Self-assembly of α-helices to form rare two-dimensional square P4mm symmetry via silica mineralization

Article abstract of DOI:10.1002/chem.201302627

Space oddity: The synthesis of a polypeptide-silica complex with a rare two-dimensional square P4mm structure is reported. The electrostatic "zipping" of the interacting amino and carboxylate groups along the α-helices and the diagonal formation of the si

Full text of DOI:10.1002/chem.201302627

COMMUNICATION
DOI: 10.1002/chem.201302627
Self-Assembly of a-Helices to Form Rare Two-Dimensional Square P4mm
Symmetry via Silica Mineralization
Yuan Yao, Dong Wang, Lu Han, and Shunai Che*[a]
Proteins are one of the most important types of biomacro-
molecules in living organisms and play specific and irre-
placeable roles in human metabolism, especially during bio-
chemical processes. In addition to their specific amino acid
sequence proteins have a sophisticated hierarchical structure
and spatial conformation.[1] The a-helix, which is a central
type of secondary structure in proteins, forms the corner-
stone of many more complex three-dimensional architec-
tures in proteins, and dominates numerous biological activi-
ties both in vivo and in vitro.[2] Because they have a rigid,
rod-like morphology and a unique conformational transfor-
mation property,[3] a-helices have been well studied and
broadly used as building blocks in the design of molecular
assemblies, generating a variety of functional materials with
notable activities, such as micelles, vesicles, hydrogels and
organic–inorganic hybrids.[4–9] These functional materials
have been applied for controlled compound release, diagno-
sis, biological simulation, hard tissue engineering and double
emulsions.[6,10–17]
Generally, highly concentrated solutions of polypeptides
form liquid crystalline phases when they adopt a-helical
conformations.[18–20] Some uncharged, rod-like polypeptides
can form cholesteric liquid crystalline phases in a diverse
range of solutions.[21,22] Most of the liquid crystalline phases
of polypeptides are two-dimensional columnar hexagonal
mesophases because they are more thermodynamically
stable in this arrangement. Recently, bioinspired mineraliza-
tion has provided a novel and facile approach to the study
of the packing behavior of polypeptides.[8,9,23,24] The secon-
dary structure of polypeptides can be utilized to control the
oxide pore architectures while the polypeptides retain their
solution conformation in nanocomposites.[25,26] Deming and
packed a-helices, and the silicification was restricted to the
surface of the preformed hexagonal PLL crystals, leading to
a layer of amorphous silica. The biomineralization of the
polypeptides replicated the PLL crystals rather than the
liquid crystal structure of a-helices. Consequently, the a-
helices in these biomineralization products adopted an ordi-
nary P6mm symmetry without diverging from the reported
two-dimensional columnar hexagonal mesophase of poly-
peptides. The self-assembly behavior might be intriguing if
the mineralization occurs at the a-helix level.
Herein, PLL was chosen as the template for the a-helix,
the average diameter and length of which were estimated to
be about 17.5 ꢁ and about 37 nm under the ideal condition,
respectively. Because PLL contains numerous ionizable
amino groups on its side chains, the carboxyethylsilanetriol
sodium salt (CES) was selected as the initiator of a-helix
and the co-structure directing agent (CSDA). At pH 9.0–
10.0, the carboxyl groups of the CES (pKa ~4.5) may interact
with the amino groups (pKa ~10.5) on the side chain of PLL,
decreasing the electrostatic repulsion between the side
chains by neutralizing the positive charges through the
charge cooperative effect and stabilizing the a-helix confor-
mation.[3] Additionally, the silane sites on CES can co-con-
dense with tetraethyl orthosilicate (TEOS) to form a silica
framework between the tightly packed a-helices and fasten
the microcosmic structure of the polypeptide–silica complex
(PSC). This novel CSDA route not only offers a new charge
balance strategy to stabilize the a-helix conformation[3,27,28]
and self-assemble a-helices to highly ordered mesostruc-
tured PSC,[29,30] but enables a facile approach for the study
of columnar liquid crystal phases with electron microscopy.
