Cover Illustrations (from top):Molecular graphics representation looking into the channel of the a hemolysin pore. Song!et al.,1996 (Figure 24).

Complexation of F-actin and cationic lipids leads to the hierarchical self-assembly of a network oftubules; shown here in cross-section. Wong 2000 (Figure 9).Depiction of an array of hybrid nanodevices powered by F1-ATPase. Soong et al., 2000 (Figure 1).

Electronmicrograph, after fixation, of neuron from the A cluster of the pedal ganglia in L. stagnalisimmobilized within a picket fence of polyimide after 3 days in culture on silicon chip. (Scale bar =20!mm.). Zeck and Fromherz 2001 (Figure 16).

Biomolecular Materials

Report of the January 13-15, 2002 WorkshopSupported by the Basic Energy SciencesAdvisory CommitteeU.S. Department of Energy

Co-Chairs:

Mark D. AlperMaterials Sciences Division

Lawrence Berkeley National Laboratoryand

Department of Molecular and Cell BiologyUniversity of California at Berkeley

Berkeley, CA 94720

Samuel I. StuppDepartment of Materials Science and Engineering

Department of ChemistryFeinberg School of Medicine

Northwestern UniversityEvanston, IL 60208

This work was supported by the Director, Office of Science, Office of Basic Energy Sciences, of the U.S.Department of Energy under contract DE-AC03-76SF00098

This report is dedicated to the memory ofIran L. Thomas, Deputy Director of theDepartment of Energy Office of BasicEnergy Sciences and Director of theDivision of Materials Sciences andEngineering until his death in February,2003. Iran was among the first torecognize the potential impact of thebiological sciences on the physicalsciences, organizing workshops andinitiating programs in that area as early as1988. More broadly, his leadership,vision, breadth of view, and dedication toscience had an immeasurable impact onthe science programs of the Departmentof Energy and the nation over the past 20years. He will also be remembered as alover of the arts and as a caring andgenerous friend.

Table of ContentsExecutive Summary........................................................................................................................................... iForeword ............................................................................................................................................................v1. Introduction ................................................................................................................................................... 12. What Does Biology Offer............................................................................................................................. 4

2.1 Introduction.................................................................................................................................... 42.1.1 Adaptation to the Environment........................................................................................ 42.1.2 Amplification of Signals..................................................................................................... 42.1.3 Atomic Level Control of Structure................................................................................... 42.1.4 Benign Processing. .............................................................................................................. 52.1.5 Color. ..................................................................................................................................... 52.1.6 Combinatorial Synthesis. ................................................................................................... 52.1.7 Computation. ....................................................................................................................... 52.1.8 Conformational Change..................................................................................................... 62.1.9 Control of Interfaces. .......................................................................................................... 62.1.10 Control of Polymer Properties........................................................................................ 62.1.11 Energy Conversion. .......................................................................................................... 82.1.12 Enzymes. ............................................................................................................................. 82.1.13 Evolution............................................................................................................................. 82.1.14 Extreme Environments..................................................................................................... 92.1.15 Hierarchical Construction................................................................................................ 92.1.16 Lightweight Materials. ..................................................................................................... 92.1.17 Lubricants. .......................................................................................................................... 92.1.18 Mass Production.............................................................................................................. 102.1.19 Materials Recycling......................................................................................................... 102.1.20 Membranes. ...................................................................................................................... 102.1.21 Model Materials for Studies of Basic Materials Physics........................................... 102.1.22 Molecular Recognition. .................................................................................................. 112.1.23 Motors, Rotors, Pumps, Transporters, Tractors, Springs, Switches, Ratchets. .... 112.1.24 Multi-Functional Materials............................................................................................ 122.1.25 Organic Synthesis............................................................................................................ 132.1.26 Optical Systems. .............................................................................................................. 132.1.27 Self-Assembly. ................................................................................................................. 132.1.28 Self Healing, Repair, Damage and Fault Resistance or Tolerance. ........................ 142.1.29 Signal Transduction. ....................................................................................................... 142.1.30 Smart Materials/Sensors. .............................................................................................. 142.1.31 Structural Materials. ....................................................................................................... 152.1.32 Systems.............................................................................................................................. 162.1.33 Template Directed Synthesis......................................................................................... 162.1.34 Transport Systems........................................................................................................... 17

3. Self-Assembled, Templated and Hierarchical Structures.................................................................... 183.1 Introduction.................................................................................................................................. 183.2 Organic Materials........................................................................................................................ 19

3.2.1 DNA-Based Materials....................................................................................................... 193.2.2 Polymer-Organic Hybrid Structures. ............................................................................ 213.2.3 Lipid-Protein-Nucleic Acid Hybrid Structures. .......................................................... 223.2.4 Biomolecular Materials in Microchannels. ................................................................... 23

3.3 Organic-Inorganic Hybrid Structures. .................................................................................... 233.3.1 Natural Mineralized Tissues. .......................................................................................... 233.3.2 Polymer-Directed Mineralized Composites................................................................. 243.3.3 DNA-Inorganic Hybrids. ................................................................................................. 253.3.4 Protein-Inorganic Hybrids............................................................................................... 27

3.4 Abiological Molecules Mimicking Biological Structures..................................................... 27

3.4.1 Colloidal Particles with a Valence. .................................................................................273.4.2 Hydrophilic/Hydrophobic Coatings on Inorganic Materials...................................28

4. The Living Cell in Hybrid Materials Systems........................................................................................294.1 Introduction..................................................................................................................................294.2 Metabolic Engineering................................................................................................................304.3 Engineering the Cell/Materials Interface. ..............................................................................324.4 Artificial Cells...............................................................................................................................354.5 Model System – Cell-Based Biosensors. ..................................................................................35

5. Biomolecular Functional Systems ............................................................................................................375.1 Introduction..................................................................................................................................375.2 Energy Transducing Membranes and Processes. ..................................................................385.3 Motors, Rotors, Ratchets, Switches. .........................................................................................395.4 Enzymes. .......................................................................................................................................425.5 Pores, Gates, Channels. ..............................................................................................................43

6. Promise and Challenges.............................................................................................................................45

References .........................................................................................................................................................47

3/20/03

i

Executive Summary

The Office of Basic Energy Sciences in theDepartment of Energy Office of Science, andthe Basic Energy Sciences AdvisoryCommittee convened a workshop in January,2002 to explore the potential impact of biologyon the physical sciences, in particular thematerials and chemical sciences.

Twenty-two scientists from around the nationand the world met to discuss the way that themolecules, structures, processes and conceptsof the biological world could be used ormimicked in designing novel materials,processes or devices of potential practicalsignificance. The emphasis was on basicresearch, although the long-term goal is, inaddition to increased knowledge, thedevelopment of applications to further themission of the Department of Energy.

The charge to the workshop was to identifythe most important and potentially fruitfulareas of research in the field of BiomolecularMaterials and to identify challenges that mustbe overcome to achieve success. This reportsummarizes the response of the workshopparticipants to this charge, and provides, byway of example, a description of progress thathas been made in selected areas of the field.

The participants agreed on severalconclusions. First and foremost, they agreedthat:

The world of biology offers anextraordinary source of molecules andinspiration for the development of newmaterials, devices and processes. Progressin research in a number of areas in thisfield has been rapid and the panel foreseesa revolutionary impact of the linkage ofbiology and materials science on scienceand technology in general, and the missionof the Department of Energy in particular.

In particular the panelists agreed that:

The interest of the Department of Energyin biomolecular materials and biologicalprocesses is very broad. There is a need forlighter and stronger materials to improvefuel economy. There is a need forfunctional materials to control transportacross membranes, to make separationsand purification processes more efficient.There is a need to increase energyefficiency by using low temperatureprocesses to make materials. There is aneed for energy producing processes thatcan convert light, carbon dioxide, andwater to high-density fuels and therebydecrease, at least to some extent, ourdependence on fossil fuels. Finally thehigh specificity of biological reactions,producing little or no side products, andthe inherent biodegradability ofbiological systems strongly suggest thatthese systems need to be explored by DOEfor their potential beneficial effects on theenvironment.

Having agreed on these principles, theparticipants stepped back to explore potentialresearch directions in the field. The world ofbiology is immense. As described in Section 2of this report, living organisms perform anextraordinary number of functions, virtuallyall of which can be seen to have relevance tomaterials, processes or devices. Some of theseimpacts have already been explored, at least tosome extent, most have not. At this stage anoutline of productive directions in the fieldcan be identified only through broad brushstrokes.

Specifically, the participants felt that a DOEprogram in this area should focus on thedevelopment of a greater understanding of theunderlying biology, and tools to manipulatebiological systems both in vitro and in vivorather than on the attempted identification ofnarrowly defined applications or devices. The

3/20/03

ii

field is too immature to be subject to arbitrarylimitations on research and the exclusion ofareas that could have great impact.

Future Directions.

These limitations aside, the group did respondto the charge and develop a series ofrecommendations. Three major areas ofresearch were identified as central to theexploitation of biology for the physicalsciences. Sections 3, 4 and 5 in this report aredevoted to those areas.

Self Assembled, Templated and HierarchicalStructures. Biology acts at the nanoscale,synthesizing and manipulating molecules withdimensions as small as tenths of nanometers.Through successive rounds of complexationand linkage of these molecules, it developsstructures on the meter length scale. All of thisis accomplished without conscious direction.Understanding and control of the processesinvolved in this self-fabrication are critical tothe successful exploitation of biology. This isdiscussed in Section 3.

The Living Cell in Hybrid Materials Systems.Despite the extraordinary advances in the pastdecade in our understanding of biologicalsystems, many remain far too complex for usto use, mimic or recreate. As a result, it mustbe expected that for well into the future, manyof the cellular functions we wish to exploit willhave to be performed by intact, living cellsthemselves. Thus, methods to incorporateliving cells or tissues into non-living structuresand devices, and to have them communicatewith those structures and devices will berequired. This area is discussed in Section 4.

Biomolecular Functional Systems. Livingsystems perform a wide variety of functionsthat could be controlled and used in vitro.Critical to this goal is the thoroughunderstanding of the molecular components ofthese systems and how they interact, leadingto our ability to manipulate those componentsand interactions. In some cases, intact cells (asfound in nature or altered by design) will need

to be used (see Section 4). However, in othercases, this will involve removal of theparticular functional system from theorganism. In still other cases it will involve therecreation or mimicking of it outside theorganism. This area is discussed in Section 5.

Workshop participants also discussed thechallenges and impediments that stand in theway of our attaining the goal of fullyexploiting biology in the physical sciences.Some are cultural, others are scientific andtechnical.

Barriers.

Cultural Challenges. Those who know thebiology, the biologists, are, more often thannot, descriptive scientists, whose goal is toidentify the molecular components ofbiological systems and understand how theywork together to produce the observedfunction. They are, in general, not focused onsynthesis or creation of these molecules orsystems, or mimics of them, nor are theyfocused on their adaptation to functionalsystems working outside the organism. Thisculture has changed somewhat in recent yearswith the focus on the molecular basis ofdisease and the identification of targets andthen lead compounds for pharmaceuticals.The number of biologists with an explicitinterest in the non-biomedical application oftheir systems however, remains small.

1. On the other hand, until recently,chemists, physicists and materialsscientists, who traditionally do have aninterest in creating materials, processesand devices, have had little formaltraining in the biological sciences. A verysophisticated understanding of a field isrequired to exploit it, thus trulyinterdisciplinary training needs to besignificantly enhanced. We are alreadyseeing this, with the organization ofdepartments and groups in “chemicalbiology” and the significant increase inthe enrollment of chemistry, physics andmaterials science students in biological

3/20/03

iii

science classes.

2. The application of biology to the physicalsciences is by definition amultidisciplinary activity requiringextensive collaboration. There have,however, historically been fewcollaborations between biologists andmaterials scientists, although there havebeen some with physicists and more withchemists.

Scientific and Technical Challenges.1. Biological systems are not generally

robust. They function best at roomtemperature, although some have beenfound in freezing or boilingenvironments. They are subject todeterioration in non-sterile environments.They generally require an aqueous milieu.Thus issues of the adaptation of biologicalsystems to the harsher environments ofmaterials, processes and devices, andtheir strengthening for long-term viabilitymust be addressed.

2. We do not, even after the revolutionaryadvances of the past few decades,understand biological systems wellenough to control and manipulate them.Basic research into the molecules,structures and processes is requiredbefore adaptation and mimicry can beachieved. Processes such as molecularrecognition, self-assembly, proteinfolding, energy transduction, nervoussystem function must be furtherelucidated.

3. Biological systems are, at their highestlevel of function, exceptionally complex,with large numbers of componentsinteracting in very specific ways. Manysystems are multifunctional and highlyresponsive to their environment. Issues ofsimplification or of precise assembly ofmulticomponent complex objects withoutsacrificing their function need to beaddressed

4. Theory, simulation and modeling havenot been applied to biological systems tothe extent that they have become routinein the materials sciences, physics andchemistry. This field must be developed.

5. Characterization tools, especially at thesingle molecule level need to bedeveloped. This is a particularlychallenging issue because the NationalInstitutes of Health, the primary federalagency for support of biological and bio-medical research has, in the past, notemphasized instrument development tothe extent that the DOE programs have.

Recommendations.

Program Relevance. In view of what hasrecently developed into a generally recognizedopinion that biology offers a rich source ofstructures, functions and inspiration for thedevelopment of novel materials, processes anddevices support for this research should be acomponent of the broad Office of Basic EnergySciences Program.

Broad Support. The field is in its early stagesand is not as well defined as other areas. Thus,although it is recommended that support befocused in the three areas identified in thisreport, it should be broadly applied. Goodideas in other areas proposed by investigatorswith good track records should be supportedas well. There should not be an emphasis on“picking winning applications” because it issimply too difficult to reliably identify them atthis time.

Support of the Underlying Biology. Basicresearch focused on understanding thebiological structures and processes in areasthat show potential for applicationssupporting the DOE mission should besupported.

Multidisplinary Teams. Research undertaken bymultidisciplinary teams across the spectrum ofmaterials science, physics, chemistry andbiology should be encouraged but not

3/20/03

iv

artificially arranged.

Training. Research that involves the training ofstudents and postdocs in multiple disciplines,preferably co-advised by two or more seniorinvestigators representing different relevantdisciplines, should be encouraged withoutsacrificing the students’ thorough studieswithin the individual disciplines.

Long-Term Investment. Returns, in terms offunctioning materials, processes or devicesshould not be expected in the very short term,

although it can reasonably be assumed thatapplications will, as they have already, ariseunexpectedly.

The workshop participants wish to thank andacknowledge Patricia Dehmer, Director of theOffice of Basic Energy Sciences; Iran Thomas,Director of the Division of Materials Sciences;and the Basic Energy Sciences AdvisoryCommittee for their vision in identifyingbiomolecular materials as an important newfield and in sponsoring this workshop.

3/20/03

v

Foreword

In 1999, the Basic Energy Sciences AdvisoryCommittee (BESAC) convened a workshop todesign a roadmap for research in complexsystems. The report of the workshop, ComplexSystems – Science for the 21st Century, outlined anexciting science agenda that both integrated thedisciplines of physics, materials sciences,chemistry, biology, and high-performancecomputing, and also could be built on thefoundations that had been put in place a yearbefore by theNationalNanotechnology Initiative.In June 2001,Dr. JamesDecker, ActingDirector of theOffice ofScience, U.S.Department ofEnergy, askedBESAC to helprefine thatresearchagenda. “Inthe worldbeyond nano,”Dr. Deckerwrote in hischarge letter tothe Chair ofBESAC, “itwill benecessary touse atoms,molecules, andnanoscalematerials as the building blocks for largersupramolecules and hierarchical assemblies. Aswas described in Complex Systems – Science forthe 21st Century, the promise is nanometer-scale(and larger) chemical factories, molecularpumps, and sensors. This has the potential toprovide new routes to high-performance

materials such as adhesives and composites,highly specific membrane and filtrationsystems, low-friction bearings, wear-resistantmaterials, high-strength lightweight materials,photosynthetic materials with built-in energystorage devices, and much more. Themagnitude of the challenge is perhaps moredaunting than any faced before by thesedisciplines. I would greatly appreciate BESAC’shelp in defining these challenges.”

BESACconsidered anumber ofworkshoptopics thatweresuggested bythis charge.One involvedtheexploration ofbiomolecularmaterials,materialsbased onbiologicalstructures andprinciples butwhose studyand useencompassesresearch at theinterfacesamong themanydisciplinesenumerated in

Dr. Decker’s charge. As a result of the rapidlyincreasing interest in research applying theprinciples and structures of biological systemsto the physical sciences, this BESAC workshopwas held in San Diego, California, January 13-15, 2002. Mark D. Alper of the LawrenceBerkeley National Laboratory and the

Table 1. Speakers

Mark Alper Lawrence Berkeley National Laboratory/University of California at Berkeley

Samuel Stupp Northwestern UniversityLia Addadi Weizmann Institute of Science

Paul Alivisatos University of California at Berkeley/Lawrence Berkeley National Laboratory

Hagan Bayley Texas A&M UniversityAngela Belcher University of Texas at Austin

Carolyn Bertozzi University of California at Berkeley/Lawrence Berkeley National Laboratory

Jean Fréchet University of California at Berkeley/ Lawrence Berkeley National Laboratory

Reza Ghadiri Scripps Research Institute

Wolfgang Knoll Max-Planck Institute for Polymer Research,Mainz

Chad Mirkin Northwestern UniversityCarlo Montemagno University of California at Los AngelesThomas Moore Arizona State UniversityDaniel Morse University of California at Santa BarbaraDavid Nelson Harvard UniversityCyrus Safinya University of California at Santa BarbaraPeter Schultz Scripps Research InstituteNadrian Seeman New York UniversityDouglas Smith University of California at San DiegoViola Vogel University of WashingtonUlrich Wiesner Cornell UniversityX. Sunney Xie Harvard University

3/20/03

vi

University of California at Berkeley and SamuelI. Stupp of Northwestern University were co-chairs. Twenty-two leaders (Table 1) in a widevariety of areas linking biology, physics,materials sciences, and chemistry were invitedto discuss progress in the field, definepromising future directions and identifybarriers to their pursuits. Thirty otherparticipants attended. The agenda for themeeting is shown in Table 2. This report of thepresentations and discussion at the workshopbegins with an introduction followed by a

discussion outlining the wide potential forresearch in the field, identifying molecules,structures and principles in biology that couldreasonably be applied to solving problemsimportant to the Department. This is followedby three sections discussing areas in this broadfield the workshop attendees felt were ofparticular interest at this time, and alsoamenable for productive research, given ourpresent knowledge of the underlying biologyand the tools and techniques existing for theirmanipulation.

