Original Paper
|
|
|
file on Synergy |
Acta Biochim Biophys
Sin 2008, 40: 133-139
doi:10.1111/j.1745-7270.2008.00379.x
Insulin analogs with B24 or
B25 phenylalanine replaced by biphenylalanine
Haijuan Du1,3,
Jiahao Shi2, Dafu Cui1, and Youshang Zhang1*
1 Institute of
Biochemistry and Cell Biology, Shanghai Institute for Biological Sciences,
Chinese Academy of Sciences, Shanghai 200031, China
2 School of
Life Science and Technology, Tongji University, Shanghai 200092, China
3 Graduate
School of Chinese Academy of Sciences, Shanghai 200031, China
Received: July 2,
2007
Accepted: October
9, 2007
*Corresponding
author: Tel, 86-21-54921237; Fax, 86-21-54921011; E-mail, [email protected]
B24 and B25
phenylalanines (Phe) play important roles in insulin structure and function.
Insulin analogs with B24 Phe or B25 Phe replaced by biphenylalanine (Bip) were
prepared by enzymatic semisynthesis. The biological activities were determined
by receptor binding assay and in vivo mouse convulsion assay. The
results showed that B25 Bip insulin has 139% receptor binding activity and 50% in
vivo biological activity, whereas B24 Bip insulin is inactive, when
compared with native insulin, suggesting that B24 Phe is crucial for insulin
activity. The structures in solution were studied by circular dichroism and
fluoremetry, and our results suggested that the insulin analogs with low
activities tend to be more tightly packed. The association properties were
studied by size exclusion chromatography. The Bip-amide replacement of B24 Phe
in deshexapeptide insulin or B25 Phe in despentapeptide insulin will cause the
monomeric B24 Phe-amide deshexapeptide insulin or B25 Phe-amide despentapeptide
insulin to associate and form dimers, whereas the mutations of B24 Phe in
insulin will make insulin dimers dissociate into insulin monomers.
Keywords insulin; biphenylalanine; circular dichroism; fluoremetry;
self-association
Insulin is a protein hormone produced in
pancreatic islet cells and stored in the form of hexamers composed of
zinc-coordinated dimers. After secreting into the bloodstream, insulin
oligomers are dissociated into monomers that bind with insulin receptors to
express the biological activities [1]. It has been known that B24 phenylalanine
(Phe) and B25 Phe are important not only in maintaining insulin conformation,
but also in expressing insulin activity. First, these two residues are highly
conserved [2]. Second, the crystal structure of insulin revealed that B24 Phe
and B25 Phe play important roles in the formation of dimers and hexamers [3].
Finally, the mutation of B25 Phe to Leu (“insulin Chicago”) and the
mutation of B24 Phe to Ser (“insulin Los Angeles”) reduced insulin
activity and caused diabetes [4].
It was found that B24 Phe mainly affected
the flexibility of its neighboring backbone, important in receptor binding of
insulin [5]. Nakagawa and Tager reported that insulin activity was almost
totally lost when B25 Phe was replaced by Ser, but when B25 Phe was replaced
by a much larger amino acid, naphthyl alanine, 50% of the receptor binding
activity and 66% of the in vitro biological activity were retained [6].
Quan et al. [7] reported the receptor binding activities of 19 insulin
analogs with B24 or B25 Phe replaced by natural or unnatural amino acids,
showing that insulin activity was highly dependent on the aromatic character of
these two residues and the position B24 was extremely restrictive to structural
modification, whereas B25 was extremely permissive.
Desoctapeptide insulin (DOI, insulin with
B23–B30 removed) and desheptapeptide insulin
(insulin with B24–B30 removed), both without
B24 Phe and B25 Phe, were inactive. Despentapeptide insulin (DPI, insulin with
B26–B30 removed), containing both B24 Phe and
B25 Phe, were active [8]. Deshexapeptide insulin (DHI, insulin with B25–B30 removed), containing only B24 Phe
obtained in our laboratory by enzymatic semisynthesis, was found to be still
active, although a little less active than DPI [9]. In addition, B25 Phe-amide
DPI with the C-terminal B25 Phe replaced by Phe-amide has higher receptor
binding activity than DPI [10].