The X-ray diffraction (XRD) pattern of the PSC exhibits
two well-resolved peaks in the range of 2q=58–7.58 (2q=
co-workers have found that polyACTHNUGRTNEUNG(l-lysine) (PLL) and its de-
rivativesꢀ main chains adopted helical conformations in the
presence of monosilicic acid and multivalent counterions,
which later self-assembled into hexagonal silica plate-
lets.[27,28] However, no silica was found between the tightly
5.058, d1 =17.5 ꢁ and 2q=7.158, d2 =12.4 ꢁ; Figure 1a). The
d-spacing ratio of d1/d2 equals to 2, which suggests the ex-
p
istence of a very rare 2D square lattice and the two peaks
can be indexed to 10 and 11 reflections with a unit-cell pa-
rameter of a=17.5 ꢁ. However, the XRD data are insuffi-
cient to confirm the structures individually; there are three
possible plane groups including P4, P4mm, and P4gm. Addi-
tional studies were necessary to confirm the mesostructure
of PSC. The solid-state diffuse-reflectance circular dichroism
(DRCD) spectra of the PSC platelets (Figure 1b) revealed
one obvious negative peak at 225 nm and another weak neg-
ative peak at 205 nm. Because silica does not display effec-
tive absorbance in the ultraviolet range, the peaks must
[a] Dr. Y. Yao,+ D. Wang,+ Dr. L. Han, Prof. S. Che
School of Chemistry and Chemical Engineering
State Key Laboratory of Metal Matrix Composites
Shanghai Jiao Tong University, 800 Dongchuan Road
Shanghai, 200240 (P.R. China)
E-mail: chesa@sjtu.edu.cn
[+] These authors contributed equally to this work.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.201302627.
Chem. Eur. J. 2013, 19, 15489 – 15492
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
15489

Figure 1. a) XRD pattern, and b) DRCD spectra of the PSC prepared at
the optimal molar ratio of PLL/CES/TEOS/H2O=1:1:10:14500.
belong to the a-helices of PLL, the signals of which typically
appear at 208 and 222 nm. The deviation might be caused
by the incomplete helical conformation and a small quantity
of random-coil content.[31]
Scanning electron microscopy (SEM) images revealed
a uniform morphology for PSC: round platelets from 100 to
600 nm with average diameter of about 500 nm (Figure 2a
and b). The thickness of the platelets is about 50 nm, which
is larger than the ideal average length of the PLL template
(ca. 37 nm), due to the incomplete helical conformation and
mismatched arrangement of the PLL strands. The platelets
grew to the longest length of peptide because the longitudi-
nal growth of empty porous silica began with the shorter
peptide and then subsequently induced a lengthening along
the peptide rod axis even without PLL (Figure S2 in the
Supporting Information). Furthermore, the additional thick-
ness (2–5 nm) was caused by the deposition of amorphous
silica.[32,33]
High-resolution
transmission
electron
microscopy
Figure 2. a,b) SEM, and c,d) HRTEM images taken from the top (c) and
side (d) of the PSC round platelet. e) The electrostatic-potential map
clearly shows that most of the silica (high electron density, corresponding
to the dark blue area) locate in the diagonal position of the helices (low
electron density, corresponding to the earthy yellow area).
(HRTEM) images and the corresponding Fourier diffracto-
grams (Figure 2c and d, inset) illustrate the structure of
PSC. The P4mm symmetry can be identified with certainty
with the highest order of rotation of fourfold and the axes
of reflection inclined toward each other by 458. The unit-cell
parameter calculated from the TEM image is 16.8 ꢁ, which
approximately equals to the XRD results (17.5 ꢁ) and the
diameter of PLL in an a-helical conformation (about
17.2 ꢁ, assuming the side chains are nearly straight lines).
According to the results of XRD, DRCD, SEM and TEM,
the fundamental unit of the PSC with ordered P4mm sym-
metry should be an a-helix rather than any other secondary
structure. Furthermore, the interaxial separation between a-
helices of this P4mm structure is very small so that almost
no silica walls are formed in its two adjacent a-helices (Fig-
ure S3 in the Supporting Information). An electrostatic-po-
tential map (Figure 2e) of the structure was constructed by
taking the inverse Fourier summation of the crystal structure
factors[34] (Table S1 in the Supporting Information). It also
shows that an extremely thin silica wall (or almost no wall)
was formed between the two nearest PLL strands.