Table 2. Agenda.Doubletree Golf Resort San Diego, 14455 Penasquitas Drive, San Diego, CA 92129Sunday, January 13, 20027:30 pm Speakers’ DinnerMonday, January 14, 2002

8:00 am Welcome Pat Dehmer, Iran ThomasDOE/BES

8:10 am Introduction Mark Alper

8:20 am Workshop Organization Samuel Stupp

Bio-Inorganic Systems — Angela Belcher , Chair

8:30 am Opportunities at the Biology/Materials Interface Paul Alivisatos

8:50 am Silicon Biotechnology: Proteins, Genes andBiomolecular Mechanisms Daniel Morse

9:10 am Control of Minerals by Organisms - Nanometers toMillimeters and More Lia Addadi

9:50 am Protein Control of Inorganic Materials Angela Belcher

10:10 am Towards a Tetravalent Chemistry of Colloids David Nelson

10:30 am Discussion

Biomimetics and Biomolecular Self Assembly — Sam Stupp, Chair

11:00 am Self-Assembly of Cell Cytoskeletal Proteins Samuel Stupp

11:20 amFunctional Materials Design, System Construction,and Computation. Adventures in Information Space &Complexity

Reza Ghadiri

11:40 am Supramolecular Assembly of Cell CytoskeletalProteins Cyrus Safinya

12:00 pm Working Lunch

1:00 pm DNA Nanotechnology Nadrian Seeman

1:20 pm Functional 2-3 Dimensional Bio-inorganicNanostructures Chad Mirkin

1:40 pm Discussion

3/20/03

vii

Biomolecular Functional Systems — Mark Alper, Chair

2:20 pm Signal Transduction and Active Transport at theNanoscale Viola Vogel

2:40 pm Dendridic Macromolecules and Bioinspired FunctionalNanoscale Assemblies Jean Fréchet

3:00 pm Engineered Protein Pores with Applications inBiotechnology Hagan Bayley

3:20 pm Providing Energy of Biomolecular Processes with anArtificial Photosynthetic Membrane Thomas Moore

3:40 pm Using Biology to Make New Materials Peter Schultz

4:00 pm Discussion

6:30 pm Working DinnerTuesday, January 15, 20028:30 am Discussion

Cell Engineering and Cells in Artificial Environments — C.!Bertozzi, Chair

9:00 am NanoEngineering Biotextiles Carlo Montemagno

9:20 am Probing Biochemical Reactions: From Single Moleculesto Single Cells Sunney Xie

9:40 am Supramolecular (Bio-) Functional InterfacialArchitectures Wolfgang Knoll

10:00 am Artificial and Biological Machines Carolyn Bertozzi

10:20 am Structure and Shape Control in Hybrid Materials Ulrich Wiesner

10:30 amManipulation and Visualization of SingleBiomolecules: Applications in Materials Science andBiophysics

Doug Smith

10:40 am Discussion

12:00 pm Working lunch

1:30 pm Individual Group Discussions

4:00 pm Group reports Chairs

6:00 pm Dinner

3/20/03

1

1. Introduction

Mankind has made use of biological materialsfor millennia. Through most of this time, theywere used as nature made them. Homes werebuilt with wood, straw, leaves; ropes werefashioned from vines; tools were shaped frombone, antler, horn; living yeast was used tocatalytically ferment alcohol or to leavenbread. More recently, mankind sought toextend his exploitation of nature by mimickingher principles, building, for example, bird-likewings to free him from the ground, andVelcro, reported to have been inspired by themechanism by which burred seed shells stickto a dog’s coat (Ball 1999).

For most of recorded time, however, nature’sliving systems were regarded as “special.” Itwas not until the 19th century that the principleof the “vital force” was finally set aside andthe concept of making biological moleculesand employing biological processes outsidethe living cell was demonstrated. Theextracellular synthesis of urea from cyanate byFriedrich Wöhler in 1828 demonstrated that“life” was not a requirement for the synthesisof molecules found naturally only in livingorganisms. [As Wöhler wrote to Berzelius, “Imust tell you that I can make urea without theuse of kidneys, either man or dog.”] Yearslater, in 1897, Buchner demonstrated that theentire 12 step/12 enzyme pathway convertingglucose to ethanol could proceed in extractsfrom yeast cells that had been killed andcompletely disrupted through grinding withsand.

The impact of these discoveries on materialsscience was immense – although not, to thisday, fully exploited. Nature, throughevolution – the extraordinary linkage ofnatural variation and selection – has, overbillions of years, learned to develop thousandsof extraordinarily sophisticated materials and

chemical processes that can serve us well inour search for the advanced materials requiredto meet our demand for improvements inproductivity, conservation, and safety. As inother fields, opportunities often lie untappeduntil the need and the tools to exploit themarise. In this area, biomedical applicationscame first driven by human healthconsiderations. But the time for applications tothe physical sciences is now here, and the pastfew years have seen a burgeoning of ourinterest in pursuing this exciting field ofendeavor.

Despite the great interest over the past decade,successful and widespread use of biology inmaterials science remains a formidablechallenge. The application of biologicalmaterials or of materials that mimic biologicalsystems lags far behind our enthusiasm forthem. Our ability to control chemical reactionswith nature’s exquisite sensitivity, to makepolymers with precise molecular weight, or toassemble macromolecules into large-scalestructures is at a very primitive, descriptivestage. We are even further from anunderstanding, much less the ability toimitate, the metabolic, catalytic, andregulatory processes that harness energy forvital processes and synthesize all vitalsubstances.

There is however reason to be optimistic thatwe will, in the not too distant future, come tounderstand the very complex physics andchemistry of biological processes. A revolutionhas taken place in biology over the past fewdecades. We now have a vastly increasedknowledge and understanding of thebiochemistry and molecular biology ofbiological materials and how their uniqueproperties arise from their structure. We nowhave a vastly increased arsenal of tools to

3/20/03

2

analyze, characterize and manipulate thesesystems and we now have theories and highlydeveloped simulations to guide and interpretexperiments. We are, as a result, developing avastly increased ability to modify biologicalmaterials and processes for our needs and tosynthesize, de novo, new materials that arebased on biological principles.

As Wöhler and Buchner demonstrated, thereare no mysterious vital forces governing thebehavior of biological systems. They are,instead, governed by the coulomb andchemical potentials that govern everything inthe universe. Quantum mechanics, Newton’slaws, and thermodynamics determine themotions of particles, mass transport, andenergy balance. Geometry influences howthings can be packed. The difficulty is that wedon’t yet understand how these relativelysimple forces can give rise to such complexphenomena. At the molecular level, we don’tfully understand the relationships amongstructures, properties, and functions. We don’tunderstand chemistry well enough to makethese complicated molecules easily. At thelevel of molecular assemblies and sub-cellularcomponents, we don’t understand how theyare organized and how they functioncollectively. The cellular level, with all of itsinteracting components, mass and energyflows, is beyond our ability to even describecompletely.

Section 2 outlines the awe inspiring array ofmolecules, structures, and processesdeveloped by living organisms and availablefor our use or modification. It is an impressivelist, providing, in effect, an “existence proof”of what can be done and challenging us toexploit it. It must be remembered, however,that even this impressive catalogue does notdescribe the upper limits. Despite their manyinteresting and useful properties, biologicalmaterials and processes evolved under severeconstraints that limited their development. Forone, nature does not optimize structures andprocesses – evolution stops when it has madestructures and processes that are “good

enough” for their specific, or narrowly definedpurpose and successfully adapt their hostorganism to its environment. Further, eachstructure or process is limited by the fact thatit must “co-exist” and interact with the otherstructures and processes on that organism. Ona more fundamental level, only a smallnumber of the 92 naturally occurring elementshave been used, and only small ranges of pH,temperature, and pressure have beenexplored. Often, the constraints do not preventour use of these materials and processes.Clearly wood is a ubiquitous structuralmaterial, and fermentation is a well-developedindustrial process. However, these processesare limited in their properties and applicationsof biomolecular materials. Wood cannotsubstitute for carbon fiber reinforcedcomposites in airplanes, and fermentation byitself will not produce absolute alcohol. Thereis a real possibility that, once we understandthe principles of nature’s construction, we willbe able to use these principles for our owndesign goals and “improve on nature.”

The interest of the Department of Energy inbiomolecular materials and biologicalprocesses is very broad. There is a need forlighter and stronger materials to improve fueleconomy. There is a need for functionalmaterials to control transport acrossmembranes, to make separations andpurification processes more efficient. There isa need to increase energy efficiency by usinglow temperature processes to make materials.There is a need for energy producingprocesses that can convert light, carbondioxide, and water to high-density fuels andthereby decrease, at least to some extent, ourdependence on fossil fuels. Finally the highspecificity of biological reactions, producinglittle or no side products, and the inherentbiodegradability of biological systems stronglysuggest that these systems need to be exploredby DOE for their potential beneficial effects onthe environment.

Following the cataloging of some of nature’sstructures and processes in Section 2, we

3/20/03

3

discuss, in Sections 3, 4, and 5, three broadareas of research that emerged at theworkshop as having significant discoverypotential because of the breadth of knowledgethat already exists in the underlying biology,because of the applicability of this knowledgeto materials research in the physical sciences,and because of the promise seen in thepreliminary research already begun. Thesethree areas are:

• self-assembled, templated, andhierarchical structures, both bio-inorganicand bio-organic,

• the living cell in hybrid materials systems,

• biomolecular functional systems.

Finally, it should be noted that this reportreflects the focus of the workshop on adiscussion of materials and processesdesigned for nonmedical applications,consistent with the mission of the EnergyDepartment as distinguished, for example,from the mission of National Institutes ofHealth (NIH).

3/20/03

4

2. What Does Biology Offer1

2.1 Introduction. Before addressing thespecific question of a proposed researchagenda for the Department of Energy, it isuseful to view the full breadth of the potentialimpact of the field of biology on materials andchemical applications.

Therefore, we look here at the humblingcatalogue of what organisms can do, andconsider how our imaginations might allow usto harness, adapt, and mimic these capabilitiesfor advanced materials or processes. This is,however, by no means a list of what can beaccomplished. Some or many of thesebiological processes may not be reproducibleoutside the living organism. Some or manythat are, may not be superior either insimplicity or effectiveness, to fullynonbiological solutions. Those that are, maytake years or decades to develop. On the otherhand, progress is surprisingly rapid. Speakingof his field of self-assembling electronicdevices, Fraser Stoddart of UCLA has beenquoted as saying “Something that I thoughtwould …[for]ever be a dream in my lifetimenow stands a good chance of becoming atechnological reality before this decade is out”(Ball 2001).

2.1.1 Adaptation to the Environment.Organisms sense their environments and altertheir properties to adapt to them. Shiftinghumans to high altitude results in thespontaneous increase in the production of themolecule 2,3 bis-phosphoglycerate. Thismetabolic product of glucose binds to theprotein hemoglobin in the blood, causing achange in its shape to decrease its affinity foroxygen. This decreased affinity allows thehemoglobin to deliver more oxygen to themuscles and brain at each cycle through theblood stream, a critical adaptation to the loweroxygen levels at high altitude. This effectprogresses over hours to days as we

acclimatize to the altitude. Other responses tothe environment, discussed below, can occurin less than a second (withdrawal from a hotsurface) or more than years (evolving lungsfor breathing air).

2.1.2 Amplification of Signals. Bloodclotting, gene expression, and the activation ofenzymes involved in the control of cellularenergy production require the amplification,by many orders of magnitude, of signalscarried by as few as one molecule, photon,electron, or ion. Amplification is a multi-steppathway. At each step one enzyme moleculeactivates a very large number of copies of thenext enzyme in the pathway, which, in turn,activates many more copies of the enzymecatalyzing the next step. This sequentialactivation/amplification continues until thefinal product is produced at sufficient levels toachieve its macroscopic function.

2.1.3 Atomic Level Control of Structure.One of the most striking capabilities of livingorganisms is their ability to produceextraordinarily complex molecules withvirtually error-free control of the selection andlocation of each individual atom. This is acritical ability because the structure, propertiesand function of molecules can dependsensitively on the type or position of a singleatom. Mirror images of the same molecule, forexample, can have drastically differentproperties. The converse also applies.Molecules can be designed at the atomic levelfor very specific structures and functionsthrough atomic level control of design. Thislevel of control has been a grand goal ofsynthetic chemists and materials scientists foryears. In fact the challenge extends beyond themolecular to the systems level. Deer antlers,for example, do not need to be shaped afterinitial synthesis. Our control of structure atthis level would allow “net shape

3/20/03

5

manufacturing,” in which the product isproduced in the shape required; thus, noexpensive and wasteful machining isnecessary.

2.1.4 Benign Processing. Biologicalprocesses are generally less hazardous andinvolve fewer toxic materials than theirsynthetic counterparts. For example, syntheticnanocrystals, which are of such great interestnow, are often synthesized at very hightemperatures with hazardous precursors.Organisms, on the other hand producemagnetic and semiconductor nanoparticles,often with great homogeneity, at roomtemperature and pressure. Teeth, shell andother ceramics are produced biologicallyunder far more benign conditions than aresynthetic ceramics.

2.1.5 Color. Certain birds (for example,peacocks) fish, snakes and butterflies appearcolorful, but without synthesizing the“traditional” light absorbing pigments.Instead, they produce overlapping scalesmade of carbohydrate that impart iridescenceby creating interference patterns. These givethe appearance of different colors dependingon the nanometer scale spacing, the thicknessof the layers, the angle of viewing, thewavelength of the light illuminating them, therefractive index of the liquid between thelayers. At near grazing angles, for example,only ultraviolet light is reflected, making thematerial virtually invisible. It should bepossible to develop materials using similarproperties that change their response to lightin the presence of applied electrical ormagnetic fields. Such materials could be used,for example, in “smart” windows that wouldreversibly reflect or transmit light, therebycontrolling the heat load in buildings.

2.1.6 Combinatorial Synthesis. Vertebratesproduce upwards of one hundred milliondifferent antibody molecules, each with aslightly different binding site. As a result, invirtually all cases, there are a few selectantibodies with a shape that allows them to

first bind to invading viruses or bacteria, andthen, after a few days time, to arrange for thesynthesis of many more copies of themselvesto overwhelm the invader and ward offdisease. In many cases, the current state oftheory and modeling is too primitive to allowthe prediction of the structure of a materialfrom the desired mix of properties. The ability,through combinatorial synthesis to developmillions of extremely similar, but criticallydifferent structures, and the ability to rapidlyscan them all for the desired properties, as isdone in the immune system, can potentiallysave enormous amounts of time and expensein the development of materials. A variety oftechniques can be employed. In phage display,variants of a given protein are presented onthe outer surface of bacterial viruses. Ininorganic “combi-chem”, ink jet printersdeposit nanoliter amounts of a variety ofcompounds in small wells that can beprocessed and analyzed in parallel. Batterymanufacturers have recently announced thattheir next generation cells will containmaterials discovered through combinatorialmethods. Combinatorial synthesis has becomea major focus in the search for better andcheaper catalysts for organic reactions.

2.1.7 Computation. The holy grail ofcomputer designers remains the developmentof a device that mimics the human brain.Nothing comes close to its ability to store andretrieve information, often from apparentlyunrelated “data entry” events. Its ability tofocus on a subset of the huge number ofsimultaneous stimuli and inputs it receives isalso unmatched, as is its ability to “reason” bycombining bits of information and weightingthem appropriately as it sums their input intothe solution of a problem. No device competesin use of spoken language. It is true thatsilicon computers are orders of magnitudefaster than the millisecond biological processesof the brain but none match the brain's abilityto process in parallel. It is interesting to notehowever, that the unit of biologicalinformation, the DNA or RNA nucleotide,occupies a volume of about one cubic

3/20/03

6

nanometer, and can be “accessed” for “readout” with exceptional specificity.

A number of biological materials are beingstudied for their ability to “compute”,although clearly, nothing approaching a“biological computer” has been developed.The protein bacteriorhodopsin has beenshown to have the capability for holographicdata storage. Adleman’s report in Science(Adleman 1994) sparked a great deal ofinterest for his use of DNA to solve complexmathematical problems. Shapiro and hiscolleagues at the Weizmann Institute have alsoreported in Nature (Benenson et al., 2001) onthe use of DNA and its restriction and ligationenzymes in a system that can do computation.Clearly an understanding of the method ofparallel processing and the mechanism of theself-assembly of the billions of neurons into afunctioning brain will be of great benefit, withmany, as yet undefinable, applications.

2.1.8 Conformational Change. Many cellularresponses to external stimuli are based on thefact that proteins can alter their structure, andtherefore their properties, in response tochanges in their environment. This response ismediated through the binding of one or moremolecules from that environment at a specificsite on the protein. This binding event cancause one part of a ~10nm protein to movemore than 1nm relative to another. Enzymeactivity is regulated in this fashion, with theprotein shifting from an active conformationto one that is less effective either in bindingthe substrate, performing the chemistry of thereaction, or both. The oxygen binding proteinhemoglobin also exhibits conformationalchanges. On a shorter time scale than that forthe increase in the concentration of 2,3 bis-phosphoglycerate (see Section 2.1.1),hemoglobin responds to a change in protonconcentration. In low proton environments, asin the lungs, it assumes a conformation thatenhances its ability to bind oxygen. In highproton environments, as in the muscles,however, these ions bind to hemoglobin andcause it to change shape, releasing oxygen for

its use in the metabolic processes that createenergy required for biosynthesis, musclecontraction and other functions.

2.1.9 Control of Interfaces. Organisms havelearned to control and exploit a wide varietyof interfaces with disparate materials. Inbiomineralized tissues such as bone, teeth,shell, inorganic materials of a variety ofcompositions are in direct contact with variousorganic materials. Proteins often direct thesynthesis of the inorganic phases. Theseprocesses serve as models that could, forexample, be applied to the controlled growthof mineralized phases for functional thin filmsor particulate applications (Klaus et al., 1999).Phage display and other such techniques havebeen used to identify proteins that bindspecifically to semiconductor and othernanocrystals (Whaley et al., 2000; Whaley andBelcher 2000; Lee et al., 2002; Seeman andBelcher 2002). The crystalline ordering of viralparticles could be used to direct the orderingof nanocrystals in arrays that might promotecollective behavior. Biocompatibility, aninherent interfacial property of biomolecularmaterials is also becoming increasinglyimportant, as we design hybrid devices thatexploit the many cellular functions we cannotyet reproduce in the absence of the living cell(Section 4). Whole cells have been attached tosurfaces in bioreactors for fermentations.Hybrid circuits with both semiconductor chipsand synaptically connected neurons have beenexplored. Nerve cells from the snail Lymnaeastagnalis have been immobilized, through non-specific linkages, on silicon chips usingpolyimides such that a voltage pulse on thechip could excite the neuron (Zeck andFromherz 2001). “Metabolic engineering” hasbeen employed to alter the surfaces of cells toimprove their binding to specific sites on non-living surfaces including metals, polymers, orceramics, while the cells maintain their naturalfunctions (Section 4.3).

2.1.10 Control of Polymer Properties. Theproperties of polymers depend on the typeand sequence of their constituent monomers.