Biphenylalanine (Bip) is a highly
hydrophobic unnatural amino acid with intrinsic fluorescence. Here, we report
the enzymatic semisyntheses of analogs of insulin with B24 or B25 Phe replaced
by Bip, and analogs of DHI and DPI with B24 Phe and B25 Phe replaced by Bip-amide.
The influence of Bip replacement on the structure and function of insulin was
studied by circular dichroism (CD), fluoremetry, in vitro receptor
binding assay, and in vivo biological assay.
Materials and Methods
Materials
Zn-free insulin was prepared from
crystalline porcine insulin purchased from Nova Biomedical (Waltham, USA) as
previously described [11]. tosylphenylalanine
chloromethyl ketone (TPCK)-trypsin and trifluoroacetic acid (TFA) were from
Sigma-Aldrich (St. Louis, USA). 1,4-butanediol was from Tokyo Chemical Industry
(Tokyo, Japan). Fmoc-amino acids, H-Thr(tBu)-2-Chlorotrityl resin and rink
amide MBHA resin were from GL Biochem (Shanghai, China). Other reagents were of
analytical grade. Sephadex G25 fine, diethylaminoethyl-Sephadex A25, and
Superdex G75 HR 10/30 columns were from Amersham Pharmacia Biotech (Uppsala,
Sweden). Human placental membrane was a gift from Prof. Youmin Feng at the
Institute of Biochemistry and Cell Biology, Shanghai Institutes for Biological
Sciences, Chinese Academy of Sciences (Shanghai, China).
Synthesis of oligo peptides
Peptides were synthesized using the Fmoc
strategy on an ABI-433A peptide synthesizer (Applied Biosystems, Foster City,
USA). The octapeptides, GlyBipPheTyr(tBu)Thr(tBu)ProLys(Boc)Thr(tBu) and
GlyPheBipTyr(tBu)Thr(tBu)ProLys(Boc)Thr(tBu) were cleaved from the
H-Thr-(tBu)-2-Chlorotrityl resin by anhydrous TFA/dichloromethane (v/v=1:99). The products were washed with water, dissolved
in 0.1% TFA/90% acetonitrile, and lyophilized. The dipeptide GlyBip-amide and
tripeptide GlyPheBip-amide were cleaved from the rink amide mBHA resin with 97% TFA. The cleaved
peptides were dissolved in water and lyophilized. Purity of all peptides was checked
by HPLC using a Waters Series 600 (Waters, Milford, USA) and mass spectroscopy
using an API2000 Q-trap mass spectrometer (Applied Biosystems).
Enzymatic semisynthesis of
insulin analogs
The preparation of DOI and the
semisyntheses of the insulin analogs were carried out as described previously
[10]. The synthetic peptide (60 mM) and DOI (6 mM) were dissolved in a
solution containing 30% dimethylformamide and 60% 1,4-butanediol with pH
adjusted to 7.0 with Tris. TPCK-trypsin (enzyme/DOI ratio 1:10 by weight) was
added and the reaction mixture was incubated at 37 ?C overnight. The crude product precipitated by acetone
at 4 ?C was purified by HPLC and lyophilized.
Insulin analogs with protected side chains were treated with anhydrous TFA for
1 h to remove the protecting groups and purified by HPLC.
In vivo biological activity assay
The in vivo biological activity was
measured by the mouse convulsion test according to the Chinese Pharmacopoeia
[12]. Briefly, ICR mice (18–20 g,
purchased from the Sino-British Sippr-BK Experimental Animal Ltd. (Shanghai,
China) were fasted overnight. From 1 mg/ml stock solutions, samples of
different dosages were prepared by serial dilutions in saline. For each dosage,
10 ICR mice were injected with sample (0.2 ml/mouse) and put into a 35 ?C chamber. The convulsion responses of mice
were observed and recorded. The activity was calculated by the ratio of the insulin
dosage with the insulin analog dosage just producing a convulsion rate over
50%.