198 nm and a weak, broad positive signal at 220 nm caused
by the charge repulsion between the side chains (Figure 3,
curve a, pH~6.5). After adding CES, two strong, negative
peaks appeared at 208 and 222 nm (Figure 3, curve b, pH~
10.4), indicating that the conformation of PLL transformed
from random coil to a-helix, because the formation of ion
pairs reduced the charge density on the side chains and fa-
cilitated the formation of the a-helix. After the addition of
TEOS with 8 min of stirring, the intensity of the two nega-
tive peaks declined (Figure 3, curve c, pH~9.1) because the
TEOS and CES hydrolyzed, which generated silicic acid and
decreased the pH value subsequently influencing the a-heli-
cal conformation. After 24 min of stirring, the CD intensity
rebounded noticeably (Figure 3, curve d, pH~9.5) because
the polycondensation rate of silicic acid became higher than
the rate of silica hydrolysis in the source, resulting in an ad-
ditional increase in pH value and the a-helix ratio of the
PLL chains. After 32 min, the solution became opaque and
the CD intensity was strongly influenced by the reflection of
the suspended PLL–CSDA–silica complex. Therefore, the
The CD spectra were utilized to study the conformational
behavior during the formation of the PSC. The original PLL
hydrobromide salt solution displayed typical random-coil
conformation characteristics with a strong negative peak at
15490
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15489 – 15492

a-Helices Self-Assembly
COMMUNICATION
teraction energy and minimizing the excluded volume,[36]
which has not been observed in polypeptide liquid crystal
phases before. It is reasonable to speculate the formation of
P4mm symmetry based on the small interaxial separation
and the azimuthal orientation for a-helix in close interac-
tion. The PLL chains experience a transient conformation
conversion from random coil to a-helix when CES was
added, whereas a helical arrangement of the anionic carbox-
yl groups form in the grooves of the e-amino side
arms;[30,37,38] this arrangement stabilizes the a-helices by neu-
tralizing the charges on the PLL side chains, which is similar
to the phosphate anion–silicic acid pairs reported previous-
ly[27] (Figure 4a and b). In this situation, the PLL and CES
molecules cannot be treated as homogeneously charged rods
because they have a relatively small interaxial separation.[39]
A small a-helix-to-a-helix interaxial separation would be
obtained by the formation of an electrostatic “zipper” (Fig-
ure 4e) with interaction between the positively charged
strands (amino groups) and negatively charged grooves (car-
boxylate groups) of opposing molecules. An ideal “zipper”
requires the mutual azimuthal orientation 08. This angle can
be obtained in geometrical calculations based on the typical
helical structure (Figure S5a in the Supporting Information).
This result agrees with the valuable calculations of Korny-
shev and co-workers.[40] They revealed that the helices will
always choose the azimuthal orientation minimizing the in-
teraction energy, in which the optimal azimuthal orientation
should be all the same.
Figure 3. CD spectra of the solution at different experimental steps and
reaction times (reagents were added following the optimal molar ratio of
PLL/CES/TEOS/H2O=1:1:10:14500): a) original PLL hydrobromide so-
lution; b) CES added; c) TEOS added with stirring for 8 min; d) stirred
for 24 min.
silica framework could maintain the primitive mesostructure
even when the conformation of the PLL molecules was par-
tially changed. TEOS was not a unique prerequisite for the
formation of a-helix conformation in this approach, but was
indispensable to promote the self-assembly and fixation of
the silica framework.
The XRD patterns of the lyophilized PSC products at dif-
ferent intervals show that the P4mm structure was formed
at the very start with a gradually enhanced unit-cell parame-
ter (Figure S3 in the Supporting Information). The amount
of CES was also varied and PSC platelets with P4mm sym-
metry appeared when CES was added to the solution in an
approximate 1:1 molar ratio of PLL/CES, which was similar
ꢀ
to the situation caused by PO4 concentration in a previous
From the interaxial separation of 16.8–17.5 ꢁ of P4mm
structure of PSC, it can be determined that the interaxial
separation between PLL strands in diagonal positions is
23.8–24.7 ꢁ. The mutual orientation of the angles in the di-
report[35] (Figure S4 in the Supporting Information).
Square columnar symmetry is very rare[29,30] because long-
range oriental order is always formed by maximizing the in-
Figure 4. Illustration of the mechanism of formation for PSC with a 2D square structure. a) A typical a-helix conformation of PLL formed at a relatively
high pH with the amino-ended side chains maintaining the helical arrangement (see the outer helix). A plane perpendicular to the parallel axes of the a-
helix depicted as circles with azimuthal angle of f. The charges (shown in different colors) are inclined to the next one by 1008 in a clockwise direction
as they become shallower one by one in a typical pitch. b) Interaction between PLL and CES. The hydrophobic alkoxysilane groups from CES bind pref-
erentially to the grooves between the amino groups of PLL; CES anionic carboxyl groups emerge on the surface, forming distinct helical strands as con-
tinuous line charges similar to the amino groups on the PLL side chains. The positive and negative charges are depicted as red and green circles, respec-
tively. The negative charges locate in the middle position between two adjacent positive charges. c) A square lattice, in which PLL columns are framed
by silica wall. d) Optimal azimuthal orientations of the PLL molecules required for an ideal zipper (08). e) The surface charge patterns of two model
PLL molecules. The positively charged strands are close to the negatively charged grooves of the opposing molecule, forming an electrostatic zipper with
an optimal azimuthal orientation angle of 08. The PLL molecules were assumed completely helical, and the red arrows represent the azimuthal orienta-
tion with positive charge as standard.