3/20/03

7

Increased numbers of types of monomersallow an increased variety of structures,properties and functions. Most syntheticpolymers are made from a single type ofmonomeric unit – for example, styrene inpolystyrene, ethylene in polyethylene. Toachieve different properties, two or even threetypes of monomers can be incorporated into asingle polymer. Alternatively, a number ofdifferent polymers can be blended. Even thesesystems are at best rudimentary analogues ofnatural polymers. Biologically producednucleic acids, on the other hand, use fourprimary monomers (ATGC in DNA, AUGC inRNA), and a few others, in smaller amounts,that are methylated or acetylated. Proteinsgenerally use 20 primary amino acids asmonomers, some of which are modified afterinitial synthesis, for example to hydroxy-proline or gamma-carboxy-glutamate, to, ineffect, create a far larger library.Polysaccharides are even more complex.Although some are made of a single monomer,others draw from many, varying in size,stereochemistry, charge, and attachedfunctional groups, each affecting theproperties of the polymer in a different way.Cyclic polysaccharides have been used forexample to insulate photoluminescentpolymer chains (Cacialli et al 2002). Recentresearch has expanded the number ofmonomers beyond those nature uses in bothproteins and nucleic acids, allowing theincorporation of amino acids or nucleotideswith an almost unlimited variety of sizes orwith specific redox, optical, electrical,magnetic, and chemical properties. Equallyimportant to the materials properties ofpolymers of biological origin is the fact thatnucleic acids, proteins or many carbohydrates,are made to a precise, uniform length, givinggreater control of properties alignment andcrystallization. Naturally occurringbiopolymers do have a limited number ofdifferent backbone structures, but research hasprogressed in broadening this range, furtherincreasing the breadth of properties that canhe achieved. Techniques are being developedto use these materials in very imaginative

ways, for example, in a technology to “write”thin lines of proteins or nucleic acids on avariety of inorganic surfaces using the tip ofan atomic force microscope as a quill, and asolution of the polymer as the ink (Demers etal., 2002).

2.1.10.1 Nucleic Acids. DNA and RNA are ofcourse involved in the storage and use ofcellular information, and, in the case of RNA,in catalysis and as scaffolding insupramolecular structures. These materials arefound base-paired in double strands in thewell-known “double helix” form of DNA or insingle strands, which often, through the samemechanism, fold back on themselves and base-pair into precisely defined 3-dimensionalstructures. Their role in information storageand retrieval is an impressive one, with, asmentioned above, one bit of informationoccupying only one cubic nanometer, yet stillallowing rigorous specific addressability(regulation of specific gene expression) andaccuracy in reading. The DNA double helixcan in fact serve not only as an informationcarrier but also as a structural material(Section 3.2.1), exhibiting the properties of along, relatively rigid rod with a persistencelength of approximately 50nm. Capitalizing onthis, nanocrystals and complex organic groupswith electronic or optical properties have beenlinked to individual single strands of DNA ofdefined sequence. These single strands thenare base paired with other defined sequenceoligonucleotides to align those functionalgroups in precise positions in space. Some ofthese structures have demonstrated energytransfer from donor to acceptor groups,presenting intriguing possibilities forelectronic materials. DNA molecules withcomplex topologies have been formed intospecific objects, nanomechanical devices andperiodic arrays, with potential ultimateapplications in nanoelectronics andnanorobotics. For example, DNA base pairinghas been exploited in the fabrication ofrudimentary structures, demonstrating and,nand, or and nor gates. Other devices,involving DNA single strands that compete

3/20/03

8

with each other for binding complementarystrands demonstrate informationmanipulation (Seeman, 1999). DNA moleculeswith complex topologies have been formedinto specific objects, nanomechanical devicesand periodic arrays, with ultimate applicationsfor nanoelectronics and nanorobotics. Theprecise structures of folded single strands ofRNA, for example in transfer RNA, are criticalto the role of these molecules in geneticinformation storage and transfer, but couldalso be exploited in using them in materialsapplications. Opportunities for the use of thesematerials abound, especially when thedifficulties in producing large amounts ofnucleic acids are overcome.

2.1.10.2 Proteins. Proteins fold into precisethree-dimensional shapes, driven by theiramino acid sequence. The range of propertiesthat these polymers exhibit, or contribute to isenormous, including lubricity, adhesion,viscosity, stiffness, toughness, flexibility,optical clarity, and, in the case of, for example,aspartame, taste. A very large number ofproteins and their functions remainundiscovered or uncharacterized. Theemerging field of proteomics will, no doubt,result in the discovery and understanding ofmany of these, some performing novel, usefulfunctions and others helping us understandhow the structure of proteins determines theirfunction.

2.1.11 Energy Conversion. Chemical energypowers organisms without the use offlammable fuels or high temperatures.Biological molecular motors involve the directconversion of chemical energy into mechanicalenergy without the inefficient production ofheat seen in conventional motors and engines.Much of this process, and also the use ofenergy for biosynthesis of cellular materials,involves the recycling of “energy carrying”molecules between their [incorrectly] so-called“high energy” and “low energy” states. Inparticular, the molecule adenosinediphosphate can be “activated” in thepresence of metabolic energy through its

binding a third phosphate group to makeadenosine triphosphate (ATP). Removal ofthat group involves a significant negative freeenergy that can “drive” energy requiringprocesses. The photosynthetic sequence oflight energy capture, production of “highenergy” molecules, and the use of thosemolecules to produce metabolic andmechanical energy is one that, if duplicated,would revolutionize energy conversion.Chemical energy can also be stored as amolecular or ionic gradient across amembrane.

2.1.12 Enzymes. Enzymes accelerate reactionsby up to 13 orders of magnitude at roomtemperature and atmospheric pressure. Mostimportant is their exquisite specificity for bothstarting materials and products, and thus theirsynthesis of the desired molecules whileproducing no byproducts. The activity ofenzymes can be controlled over several ordersof magnitude through the binding of specificeffector molecules. Selective activation ofenzymes could allow control of complex,multi-component, specific chemicalconversions such as the synthesis of fuels fromcarbon dioxide and water using sunlight orother energy sources.

2.1.13 Evolution. Materials ideally suited toperform in the environments for which theywere originally designed are often poorlysuited to new environments that have arisensince their design. Living organisms evolveconstantly to allow them to survive and in factthrive in new environments. One of the mostfrequently quoted (although recentlychallenged) cases involves the change in colorof moths in England from white to blackduring the industrial revolution and back towhite again after institution of pollutioncontrols. The principle lies in what might becalled error prone manufacture. Each unit(organism) manufactured (born) has anaturally occurring slight variation (mutation)from the norm in one or more of its thousandsof characteristics. In some cases, this alterationleads to death or decreased fertility. In most

3/20/03

9

cases, these alterations are unnoticed innormal use (life), but when conditions change,those individuals with the particular mutationthat improves their survival, and hencefertility in that new environment take over thepopulation. In many cases, this process occursover several generations, each one slightlydifferent from its parents. Evolution has alsobeen employed in the laboratory in the“maturation” and optimization of antibodiesand other proteins where the continualproduction of variants and the selection of the“better” variants, has led to better functioningproducts (Yin 2001). Development of materialsthat themselves evolve is a very long termgoal, perhaps beyond our grasp. However, theuse of the principles of evolution to optimizematerials, as in the antibody work, is anexcellent example of biologically inspiredmaterials science.

2.1.14 Extreme Environments. In general,biological systems are regarded as impracticalfor many applications because of theirsensitivity to extremes of environment.Organisms have, however, adapted to live inthe below freezing waters of Antarctica or theocean depths, and in the near boilingconditions of hot springs or deep oceanthermal vents. Those living at the oceandepths are protected against exceptionallyhigh pressures, those living in salt flats areprotected from high osmotic pressure, cellsthat line the stomach are adapted to extremeacidity (pH 1) and others have been shown tobe exceptionally resistant to ionizing radiation.In each case, specific protective mechanismshave evolved. Transferring these capabilitiesto organisms or systems that are employed forspecific functions could protect them fromthese “biologically extreme” environments,although it must be kept in mind that what isregarded as extreme biologically is not at allextreme in conventional materials synthesis.

2.1.15 Hierarchical Construction. Biologicalstructures are extraordinarily complex, farmore so than synthetic systems, and theirsophisticated properties reflect that

complexity. However, to a great degree, theirsynthesis is far less complex, relying onsequential hierarchical construction principles.For example, the synthesis of collagen fibers,whose thickness can be measured inmillimeters, can be described as a sequence ofrelatively simple steps starting with theassociation of groups of atoms and graduallybuilding in complexity to generate advancedproperties. The groups of atoms in aminoacids are assembled in a linear polypeptidepolymer through the DNA-directed, proteinsynthesizing machinery. This single peptidechain then folds with two others of similarstructure into a triple helix collagen molecule.Collagen molecules than associate in aspontaneous but highly controlled process toproduce fibrils, which associate in a similarmanner to, eventually, produce the final,exceptionally strong collagen fibers. Each stepalong the path is programmed into thestructure of the material, allowing, throughkinetically and thermodynamically driven selfassembly, the development of great structuralcomplexity with minimal complexity ofdesign.

2.1.16 Lightweight Materials. Living systemsare inherently light-weight. They use, almostexclusively, carbon, hydrogen, nitrogen,oxygen, with lesser amounts of phosphorusand sulfur and very small amounts of others.They also produce low density hydrogels andrelated structures. Our ability to mimic livingsystems and develop lightweight structuraland functional materials would lead toenormous reductions in weight and fuel usein, for example, automobiles.

2.1.17 Lubricants. Enormous inefficiencies,loss of function and expense result frominadequate lubrication of the contact surfacesof moving parts. Living systems must solvethe same problems and have evolvedmolecules to lubricate joints, portions of theeye, and internal organ surfaces. These usuallyhighly charged molecules could serve as amodel for biomimetic lubrication.

3/20/03

10

2.1.18 Mass Production. Large scaleproduction of materials can often beexpensive. Organisms can, and in fact already,serve as factories. Their regulatorymechanisms also allow for the control of thelevels of each molecule made. Certain proteinscan be present in as few as 10 copies per cell.Alternatively, activators and “strongpromoters” can lead to very high levels ofproduction of defined products, often atamounts approaching tens of percent of totalcell volume. Genes for the naturally occurringplastic polyhydroxybutyrate have beentransferred into plants. Acres of farmlanddevoted to these transgenic plants could beinexpensively harvested and the polymerextracted. Genes for proteins are now beinginserted into goats or cows in a manner thatleads to their secretion into the easily collectedmilk, which, of course, can be “grown” for thecost of animal feed.

2.1.19 Materials Recycling. Biosynthesis isaccomplished almost exclusively by enzymecatalysis of chemical reactions that makebonds between small molecules to make largerones. These enzymes increase the rate ofreactions that would be otherwise far too slowto support life. Other enzymes, produced by,for example, soil bacteria, and widely presentin the environment, break these bonds, “bio”-degrading the molecules, also at rates farexceeding those of the uncatalyzed reaction.Energy input is, of course, required in onedirection, usually the synthetic direction, sincebonds are being made and entropy is beingdecreased. As a result, the degradativereactions usually proceed exergonically, andwith relatively low activation energies. Easilydegraded structures are of increasingimportance. Industries are being required tobe responsible for the entire product cycle. Themanufacturing process will have to includerecycling: cradle-to-cradle (new product),rather than cradle-to-grave responsibility willbecome the norm. The biological model couldserve well.

2.1.20 Membranes. Cellular membrane are

extraordinary multi-functional structures.They define the boundaries of cells and of sub-cellular organelles. Cell surface membraneshelp, through the use of embedded proteinsand carbohydrates, to identify the cell to theoutside world and receive signals from thatworld. They house transport systems, motors,rotors, energy transduction devices andexquisitely sensitive and selective sensors.They create non-polar compartments in themidst of a fully aqueous environment. Theyare self-healing, self-assembling, can grow asthe cell it surrounds grows, and can split intotwo as the cell divides. Membranes are highlyflexible and can adapt their shape to a varietyof structures and also to perturbations in thosestructures as the cells progress through thevarious stages of their lives or perform theirmyriad functions. They are also quite robust,despite the fact that the individual componentmolecules are not covalently linked to eachother. A great deal of effort has gone into themimicking of the cell membrane and muchsuccess has been achieved. Artificial, self-assembled monolayers serve in a wide varietyof efforts to study self-assembly or othermembrane associated properties. Membranemimics are made with artificial molecules,mirroring the self-assembling amphiphilicproperties of membrane lipids, butincorporating greater rigidity through crosslinking, or functionality through lightabsorbing chromophores, inserted channels(Section 5.6) or molecular recognition groups.Use of other molecular components allows formultilayered membranes with their own setsof properties.

2.1.21 Model Materials for Studies of BasicMaterials Physics. Basic physical laws suchas those of Newtonian mechanics, statisticalmechanics and quantum mechanics provideour only rigorous foundation forunderstanding the properties of materials.However, factors such as complexity,nonlinearity, and many-body interactions tendto frustrate our ability to make connectionbetween the microscopic physical laws andmacroscopic materials properties.

3/20/03

11

Fundamental experimental studies ofmaterials science must be aimed at elucidatingthe basic connections between physical lawsand materials properties; but, as a practicalmatter, an experimentalist has to pick exampleor “model” materials to study. Biologicalmaterials can act as model materials thatrevolutionize such studies since they haveproperties that can currently be controlled to afar greater extent than synthetic materials. Anexample of this is in the field of polymermaterials and polymer rheology. Half acentury of theoretical and experimentalresearch had been dedicated to understandingthe connection between microscopic molecularand macroscopic materials properties ofpolymeric fluids without having theexperimental ability to directly control orvisualize the molecular dynamics. The fieldwas revolutionized in the early 1990s by theintroduction of techniques for directlyvisualizing and manipulating single DNAmolecules using optical tweezers andfluorescence microscopy (Perkins et al., 1994).Use of DNA allows the preparation of apolymer solution in which every polymer inthe solution has exactly the same length, andstructure. Optical tweezers allow individualmolecules to be mechanically manipulatedand the forces acting on the individualmolecules to be directly measured.Fluorescence microscopy allows the molecularconformation of individual polymers to bedirectly visualized. For the first time thisallowed rigorous testing and refinement ofmolecular theories by direct comparison withmolecular measurements (Smith et al., 1999;Babcock et al., 2000).

2.1.22 Molecular Recognition. Manybiological macromolecules have theextraordinary ability to recognize specificother molecules in an environment containinglarge numbers of very similar structures. Thisselectivity is the basis for the functioning ofthe cell membrane as a biosensor, the self-assembly of collagen and other structures, thestructure and replication of DNA as a geneticor structural material, the specificity of

enzymes for reactants and products. Themechanism of this recognition involves severalfactors. The first is a geometric fit similar tothat between two pieces of a jigsaw puzzle.Individual atoms in a molecule are in suchprecisely defined positions that each moleculeassumes a defined shape, which iscomplementary to the shape of the molecule towhich it will bind. In addition, groups ofopposite charge line the surfaces of two boundmolecules. Other binding forces including theso-called hydrophobic interactions and polarinteractions (hydrogen bonding, salt bridges)are also found along the interface. Complexityis added by the fact that proteins are not fixedin shape but rather are quite flexible, rapidlyand spontaneously shifting among a set ofpossible conformations. Generally, only oneconformation of a given protein binds aparticular target molecule, and the binding ofthat molecule locks the protein into thatparticular configuration. This is a powerfultool in protein design and function but it doesadd complexity to the task of designingproteins for specific molecular recognitionfunctions.

2.1.23 Motors, Rotors, Pumps,Transporters, Tractors, Springs, Switches,Ratchets. Cellular function depends to a greatextent on a variety of motors, rotors andrelated devices. These transport systems usechemical energy to move ions,macromolecules, organelles, chromosomes,and even whole cells. A newly characterizedmotor packs DNA into the heads of viruses asthey are produced in cells. Using a motorgenerating 60 piconewtons of force to counteran internal pressure 10 times that of achampagne bottle (60 atm), the DNA ispackaged, perhaps to provide the “spring” torelease it into the next cell that is attacked(Smith 2001). The enzyme ATP synthase, andthe various components of the electrontransport chain, which are responsible for theproduction of most of the cellular ATP fromADP and phosphate, are themselves molecularpumps and motors. Energy released by theoxidation of nutrients is used by the electron

3/20/03

12

transport chain to pump protons across themembrane. These protons “flow” backthrough the synthase, causing it to rotate anddrive the synthesis of ATP. Thermodynamicsrequires that this process can be run in reverse.Thus, the release of energy accompanying theconversion of ATP to ADP and phosphatecauses the rotation of the protein and thepumping of protons out across the membrane.The synthase converts chemical to mechanicalenergy (or the reverse) with nearly 100%efficiency (Yasuda et al., 1998).

Another rotational motor imbedded in the cellmembrane drives the high-speed rotation offlagella, the tails of bacterial cells that enablethem to “swim” towards nutrients and awayfrom repellants. Linear motors such as myosin,kinesin, DNA polymerase also move withinthe cell using the direct conversion of chemicalenergy to mechanical energy. To a greatextent, the nature and structure of themolecules involved in these functions areknown. Some could replace micro-mechanicaldevices now made by lithographic techniques.For example, single steps in the motion of

these motors have been analyzed and resolvedto be on the order of 10 nanometers. Somemanipulation has also been achieved. Rigid

rods have been attached to the synthase andobserved in a microscope to rotate in a circle,driven by the rotation of the protein (Figure 1).Techniques, such as single moleculespectroscopy allow the study of individualmotors and rotors.

Biomolecular ratchets and springs store orrelease energy and rectify motion. The energyfor springs is provided by hydrolysis of anucleotide or binding of a ligand. Ratchets arepowered by Brownian motion in polymerizingfilaments. For example, the spasmoneme ofthe vorticellid contracts upon exposure tocalcium, by 40% of its length (2.3 mm) inmilliseconds, at a velocity approaching 8cm/sec. No energy source is required (Amos1971, Moriyama et al., 1999, Mahadevan andMatsudaira 2000). Small changes in proteinsubunits, amplified by linear arrangements inthe filaments can lead to structures that storeenergy and then release it on demand, creatingmovement. Other structures act as molecularswitches or “dimmers”. For example, enzymeactivity can be controlled on a continuum fromfull activity to complete inactivity by a varietyof effectors (Zhou et al., 2001). Ion channelscan be controlled over the full spectrum ofactivity by an equally diverse group of ionsand molecules. Some have speculated on theinterfacing of millions of these efficientbiological devices to produce “macro” levelsof power. Nature again shows the way,bundling actin and myosin molecules to makemuscles. Jiménez and colleagues (Jiménez2000) have mimicked the system by designinga molecular assembly in which two synthetic“filaments” mimicking actin and myosin slidealong each other to contract or stretch. Othershave done pioneering work to achieve self-assembly of these motor systems from theircomponents and, for example, to control themotion of kinesin using defined micro-butuletracks (Nêdêlec et al., 1997; Hiratsuka et al.,2001).

2.1.24 Multi-Functional Materials. Manybiological materials perform several functionssimultaneously. Skeletal plates of calcite

Figure 1. Depiction of an array of hybridnanodevices powered by F1-ATPase. Soong et al.,2000.

3/20/03

13

(calcium carbonate) on the arms of thebrittlestar Ophiocoma wendtii provide notonly structure and protection but also containthousands of lenses, which serve to focus lighton nerve receptors beneath, somewhat like acompound eye (Figure 2). The plates, whichare clear, birefringent single crystals of calcite,allow the organism to detect changes in lightintensity on its body surfaces and change colorfrom night to day (Aizenberg et al., 2001).Multifunctional scales coat the wings ofbutterflies and simultaneously aid in theaerodynamics of the wing, assist intemperature control, and provide colors andpatterns on the wing, serving as an avoidancedefense mechanism against predators.

2.1.25 Organic Synthesis. Living systems areperhaps the ultimate factories. Extraordinarilycomplex and large materials are synthesizedfrom an exceptionally small list of simpleprecursors. The key to this lies in the use of acomplex of interlinked metabolic pathways,charting a sequence of comparatively simpleorganic chemical reactions that convert thestarting materials, through as many as 20reactions or more, to the product. Enzymaticcontrol of the reactions insures that nobyproducts are produced. A large number ofcommon intermediates insures that aminimum amount of duplication is involved.