The in vivo biological activity was
measured by the mouse convulsion test according to the Chinese Pharmacopoeia
[12]. Briefly, ICR mice (18–20 g,
purchased from the Sino-British Sippr-BK Experimental Animal Ltd. (Shanghai,
China) were fasted overnight. From 1 mg/ml stock solutions, samples of
different dosages were prepared by serial dilutions in saline. For each dosage,
10 ICR mice were injected with sample (0.2 ml/mouse) and put into a 35 ?C chamber. The convulsion responses of mice
were observed and recorded. The activity was calculated by the ratio of the insulin
dosage with the insulin analog dosage just producing a convulsion rate over
50%.
In vitro receptor binding assay
Receptor binding assay was carried out
using human placental membrane as described previously [13]. The membrane
containing approximately 0.2 mg protein was incubated at 4 ?C overnight with 125I-labeled insulin (approximately 105 cpm) plus a selected amount of insulin or
analogs in 0.6 ml of 50 mM Tris-HCl buffer (pH 7.5) containing 1% bovine serum
albumin. The unbound 125I-labeled insulin
was removed by centrifugation and the precipitate was washed with the same
buffer pre-cooled on ice. The radioactivity in the precipitate was counted. The
receptor binding activities of the insulin analogs were expressed as the ratio
of the insulin IC50 with the insulin
analogs IC50. Each determination was carried out in
duplicate.
Fluoremetry
Fluorescence spectra from 290 to 400 nm
were recorded on a Hitachi model F-2500 FL spectrophotometer (Hitachi
Instruments, San Jose, USA), using a 4 ml cell with 1 cm path length, with the
exciting wavelength of 264 nm. Samples were dissolved in 10 mM phosphate-buffered
saline (PBS), pH 7.4, to final concentrations of 17 mM. Anisotropies (r) were calculated from the
maximal emission intensity I and the instrumental correction factor G
according to the following equation (Equation 1):
Eq. 1
where G=IHV/IHH, IHV is the vertical emission intensity when excited with
a horizontally polarized light, IHH is the horizontal emission intensity when excited
with a horizontally polarized light, IVV is the vertical emission intensity when excited with
a vertically polarized light, and IVH is the horizontal emission intensity when excited
with a vertically polarized light source. Data were expressed as averages of
three scans [14].
CD analysis
Samples were dissolved in PBS (pH 7.4) and measured
in a Jasco-715 spectropolarimeter (Jasco, Tokyo, Japan) at room temperature
[15]. The far-ultraviolet (UV; 250–190
nm) spectra were recorded at a concentration of 0.2 mg/ml and cell path of 0.1
cm. The near-UV (300–245 nm) CD spectra
were recorded at a concentration of 0.5 mg/ml and cell path of 1.0 cm.
Size exclusion chromatography
Superdex G75 column (HR 10/30) was used for
the size exclusion chromatography [16]. Sample (0.04 ml) with different insulin
concentrations (38 mM, 75 mM, 150 mM, and 300 mM
in PBS, pH 7.4) was loaded onto the column and the column was eluted at room
temperature with PBS (pH 7.4) at a flow rate of 0.5 ml/min. The absorbance at
230 nm was monitored.
Results
Characterization of insulin
analogs
Insulin analogs prepared by enzymatic
semisynthesis and purified by HPLC were analyzed by mass spectrometry (Table
1) and pH 8.3 native polyacrylamide gel electrophoresis (Fig. 1).
The molecular masses of insulin analogs are consistent with their theoretical
values. In native polyacrylamide gel electrophoresis, each insulin analog
shows a single band.