Chem. Eur. J. 2013, 19, 15489 – 15492
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
15491

S. Che et al.
[3] H. Lu, J. Wang, Y. Bai, J. W. Lang, S. Liu, Y. Lin, J. Cheng, Nat.
Commun. 2011, 2, 206.
[4] T. Li, J. Lin, T. Chen, S. Zhang, Polymer 2006, 47, 4485–4489.
[5] E. G. Bellomo, M. D. Wyrsta, L. Pakstis, D. J. Pochan, T. J. Deming,
Nat. Mater. 2004, 3, 244–248.
[6] A. P. Nowak, V. Breedveld, L. Pakstis, B. Ozbas, D. J. Pine, D.
Pochan, T. J. Deming, Nature 2002, 417, 424–428.
[7] E. P. Holowka, D. J. Pochan, T. J. Deming, J. Am. Chem. Soc. 2005,
127, 12423–12428.
[8] M. B. Dickerson, K. H. Sandhage, R. R. Naik, Chem. Rev. 2008, 108,
4935.
[9] C.-L. Chen, N. L. Rosi, Angew. Chem. 2010, 122, 1968–1986;
Angew. Chem. Int. Ed. 2010, 49, 1924–1942.
[10] J. Lin, J. Zhu, T. Chen, S. Lin, C. Cai, L. Zhang, Y. Zhuang, X.-S.
Wang, Biomaterials 2009, 30, 108–117.
[11] C. Kojima, S. Tsumura, A. Harada, K. Kono, J. Am. Chem. Soc.
2009, 131, 6052–6053.
[12] J. W. Bryson, S. F. Betz, H. S. Lu, D. J. Suich, H. X. Zhou, K. T.
OꢀNeil, W. F. DeGrado, Science 1995, 270, 935–941.
[13] S. Zhang, Nat. Biotechnol. 2003, 21, 1171–1178.
[14] J. D. Hartgerink, E. Beniash, S. I. Stupp, Science 2001, 294, 1684–
1688.
[15] A. George, S. Ravindran, Nano Today 2010, 5, 254–266.
[16] A. Firth, A. Aggeli, J. L. Burke, X. Yang, J. Kirkham, Nanomedicine
2006, 1, 189–199.
agonal position would be negated by the larger distance and
by formation of the silica wall, which prevents their interac-
tion and leads to formation of the P4mm structure. All
other two-dimensional structures were unreasonable be-
cause close contact between identical charges causes strong
electrostatic repulsion.
In summary, the synthesis of PSC platelets was achieved
through a co-structure-directing route using PLL as the tem-
plate and CES as both the initiator of a-helix and CSDA.
Subsequent analyses confirmed that the PLL chains adopt
an a-helical conformation within a rare two-dimensional,
square P4mm symmetry within the PSC platelets. This
result, when combined with previous reports on DNA bio-
mineralization,[30,37] provides deep insights into the helical
conformation of biomacromolecules and enriches the details
of the phase diagram. This route may also be extended to
utilize other polypeptides with functional groups on the side
chains by using appropriate CSDAs and maintaining the or-
dered secondary structure of the polypeptides. These experi-
ments open up new methods for protein biomineralization
and analysis of the resultant crystal structures.
[17] J. A. Hanson, C. B. Chang, S. M. Graves, Z. Li, T. G. Mason, T. J.
Deming, Nature 2008, 455, 85–88.
[18] K. Czarniecka, E. T. Samulski, MCLC S&T Sect. B 1981, 63, 205–
214.
[19] J. Watanabe, Y. Takashina, Macromolecules 1991, 24, 3423–3426.
Experimental Section
[20] F. Livolant, Y. BouACTHNUGRTNENUGliACHUTGTNRENNUGgand, J. Phys. France 1986, 47, 1813–1827.
[21] T. Sato, J. Nakamura, A. Teramoto, M. M. Green, Macromolecules
1998, 31, 1398–1405.
[22] E. G. Bellomo, P. Davidson, M. Impꢄror-Clerc, T. J. Deming, J. Am.