2.1.26 Optical Systems. Organisms havedeveloped a variety of optical systems, not the

least of which is the eye of higher organisms.As described above (2.1.24) O.!wendtiiproduce calcite micro lens arrays. This concepthas already been adapted for directionaldisplays and in micro-optics. Opals andbutterfly wings also manipulate light in amanner that could serve as a model forphotonic systems (Sambles 2001).

2.1.27 Self-Assembly. Perhaps the mostpowerful of properties of biological systems istheir ability to assemble individual moleculesinto large, complex, functional structures. Theinformation for the assembly lies in themolecular structure of the components, theirgeometry and their precise alignment ofhydrogen-, ionic-, polar- and “hydrophobic”bonding groups (Section 2.1.22). Membranes(Section 2.1.20) assemble themselves becausethe lowest energy state of their componentamphipathic molecules is the membrane lipidbilayer. Proteins (Section 2.1.10.2), composedof multiple individual subunits, self-assemble,aligning the individual subunits precisely withrespect to one another to perform a function asdependent on the relationship of theindividual molecules as a watch is dependenton the meshing of its gears. The ribosome isself-assembled from 80 proteins and 4 piecesof RNA. Hierarchical construction (Section2.1.15) is simply a process involvingsequential, increasingly complex, self-assembly steps.

Viral self-assembly is another remarkableexample. The bacteriophage phi29 has about20 genes coding for proteins out of which thevirus is constructed and a variety of otherproteins and RNA which transiently aid thatconstruction process. The phage’s capsid, orshell, self-assembles about a molecularscaffold which later disassembles itself. Theempty capsid is then filled with DNA by theaction of a transiently formed molecular motorpowered by ATP (Smith et al., 2001). Once theDNA is packaged, the motor falls apart. Whilethis assembly normally occurs inside the cell,molecular biologists have identified andcloned all of the genes of the proteins needed

Figure 2. Scanning electron micrograph of a dorsal armplate from a light-sensitive brittle star. The arm-plate,which is made up of single crystal calcite, is decoratedby myriads of microlenses that concentrate light on theunderlying nerve bundles. Courtesy of L. Addadi.

3/20/03

14

to assemble the virus from purifiedcomponents in vitro (Guo et al., 1986). In a 20microliter test tube reaction each of 100 billionDNA molecules (end to end, 400 miles ofDNA) are packaged into 100 billion viralcapsids in approximately 5 minutes. Theseviruses are all infectious and each one has thepotential to make an infinite number of copiesof itself by repeating the process. A variety ofsuccessful attempts at self-assembledstructures mimicking these sorts of biologicalsystems have been described (Ball 2001).

2.1.28 Self Healing, Repair, Damage andFault Resistance or Tolerance. Livingorganisms are, of course, capable, to varyingdegrees, of self-repair and healing. Simpleorganisms, or very young organisms canreplace entire sections of their bodies. Morecomplex or older organisms are more limitedin this, although the human liver, for example,can regenerate itself even after much of it hasbeen removed. On a more molecular level,membranes have the capability to repair holes,and proteins can refold after being denatured.DNA polymerase, which copies DNA, reviewsits own work, and excises errors, replacingthem with the correct base. The application ofthis principle to non-living materials anddevices is almost as difficult to imagine as it isto calculate the energy and cost savings thatwould result if it could be achieved.

2.1.29 Signal Transduction. Livingorganisms detect changes in their chemicaland physical environment and rapidlyrespond to them. The process involvesmolecular recognition and a resultingconformational change in “receptormolecules.” This change in shape can lead toenhanced or reduced enzyme catalysis, ortransport of ions or molecules. Signalamplification (Section 2.1.2) is usually a criticalcomponent, as these systems often transduce achange in tens or hundreds of molecules into aphysiologically significant response. Manygroups have mimicked this stimulus-drivenconformational change (Krauss 2000).

2.1.30 Smart Materials/Sensors. Smartmaterials are those that alter their structureand properties in an almost immediateresponse to a change in their environment,thus, on a much shorter time scale thanadaptation and evolution. In many cases theseare reversible changes, and in some cases, theextent of the change is controlled to reflect thedegree of change in the environment. In somecases, individual molecules are ‘smart’(Section 2.1.8). In other cases a system ofmolecules responds. An individual moleculeof the enzyme glutamine synthetase, forexample, monitors the level of nineindependent factors in its environmentsimultaneously and adjusts its rate of catalysison the basis of a summation of these positiveand negative inputs. On a more complex level,entire metabolic pathways respond to singlemolecules, such as hormones, growth factorsor pathway products or other metabolicintermediates. In some cases the response tothe stimulus is a functional one. In other cases,it is simply a record of that stimulus. In cells,“analytes” are detected by “surface mounted”protein and carbohydrate receptors whosestructure is defined at the level of each atomand the spatial relationship between thoseatoms, allowing molecular recognition, withhigh affinity and specificity. In most cases,these responses are triggered by alterations inthe shape of the receptor resulting from thebinding of its target.

The ability to change chemical activity, color,electrical conduction, mechanical properties inresponse to a change in the environmentwould be quite valuable in a variety ofmaterials applications. Perhaps the mostadvanced smart materials at this time aresensors, which translate their detection ofdefined targets into measurable optical,electrical or mechanical signals. Biologicalsystems provide a very high standard toattain, demonstrating discrimination,sensitivity and adaptability that can approachthe detection of single molecules or photons.Dogs can distinguish individual humans bysmell. Honeybees have been trained to detect

3/20/03

15

explosives at levels as low as tens of parts pertrillion (Rodacy 2002). Other organisms haveexceptional senses of taste, touch, hearing,sight. The Melanophila beetle senses infraredemission of a forest fire at a distance of 50kilometers. A dish-like organ under its wingscontains structures that are tuned to absorbthe appropriate wavelengths and then increasetheir volume and apply pressure to structuresthat trigger nerve impulses. Vipers, pythonsand other snakes identify warm bloodedtargets objects by detecting the minutedifferences in radiated temperature, withexceptional discrimination levels. Asminiaturization progresses, and micro- andnano-scale devices are developed, sensors forextremely small forces will be required. Recentwork (Liphardt et al., 2001) has shown that theunfolding of single strands of RNA, whichinvolves only the breaking of hydrogen bonds,can be measured using optical tweezers,allowing speculation that these measurementtools could be adapted as (nano-) mechanicalsensors. Cantilevers have been shown to allowthe detection of the change in energy resultingfrom the binding of very small numbers ofmolecules, again allowing speculation aboutnew, ultrasensitive mechanical sensors.Although investigators have had difficultyadapting these and other such systems toworking devices, it cannot be assumed thatthese problems will not be resolved.

2.1.31 Structural Materials. Nature’smaterials have exceptional strength andtoughness (Smith, B.L. et al., 1999; Hinman etal., 1993; Waite et al., 1998; Curry 1977;Jackson1988). Spider dragline silk, synthesizedprimarily from carbon, hydrogen, oxygen andnitrogen, has long been envied for its fractureenergy which is two orders of magnitudegreater than that of high tensile steel (Hinmanet al., 1993; Heslot 1998). Its use has beenhampered by the availability of the material,(which is more difficult to obtain than theweaker material from easily grown silkworms.) Recently, however, researchers havebeen able to splice portions of the genes forspider silk into cells from a variety of other

organisms that can be grown in tissue culture(Lazaris et al., 2002). Systems also exist whichwould allow the transfer of the genes to goatsthat have been bred to produce the silk in theirmilk. The proteins produced are not yetidentical to natural silk; synthesis of only oneof the two proteins has been achieved, andeven that one is produced in a form that isshorter and weaker than the natural product.Unfortunately, we do not yet fully understandthe structure of the silk proteins, themechanism by which the spider processes theproteins into fibers, or the techniques requiredto manipulate the extremely long pieces ofDNA that code for the large silk proteins.

Other natural structural fibers such as collagenand keratin, each with their own exceptionalproperties could find applications if theirsynthesis were possible. Research is alsoprogressing in the use of synthetic materials tomimic the biological models, for examplehydrogels that can reversibly bind water tomimic the ability of collagen to absorb shock.

Extensive work has been proceeding fordecades to improve our understanding andability to mimic the “hard” biologicalstructural materials such as bone, teeth, shell,which have exceptional combinations ofmechanical properties and light weight.Abalone shell, a composite of CaCO3 andorganic polymers is 3000 times more fractureresistant than a single crystal of CaCO3 (Curry1977; Jackson 1998) and artificial composites ofceramics and adhesives are also inferior to thenatural material (Smith B.L. et al., 1999;Almquist et al., 1999). Mineralizedcomponents are, as are the organiccomponents, usually made of a small group ofsimple compounds, in this case hydroxyapatite or calcium carbonate or phosphate.They do, however, form a wide variety ofstructures with these building blocks, thusachieving a wide variety of properties whichcan perform a wide variety of functions. Thisvariability in crystal structure arises becausethe organic protein phase controls the crystalstructure of the mineral and a variety of such

3/20/03

16

structures are possible. The synthesis of thesilica needles in a marine sponge, for example,is achieved by the enzymatic condensation ofsilicon alkoxides. Enzyme molecules alignedin linear repeated assemblies in the form of arod, allow the deposition of the mineralizedspines of defined shape around them (Morse2001).

2.1.32 Systems. On a larger scale, theperformance of biological systems oftenexceeds that which is attainable with currenttechnology. Shark skin, for example, exhibitsbetter hydrodynamic behavior than polishedsurfaces, an effect attributed to theorganization and structure of the surface. Ithas served as a model for drag reducingcoatings (Bechert, D., in Ball 1999) (Figure 3).Lotus leaves are remarkable in their ability toreject dirt. Their fine surface roughnessprevents tight binding of dirt particles (andeven glues), which are then easily washedaway by water which is itself repelled by thewaxy coating. Structural surfaces on insectlegs allow for reversible adhesion to surfacesas do the microscopic setae on the feet ofgeckos, which can hang upside down and onvertical walls depending on van der Waalsbonds (Autumn et al. 2002). Valves in veinsare designed to allow blood flow in only onedirection; flow in the opposite direction forcesthe two flaps of the valve together, closing thechannel. Yu and coworkers (Yu 2001) designeda hydrogel valve that mimics these checkvalves and could find use as an actuatedcontrol in microfluidic systems.

2.1.33 Template Directed Synthesis. Muchof biological synthesis occurs through enzymecatalyzed reactions, with the great specificityof these catalysts providing the high level ofefficiency and minimization of byproducts.Equally high fidelity is achieved throughtemplated synthesis, with the productproduced through its specific match to apreexisting model. DNA and RNA synthesisuses the sequences of bases on an existingsingle strand to determine the sequence ofbases to be organized in the daughter strand.

Deposition of mineral phases is directed by theshape of the underlying proteins. Thissynthetic strategy could be valuable in the

Figure 3. The riblets on shark skin (top) providedthe inspiration for modeling studies of the dragreduction they confer, and eventually led to trialson an aircraft coated with a plastic film with thissame microscopic texture (bottom), where an upto 8% reduction in drag was observed. Ball 1999.

Figure 4. Model of a voltage-dependent K+

channel. Zhou et al., 2001.

3/20/03

17

synthesis of many other types of materialsprovided that the design of the appropriatetemplate can be achieved.

2.1.34 Transport Systems. Living organismsmust transport a wide array of molecules andstructures both in and out of the organism orits constituent cells, or to various specificlocations within these organisms, cells ororganelles. Membranes are extraordinarilyselective in allowing materials to penetrate. Insome cases, even the hydrogen ion is

prevented from passing. On the other hand,membranes can develop systems tospecifically allow the passage of molecules ofchoice (Figure 4). In many cases these“channels” open or close in response to eithervoltage changes or the presence of othermolecules. In other cases, chemical or voltagegradients are used as energy sources to drivetransport against a concentration gradient(e.g., Gulbis et al., 2000). Separationstechnologies would benefit greatly from thesecapabilities.

3/20/03

18

3. Self-Assembled, Templated and Hierarchical Structures2

3.1 Introduction. Nature, through the courseof evolution, has developed techniques toconstruct increasingly complex systems,reaching her most glorious achievement in theliving organism and, in particular, the humanbrain. Perhaps as impressive as the productitself is the means employed to achieve it.Nature begins with a surprising small set ofbuilding blocks, modifies some, producesvariants of others, and then organizes tens,hundreds, thousands or millions of them, inmost cases in precisely defined three-dimensional functional structures. Toaccomplish this, it uses the processes of self-,templated-, and hierarchical assembly,manufacturing techniques that are driven onlyby the laws of kinetics and thermodynamicsacting on the structural and electronicproperties of the assembled building blocks. Itis as if all the parts of an automobile werethrown in a box and, within minutes, the fullyfunctioning car drove off – on its own.

The long term goal of the study ofbiomolecular materials is the development ofsystems employing or based on biologicalprocesses that function independent of theliving organism. Construction is a criticalissue. At the nanoscale we cannot manipulatebuilding blocks one by one, placing each in itsrequired location relative to the others.(Techniques have been developed to usescanning probe microscopes to manipulateand arrange individual atoms one-by-one (e.g.,Eigler and Schweizer 1990; Avouris et al.,1996)but these cannot, at least now, push moleculesor structures into place, or do more than one ata time. Thus, the concept of using nature’sself-assembly principles outside the livingorganism to construct materials, structuresand devices in non-living systems hascaptivated imaginations for centuries. Now,however, with our newly achieved

understanding of these biological processes,and the structure/property/functionrelationships of the building blocks we wish touse, we can reasonably hope to achieve thisgoal.

The building blocks in this process can bebiological molecules or abiological mimics ofthem. In the first case, we take advantage ofthe inherent self-assembly “guides” thatnature has built into these molecules. In thelatter case we take advantage of the syntheticskills of the organic chemist, allowing us analmost infinite variety of structures andtherefore potential properties and functions.Taken together a linkage of chemistry andbiology can provide multiple avenues andopportunities to build assembled systems thatcan be used to advance energy production andstorage, photonics, electronics, compositedesign, catalysis, diagnostics, andcomputation. In some cases the properties andeventual functions can be predicted from thechoice of building blocks and the nature of theself-assembly process. In other cases, we seeonly a set of unique new structures whosefunctions remain to be discovered, as have somany before, by serendipity playing before theprepared mind.

The key to self-assembly lies in the structure ofthe building blocks. They must accomplishtwo tasks – their desired functionality and theability to link to other components in apredetermined fashion. The breadth ofpossibilities is breathtaking. First, natureprovides us with a wide choice of lengthscales. Amino acids (for proteins), nucleotides(for nucleic acids), sugars (for carbohydrates)and simple lipids (for lipid assemblies) havedimensions of the order of nanometers.Proteins, carbohydrates and complex lipidsare one to two orders of magnitude larger.

3/20/03

19

DNA, viruses, liposomes, cellular machines,while larger still, are orders of magnitudesmaller than micrometer size cells andorganelles. The power of variability at each ofthese length scales is also easily seen. Theproperties of polymers depend on their lengthand precise sequence. 420 distinct 20 baseoligonucleotides (DNA or RNA) can be madeusing the four major nucleotide monomers.2020 peptides of that length can be constructedfrom the constituent amino acids. Orders ofmagnitude more oligosaccharides can be madefrom their many sugar monomers, manyarising from the fact that these polymers canbe branched at any of a large number of sites.Moreover, as mentioned above, one can easilymodify nucleotides, amino acids and sugarsthrough chemical synthesis to make even thesenumbers appear small. An oligomer withvirtually any desired chemical functionality,surface binding group, chromophore, redox-active moiety, catalytic agent, andspectroscopic probe is thus accessible. Finally,new backbones can be employed, whilemaintaining the uniquely biological propertiesof sequence specificity, defined length, anddefined three-dimensional structure. Usingthese monomers or polymers as buildingblocks coupled with the ability to doautomated synthesis, we have a “toolkit” ofunmatched variability, allowing us to developstructures with an almost limitless range offunctions. In all cases however, the structuraland electronic needs of self-assembly mustalso be achieved.

Self-assembly can involve several types ofmaterials.

• Organic Materials. Use of proteins ornucleic acids with or without syntheticpolymers in novel structures based ontheir inherent H-bonding, polar, ionic,and hydrophobic self-assemblyproperties.

• Organic-Inorganic Hybrid Structures.Bone teeth, shell are highly-evolvedorganic/inorganic structures with

exceptional mechanical properties. Agreat deal of work involves attempts tomimic these structures to achieve theseproperties in artificial systems. Otherwork involves the use of bio/organicstructures to bind functional inorganiccrystals in novel photonic or electronicdevices.

• Abiological Molecules MimickingBiological Structures. The principles ofbiological structure can be used as thebasis for self-assembly of fullynonbiological materials. The sp3 geometryof carbon bonding can, for example serveas the conceptual basis for tetravalentcolloid crystals. Similarly thehydrophilic/hydrophobic forces thatdrive water, protein, and membranestructure could be the basis for thesupramolecular assembly of complexfunctional assemblies of a wide variety ofnonbiological materials.

3.2 Organic Materials

3.2.1 DNA-Based Materials. The specificity ofthe intra- and intermolecular interactions ofnucleic acids underlies the basis of geneticinformation. Self-assembly, in nature, ofstructures such as duplex DNA, DNA-RNAhybrid double helices, L-shaped tRNA’s, andsingle stranded RNA hairpin and loopedstructures allows us to exploit these moleculesas the basis for rationally programmedsynthesis of complex three dimensionalstructures. One aspect of this “DNAnanotechnology” is predicated on theconstruction of DNA motifs that containbranched molecules or their generalizations.These motifs can then be combined into morecomplex structures exploiting “sticky ends”which allow the base pairing of twoindependent pieces of nucleic acid which havecomplementary, unpaired ends (Figure 5).These linkages involve the same self-assemblyspecificity as DNA or RNA duplexes and canthus be equally easily programmed into theprimary structure of these molecules. The local

3/20/03

20

product structures that result when the twocomponents cohere are B-DNA, theconventional double helical structure. Thisassembly has been shown to be capable ofproducing complex objects (Chen et al., 1991)such as, for example. a DNA cube, (Figure 6),periodic arrays (Winfree et al., 1998), andnanomechanical devices, such as, for example,the structures shown in Figure 7: one, (a) adevice based on the B‡Z right-handed to left-handed DNA transition, and the other, (b) adevice based on the switching of base pairingbetween oligomers of identical sequences(Mao et al., 1999; Yan et al., 2002).

The first of these examples, the B-Z device(Figure 7a) operates by switching between theconventional right-handed B-DNA structuresand the unusual Z-DNA structure. Z-DNA hastwo requirements, an appropriate sequence(usually something like [CG]n) and propersolution conditions. The appropriate sequenceis at the center of the device, and consists of 20nucleotide pairs whose ends rotate 3.5 turnsrelative to each other when they switch to Z-

DNA from B-DNA. The Z-state is achieved byaddition of Co(NH3)6 Cl3, which facilitates itsformation. Removal of this activator returnsthe system to the B-state. The dangling filledcircles in the figure represent a pair of FRETdyes to monitor the transition.