Receptor binding activity and in
vivo biological activity
The receptor binding activities and in
vivo biological activities are shown in Table 2 and Fig. 2.
B25 Bip insulin has 139% receptor binding activity and 50% in vivo
biological activity when compared with native insulin, whereas B24 Bip insulin
has almost no receptor binding activity or in vivo biological activity.
That is, B25 Phe but not B24 Phe could be replaced by the unnatural amino acid
Bip. In our earlier studies, the coupling of B23 Gly and B24 Phe to inactive
DOI could convert it into active DHI [9]. Furthermore, in insulin from
different species, B24 Phe is highly conserved. These results indicate that B24
Phe is indispensable for insulin activity, whereas B25 Phe only helps to
reinforce the insulin activity. In DPI and DHI analogs with Bip-amide
replacement, B25 Bip-amide DPI showed 50% in vivo biological activity,
similar to B25 Bip insulin, whereas B24 Bip-amide DHI still retained 20% in
vivo biological activity.
Fluorescence analysis
Bip has strong fluorescence, so these
insulin analogs can be studied by fluoremetry. The results are shown in Table
3 and Fig. 3. Compared to B24 Bip insulin and B24 Bip-amide DHI, the
fluorescence intensity of B25 Bip insulin and B25 Bip-amide DPI is larger,
indicating that B25 Bip is more exposed than B24 Bip. This is consistent with
the crystal structure of native insulin [3]. Compared to B24 Bip insulin, the r
value of B24 Bip-amide DHI is smaller, indicating that B24 Bip is a little less
tightly packed with the removal of B25–B30.
Compared to B25 Bip insulin, the r value of B25 Bip-amide DPI is larger,
indicating that B25 Bip is more tightly packed with the removal of B26 to B30
[14].
CD analysis
B25 Bip insulin and insulin have similar CD
spectra with troughs at 222 nm and 208 nm, respectively (Fig. 4),
indicating that they have similar conformations in solution. The CD spectrum of
B24 Bip insulin also has double troughs but with increased absolute value of
the mean residue molar ellipticity |q|
at 222 nm and 208 nm, indicating that B24 Bip insulin has a more compact
structure [Fig. 4(A)] [17,18]. The CD spectra of B24 Bip-amide DHI and
B24 Phe-amide DHI are similar [Fig. 4(B)], whereas |q|222 nm and |q|208 nm of B25 Bip-amide DPI are larger than those of B25 Phe-amide DPI,
indicating that B25 Bip-amide DPI has a more compact structure [Fig. 4(C)].
Fig. 4(D) is the near-UV CD spectra of insulin and insulin analogs with
B24 Bip or B25 Bip. Their trough values show the following differences:
|q|B24 Bip insulin<|q|insulin<|q|B25 Bip insulin
indicating that the association tendency of
B24 Bip insulin is lower, but that of B25 Bip insulin is higher, than that of
insulin [19–22].
Association behavior analyzed
by size exclusion chromatography
The association behaviors of insulin
analogs studied by size exclusion chromatography are shown in Fig. 5.
From the profiles of size exclusion chromatography, it can be seen that the
Bip-amide replacement of B24 Phe in DHI or B25 Phe in DPI will cause the
monomeric B24 Phe-amide DHI or B25 Phe-amide DPI to associate and form dimers,
whereas the replacement of B24 Phe in insulin will make insulin dimers
dissociate into insulin monomers. The dissociation of insulin by Bip
replacement at B24 is consistent with an earlier report that B24 Ala insulin
is monomeric [23], and also consistent with the CD spectrum of B24 Bip insulin
that has the smallest |q| in the
near-UV region [Fig. 4(D)]. B25 Bip-amide DPI tends to form dimers,
whereas B25 Phe-amide DPI is monomeric at the same concentration, indicating that
the increased hydrophobicity of B25 Bip facilitates the dimer formation.
