This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
doi: 10.1098/rsta.2006.1848, 2709-2719364 2006 Phil. Trans. R. Soc. A
This article cites 59 articles, 3 of which can be accessed free
Email alerting service herein the box at the top right-hand corner of the article or click Receive free email alerts when new articles cite this article - sign up
http://rsta.royalsocietypublishing.org/subscriptions go to: Phil. Trans. R. Soc. ATo subscribe to
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
Bioinspired supramolecular liquid crystals
BY VIRGIL PERCEC*
Roy & Diana Vagelos Laboratories, Department of Chemistry, Universityof Pennsylvania, Philadelphia, PA 19104-6323, USA
A brief account on the historical events leading to the discovery of self-assemblingdendrons that generate self-organizable supramolecular dendrimers, or supramolecularpolymers, and self-organizable dendronized polymers is provided. These building blockswere accessed by an accelerated design strategy that involves structural and retro-structural analysis of periodic and quasi-periodic assemblies. This design strategymediated the discovery of porous helical supramolecular structures that self-assembledfrom dendritic dipeptides. Helical porous columns are the closest mimics of biologicallyrelated structures, such as tobacco mosaic virus coat, porous transmembrane proteins,porous pathogens and antibiotics. It is expected that this concept will allow one toinvestigate the structural origin of functions in synthetic supramolecular materials.
Keywords: self-assembly; helical pores; biological and bioinspired liquid crystals
Thi
On
*pe
1. Early history of biological supramolecular liquid crystals andinspiration for concept development
A brief historical review of complex molecular and supramolecular liquidcrystal(s) (LC(s)) of biological and synthetic origin is available (Percec 1997) andthe present short account will focus on the events that are most relevant for thetopic to be presented here.
The fascination with biological supramolecular LCs started in 1857 whenMettenheimer reported the lyotropic behaviour of myelin (Percec 1997). Duringthe 58th General Discussion of the Faraday Society, organized on 24 and 25 April1933, Rinne (1933) discussed the biological supramolecular LC systems, such aslipids, myelin, sperm, muscles and nerves. At the same meeting, Bernal (1933)became interested in biological LCs, and soon after, in 1936, he reported thediscovery of the columnar hexagonal lyotropic phase of tobacco mosaic virus(TMV; Bawden et al. 1936; Bawden & Pirie 1937).
In the autumn of 1982, at a time around the date the Nobel Prize in Chemistrywas to be announced, Aaron Klug gave a lecture on the structural analysis ofbiological assemblies at Case Western Reserve University in Cleveland. I wasinvited to join him together with a few colleagues for the dinner after his lecture.Discussions during his lecture and at the dinner in the French restaurant focused
Phil. Trans. R. Soc. A (2006) 364, 2709–2719
doi:10.1098/rsta.2006.1848
Published online 21 August 2006
s account is dedicated to Sir Aaron Klug.
e contribution of 18 to a Discussion Meeting Issue ‘New directions in liquid crystals’.
Figure 1. The structure of the supposed monotropic biaxial nematic liquid crystal based on acombination of half-disc and rod-like fragments (Malthete et al. 1986) and its corresponding side-chain polymethacrylate (Percec & Heck 1989).
V. Percec2710
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
on the self-assembly of complex biological assemblies, including TMV. At thattime, I was mostly impressed by the dinner in the French restaurant. Soon after,part of his lecture was presented at the Nobel prize ceremony in the same year(Klug 1983). Mimicking TMV, the best understood biological supramolecularsystem, with non-biological building blocks seemed to me as an interesting newresearch challenge. TMV is self-assembling, self-nucleating and self-checking(Klug 1983, 1999). I expected that the same principles would apply to the self-assembly of simple synthetic mimics. However, for quite a number of years I hadno idea how to approach it.
During the late 1980s, Saupe came to my office with two publications: onereporting the first lyotropic biaxial nematic LC (Yu & Saupe 1980) and the otherdescribing what was thought to be the first thermotropic biaxial nematic LC(Malthete et al. 1986). The paper by Malthete reported a very narrow monotropicbiaxial nematic phase, shown, subsequently, to be a smectic C phase (Hughes et al.1997) generated fromamolecule based ona combinationof a half-disc and a rod thatwas designed according to a concept advanced by Chandrasekhar (1985). Saupepreferred an enantiotropic biaxial nematic phase for physical measurements. Hementioned that Ringsdorf advised him to talk to me about it. I assigned thegraduate student Jim Heck a short project to transform the monotropic biaxialnematic LC into an enantiotropic one by the synthesis of a side-chain liquidcrystalline polymer containing the combination of half-disc and rod-like mesogenelaborated by Malthete as side groups (Malthete et al. 1986; figure 1).