The second example, the sequence-specific(Figure 7b) device is based on the differentialhybridization topology of two pairs of strands,shown as green strands on the left (the PXstate) and as purple strands on the right (theJX2 state). The green strand pairs and thepurple strand pairs associate with 1.5 turns

Figure 6. Schematic of a DNA cube.Adapted from Chen and Seeman1991.

B'

A'

B

A

B'B'

A

A

B B

A'

A'

Figure 5. Linkage of nucleic acid oligomers usingsticky ends. The unpaired regions of A’ and B’ arecomplementary to those at A and B, allowingspontaneous linkage as shown on the right.Adapted from Seeman 1982.

B-ZZ -B

Z - D N A

B - D N A

+Co(NH3 ) 6Cl 3-Co(NH3 )6 Cl 3

( a)

A B

C D

A B

C D

A B

D C

JX2

A B

D C

PX

I II

IV III

(b)

Figure 7. (a) A DNA device that flips betweenconformation as conditions convert the doublehelix from the B, (right-handed) to Z, (left-handed)DNA forms. Mao et al., 1999. (b) A sequence-dependent device. Yan et al., 2002.

3/20/03

21

each of their devices, but that they haveextensions drawn horizontally in the figures.Starting with the PX state, if one adds the full-length Watson-Crick complements to thegreen strands, including the extensions, thegreen strands will be removed from the rest ofthe device. If the complements contain biotingroups (black dots) they can be removed fromsolution by streptavidin-coated magneticbeads. The unstructured intermediate at thetop of the figure can be converted to the JX2structure by the addition of the purple strands.The tops of the PX and JX2 structures are thesame, but the bottoms are rotated a half-turnrelative to each other. The JX2 structure can beconverted back to the PX structure by thesame procedure, as shown at the bottom of thediagram. The green and purple strands andthe segments to which they arecomplementary can be varied, leading to avariety of devices that can be addressedindividually.

DNA base/pairing specificity is the basis ofthe growing field of DNA-based computation(Adleman 1994). Methods to achieve thisinvolve parallel methodologies that includeDNA-based logic gates and computation byself-assembly (Winfree 1995; Mao et al., 2000).The importance of computation by self-assembly is that this area can lead toalgorithmic self-assembly: whereas theincorporation of a large number of distinctmolecular tiles, N, in an array would naivelybe seen to require N distinct tile types and thecosts associated with their individualpreparation, algorithmic assembly, in whichthe pattern of each surface leads to theappropriate linkage of the component pieces,can result in the production of similar arraysfrom far fewer tiles, thereby saving expense.Likewise, algorithmic assembly can ultimatelylead to highly precise dimensions forconstructs in one or more dimensions, inemulation of the ways in which biologicalsystems make polymers of specific lengths(Winfree 2000).

3.2.2 Polymer-Organic Hybrid Structures.While DNA and, for that matter, proteins, areremarkably valuable for their ability toprovide encoded as well as directed self-assembly, synthetic polymers built on theirmodel possess a great deal of versatility offunction. For example, random copolymersobtained from a generic vinyl monomer and adendritic macromonomer might provide aunique template along the polymer backbonecapable of spatial assembly and organizationof functional structures that have been linkedto them. The preparation of a random styreniccopolymer with dendritic pendant groups hasbeen demonstrated (Hawker 1992) as has thepreparation of similar copolymers withpendant dendrons functionalized with easilyremovable protecting groups (Tully et al.,1999) for further synthesis.

Another approach to the precise assembly andorganization of functional structures into asupramolecular entity involves their ligationto the periphery of shape persistentdendrimers. The preparation of several typesof complex dendritic assemblies (Malenfant etal., 1998; Andronov and Fréchet 2000a,b) hasbeen demonstrated. A rigid rod-like core in adumbbell-type structure (Malenfant et al.,1998) with dendrons at both ends effectivelyacts as a fixed spacer between these dendrons,thereby preventing van der Waals contactbetween them. The shape-persistent propertiesof these and other dendritic assemblies can beexploited in order to install a well-definednumber of structures on the same molecule,while preserving a minimum separation. Thistype of ensemble would provide uniqueinsight into the properties of complex butwell-defined nanostructured assemblies.

Figure 8. Dendrons with strong ligands linked viarigid rod-like oligomers. Malenfant et al., 1998.

3/20/03

22

Building blocks possessing dendrons withstrong ligands at their periphery (Figure!8) canalso be used. The focal point of the dendrimercan connect to a well-defined rigid rod-likeoligomer (Martin and Diederich 1999) of thedesired length. Two specific examples includean oligothiophene similar to those alreadyprepared (Andronov and Fréchet 2000b) and avariety of phenylacetylene oligomers (Moore1997; Tour 1996).

3.2.3 Lipid-Protein-Nucleic Acid HybridStructures. Studies of the biologicalmembrane, a lipid-protein complex thatperforms a myriad of functions, (Section2.1.20) have led to the development ofsupramolecular assemblies in reconstitutedbiological polymers which exhibit hierarchicalself-assembly on length-scales spanningsubnanometer to microns. Use of filamentousactin (F-actin) (a component of muscleprotein) and membrane lipids has led to theassembly of a network of connected tubules onthe mesoscale (>1!micron) (Wong et al., 2000).On the nanometer scale, these tubules consistof stacked (F-actin-lipid-F-actin) multilayers(Figure 9). Since the physical and chemicalconcepts leading to the self-assembly aregeneral, by replacing the biologicalcomponents with synthetic analogs (e.g.,replacing lipids with diblock co-polymers, orreplacing F-actin with synthetic polymers)

“plastic” mesoscale tubules and tubulenetworks could be realized for possibleapplications in chemical delivery, nanoscaletemplating, and nanoscale mask development.

Cationic lipids (CLs) are currently used inclinical applications as vectors to delivergenes. For example, a significant fraction ofworldwide clinical trials on developing cancervaccines use CLs (Henry 2001). Recent workhas found that CL-DNA complexes exhibit anunexpectedly large degree of order on the sub-to-many nanometer length scales. The two keyCL-DNA structures are inverted hexagonal(

†

IICH ),where DNA is coated by inverted

structured lipids arranged on a hexagonallattice, and lamellar (

†

aCL ), with alternating

lipid-bilayer and DNA monolayers (Raedler etal., 1997; Koltover et al., 1998) (Figure 10).

The lamellar lipid-DNA complexes have novelmaterials applications. It was found thatcertain electrolytes (e.g. Mg2+, Mn2+, Co2+...) canmediate unusual attractions between DNA

chains bound to lipid membranes, forcing thechains to form the most compact state of DNAon a surface, with the electrolytes trappedbetween the DNA chains (a precursor of nano-wires) (Koltover 2000). Potential applicationsof the DNA-lipid-electrolyte hybrid materialare in high-density storage and retrieval ofgenetic information, and in parallel processingof nanometer scale wires. In many of thesehybrid structures, the mechanical properties ofthe biological are also of great interest. The

Figure 9. Complexation of F-actin and cationic lipidsleads to the hierarchical self-assembly of a network oftubules; shown here in cross-section. Wong 2000.

Figure 10. Structures of the lamellar (L) and hexagonal (R)complexes of DNA and cationic lipids derived fromsynchrotron radiation data. Raedler et al., 1997; Koltover etal., 1998; Koltover et al., 2000.

3/20/03

23

protein components of viral capsids self-assemble as containers for the viral DNA orRNA. These structures are able to resist strongforces, even when loaded with tightly packed(and highly charged) strands of DNA andRNA. A theoretical understanding of theability of capsids to resist being blown apartby these forces might suggest new ideas formaking strong materials.

3.2.4 Biomolecular Materials inMicrochannels. Self-assembly ofbiomolecules within confined geometries hasthe potential to control lengths on scalescomparable to the physical confinement.Micro-channels are able to control orderingfrom the sub-micrometer to the many-micrometer scales. Several applications ofmicrochannel-based confinement andalignment of macromolecules can beenvisioned. First, highly oriented biopolymersamples can be used as templates for makinginorganic materials with desiredmicroporosity. Second, patterned surfacesconsisting of ordered monodisperse defectsorganized in two dimensions may bedeveloped with applications in templating andchemical delivery (Pfohl et al., 2001).

3.3 Organic-Inorganic Hybrid Structures.Nature has provided a truly inspiring set ofexamples of the use of self-assembly andtemplated synthesis in organic/inorganiccomposites.

These materials serve a wide variety offunctions. Each has a complex shape, structureand organization that has been tuned forfunction over millions of years of evolution. Inthe course of this evolution, a number offundamental problems in basic materialschemistry have been solved. The most strikingis the ability to control self-assembly at severalsuccessive hierarchical levels, buildingstructures step-by-step from the Ångstrom tothe macroscopic length scale.

At the molecular level, these compositesinvolve biological macromolecules that serve

as templates for the synthesis of mineralsurfaces, controlling their nucleation, growth,orientation, microenvironment and structure(Belcher et al., 1996; Falini et al., 1996; Cha1999; Meldrum et al., 1993). The heart of theproblem thus lies in the rules of molecularrecognition between proteins and crystalsurfaces, a subject that has been studied insome depth. Despite this work, however,detailed mapping of the active sites ofinteraction at organic-inorganic interfaces, interms of protein sequences, structure andassembly is critical information that is justnow starting to emerge and a completeunderstanding remains one of the primarychallenges. Achievement here will then leadus to the subsequent challenge: the need totranslate this information from naturalsystems to artificial materials with the samelevel of control.

In some cases, imitations will be very closelyrelated to the structures of the naturalproducts such as bone, teeth, shell. In othercases, artificial systems will be developedbased on the two underlying principles: one,the rules of organization learned by studyingthese systems and two, the concept that thedesirable properties of these two disparatetypes of materials can be combined in amaterial whose properties are inaccessible tostructures containing only organic or onlyinorganic materials. Orthogonal to thisapproach is the line of research that separatesthe study of mineralized biological structuresinto those that are structural and those that arefunctional. Successful work has already beenreported in all these areas (Colvin et al., 1992;Brust et al., 1995; Li et al., 1999; MannAlivisatos et al., 1996; Mirkin et al., 1996;Brown 1992).

3.3.1 Natural Mineralized Tissues.Organisms use three conceptually differentstrategies to build their skeletal parts. Eachcould serve as the basis for biomimeticorganic/inorganic synthesis. The easiestapproach to the filling of a given shape withsolid material applies when there is no

3/20/03

24

internal order or structure, i.e., when thematerial is amorphous. The most strikingexamples are found in silicaceous spongespicules and diatoms. The recently determinedsequence of a family of proteins that catalyzethe polymerization of silica is providinginformation on the mechanism of assembly ofthis natural biological material. It is hoped thatthis understanding will be applied to thedevelopment of new synthetic pathways. Forexample, these findings, coupled with geneticengineering to identify the structuraldeterminants responsible for the catalysts,have led to the predictive synthesis of a‘biomimetic’ self-assembling peptide thatsimultaneously catalyzes and directs thestructures of silicas and silicones from thecorresponding precursors (Morse 2001).However, not all amorphous phases used byorganisms, for example calcium carbonate, arestable under non-biogenic conditions. Here,specific macromolecules impose a structuredmicroenvironment on the forming amorphousphase, continuously inhibiting the formationof crystallization nuclei. Thus amorphousmaterials must have some structure that, it issuspected, encodes their fate.

A second solution involves building a materialas a single crystal, deposited from anamorphous phase that very slowly transformsin a controlled manner. The regulation of thetransformation is crucial to the shape,structure and microtexture of the finalmaterial. Examples of single crystal skeletalelements are the calcitic sea-urchin larvalspicules Figure 11 and adult spines. Thestriking calcitic microlenses found in thedorsal arm plates of light-sensitive brittle starsalso belong to this group (Aizenberg et al.,2001).

A third solution involves building skeletalmaterials as polycrystalline assemblies formedinside self-assembled matrixes composed ofbiological macromolecules such as proteinsand polysaccharides. Control of themicroenvironment of crystal formation inthese systems, which include bone, teeth and

mollusk shells, is crucial.

Each of these approaches provides inspirationfor the design and construction oftechnologically important materials based ontotally new concepts. Thus, they openopportunities beyond those arising by simplymimicking specific materials.

3.3.2 Polymer-Directed MineralizedComposites. Inspired by thebiomineralization phenomena discussed in theprevious section, various synthetic approachestoward organic-inorganic hybrid materialshave recently been pursued. Among these, thestudy of amphiphilic polymer based polymer-ceramic hybrid materials is a particularlyexciting emerging research area offeringenormous scientific and technologicalpromise. Through the choice of theappropriate synthetic block copolymersystems (e.g., poly(isoprene-block-ethyleneoxide, PI-b-PEO) as well as ceramic precursors(e.g., organically modified ceramic precursors,ORMOCER¢S), unprecedented morphologiccontrol on the nanoscale is obtained in theblock copolymer directed synthesis of silica-type materials (Figure 12) (Simon et al., 2001).As in natural systems, control lies in theunique polymer–ceramic interface; thehydrophilic portions of the block copolymerscan be completely integrated into the ceramicphase. The resulting composites are thendescribed as ‘quasi two-phase systems’

Figure 11. Scanning electron micrograph of a larvalspicule from L. pictus (age 72 hours) mounted on aglass fiber. The whole spicule is a single crystal ofcalcite. Of note is the elegance, sculpted shape andsmoothness of the spicule, relative to the man-made glass fiber. Courtesy of L. Addadi.

3/20/03

25

allowing for a more rational hybridmorphology design based on the currentunderstanding of the phase behavior ofdiblock copolymers and copolymer-homopolymer mixtures. The structuresgenerated on the nanoscale are a result of afine balance of competing interactions, anotherfeature of complex biological systems. Inaddition to morphologies known fromconventional block copolymer studies newphases may be discovered. For examples, theexistence of a bicontinuous, cubic ‘Plumber’sNightmare’ phase was recently suggested forthese block copolymer-ceramic hybrids(Finnefrock et!al., 2001). This indicates subtle,not yet understood differences to conventionalblock copolymer systems and emphasizes theneed for future interactions with theory andsimulations. Nano-engineering of suchhybrids towards applications has beendemonstrated in the area of nano-objects ofpredetermined sizes, shapes and compositionsfor nanobiotechnology as well as mesoporousmaterials for separation technology andcatalysis (Simon et al., 2001).

The potential of this approach for newmaterials lies in the versatility of the polymerchemistry as well as that of the sol-gelchemistry that can be exploited in thematerials synthesis. Focusing on the polymerside, Figure 13 shows a complexity diagramfor blocked (compartmented) macromolecules

(Simon et al., 2001). It illustrates that when thenumber of building blocks along the chain isincreased, the complexity of the resultingstructures is elevated significantly. For thecase of passing from AB diblock copolymers toABC triblock copolymers this has already beendemonstrated (Stadler et al., 1995; Breiner etal., 1997). A whole range of new morphologieshas been found for ABC triblocks. Theunderstanding of their detailed phasebehavior is a current area of intensiveresearch. It will be an interesting challenge totry to use those polymer systems as structuredirecting agents for the generation of polymer-inorganic hybrid materials. In this way, forexample controlled access to inorganic nano-objects in the form of rings or helices shouldbecome accessible. This is only one possibleresearch pathway, however, since the variety

of the polymer chemistry as well as theinorganic sol-gel chemistry is only limited byone’s imagination (Figure 13). Clearly, simpletheoretical models of how proteins control thekinetics of biomineralization and affect thestrength of resulting material would be veryvaluable.

3.3.3 DNA-Inorganic Hybrids. The use ofDNA as a template for nanometer scalemolecular organization was first proposed in1982 (Seeman 1982). The self-assembly andorganization of nanocrystals and nanoparticleson the molecular level using DNA followed(Figure!14) (Loweth et al., 1999; Mahtab 1996;Alivisatos 1996; Mirkin 1996; Elghanian 1997;

Figure 13. Complexity diagram of blockedmacromolecular systems. Simon et al., 2001.

Figure 12. Transmission electron micrographs of blockcopolymer directed polymer-silica hybrids withdifferent proportions of inorganic phase, each resultingin a unique morphology. Scale bars in B., C., and E.,as in A. Simon et al., 2001.

3/20/03

26

Storhoff 1999; Robinson and Seeman 1987).

These methods employed rely directly onhybridization of the two complementary DNAstrands (Section 3.2.1). Because the DNAduplex acts as a rigid rod over comparativelylong distances, the approach offers thepotential for control over particle composition,particle periodicity, interparticle distance, andcomposite stability which can be achievedthrough careful selection, and often design, ofnanoparticle and biological building blocks.This in turn would allow the development of awide variety of devices such as sensors andspectroscopic enhancers and advances inmicroimaging methods (Mirkin et al., 1996;Alivisatos et al., 1996; Robinson and Seeman1987). Although the strategy has beenextensively developed for gold nanoparticles,it has been explored with several othernanoparticle compositions as well, including,Ag, CdS, core-shell inorganics, liposomes, andpolymer particles. In fact, individual linkagesare controlled to hybridize (join) or melt(release) at specific temperatures, throughcontrol of the relative proportions of A-T andG-C pairs.

The incorporation of other molecularelectronic components into 3D DNA arrayscould lead to new computation capabilities,both in terms of density and speed (Robinsonand Seeman 1987). The incorporation of

nanomechanical devices within such arrayscould lead to the multiple states necessary fornanorobotics based on arrays of molecularscale devices with controlled mechanicalmovement (Yan et al., 2002). Algorithmicassembly in 2 and 3 dimensions, particularlywhen combined with nanodevices could leadto extremely smart materials. Similarly, theaddition of functional protein systems to thesecomplexes could lead to novel combinations ofenzymatic or binding activities. It seems likelythat extension of individual constructs frommicron to larger scales will requirehierarchical techniques, similar to Reif frames(Reif 1999). Ultimately, DNA also offers theattraction of self-replicating systems, althoughat this time self-replication of branched DNAmolecules is likely to be somewhat oblique(Seeman 1991). It is also of some interest tounderstand theoretically the strength of thesematerials, and quantities such as the criticalforce for DNA pull-out which causes thesestructures to tear apart. The optimal overlaplength for complementary base pair sequencesand the effect of sequence heterogeneity alsoneed to be understood.

DNA scaffolds are also used to induce thedimerization of proteins. More recently, asmentioned, an AFM tip has been used to“write” DNA with attached nano-particles onsurfaces of metals and oxides with preciselycontrolled feature size. The tip is coated withcharged molecules to which the DNA sticksuntil transferred to the substrate. Feature sizeis controlled by varying the humidity andsurface-tip contact time. Again, base pairingcan allow specific binding of complementaryDNA strands carrying desired activestructures, to create circuits, photonic crystals,or catalysis (Demers et al., 2002).

DNA could itself be an electronic component.Its base pairing properties have also been usedin studies of its potential role as a conductingmaterial that can be used in targetedattachment of functional wires (Braun et al.,1998). Oligomers are attached to twoelectrodes and a DNA strand that can base

Figure 14. Nanocrystals aligned through theirattachment to DNA strands. Alivisatos et al.,1996; Loweth et al., 1999.

3/20/03

27

pair with both oligomers is hybridized toconnect the electrodes. The DNA is thenmetalized with silver through ion exchange.

