Wijaya, D. & Ng, F.M. (1993) Effect of an anti-

lipogenic fragment of human growth

hormone on glucose transport in rat

adipocytes. Biochem. Mol. Biol. Int. 31, 1403±

1407.


J. Peptide Res. 56, 2000 / 388±397 | 397

the conformational and biological analysis of a cyclic anti-obesity peptide from the c-terminal...

Documents