Chem. Soc. 2004, 126, 9101–9105.
[23] S. Mann, G. A. Ozin, Nature 1996, 382, 313–318.
[24] H. Cçlfen, S. Mann, Angew. Chem. 2003, 115, 2452–2468; Angew.
Chem. Int. Ed. 2003, 42, 2350–2365.
[25] P. A. Mirau, J. L. Serres, M. Lyons, Chem. Mater. 2008, 20, 2218–
2223.
[26] K. M. Hawkins, S. S. S. Wang, D. M. Ford, D. F. Shantz, J. Am.
Chem. Soc. 2004, 126, 9112–9119.
[27] M. M. Tomczak, D. D. Glawe, L. F. Drummy, C. G. Lawrence, M. O.
Stone, C. C. Perry, D. J. Pochan, T. J. Deming, R. R. Naik, J. Am.
Chem. Soc. 2005, 127, 12577–12582.
[28] E. G. Bellomo, T. J. Deming, J. Am. Chem. Soc. 2006, 128, 2276–
2279.
In a typical synthesis, PLL (6 mg) was dissolved in deionized water (10 g)
while being stirred at room temperature for 20 min. Subsequently, CES
(30 mg; 25% in water) and TEOS (80 mg) were added to the solution
(molar ratios of PLL/CES/TEOS/H2O=1:1:10:14500), followed by stir-
ring at room temperature for 1 h. The mixture was allowed to react at
room temperature under static conditions for 3 days. The precipitate was
collected by centrifugation, washed with deionized water (2ꢃ) and the
final product was obtained by lyophilization at ꢀ548C.
Powder XRD patterns were recorded on a Rigaku D/Max 2000 powder
diffractometer equipped with a CuKa radiation source. SEM images of
each sample were taken with a JEOL JSM-7401F microscope at 1.0 kV.
HRTEM observations were performed with a JEOL JEM-2100 micro-
scope operating at 200 kV. CD spectra were collected with a JASCO J-
815 spectropolarimeter, and DRCD were taken with an extra DRCD ap-
paratus.
[29] B. Chen, X. Zeng, U. Baumeister, G. Ungar, C. Tschierske, Science
2005, 307, 96–99.
Acknowledgements
[30] C. Jin, L. Han, S. Che, Angew. Chem. 2009, 121, 9432–9436; Angew.
Chem. Int. Ed. 2009, 48, 9268–9272.
[31] T. K. Chowdary, E. E. Heldwein, J. Virol. 2010, 84, 4923–4935.
[32] H. Chen, L. Xia, W. Fu, Z. Yang, Z. Li, Chem. Commun. 2013, 49,
1300–1302.
[33] C. Fan, H. Qiu, J. Ruan, O. Terasaki, Y. Yan, Z. Wei, S. Che, Adv.
Funct. Mater. 2008, 18, 2699–2707.
[34] Y. Sakamoto, M. Kaneda, O. Terasaki, D. Y. Zhao, J. M. Kim, G.
Stucky, H. J. Shin, R. Ryoo, Nature 2000, 408, 449–453.
[35] L. Xia, Z. Li, Langmuir 2010, 26, 1116–1122.
[36] R. Hołyst, P. Oswald, Macromol. Theory Simul. 2001, 10, 1–16.
[37] L. Han, C. Jin, B. Liu, S. Che, Chem. Mater. 2012, 24, 504–511.
[38] S. Krimm, J. E. Mark, M. L. Tiffany, Biopolymers 1969, 8, 695–697.
[39] G. S. Manning, QRBio 1978, 11, 179–246.
This work was supported by the 973 project (2013CB934101), the Nation-
al Natural Science Foundation of China (21201120 and 21101106), the
Youth Natural Science Foundation of Shanghai (12ZR1445100) and
Evonik Industries. We thank the Instrumental Analysis Centre of Shang-
hai Jiaotong University for their collaboration on the DRCD measure-
ments.
Keywords: alpha-helix · electrostatic zipper · polypeptides ·
self-assembly · silica mineralization
[40] A. Kornyshev, S. Leikin, J. Chem. Phys. 1997, 107, 3656–3674.
[1] C. Branden, J. Tooze, Introduction to Protein Structure, Vol. 2, Gar-
land, New York, 1991.
[2] L. Stryer, Biochemistry, 4th ed., Freeman, New York, 1995, pp. 483–
Received: July 6, 2013
508.
Published online: October 15, 2013
15492
www.chemeurj.org
ꢂ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 15489 – 15492