There also are many opportunities to use thesematerials to prepare high surface area“designer” catalysts, nanoelectronic devices,power structures (solar energy conversion andbatteries), plasmon wires, and colloidalcrystals for photonic applications (e.g.photonic band-gap structures, LEDs, lasers).However, to realize the full potential of thesestructures, we first must develop a firmfundamental understanding of the chemistryrequired to biofunctionalize the hundreds ofdifferent types of nanoscale building blockscurrently available; the properties (electrical,mechanical, chemical, optical, and structural)of the almost infinite number of possiblecomposite structures that can be prepared viathe strategy; and the ways to modifyanisotropic particles (e.g. rods, arrowheads,and prisms), in spatially well-defined ways.

3.3.4 Protein-Inorganic Hybrids. Analternative approach to study the rules ofmolecular recognition between biological(macro)molecules and organized inorganicsurfaces has been illustrated in the use ofcombinatorial selection either throughantibodies or proteins exhibited in techniquesknown as phage and bacterial display. Thesetechniques are particularly useful in the searchfor organic molecules that interact withinorganics with important electronic or opticalproperties, because these appear infrequentlyin nature and thus have no evolved models.Large libraries of peptides with random aminoacid sequences are produced and screened forbinding to selected inorganic crystals. Proof ofprinciple has been provided for both methods,identifying peptides that bind, with highspecificity, a variety of semiconductors (Brown1992; Brown 1997; Brown 2001; Naik et al.,2002; Lee et al., 2002) (Figure 15). Anexpansion of the use of these techniques, bothin breadth (to many more types of materials)and in depth (to reach understanding of theinteraction involved at the molecular level) is

expected. Use of peptides with multiplebinding sites mimicking multidomain proteinssuch as fibronectin or, motor proteins such askinesin could be exploited (Figure 15) totransport inorganic components of electronicstructures to their positions and orientation incomplex structures (Ball 2000). In fact, Schmidtand colleagues (Winter 2001) have linked CdSnanocrystals to neurons using customdesigned, nanometer length octa-peptidesthat, at one end have sulfur containingcysteines to bind the semiconductor, and at theother end, designed RGDS sequences to bindthe cell surface integrin av subunit on theneurons. It should also be noted that syntheticpolymers can be used to organize inorganicfunctional nanostructures (Andronov andFréchet 2000b).

(a)

(b)Figure 15. Phage recognition of semiconductorheterostructures. In (a), phage (yellow), selected torecognize GaAs were placed on a substrate patternedwith concentric rings of 1-mm GaAs lines (red)separated by 4mm SiO2 spaces (blue). When viewedfrom above (b) the fluorescently labeled phagedemonstrate their selective binding to the GaAs rings.Whaley et al., 2000.

3/20/03

28

3.4 Abiological Molecules MimickingBiological Structures.

3.4.1 Colloidal Particles with a Valence. Therichness of variety found in biologicalmolecules is derived in large part from the factthat the central atom, carbon, is able to bind tofour other atoms, in tetrahedral, sp3geometries. Self-assembly and manipulation ofmicron or sub-micron colloidal particles havemany potential uses, including particle basedassays. and photonic band gap materials.Dense glassy and crystalline particle arrays(possibly involving DNA linker elements)usually display the high coordination numbers(Z = 12-14) characteristic of an isotropic pairpotential. It is therefore of some interest todevise a means by which micron scale colloidscould link with, for example, a four-foldvalence, similar to the sp3 hybridized chemicalbonds on an Ångstrom scale associated with,for example, carbon, silicon and germanium.In contrast to close packed fcc and bcc colloidarrays, a tetravalent colloid crystal with adiamond lattice structure and appropriatedielectric constant is predicted to have a verylarge photonic band gap. Efficient creation ofcolloidal particles with a 1-, 2-, 3- or 4-foldvalence would also find useful applicationsranging from anchoring small catalyticparticles to surfaces to linking a precisenumber of biomolecules such as kinesin, RNAor DNA to particles which could then bemanipulated magnetically or with opticaltweezers. Theoretical analysis of proposals toproduce the equivalent of directional bonding

on a micron scale (perhaps involving orderedstates of surface active molecules on spheres)would be very useful here (Nelson 2002).

3.4.2 Hydrophilic/Hydrophobic Coatings onInorganic Materials. Whitesides and hiscolleagues have reported a number ofpioneering studies in which selected surfacesof inorganic materials such as gold plates(Clark et al., 2001) plastic “logs” (Oliver et al.,2001) and plastic “chips” with wires andimprinted devices (Gracias et al., 2000). arecoated with hydrophobic groups andimmersed in water to drive directed self-assembly into multicomponent structures.

We know the rules governing molecularrecognition and assembly: geometry and theweak bonds, hydrogen, electrostatic, polar and“hydrophobic”. What is far more difficult isthe task of constructing molecules that notonly perform their required function but alsohave structures that will self-assemble inprecisely defined complexes. It is easier tomake “sticky” structures that will link inrandom or multiple ways. Specificity is theproblem but is fundamentally the sameproblem faced more broadly: how does onearrange the appropriate atoms in a structurethat will perform in a totally predictablemanner, both in self assembly and in function.Organic/inorganic self assembly presents aslightly different challenge. Here it is not thetailoring of two surfaces for interaction butrather of one, to control the organization of theother, again an area whose rules are yet to bewritten.

3/20/03

29

4. The Living Cell in Hybrid Materials Systems

4.1 Introduction. Materials transformationsare a central function of living organisms.Bacteria, for example, can convert the simplesix carbon sugar glucose, in the presence ofappropriate ions such as sodium, chloride,phosphate, ammonium, into the literallythousands of diverse materials required fortheir growth: carbohydrates, proteins, lipids,nucleic acids, vitamins. They can fabricatestructures such as membranes, and controlelements such as receptors, repressors andinhibitors. These metabolic conversionsdepend on two cellular capabilities. The first isthe ability to create “pathways”– longsequences of relatively simple reactions, eachof which converts the product of the previousreaction to the starting material for the next.The net effect of a pathway can be a radicalchange in the structure and therefore functionof the material. Metabolic conversions alsodepend the ability to control those pathwaysso that other chemical reactions that arekinetically and thermodynamically possible,but would divert pathway intermediates toundesirable, wasteful or, toxic products, arenot allowed to proceed.

Cells also perform functions: sensing; energycapture, storage and transduction; chemotaxis;transport; for example. These are performedby complexes of molecules which self-assemble into functional units. Through abillion years of natural variation and selectionthese pathways and processes have evolved toan extraordinary level of efficiency andselectivity so that now they make highlyefficient use of energy, perform functions, andproduce valuable resources that are extremelydifficult or impossible to recreate viatraditional synthetic methods. In addition,they can be engineered to perform functionsthey do not perform naturally. Thus, theiradaptation to DOE mission goals, for example,

environmental bioremediation, energyharvesting, industrial fermentation, biosensingis both possible and potentially valuable. As aresult, the workshop participants placed, as ahigh priority, the goal of developing a greaterunderstanding of the biology underlying thisvariety of processes and the development oftechnologies that employ the cells to capitalizeon them. One might, in principle, expect thatin the future, individual processes of interestcould be isolated from cells and the rest of thecellular machinery and used in an abiologicalenvironment. This was judged, in many areas,to be sufficiently beyond current reach torequire that we need, in the short run, todevelop techniques that achieve these endsthrough the use and maintenance of the entireliving cell. (See, for example, Zeck 2001, Figure16.) It is also important to note that this would

involve the use of a wide variety of cells, fromorganisms ranging from viruses throughplants, bacteria, insects and humans, sinceeach type offers its own advantages andlimitations. In each case, the optimal cell forthe goal at hand would have to be identified,adapted to the need, and then used. The panel

Figure 16. Electronmicrograph, after fixation, ofneuron from the A cluster of the pedal ganglia in L.stagnalis immobilized within a picket fence ofpolyimide after 3 days in culture on silicon chip.(Scale bar = 20 mm.). Zeck and Fromherz 2001.

3/20/03

30

identified three important research areas inthis field:

• Metabolic Engineering. Livingorganisms evolved molecules andprocesses to allow them to survive andreproduce in their particular niche in theenvironment. Thus, not infrequently,naturally occurring molecules orprocesses will not be ideally suited to aDOE mission need. One solution thisproblem would be to “engineer” cells sothat they do produce the molecule orprocess required.

• Engineering the Cell-MaterialsInterface. Living cells have evolved verycomplex outer surfaces to allow them tosurvive in their environment, makingspecific contacts with other cells andextracellular proteins and communicatingwith them. Interfaces with inorganicsurfaces are generally not natural;research is required to design the cellsurface and the substrate to insurebinding, maintenance of function andcommunication.

• Artificial Cells. Ideally individual cellularmolecules or processes would beincorporated intobiological/nonbiological devices andstructures. As noted, in some cases it willbe difficult to purify the specificstructures from cells while maintainingtheir activity. Use of whole cells couldserve many purposes, but it is likely thatthe complexity of cells and the myriad ofmolecules and structures in them couldinterfere with the single functionrequired. Development of artificial cellsthat maintain function in a cellularenvironment but include only thosefunctions required could be a solution tothis problem.

• Model System-Cell-Based Biosensors.The discussion of the cell/materialsinterface led directly to a timely

application of research in this area: Thedevelopment of biosensors that use wholecells as the detection element andinorganic structures as the signaltransduction and display elements.

4.2 Metabolic Engineering. The ability toengineer cellular processes for improvementsin efficiency, or to establish entirely newmetabolic systems, is of considerable and verybroad interest for energy sciences. Theavailability, as a result of the recent majorDOE and NIH genome programs, of completegenome sequences for humans, othereukaryotes and numerous microbes nowprovides the blueprint for establishing thegenetic basis of metabolism. Accordingly, inthis post-genomic era, metabolic engineeringvia genetic approaches will be greatlyfacilitated.

Two methods of metabolic engineering havebeen established: rational design andevolutionary selection. “Rational” approachesseek to define the specific genes, enzymes(primary gene products) and metabolites(secondary gene products) involved in aspecific pathway and to alter these inpredetermined ways to cause predictablechanges in a cellular process. New pathwayscan be created in a cell by introducing both thegenes encoding the enzymes in thesepathways, and also the appropriate metabolicsubstrates. Applications to the production ofbiopolymers, both natural and unnatural, areparticularly interesting. For example, bacterialcells can be modified to producepolysaccharides on a large scale by firsttransfecting them with the genes encoding theappropriate glycosyltransferases and thenfeeding those cells simple sugars as buildingblocks (Mahal et al., 1997). The chemicalsynthesis of such polymers outside the cellularcontext would be very costly and energyconsumptive. Proteins are already beingproduced by cells through recombinantoverexpression, but the full exploitation ofthese systems is limited by the fact that mostcells can use only the 20 naturally occurring

3/20/03

31

amino acids in protein synthesis. Recentbreakthroughs(Section 5.4) have enabled theincorporation of unnatural amino acids intoproteins during large-scale overexpression incells, allowing the introduction of a widerange of new optical, electronic, chemical andstructural capabilities (Kiick et al., 2002). Thishas been accomplished by rational design ofnew enzymes for amino acid activation andthe design of new genetic codes for translationof modified genes into modified proteins. Afour-base codon allows one to expand thegenetic code beyond the 20 natural aminoacids to include almost any synthetic aminoacid of interest. It may also be possible toengineer the cellular protein biosyntheticmachinery to generate novel polymerbackbones. Polyester synthesis, for example,generally unknown in biology, would allowthe biosynthetic generation of novelfunctionalized polymers that fold into discreteconformations, enact specific functions, yetubiquitously biodegrade via ester hydrolysisin the environment.

Given the vast information content of anorganism’s genome and the complexity ofmost metabolic systems, rational metabolicengineering can be very difficult. Identifyingthe elements to alter and engineering themwhile leaving other systems intact requiresgreat knowledge and technical skill. In somecases there may be several genotypes thatproduce the target phenotype and alteringonly one might not have any effect. To accessnew metabolic properties in rapid fashion,“evolutionary” approaches to cell engineeringare taking a prominent position. Theseapproaches in effect allow the organism toselect the method by which it is altered. Theyproceed by iterative cycles of mutagenesis tocreate genetic diversity followed by selectionfor a given property of interest. The selectioncriteria must be carefully designed so that onlycells with the desired novel metabolicproperty have the required reproductiveadvantage. After multiple rounds of suchselection, cells with optimized properties canbe characterized at the genetic and

biochemical level to determine the molecularbasis of the improvement or novel property.

This evolutionary approach has been used toselect cells that overproduce certaincarbohydrate structures on cellularglycoproteins in order to extend their lifetimesfor biotechnology applications.Diversity/selection methods have also beenapplied to develop novel protein biosyntheticenzymes that prefer unnatural amino acids.Combined with the genetic coding strategiesdescribed above, this technique can be used toengineer new organisms that may acquire acompetitive advantage in the presence ofunnatural amino acids. Applications to theproduction of novel protein-like materials andenvironmental remediation can be envisioned.Moreover, new catalysts that exploit unnaturalfunctional groups can be evolved for specificapplications.

Despite recent advances, there is considerableroom for the development of new techniquesfor metabolic engineering. A major stumblingblock at present is the lack of knowledgeregarding the integration and regulation ofmetabolic flux within cells. Genomics andproteomics programs are now well-establishedaround the world, but, by contrast,“metabolomics” (the complete inventory of allmetabolic pathways and intermediates) is stillin its infancy. In order to accelerate metabolicengineering efforts, metabolic profiling effortsmust be undertaken so that the fundamentalcircuitry of cells is established. High-throughput mass spectrometry methods arepredicted to play a major role in these efforts.

Computational modeling and theoreticalpredictions will play a major role in futuremetabolic engineering efforts. Experimentalsystems need to be designed so that therelationships between environmental stimuliand changes in metabolic flux and output canbe elucidated. These experimental parameterscan be used to generate kinetic andthermodynamic models of metabolicnetworks. Ideally, metabolic modeling would

3/20/03

32

provide algorithms for rational engineering ofnew systems. Mutagenesis techniques can beapplied to the generation of vast arrays ofcellular variants with myriad metabolicproperties. High-throughput screening ofthese can identify a range of properties, andsubsequent metabolic and genetic profiling ofthe variants will provide a basis set fromwhich models can be generated. In addition,there may be a role for theory in the design oflibraries for evolution-based experiments sothat sufficient “diversity space” is covered.

Methods for probing cellular processes in vivoand in real time are still quite limited despitetheir paramount importance for cellengineering. New techniques for taggingspecific proteins or metabolites withfluorescent probes will be particularly usefulin this regard. The above methods forexpressing proteins with unnatural aminoacids are predicted to contribute significantly,as they will allow the delivery of fluorescentamino acids to a specific site in a protein ofinterest. The development of novel chemistriesthat permit highly selective proteinmodification within a cellular context isanother important goal.

Non-invasive magnetic resonance and infraredspectroscopy methods have enormouspotential for tracking metabolic pathwayswithin cells. Probes for these techniques,including fluorine atoms and azido groupsrespectively, have minimal structural impacton their host biomolecules and can be detectedbased on their orthogonality to native cellularcomponents. Already, the coherent anti-StokesRaman spectroscopy (CARS) technique hasbeen used to image endogenous biomoleculessuch as lipids and DNA within living cells(Figure 17) (Cheng et a., 2001; Volkmer et al.,2001; Zumbusch 1999). Further application ofthis and related methods should elucidate thedistribution of metabolites within cells in atemporal manner. Since biological work isaccomplished a single molecule at a time,spectroscopic methods that probe singlemolecules are of significant interest. The

ability to probe the action of a biomoleculewithin its native cellular context maytransform our understanding of processessuch as vesicle transport, gene transcription,and protein translocation. In general, newmethods for probing cellular processes withspatial and temporal resolution, and in realtime, are urgently needed.

4.3 Engineering the Cell/Materials Interface.In their native environments, animal cells arefound in tissues where they attach to othercells and to the extracellular matrix. Celladhesion is mediated by multiple, reversiblereceptor-ligand interactions that provide highavidity and specificity when engaged inconcert. Several groups have sought toengineer the adhesion of cells to artificialsubstrates by mimicking natural cell adhesionprocesses. These efforts have largely focusedon decorating the material surface withextracellular matrix-like structures that serveas specific recognition determinants for cellsurface receptors. The integrin family ofadhesion molecules plays a primary role inattachment of cells to extracellular matrix

Figure 17. Coherent Anti-Stokes Raman Microscopy(CARS) allows sensitive 3D imaging of live andunstained cells based on vibrational spectroscopy. Itprovides a point-by point chemical map of cellularconstituents, such as protein, DNA and lipid. ThisCARS image is of NIH3T3 cells using the Raman shiftof 2870 cm-1, the frequency of the aliphatic C-Hstretching vibration. The bright spots are due tomitochondria and other organelles that are rich in C-Hbonds. Cheng, et al., 2001.

3/20/03

33

components such as fibronectin, vitronectinand laminin. A minimal epitope for integrinbinding that is shared by these matrixcomponents is the Arg-Gly-Asp (RGD)tripeptide. The RGD motif has beenincorporated into polymer matrices which canthen support the adhesion of cells in abiospecific manner. Alternatively, wheat germagglutinin, a lectin that binds the cell surfacesugars N-acetyl glucosamine and N-acetylneuraminic acid can be linked to polystyrenemicrospheres to mediate the attachment ofliving cells to the microspheres. Opticaltweezers can be used to bring the cells andmicrospheres into contact in controlledgeometries in a process known as “light-driven microfabrication.”

The defined molecular architecture of self-assembled organic monolayers (SAMs) hasattracted significant attention to their use inmediating the adhesion of living cells. SAMscan be introduced on a variety of substrates(i.e., glass, gold or silicon). Those thatspecifically adsorb or repel extracellularmatrix proteins have been patterned inspatially defined arrays usingmicrolithography techniques to affordmicrodomains capable of supporting celladhesion. The adhesive domains werecomposed of hydrophobic hydrocarbon chainsor positively charged amino-terminatedhydrocarbons, both of which promote thedeposition of fibronectin and other proteins.Protein-repellent domains were composed ofpolyethylene glycol groups or perfluorinatedalkanes. The lack of matrix protein depositionon these domains rendered them incapable ofsupporting adhesion. Using these tools,defined micrometer-scale arrays of adherentcells segregated by non-adherent domainscould be engineered on artificial surfaces(Groves 2001). One of the limitations to thisapproach is the poorly defined nature ofadhesion at the molecular level. Theextracellular matrix proteins that supportedcell adhesion were secreted by the culturedcells and heterogeneously deposited on theSAMs, precluding their precise definition.

Furthermore, the interface between surfaceand cell may have been composed of manyprotein layers. In a biosensor device, forexample, the transmission of information fromthe cell surface to the material surface mightbe impeded by such an intervening layer.

The current ‘biomimetic’ strategies forengineering cell adhesion focus on the

O

OO

O NH

NH2

O NH

NH2

O NH O NH

O

NN

ManLev-treated cell

Hydrazide-functionalized material

Engineered adhesionof cells to material

ketone Cell

CellOHO

HO

HNHO

OH

OCO2

–HO

OHOH

HNHO

O

O

O

O

O

N-Levulinoylmannosamine(ManLev)

ketone

Figure 18. Chemical attachment of cells to syntheticmaterials. Cells are coated with ketone groups bymetabolism of unnatural ketosugars such as ManLev.Ketones react selectively with hydrazide groups installedon material surface coatings. Mahal et al., 1997.