This simple experiment was expected to transform the monotropic biaxialnematic phase into an enantiotropic one by the so-called polymer effect (Percec &Keller 1990). This project was a failure. The corresponding polymer and manyadditional related structures synthesized by Jim Heck produced only a hexagonalcolumnar LC phase and, therefore, I considered that a failure should not besubmitted to prestigious journals (Percec & Heck 1989, 1990, 1991a–e, 1992;Percec et al. 1992a,b, 1993a,b). However, my decision was incorrect since thisfailure was, in fact, a discovery that led to self-organizable dendronized polymers(Percec & Heck 1989, 1990, 1991a–e, 1992; Percec et al. 1991, 1992a–c, 1993a,1998a,b, 2005e) and self-assembling dendrons that produce supramolecularpolymers (Percec & Heck 1991d, 1992; Percec et al. 1993a,b, 1996, 2001, 2002,2004b). Additional support received from the work of Lehn on so-called ‘tubular’mesophases (columnar phases of macrocyclics; Lehn et al. 1985; Lehn 1988)encouraged the inspiration for a synthetic approach to the self-assembly of helicalrod-like and icosahedral virus-like supramolecular architectures (Percec & Heck1992; Percec et al. 1992a,b, 1994a,b, 1998a, 2004b).
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
2. An accelerated design strategy to libraries of self-assembling dendrons
Figure 2 outlines the accelerated rational design strategy to libraries of quasi-equivalent self-assembling dendrons elaborated in our laboratory by using theretrostructural analysis of their periodic and quasi-periodic assemblies (Percecet al. 1996, 1998a,b, 2001, 2004b, 2006a).
The structural and retrostructural analysis involves combinations oftechniques that consist of small- and wide-angle X-ray diffraction (XRD)experiments (performed on powder, oriented fibres and large monodomains ofLCs), transmission electron microscopy, electron diffraction, electron densitymaps and reconstruction of the XRD data. The cubic Pm3n (Balagurusamy et al.1997; Hudson et al. 1997; Percec et al. 1998a) and Im3m (Yeardley et al. 2000),the tetragonal P42/mnm (Ungar et al. 2003), the liquid quasicrystals (Zeng et al.2004), the tubular hexagonal (Percec et al. 2004b) and the tubular rectangular(Peterca et al. 2006) LCs were discovered during these investigations. Librariesof self-assembling dendritic building blocks based on a diversity of dendronarchitectures were elaborated via this accelerated design strategy (Percec &Kawasumi 1992; Percec et al. 1994a, 1995, 2001, 2004a,b, 2005a, 2006a). Morecomplex LC lattices, such as superlattices, were discovered with the aid of twin-dendritic benzamides (Percec et al. 1999, 2003) and semi-fluorinated Janus-dendritic benzamides (Percec et al. 2005a; figure 3).
The ultimate goal of this work is to assess the structural origins of functions bydesigning the three-dimensional tertiary and quaternary structure that isrequired to provide the function. This idea was inspired by Klug’s work thatelucidated in the field of biological macromolecular complexes the mechanism viawhich ‘structure determines the function’ (Klug 1983, 1999). So far, there are twoexperiments that indicated the potential to design functions through three-dimensional supramolecular assemblies. The first one refers to electronicfunctions (Percec et al. 2002) and will not be discussed here. The second oneinvolves the self-assembly of dendritic dipeptides into helical porous or tubularcolumns (Percec et al. 2004b) and will be summarized in §3.
3. Self-assembly of dendritic dipeptides into helical porous or tubularcolumns
In the absence of nucleic acid, the self-assembly of TMV viral coat provides, mostprobably, the best elucidated helical porous protein (Klug 1983, 1999). Manyother porous proteins act as pathogens (Gouaux 1998), transmembrane channels(Agre 2004; MacKinnon 2004) and antibiotics (Wallace 1986), and provide accessto stochastic sensing (Bayley & Cremer 2001). Therefore, the elaboration ofsynthetic porous protein mimics would be of interest for gene transfer therapy,channels, antimicrobials, separation processes, encapsulation and release andstochastic sensing. Ultimately, this may lead to the reconstruction of the cellmembrane and its functions. Recently, we have reported that amphiphilicdendritic dipeptides provide a general strategy to the self-assembly of helicalporous protein mimics (Percec et al. 2004b). The general self-assembly process isinvestigated in bulk state by the methods used in figure 2 and in solution by acombination of NMR, UV and circular dichroism experiments (Percec et al.