Discussion
B25 Bip insulin is highly fluorescent and
its structure in solution is similar to insulin as shown by their CD spectra.
The replacement of B25 Phe of insulin by Bip increases its receptor binding
activity and retains 50% in vivo biological activity. Bip at B25 is a
good probe in fluorescent analysis of insulin [24], and we expect Bip might be
a useful probe for other proteins as well. The fluorescence intensities of
insulin analogs with B25 Phe replaced by Bip are higher in comparison with
insulin analogs with B24 Phe replaced by Bip, indicating that B25 Bip is more
exposed than B24 Bip (Fig. 3). The CD spectra of B25 Bip insulin and
insulin are similar, indicating that they have similar structures in solution.
The solution structures of B24 Bip-amide DHI and B24 Phe-amide DHI are also
similar, as shown by their CD spectra (Fig. 4). The negative values of
peaks at 208 nm and 222 nm of B24 Bip insulin and B25 Bip-amide DPI,
respectively, are higher than those of insulin and B25 Phe-amide DPI,
indicating that they have more compact structures [17,18]. The structures of
Bip-replaced insulin analogs in solution are consistent with their receptor
binding activities (Table 2). The fluorescence of B24 Bip insulin is
the lowest, as shown in Fig. 3. Its structure is more compact than that
of insulin, as shown in its CD spectrum, so its buried Bip can not bind with
insulin receptor, resulting in its very low binding activity. The solution
structure of B24 Bip-amide DHI is similar to that of B24 Phe-amide DHI. Its
fluorescence intensity is higher than that of B24 Bip insulin but not as high
as B25 Bip-replaced analogs, indicating that its Bip is more exposed but not
enough to show significant receptor binding. From our results, we think in
insulin and Bip-replaced analogs, the exposure of the aromatic side chain is
very important for receptor binding. B24 Bip-amide DHI has 0.74% binding activity
and 20% in vivo biological activity. This difference was also observed
in DHI, indicating that higher biological activity can be expressed with lower
receptor binding because of the presence of spare receptors [25].
Acknowledgements
We are grateful to Ms. Xiaoxia Shao (Institute of Protein Research, Tongji
University, Shanghai, China) for
determining the molecular weight by mass spectroscopy.
References
1 Blundell TL, Dodson GG, Hodgkin DC, Mercola D.
Insulin: The structure in the crystal and its reflection in chemistry and
biology. Adv Protein Chem 1972, 26: 279–402
2 Xu B, Hu SQ, Chu QC, Huang K, Nakagawa SH,
Whittaker J, Katsoyannis PG et al. Diabetes-associated mutations in
insulin: Consecutive residues in the B chain contact distinct domains of the
insulin receptor. Biochemistry 2004, 43: 8356–8372
3 Baker EN, Blundell TL, Cutfield VF, Cutfield
SM, Dodson EJ, Dodson GG, Hodgkin DMC et al. The structure of 2Zn pig
insulin crystals at 1.5? resolution. Philos Trans R Soc Lond B Biol Sci 1988,
319: 369–456
4 Shoelson S, Haneda M, Blix P, Nanjo A, Sanke
T, Inouye K, Steiner D et al. Three mutant insulins in man. Nature 1983,
302: 540–543
5 Mirmira RG, Tager HS. Role of the phenylalanine
B24 side chain in directing insulin interaction with its receptor. J Biol Chem
1989, 264: 6349–6354
6 Nakagawa SH, Tager HS. Role of the
phenylalanine B25 side chain in directing insulin interaction with its
receptor. Steric and conformational effects. J Biol Chem 1986, 261: 7332–7341
7 Quan B, Smiley DL, Gelfanov VM, Di Marchi RD.
Site specific introduction of unnatural amino acid at sites critical to insulin
receptor recognition and biological activity. In: Blondelle SE ed.