Figure 19. Generation of hydrazide-functionalizedsupported bilayers and adhesion of cells expressingketone groups. Hydrazide-functionalized lipids andphosphatidylethanolamine bulk lipids generated asupported bilayer on glass to which HeLa cells adhered.Bruehl et al., 2002.

3/20/03

34

modification of materials with antibodies thatbind naturally occurring epitopes on cellsurfaces. This narrow focus limits both thetypes of materials and the types of cells thatcan be interfaced with each other. An idealstrategy for engineering cell adhesion wouldenable one to control the chemical compositionof both material and cell surfaces andrationally define the nature of the interface.Toward this end, metabolic engineeringapproaches have been applied to introducenovel chemical reactivity onto cell surfaces.Reactive electrophiles such as ketones andazides can be incorporated into cell surfacepolysaccharides by simply feeding cells thecorresponding keto- or azido-monosaccharidesubstrates (Saxon 2000). These functionalgroups are uniquely reactive with hydrazidesand phosphines, respectively, but areessentially unreactive with any native cellularcomponents (Figure 18). Cells engineered toexpress these electrophiles are primed to reactwith material surfaces functionalized with thecomplementary nucleophiles, therebyanchoring cell to material with a chemically-defined linkage. Such molecular-level controlover the cell-material interface may facilitateintegration into a device environment (Figure19).

It should be noted, however that while theadhesion proteins that nature uses to mediatecell-cell interactions are typically a fewhundred nanometers long, with manyfunctional groups, most current technologiesfocus on the control of cell-surface interactionsby using small fragments of these molecules.While these may be valid approaches tooptimize one or another cellular response,they are far too simplistic to truly mimic thecomplexity of the natural environment of cells.In their native environment, cells are anchoredto complex extracellular matrices (ECM) ofmultifunctional proteins. One such ECMprotein, fibronectin is a large (>100 kDa),molecule composed of repeating, structurally-defined modules which are often less than 100amino acids longs. Some modules carry celladhesion sites, while others carry sites

required for binding other ECM proteins andfor matrix assembly. Structural models havebeen derived to describe how cells can stretchfibronectin and thereby switch the function(see also Section 5.3) of its recognition andbinding sites (Vogel et al., 2001). Such proteinconformation changes which can occur duringnormal cellular activities also play a key rolein cell signaling. Systems have been designedto display these protein unfolding events asvisible color changes using fluorescenceresonance energy transfer (FRET). Since thisphenomenon is dependent on the distancebetween a donor and acceptor, the degree towhich the protein unfolds, increasing thedistance between component parts labeledwith donors and acceptors results in a colorchange (Figure 20; Baneyx et al., 2001). Similaralterations in molecular binding events as aresult of protein conformational changes havebeen shown in the FimH protein of bacteria(Thomas et al., 2002, Section 5.3).

The strict environmental demands ofmammalian cells in order to maintain viabilityhave limited their utility in artificial devices.Thus, methods that increase the robustness ofcells are of paramount importance. Selection-based cell engineering methods might beapplied to this problem. Encapsulationstrategies might also be explored. There areheartier cells from other organisms such asplants, which have cell walls, that might be

Figure 20. Imaging different protein conformations ofthe fibroblast cell adhesion protein fibronectin. Thefibronectin was labeled with donor and acceptorfluorophores. In the compact state on the cell surfaceit appears red (FRET energy transfer). In the extendedstate in matrix fibers, the donors and acceptors aretoo far apart for FRET and they appear green. Baneyxet al., 2001.

3/20/03

35

used as hosts for metabolic and signalingmachinery from other species. It will be ofinterest to reconstitute target metabolicpathways from mammalian or bacterial cellsinto plant cells for large scale production. Forexample, analyte detection and signaltransduction mechanisms might berecapitulated in plant cell hosts for biosensing.Advances in plant cell genetics and metabolicmapping will be important components ofsuch a program.

Finally, expanding the scale of cell-baseddevices from the single cell to organizedcellular arrays or tissues is a major frontier.This will require that the impact of a cell’senvironment on its physiology be understoodin some detail. Also, new methods fororganizing multicellular systems into 2-D and3-D architectures will be required. Theinvention of new biocompatible polymers andsurfaces will be a major part of efforts in thisdirection.

4.4 Artificial Cells. Given the limitations ofliving cells with respect to robustness, theneed for continuous nourishment, and otherenvironmental demands, the notion ofcreating synthetic assemblies that performspecific complex functions is particularlyattractive. Already, vesicle-based systems havebeen designed that harvest light to generateATP (Section 5.2). The essential features of thesystem include a synthetic chromophore that,upon light absorption, creates a protongradient across the membrane. Whenembedded in the vesicle membrane the ATPsynthase, isolated from chloroplasts, uses theproton motive force to synthesize ATP fromADP and inorganic phosphate. This self-contained energy transducing system might beincorporated into a higher-order assembly todrive ATP-dependent chemical or physicalprocesses.

In addition to vesicles, supported/tetheredbilayers can be used to organize functionalassemblies. Membrane proteins can beembedded in these structures, where they

maintain their normal receptor or transporterproperties. For example, translocation of ionsacross these bilayers can be accomplished bythe action of an ATP-dependent ion pumpembedded in the membrane. In principle, theresulting ion gradient could be used to drivecell-like processes.

4.5 Model System – Cell-Based Biosensors.Cells have shown great utility as “factories”for important products, both naturallyoccurring and designed. They also performspecific functions, many of which are ofinterest in the physical sciences. One suchfunction allows their use as biosensors.Successful defense against chemical andbiological warfare hinges on the ability todetect toxic substances (noxious gases,biological toxins and pathogenic organisms)rapidly at distant sites. The ability to analyzefood and beverage samples prior to popularconsumption is a major component of qualitycontrol and can minimize the occurrence ofpublic health crises. The potential of cell-basedbiosensors is enormous when one considersthat cells are naturally endowed with multiple,complex analyte recognition and signalamplification mechanisms that could beexploited for this purpose. Even at this earlystage of development, several cell-basedbiosensors have been described. Many of theearliest examples utilized relatively robustbacterial cells as producers of enzymes capableof converting the analyte of interest to adetectable species.

More recently, cultured animal cells have beenrecruited for advanced sensor devices. Theirdiverse repertoire of recognition systems makethem uniquely capable of detecting myriadsubstances, including bacteria and viruses.Their signal amplification mechanisms includeelectrochemical, enzymatic and morphologicalcascades, offering the opportunity to interfaceanimal cells with a range of transducingdevices. As an example, mast cells have beenincorporated into an antigen-detecting device.The mast cells degranulate in response toantigen binding at the cell surface, releasing

3/20/03

36

heat into the environment that is detected in acalorimeter. Single neurons have been used todetect the presence of the neurotransmitteracetylcholine that, upon receptor binding,triggers a detectable change in membraneconductance and intracellular calcium levels.

Despite these achievements, the developmentof cell-based biosensors still presentssignificant design challenges. Theserequirements are reflected in all three areas ofresearch identified by the panel in this section.One involves the engineering of the cellularmetabolic machinery, for example modifyingexisting cellular amplification, signaltransduction or receptor systems. Althoughcells have numerous endogenous analytedetection/signal transduction pathways thatcan be exploited in a biosensing device, thereare many circumstances where the cell’snatural machinery does not suffice. One mustalso engineer the cell surface to adapt the cellto bind the particular target of interest.Existing cellular receptors may need to bemodified or new ones incorporated in themembrane to allow the cell to serve as a sensorfor organisms or molecules it does notnaturally respond to. A detailedunderstanding of cellular signaling pathwaysand receptor-ligand interactions andspecificity is crucial for this.

Second, these systems also require theengineering of the interface between the cellsand the microelectronics of the device. This isof paramount importance, in that the sensorresponse time and sensitivity are oftendependent on the intimacy of cell-transducerinteractions. In addition, animal cells oftenhave stringent adhesion requirements foroptimal viability. Bacterial cell-basedbiosensors have been constructed byimmobilization of cells through polymerencapsulation, non-specific adsorption at asurface and physical containment within asemi-permeable membrane. These simplemechanisms are not readily imported toanimal cell-based biosensors since animal cellsare orders of magnitude more sensitive thanmicrobial cells to their local environment.Consequently, significant effort has beendirected to new strategies for controlling theattachment of animal cells to syntheticmatrices with molecular level definition. (Thissame issue arises during attempts to integratecells into other artificial environments, such asbioreactors and bioremediation filters.)Finally, there is the challenge of isolating thesensing/signal transduction machinery of thecell from non-essential components in an“artificial cell” to develop a far simpler, andtherefore more easily managed and supportedsystem.

3/20/03

37

5. Biomolecular Functional Systems

5.1 Introduction. Discussions ofnanotechnology most often focus onminiaturized circuits for advanced computers.Discussions of biotechnology most often focuson metabolism, regulation, biochemicalreactions and the production of smallmolecules, proteins, nucleic acids, reagents orpharmaceuticals. Discussions of living cellsmost often focus on systems producingmetabolic products, degrading materials,growing and duplicating. Less thought has,until recently, been given to the fact thatwithin the nanometer-size biological living celllie complex structures that perform verysophisticated mechanical or energytransduction functions. Some capture light andconvert it to chemical energy. Others transportsubstances across otherwise impermeablemembranes. Others act as molecular trucks,moving structures and molecules from placeto place within the cell. Others act as motors,rotors, ratchets, amplifiers, switches and otherdevices that perform the sorts of functions wenormally associate with meter-size machinesrun by gasoline, oil or electric engines. There isincreasing interest in these biological devicesand increasing awareness that they couldserve or be mimicked to meet needs forfunction either at the nanoscale level, or, inlarge-scale arrays, at the macro-scale level.

Four broadly drawn biological functions weresingled for fruitful exploration of theirpotential impact on the physical sciences.

• Energy Transducing Membranes andProcesses. Perhaps the most well knownand best studied of these are the lightharvesting systems in photosynthetic andrelated organisms. Photons are capturedand focused to provide energy for cellularfunction. Another energy transductionsystem, oxidative phosphorylation,

generates cellular energy through thereoxidation of electron carriers that arereduced in cellular metabolism. In mostcases the energy is captured and used inits chemical form, ATP.

• Motors, Rotors, Ratchets, andSwitches. Molecules such as kinesin andmyosin are able to convert chemicalenergy into motion, as they transportother structures from place to place withinthe cell along “rails” of microtubules andactin, respectively. Bacterial flagella rotateand propel the organism through fluids,either directly towards an identified goalor randomly in search of a target if none isin “sight”. Other biological molecules,such as enzymes, respond to signals asswitches, activating or inactivatingthemselves or other molecules. Ratchetsallow controlled motion in a singledirection.

• Enzymes. These proteins (along withcatalytic RNAs) accelerate reactions by upto 16 orders of magnitude, converting onespecifically defined molecule into anotherequally well-defined product. Theyfunction through selective binding to theirsubstrates, and then, by creating theirown chemistry, lowering the activationenergy along the path to products.

• Pores, Gates, and Channels. Thesemembrane structures allow the controlledspecific passage of ions or molecules intoor out of cells or organelles. Often theseproteins are highly regulated, respondingto the presence of other molecules (ligand-gated) or to a potential (voltage-gated).Often, they can use energy to concentratespecies against a gradient.

3/20/03

38

5.2 Energy Transducing Membranes andProcesses.

“On the arid lands there will spring upindustrial colonies without smoke andwithout smokestacks; forests of glasstubes will extend over the plains andglass buildings will rise everywhere;inside of these will take place thephotochemical processes that hithertohave been the guarded secret of theplants, but that will have beenmastered by human industry whichwill know how to make them beareven more abundant fruit than nature,for nature is not in a hurry andmankind is.”

– Giacomo Ciamician 1912

Aerobic cells contain mitochondria or activemembranes which pump protons across alipid bilayer and establish the proton motiveforce (PMF). To accomplish this, they use thefree energy made available by thespontaneous reoxidation by oxygen of thereduced redox carriers generated by themetabolic oxidation of nutrients. Anaerobiccells use ATP to pump protons for the PMF.Photosynthetic membranes in chloroplastsexploit the capture and concentration of lightenergy as the energy source to produce thePMF. Halobacteria, living in salt ponds, uselight energy to establish a PMF independent ofphotosynthesis. However created, the PMFitself serves as an energy source, driving,through a variety of mechanisms, thesynthesis of carbohydrates and othermolecules; the production of chemical energy;the transport of sugars and other nutrients; therotation of the flagellum, the cellular propeller;and other energy requiring cellular processes.Most often, this is achieved through theproduction of ATP, produced when thephotons flow back across the membranethrough the ATP synthase, a 12 nm diameterenzyme, a motor which rotates at 8000 rpmgenerating 80-1000 pN•nm of rotary torque(Noji et al., 1997) and catalyzes the formationof the phospho-anhydride bond linking ADP

to Pi.

The mimicking of the capture and use of solarenergy through systems like thephotosynthetic centers or the halobacterial“purple membranes” has been a target ofresearch for years. Similar efforts have focusedon energy capture in metabolic oxidations thatoccur in aerobic cells or PMF formation inanaerobes. The primary challenge is to link theenergy capturing system, whether it be opticalor chemical, with the energy requiringfunction.

The two forms of cellular energy, the chemicaland voltage potential in the PMF and the“activated” molecule ATP are in effectinterchangeable, in that either can beharnessed to produce the other.

The PMF acts across membranes, ATP acts insolution, thus their applicability to artificialsystems will depend on the particular needs ofthe system. One research direction involvesthe production of a “synthetic” PMF inartificial membranes that can be used as anenergy source to power a variety of energyrequiring processes of interest that havethemselves been inserted into thosemembranes. These processes could include,for example, molecular motors, transporters,separation systems, pumps, biochemicalreactions. In some cases the PMF could belinked directly to the process, in other cases itcould be linked via the production andtemporary storage of ATP. A considerablechallenge lies in the isolation and use of thenatural photosynthetic antennae or in thesynthesis of artificial mimics of that system. Inone example, highly branchedphenylacetylene dendrimers have beensynthesized and shown to act, with very highefficiency as photon energy traps (Shortreed etal., 1997). In another example, an artificialphotosynthetic reaction center of lightharvesting pigments was used. Capturedphotons were used to pump protons across themembranes of synthesized liposomes (hollowspherical lipid vesicles of predetermined size).

3/20/03

39

A PMF equivalent to that found in naturalphotosynthetic membranes was produced(Steinberg-Yfrach et al., 1997; Steinberg-Yfrachet al., 1998; Gust et al., 2001). and successfullylinked to ATP synthase molecules embeddedin the membrane. The 150nm diameterliposome, was manufactured through self-assembly from the known membranecomponents, phosphatidyl choline,phosphatidic acid, cholesterol and the ATPsynthase was isolated from spinachchloroplasts (Figure 21).

These systems represent major advances, butare still quite rudimentary. Many barriersremain. Of primary importance is the need toincrease the rate of proton transport andimprove the robustness of the system. Theartificial reaction center would also performbetter if it could be inserted into planarmembrane systems rather than the sphericalliposomes so that a wider range of transducerscould be powered and explored. Artificialpumps for ions other than protons should bedeveloped to broaden the applicability of thesystem, since the difference in chemicalpotential of any ion across a membranerepresents an energized membrane that can becoupled to a transducer to perform work.Nature for example, also uses sodium to driveATP synthase, and sodium and potassium

pumps are critical for a variety of otherbiological functions. Success will lead toenergy capture systems that are nonpollutingand do not rely on fossil energy.

5.3 Motors, Rotors, Ratchets, Switches.Design of structures for controlling transportprocesses is tightly linked to the size scaleinvolved. While sophisticated machines anddevices have been developed for themacroscopic world, new principles have to beemployed to control movement of objects atthe nanoscale. Nature again provides a model,having evolved specialized molecules toactively transport molecules and structuresover long distances (on the cellular scale), tospecific destinations and against concentrationgradients, all with very high (sometimesapproaching 100%) efficiency. The goal forbiomolecular research is to use these biologicalprinciples to create novel hybrid materials thatare capable of dynamic spatial self-organization.

Specific inspiration for this work comes fromcells that use motor proteins such as kinesin(that walk along microtubules- longer-polymerized protein filaments), to activelytransport cellular components to definedlocations. The structure of two functionalstates of kinesin, one bound to ADP and theother bound to a non-hydrolyzable analogueof ATP, has been determined at the AdvancedLight Source and the Stanford SynchrotronRadiation Laboratory, two BES user facilities(Kikkawa 2001) greatly assisting in theelucidation of the mechanism of action of thistransporter. Challenges for the construction ofmolecular shuttles using biological motorsinclude (a) guiding the direction of the motionon manufactured tracks, (b) controlling thespeed, and (c) directing the loading andunloading of specific cargo. Guiding themovement of molecular shuttles on surfaceshas been accomplished by various techniquesrelying on surface topography and chemistryas well as flow fields and electric fields (Hessand Vogel 2001). The high stiffness andpersistence length of microtubules, for

Figure 21. Liposome-based artificialphotosynthetic membrane. Protons pumped intothe liposome by a light-driven electron transfersystem (C-P-Q) drive the production of ATP bythe CFoF1-ATP synthase. Steinberg-Yfrach, etal., 1998.

3/20/03

40

example, enables the control of theirplacement by topographic surface features(Hess et al., 2002a) and makes possible thefabrication of microscopic “rectifiers” whichsort microtubules according to their directionof motion (Hiratsuka et al., 2000). Approachesto loading incorporate biotinylated tubulininto microtubules which then bindsspecifically to streptavidin-coated cargo (Hesset al., 2001). The cargo itself could bemolecules or micro- or nanofabricated devices.The sliding of myosin along actin fibersrepresents another type of motor, allowing themovement of structures attached to theseproteins. Rates of motion vary considerably,from tenths to tens of micrometers/second.Mimicks have been made (Jiménez 2000).

The ATP synthase, discussed above in thecontext of converting solar energy to chemicalenergy can also be considered a molecularmotor. Consistent with the law of“microscopic reversibility”, addition of ATPwill drive the rotation of the system, as ithydrolyzes ATP to produce ADP and Pi (Nojiet al., 1997). This chemically driven rotatingmotor has great potential and has also beenthe focus of research. Through a combinationof intra-biomolecular and bio-inorganicbonding strategies, these motor proteins havebeen self-assembled with 15 nm precision atspecific locations on inorganic substratespatterned using electron beam lithography(Montemagno and Bachand 1999). Peptidelinkers having an affinity for nickel wereattached to the synthase molecules which werethen placed on 200nm high nickel posts. 750-1400nm-long nickel “propellers” were alsoattached (Figure 1, page 12) to the rotors.Using conventional optical microscopy,rotation at speeds up to 7 Hz was observedfollowing the addition of ATP (Soong et al.,2000). Alternatively, fluorescently labeledactin filaments have been attached and theirrotation followed in fluorescence microscopy(Noji 1998).

The molecular motors based on the ATPsynthase discussed above are also switches,

with the capability of being turned on or off.Since they require input energy in the form ofATP to function, control and switching hasbeen achieved through the control of theconcentration of ATP available to the motor.Other artificial systems with motors have beenconstructed and turned on and off by lightthrough photo-induced release of caged ATP(Hess et al., 2001). Switching has also beenaccomplished in the artificial system in whichactin rods were attached to the rotatingsubunits of the ATP synthase and ATP addedor removed. This type of control has itsdrawbacks, however. Any control mechanismsshould be independent of regulation by ATP ifused in a cellular context so that any othercellular ATP-dependent processes will beunaffected.