Figure 2. Self-assembly, structural and retrostructural analysis of periodic and quasi-periodic arrays of supramolecular dendrimers (Percec et al. 2001,2004a, 2006a).
Figure 4. Self-assembly of (4-3,4-3,5)nG2-CH2-Boc-L-Tyr-L-Ala-OMe, with nZ6–16, into helicalpores (Percec et al. 2004b).
hydrogenated
Nc
p6mm p6mm
2 × a1
2 × a1a2
a1 ≅ a2
p6mmp6mm
super-lattice
fluorinated Janus
Figure 3. Self-assembly and co-assembly of twin-dendritic benzamides (left-hand side and middle)and of semi-fluorinated Janus-dendritic benzamides (right-hand side) (Percec et al. 1999, 2003,2005a).
2713Bioinspired liquid crystals
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
2004b). Figure 4 outlines a dendritic dipeptide and its self-assembly into a helicalporous column that self-organizes into tubular hexagonal columnar LCs.Structural details of the supramolecular helical pores that self-assembled fromdendritic dipeptides are available in the references cited.
Various dendron architectures and a large diversity of dipeptides can beused to construct dendritic dipeptides that self-assemble into helical pores(Percec et al. 2004b, 2005b). However, the strength of this concept is in the fact
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
that a single combination of dipeptide and dendron structure can mediate theengineering of the pore structure, size and stability via the stereochemistry ofthe amino acids from the dipeptide (Percec et al. 2004b, 2005b,c), theprotective groups of the dipeptide (Percec et al. 2005d), the number of themethylene groups from the alkyl tails in the periphery of the dendrons (Percecet al. 2006b) and the connecting group between dendron and dipeptide. Thisself-assembly process follows the principles of self-assembly of proteins, since itinvolves cooperativity and allosteric regulation (Percec et al. 2005b,c, 2006b).However, the resemblance of the natural systems complicates the elucidation ofthe principles (Percec et al. 2006b) of this self-assembly process, since some ofthese principles are not yet understood even in biological complexes. Recentresults have shown that both circular and elliptical porous structures can beself-assembled from dendritic dipeptides (Peterca et al. 2006), and that porouscolumns can also be designed from dendronized polymers (Percec et al. 2005e).Last but not least, these porous protein mimics exhibit an intracolumnar three-dimensional order (Percec et al. 2005b–d, 2006b,c) that resembles thoseobserved for the first time by Bernal in the case of TMV (Bernal & Fankuchen1937). Also, these porous protein mimics have been shown to mediate thetranslocation of protons through impermeable vesicle walls (Percec et al.2004b). We believe that this accelerated design strategy provides primitiveexamples of biologically inspired supramolecular LCs and higher orderedstructures that upon further perfection will have the potential to address thestructural origins of functions in synthetic supramolecular materials. It couldultimately be that chemical self-assembly follows the same principles found byKlug (1983, 1999) for the case of TMV particle that is self-assembling, self-nucleating and self-checking.
Financial support by the National Science Foundation and the Office of Naval Research isgratefully acknowledged. The research described here would not have been possible without thededication of the collaborators mentioned as co-authors in the list of references. I am indebted toall of them. This short account is based on a lecture presented at the Royal Society on 5 December2005. I would also like to thank Prof. Duncan W. Bruce from the University of York, UK for theinvitation to give this lecture and for insisting that I should summarize it in this publication.
References
Agre, P. 2004 Membrane proteins: aquaporin water channels (Nobel Lecture). Angew. Chem. Int.Ed. 43, 4278–4290. (doi:10.1002/anie.200460804)
Balagurusamy, V. S. K., Ungar, G., Percec, V. & Johansson, G. 1997 Rational design of the firstspherical supramolecular dendrimers self-organized in a novel thermotropic cubic liquid-crystalline phase and the determination of their shape by X-ray analysis. J. Am. Chem. Soc.119, 1539–1555. (doi:10.1021/ja963295i)
Bawden, F. C. & Pirie, N. W. 1937 Liquid-crystalline preparations of cucumber viruses 3 and 4.Nature 139, 546–547.
Bawden, F. C., Pirie, N. W., Bernal, J. D. & Fankuchen, I. 1936 Liquid crystalline substances fromvirus infected plants. Nature 138, 1051–1052.