Understanding Biology Using Peptides. Proceedings of the 19th American Peptide
Symposium 2006
8 Shvachkin YP, Shmeleva GA, Krivtsov VF,
Fedotov VP, Ivanova AI. Preparation and properties of
des-(pentapeptide-B26-30)-insulin. Biokhimiya 1972, 37: 966–973
9 Cao QP, Cui DF, Zhang YS. Enzymatic synthesis
of deshexapeptide insulin. Nature 1981, 292: 774–775
10 Zhu SQ, Cui DF, Fan L, Huang YD, Zhang YS.
Despentapeptide (B26-30) insulin amide: An insulin analogue with higher
receptor binding capacity than native insulin. Acta Biochim Biophys Sin 1988,
20: 292–297
11 Cui DF, Cui HR, Xu MH, Kou H, Zhu SQ.
Enzymatic syntheses of [B31-Lys-NH2]-insulin and its prolonged
effect on the reducing blood glucose. Acta Biochim Biophys Sin 1992, 24: 160–164
12 Qian XZ, Man Y, Zhou JH. Chinese
Pharmacopoeia. Beijing: Peoples Medical Publishing House 1985
13 Feng YM, Zhu JH, Zhang XT, Zhang YS. Studies
on the mechanism of insulin action VIII. Acta Biochim Biophys Sin 1982, 14: 137–143
14 Pittman I 4th, Nakagawa SH, Tager HS, Steiner
DF. Maintenance of the B-chain beta-turn in [GlyB24] insulin mutants: a steady-state fluorescence anisotropy
study. Biochemistry 1997, 36: 3430–3437
15 Guo ZY, Wang S, Tang YH, Feng YM. Mutagenesis of
the three conserved valine residues: Consequence on the foldability of insulin.
Biochim Biophys Acta 2004, 1699: 103–109
16 Ding JG, Cui DF, Zhang YS. Monomeric B27 Lys
destripeptide insulin: semisynthesis,
characterization and biological activity. Acta Biochim Biophys Sin 2002, 35:
215–218
17 Naeem A, Fatima S, and Khan RH.
Characterization of partially folded intermediates of papain in presence of
cationic, anionic, and nonionic detergents at low pH. Biopolymers 2006, 83: 1–10
18 Greenfield N, Vijayanathan V, Thomas TJ, Gallo
MA, Thomas T. Increase in the stability and helical content of estrogen
receptor alpha in the presence of the estrogen response element: analysis by circular dichroism
spectroscopy. Biochemistry 2001, 40: 6646–6652
19 Morris JWS, Mercola D, Aquilla ER. An
analysis of the near ultraviolet circular dichroism of insulin. Biochim Biophys
Acta 1968, 160: 145–150
20 Goldman J, Carpenter FH. Zinc binding,
circular dichroism, and equilibrium sedimentation studies on insulin (bovine)
and several of its derivatives. Biochemistry 1974, 13: 4566–4574
21 Wood SP, Blundell TL, Wollmer A, Lazarus NR,
Neville RW. The relation of conformation and association of insulin to receptor
binding; x-ray and circular-dichroism studies on bovine and hystricomorph
insulins. Eur J Biochem 1975, 55: 531–542
22 Strickland EH, Mercola D.
Near-ultraviolet tyrosyl circular dichroism of pig insulin monomers, dimers and
hexamers. Dipole-dipole coupling calculations in the monopole approximation.
Biochemistry 1976, 15: 3875–3884
23 Chen H, Shi M, Guo ZY, Tang YH, Qiao ZS, Liang
ZH, Feng YM. Four new monomeric insulins obtained by alanine scanning the
dimer-forming surface of the insulin molecule. Protein Eng 2000, 13: 779–782
24 Lakowicz JR. Principles of Fluorescence Spectroscopy.
2nd ed. New York: Plenum Press 1999
25 Cui DF, Cao QP, Zhu SQ, Zhang XT, Zhang YS. Enzymatic synthesis of
deshexapeptide insulin and its analog. Sci Sin (B) 1983, 26: 8147–153