Bayley, H. & Cremer, P. S. 2001 Stochastic sensors inspired by biology. Nature 413, 226–230.(doi:10.1038/35093038)
Bernal, J. D. 1933 Referring to the late Geheimrat Rinne’s report from page 1016. Trans. FaradaySoc. 29, 1082–1083.
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
Discussion
A. KORNYSHEV (Imperial College London, UK ). Can you vary the radius of ananochannel?
V. PERCEC. Yes. Currently, we have been able to design supramolecularnanochannels with internal diameters ranging from as low as few angstroms upto 16 A. We also have an example of a nanochannel with an internal diameter of24 A. We do not know yet how to bridge between 16 and 24 A and above 24 A.
A. KORNYSHEV. In which way are your systems better than nanotubes?
V. PERCEC. Our supramolecular porous structures are complementary instructure, functions and in potential utility to nanotubes. Nanotubes aremolecular objects while our structures are supramolecular nanotubes that arecreated by self-assembly and therefore, depending on the structure of solvent,temperature and concentration, in solution they exhibit dynamic or stablestructures. In bulk, they are stable over a broad range of temperatures. Theinternal diameter of our pores can be designed as mentioned above and so is theirstructure. They can be made hydrophobic or hydrophilic and their length isdetermined by the kinetics of self-assembly. They can be generated as individualpores in vesicles and also processed as free-standing films or films on varioussubstrates. Under certain conditions, the resulting film could be similar to asingle crystal or a single crystal liquid crystal forming a large domain aligned in ahomeotropic or planar way. This arrangement can be mediated by the nature ofthe substrate (e.g. fig. 3e,f in Percec et al. (2004b)).
C. F. J. FAUL (School of Chemistry, University of Bristol, UK). In nature, the sizeof channels as well as the absolute size of the gates are controlled. You haveshown that you can control the size of the channels to some degree in your self-assembled structures. Can you also control the stack height/length of theassembled structures?
V. PERCEC. Yes, under certain conditions we can control the length of the porousassemblies. For example, when they are self-assembled in phospholipid bilayersor vesicles, their length seems to be determined by the bilayer length (e.g. fig. 5in Percec et al. (2004b) and its supporting information). We believe that thehydrophilic contour created by the dipeptide at the end of the channel (seestructure shown in fig. 4d of the previously mentioned publication) provides thedriving force required to span between the inner and outer water phases of thevesicle, while the alkane structure of the phospholipids provides the requiredsolvophobic solvent-like environment that mediates the self-assembly. In bulk,their length seems to be determined by the width of the film. Of course, this isthe case for a perfect monodomain. Alternatively, disclinations provide theexpected defects. However, the hexagonal columnar liquid crystal latticetransforms into a three-dimensional crystal that maintains the same hexagonalsymmetry. Under these conditions, the channel structure from the liquid crystalstate is templated and maintained in the crystal state (to be published).Recently, we have reported the first example of supramolecular pores that act asbeing gated by opening and closing the pore (Percec et al. 2005e) or by changingthe shape of the pore (Peterca et al. 2006).
on December 17, 2011rsta.royalsocietypublishing.orgDownloaded from
C. R. SAFINYA (Department of Materials, University of California at SantaBarbara, USA). Can you make a synthetic version of bacteriorhodopsin (a protonpump), which consists of both a hydrophobic and hydrophilic channel (thehydrophobic pocket forms a reservoir of ‘protons’ and are critical to the protonchannel operation)?
V. PERCEC. This is a very good question that addresses a very challengingsynthetic problem. In principle, our methodology combined with polymerizationtechniques (already demonstrated; e.g. Percec et al. 2005e; Peterca et al. 2006)can be used to make a synthetic mimic of bacteriorhodopsin. I can imaginestrategies to make the seven interconnected transmembrane channels containinghydrophobic segments for their parts that cross the membrane and hydrophilicparts that would both drive their folding and provide the mechanism to returnfrom the water region back in the membrane. My answer is yes. We can make thesynthetic mimic and, in fact, we have something related in progress in ourlaboratory. However, from here to the bacteriorhodopsin we may have some wayto go.
C. R. SAFINYA. Can you make ‘gated’ synthetic channels?
V. PERCEC. To date, we have synthesized a primitive ‘temperature-gated’channel (Percec et al. 2005e) and work is in progress to gating channels via otherexternally regulated techniques and synthetic strategies. For example, we haverecently reported (Peterca et al. 2006) mechanisms to change reversibly andirreversibly the shape of the channel, and this is another mechanism to a gate.