Some recent work with the synthase illustratesthe power of existing biochemical techniquesto solve problems such as this. Geneticmodification of the ATP synthase has allowedthe engineering of pockets in the enzyme withhigh affinity for binding zinc ions. When thezinc binds the protein, it becomes incapable ofperforming the conformational changesnecessary for ATP hydrolysis and theaccompanying rotation of its central shaft. Itthus stops turning. Operation of this mutantmotor protein demonstrated switching as zincwas supplied or removed (Figure 22). With afluorescent actin filament attached to the rotorallowing single-molecule observation, rotationproceeded following the addition of ATP but

0

5000

10000

15000

20000

25000

0 20 40 60 80

Time(sec)

An

gle

(d

eg

)

Figure 5-4: Controllable activity of F1-ATPase. The motor ceasesrotation upon addition of Zn, but restarts when it is removed.Figure 22. Controllable activity of F1-ATPase.

The motor ceases rotation upon addition ofZn, but restarts when it is removed. Liu et al.,2002.

3/20/03

41

stopped in the presence of Zn. The motionrestarted following addition of a Zn chelatorand ATP.

Adhesion proteins found in the extracellularmatrix of mammalian cells or on the surfacesof bacteria are also of interest. Some are able toswitch their function when mechanicallystretched (Vogel et al., 2001). The bacterial

adhesion protein fibronectin, for example,switches its ligand affinity from low to high ifstretched under shear flow (Thomas et al.,2002; Figure 23).

The fibronectin system (see also Section 4.3)also provides a glimpse of the value ofcomputational tools to the study of proteinfunction. Models have been constructed tounderstand how mechanical unfolding ofthese proteins may affect their structure andthus function (Craig et al., 2001; Krammer etal., 1999). It was found, for example, that the

fibronectin module containing the cell bindingsite is among the first to unravel if stretched.Further, the mechanical stability of thesemodules can be altered several fold by singleamino acid substitutions suggesting that“engineering” of these structures for particularfunctions is possible.

Enzymes, discussed in more detail below ascatalysts, also act effectively as switches,responding to the environment, reversiblyconverting from active to inactive shapes.Binding to activator molecules alters theirconformation to a structure that is capable ofbinding substrate and converting it to product.Binding to inhibitors alters their conformationto a structure that is incapable of bindingsubstrate or incapable of converting it toproduct. In fact, regulation is not all ornothing, but can be graded from no activity tofull activity. Further, these switches canrespond simultaneously to many activatorsand inhibitors, summing the strength of eachof these signals to arrive at an appropriatelevel of activity. Enzymes also play a role inswitching other functions by controlling theproduction of actuators of those functions.

Finally, switching is also seen in the control ofthe use of genetic information. Transcriptionfactors can bind to DNA to regulate the degreeto which the information in the base sequenceis used to produce protein. Again, acontinuous spectrum of control is exercised,between completely repressed to fully active.

These are all processes that can be envisionedas components of nanoscale machinesproviding energy transduction, motion,catalysis or transport, in much the samefashion seen in the various components of themacroscale machines of today. Of particularnote is that most biological systems achievetheir exceptional level of efficiency byincorporating multiple functions into a singlestructure. This scheme could serve well as amodel for artificial systems. Again, however,we encounter the same barriers as before: ourincomplete knowledge of the structure and

Figure 23. Sheer forces cause a conformationalchange in the bacterial adhesin (the lectindomain of FimH) of E. coli. Mechanical forceseparates the terminal b strand purple balls)connecting the adhesin to the bacterium. Thisin turn causes a structural perturbation thatpropagates to the binding site (gold balls)switching it from low to high affinity. Thomas etal., 2002.

3/20/03

42

function of the biological systems, difficultiesinherent in isolating and then combining theircomponents in a self-assembling, controlledand functional manner, the need to make themmore robust.

Many challenges remain before these motorsystems can be incorporated into hybridsystems. As before, they must be made morerobust. Their complexity suggests difficulty inlarge scale manufacture, but they do self-assemble and, in fact, can easily be producedin large amounts through large-scale growthof the appropriate cells followed by efficientisolation and purification.

From an engineering perspective, motorproteins are an attractive alternative to MEMSactuated devices because of their small sizes(length scales hundreds of times smaller thanthose of equivalent MEMS devices), highspeeds, high efficiencies, long lifetimes, andease of mass-production. Hybrid combinationsof biological and inorganic components infabricated micro- and nanodevices promiseunique advantage. In fact a motor-proteinbased device, using motor proteins aspicoNewton force meters has been reported(Hess et al., 2002b).

5.4 Enzymes. Enzymes have long beenregarded as the highest form of biologicalfunctional achievement, although perhapsonly because the roles of other, more complexsystems have yet to be elucidated. Regardless,these proteins are able to select a singlespecific substrate from a mix of thousands,interact with it chemically, or just structurally,and convert it to a specific product even if thelaws of chemistry allow the production ofseveral. Enzymes do not alter thethermodynamics of the process: the freeenergy change and equilibrium constant areunaltered, but these nanoscale objects canaccelerate reactions by up to 16 orders ofmagnitude, equivalent to taking a reaction thattakes 300 years and performing it in amicrosecond. Some enzymes are “perfect, theyaccelerate reactions to a rate that is limited

only by diffusion.

Further, enzymes can be regulated (seeswitches, Section 5.3). Binding of specificinhibitors or activators can increase ordecrease their catalytic effectiveness by ordersof magnitude. In fact, in many cases a givenenzyme responds to multiple regulatoryeffectors. Glutamine synthetase, for example,responds to nine, in effect summing thepositive and negative effects to determine alevel of activity consistent with the balance ofthose effectors. This regulatory feature canalso be used as a switch, as described above.

The difficulty inherent in the use of enzymesin devices or in materials synthesis is thatnature designed only those enzymes requiredto catalyze the reactions necessary for life.Their specificity for substrate and productthus limits their usefulness in non-livingapplications.

The revolution in biology and biotechnologyover the past few decades including thedevelopment of cloning, genetic and proteinengineering, the deciphering of completegenomes, has opened the door to ourovercoming that barrier. With help from otherfields, for example, new tools such as theDOE/BES-managed synchrotron light sourcesand advanced nuclear magnetic resonancespectroscopy, the effort and time required todetermine the structure of an enzyme has beengreatly reduced, allowing us to understand itsfunction, a basic element in our ability to tailorit to our needs. We can now use techniquessuch as cloning and mutagenesis to alterenzymes so that they catalyze reactions ofinterest. Active sites are being redesigned,although progress is slow and this is by nomeans a routine procedure. Progress ishampered by the fact that enzymes changetheir conformation, both before and duringcatalysis, so there is no “one” structure, witheasily definable effects of a single change atpoint A on structure or function at point B. Inaddition recent results of “single-molecules”studies show that the activity of individual

3/20/03

43

enzyme molecules is quite different from whatwe see in ensemble studies (Lu et al., 1998).

Of great interest is the work in introducingnon-natural amino acids into proteins atdesired positions (Wang et al., 2001; Van Hestet al., 2000). This allows the incorporation intothe protein of a far broader set of optical,electronic and chemical capabilities, withfluorescent, photoactivatable groups, orgroups that introduce new chemistries. It alsoallows efforts to increase the stability ofenzymes, a critical need if they are to be usedoutside the cell. [Of interest here is the factthat this technique, originally thought to be asolely human redesign of the natural system,using a “stop” codon to introduce the non-natural amino acid, has recently beenobserved in a rare methane producingbacterium organism (Srinivasan et al., 2002;Hao et al., 2002; Böck et al., 1991). Thissynthesis work has progressed to the pointwhere the normal three-letter nucleic acidcode has been substituted for by a four basecode, allowing a much broader range ofenzymes to be produced with a broader rangeof non-natural amino acids, with multiplesubstitutions in the same protein (Maglieri etal., 2001). Other efforts involve thedevelopment of artificial enzymes, generatedthrough the selection of antibodies raisedagainst carefully designed antigens, and thenfine tuned through “molecular evolution.”Non-protein enzymes, designed to attract andbind the substrate, subject them to chemicalchange, and then repel and release the productare being explored (Fréchet 2002).

5.5 Pores, Gates, Channels. Proteins alsofunction as constituents of biologicalmembranes in roles such as receptors,transporters and channels. While a great dealof work has been done on the engineering ofsoluble proteins, notably enzymes andantibodies, relatively little work has been doneon the engineering of membrane proteins, inpart because of the complexity ofmanipulating systems that are not soluble inwater.

The membrane pore-forming protein a-hemolysin from Staphylococcus aureus hashowever been explored with great success inattempts both to push the frontiers of proteinengineering and to develop new componentsfor use in biotechnology, an excellent exampleof the tight link between basic science andtarget directed science (Bayley and Cremer2001; Figure 24) and of the manipulation ofbiomolecular materials for in vitroapplications.

a-Hemolysin is a 293 amino acid polypeptidethat is secreted by the bacterium and thenassembled into heptameric pores in the lipidbilayers. The crystal structure of the heptameris known and has provided an extremelyuseful template for designing engineeredforms of the protein. For example, work hasprogressed on the re-engineering of this pore-forming protein as a sensor for a wide varietyof analytes: divalent metal cations, variousanions, organic molecules, proteins andnucleic acids (Bayley et al., 2001).

Analyte molecules modulate the ionic currentdriven through a-hemolysin pores by thetransmembrane potential. Stochastic sensing,which uses currents from single engineeredpores, is an especially attractive prospect. Thisapproach yields both the concentration andidentity of an analyte, the latter from itsdistinctive current signature. Further, several

Figure 24. Molecular graphics representationlooking into the channel of the a hemolysinpore. Song et al., 1996.

3/20/03

44

analytes can be detected simultaneously witha single sensor element. The proteinmodification work has exploited recentadvances in protein engineering that allow thealteration of entire protein domains by geneticengineering, and specific but drastic localizedsites targeted chemical modification. The latterincludes chemical modifications of the internalsurface of the pore: for example, theattachment of poly(ethylene glycol) and itsderivatives. It also involved a new twist inprotein modification: non-covalentmodification with adapters, includingcyclodextrins, which continue to exhibit host-guest chemistry when lodged inside the pore.In an extension of the adapter concept, thepore has been genetically engineered toaccommodate two or three adapters. Further,small molecules can be trapped in the“nanocavities” between the adapters.

As in other areas discussed, a major remainingchallenge is to place the engineered pores in arobust environment to form components thatcan be used in practicable sensing devices.Possibilities include supported bilayers,bilayers across nanofabricated apertures andpolymerized bilayers. A second challenge is tomake arrays containing the protein for use insensors or in high-throughput screening.Stochastic sensing devices based on alternativesingle molecule detection techniques can beenvisioned, requiring physicists to join the

groups or teams of chemists and biologistsworking on these proteins. These techniquesinclude approaches based on fluorescence andforce measurements. Because the sensorelements are of nanometer dimensions, it willbe possible to build them into complex devicesincorporating other components discussed inthis report.

As was the case with the use of enzymes andpores, protein engineering will continue tocontribute to numerous areas of biomolecularmaterials. Advances in our ability to make“designer” protein-based materials such ascrystals, fibers, elastomers and adhesives canbe expected.

An alternative approach to the construction ofpores and channels is reflected in the work ofGhadiri and co-workers (Ghadiri et al., 1993).Hollow organic nanotubes hundreds ofnanometers in length, with internal diametersof 0.7-0.8 nm, were self-assembled fromrationally designed cyclic octa-peptides ofalternating D- and L-amino acids. These caninsert in natural or artificial membranes in asequence-dependent fashion, allowingselective transport across those membranes(Fernandez-Lopez et al., 2001). Potentialapplications in catalysis, molecular electronics,molecular separation technology can beenvisioned.

3/20/03

45

6. Promise and Challenges

There is a great deal of excitement around theworld about the application of biology to thephysical sciences, in particular, the materialsand chemical sciences. First, as outlined inSection 2, the world of biology holds muchpromise for applications outside the organism.The list of molecules, structures, properties,concepts of biology that could have an impactis a long one. Second, this excitement stemsfrom the fact that the dimensions accessible tothe physical sciences have progressivelyshortened and now extend to the nanometerrange, while those accessible to the biologicaland organic chemical sciences haveprogressively lengthened to that same range.Finally, there is excitement because the tools ofmolecular biology and biochemistry nowallow a fuller understanding and a start atmanipulation of biological systems.

There are, as has been discussed above, anumber of obstacles that stand between us andthe ready application of biology to thematerials and chemical sciences. The primarybarrier remains our incomplete understandingof biological systems themselves. Until asystem is fully understood, with a cleardefinition of its component molecules andhow they interact to create function, it is verydifficult to manipulate. Despite the greatadvances in biology over the past few decades,an enormous amount remains beyond ourgrasp, not only in the description of specificsystems but, and perhaps more importantly, inthe use of those systems to develop the generalunderlying principles. Exploration of theseareas must be a high priority.

Tool development is another area of challenge.It has been argued that the major advances inresearch depend on the development of tools.Clearly the impact of synchrotron radiation onthe field of structural biology has been

enormous and will continue to be so for quitesome time. The tools developed as part of theDOE and NIH led Human Genome Projectenhanced our ability to explore genes andproteins beyond what might have beenimagined. The tools that will be developed inthis new age of proteomics will be equallyvaluable. This too is an area that must be givenhigh priority.

Unlike physics and chemistry, biology has not,until relatively recently, had as much supportfrom theory, computation and modeling as itshould. However, it is clear that thecomplexity of metabolic processes and theorganization of the cell is such that thisapproach could bring great new insights.Theory does support efforts to understandprotein folding; it needs to be extended to thestudies of the structure of other molecules.Theory has also, for example, been applied tomodel simplified systems of microtubules andmotor complexes (Lipowsky 2001). The abilityto design structures for self assembly ortemplated or hierarchical assembly would alsobenefit greatly. The support of theory forbiology must also be a priority.

Cultural issues were also seen to impedeprogress in the field. As with mostinterdisciplinary fields, differences interminology and jargon must be overcome.Biology has traditionally been a field ofdescription, answering the question “whatmolecular components are involved and howdo they interact to create the function beingstudied.” The interest and skills of thesynthetic chemists and physicists must bebrought to bear, so that biologists focus moreattention on synthesis and manipulation. Atthe same time, physics and chemistry studentsmust receive more training in biology. As withany field, extensive study is a prerequisite for

3/20/03

46

grasping the required intuitive sense of anarea of study. Whether this is done through achange in the course and researchrequirements for doctoral degrees or throughthe fostering of more interdisciplinarycollaborations will depend on the specificresearch in question. Clearly, however, thetraining of the next generation of scientistsmust be broadened, perhaps through jointsupervision by mentors in different fields.

Finally, there are properties inherent inbiological systems that appear to limit theirapplication outside the living organisms.Perhaps most important is the fact that theyare not as robust as most inorganic structures.Compared to metals, ceramics, polymers,biological systems cannot withstand extremesof temperature, pH, ionic strength, pressure.In fact, in many cases it is extremely difficultto create systems in which they can functionoutside the living organism, under anyconditions. Of further concern is themaintenance of activity when biologicalstructures or cells are linked to inorganics indevices. On the other hand, systems could bedeveloped to shield and protect biological

structures, and they themselves can probably,at least to some extent, be made sturdier.Nature, after all did not design them forextremes – just for the environment the livingthings found themselves in. The ability tosubstitute new monomers in polymers, forexample could aid in this effort. On a morebasic level it is likely that our ultimateunderstanding of the fundamental physics andchemistry of biological processes might lead tothe extension of these principles to differentdoses of molecules and materials that wouldbe more stable and robust and byincorporating more of the predictable, moresuited to nonbiological needs.

These are not insuperable obstacles. In factthey are already being overcome in selectresearch groups around the world and inuniversities establishing departments, groups,or programs in “Chemical Biology” orChemisty and Biology” and new areas ofbiophysics and bioengineering. Once they aremore broadly overcome, and the rich world ofbiology becomes accessible to the physicalsciences, we will have nothing short of anotherrevolution in science.

3/20/03

47

References

Adleman, L., 1994 Science 266 1021.Aizenberg, J., A. Tkachenko, S. Weiner, L.

Addadi and G. Hendler 2001 Nature 412819.

Alivisatos, A.P.A., K.P. Johnsson, X. Peng, T.E.Wilson, C.J. Loweth, M.P. Bruchez, Jr. andP.G. Schultz 1996 Nature 382 609.

Almquist, N., N.H. Thomson, B.L. Smith, G.D.Stucky, D.E. Morse and P.K. Hansma 1999Mater. Sci. Eng. C 7 37.

Amos, W.B., 1971 Nature 299 127.Andronov, A. and J.M.J. Fréchet 2000a Chem.

Commun. 18 1701.Andronov, A., and J.M.J. Fréchet 2000b

Chemistry of Materials 12 2838.Autumn, K., M. Sitti, Y.A. Liang, A.M. Peattie,

W.R. Hansen, S. Sponberg, T.W. Kenny, R.Fearing, J.N. Israelachvili and R.J. Full 2002Proc. Nat. Acad. Sci. 99 12252.

Avouris, P., R.E. Walkup, A.R. Rossi, H.C.Akpati, P. Nordlander, T.-C. Shen, J.W.Lyding and G.C. Abeln 1996 Surface Science363 368.

Babcock, H.P., D.E. Smith, J.S. Hur, E.S.Shaqfeh and S. Chu 2000 Phys. Rev. Lett. 852018.

Ball, P., 1999 Nature 400 507.Ball, P., 2000 Nature 408 904.Ball, P., 2001 Nature 413 668.Baneyx, G., L. Baugh and V. Vogel 2001 Proc.

Nat. Acad. Sci. 98 14464.Baneyx, G., L. Baugh and V. Vogel 2002 Proc.

Natl. Acad. Sci. 99 5139.Bayley, H. and P.S. Cremer 2001 Nature 413

226.Bechert, D., as reported by P. Ball 1999.Belcher, A.M., X.H. Wu, R.J. Christensen, P.K.

Hansma, G.D. Stucky and D.E. Morse 1996Nature 381 56.

Benenson, Y., T. Paz-Elizur, R. Adar, E.Keinan, Z. Livneh and E. Shapiro 2001Nature 414 430.

Böck, A., K. Forchhammer, J. Heider, W.Leinfeld, G. Sawers, B. Veprek, and F.Zinoni, 1991 Mol. Microbiol. 5 515.

Braun, E., Y. Eichen, U. Sivan and G. Ben-Yoseph 1998 Nature 391 775.

Breiner, U., U. Krappe, V. Abetz and R. Stadler1997 Macromol. Chem. Phys., 198 1051.

Brown, S., 1992 Proc. Nat. Acad. Sci. 89 8651.Brown, S., 1997 Nature Biotechnology 15 269.Brown, S., 2001 Nano Letters 1 391.Bruehl, R.E., J.T. Groves and C.R. Bertozzi

2002 JACS in prep.Brust, M., D. Bethell, D.J. Schiffrin and C.J.

Kiely 1995 Adv. Mater. 7 795.Cacialli, F., J.S. Wilson, J.J. Michaels, C. Daniel,

C. Silva, R.H. Friend, N. Severin, P.Samaori, J.P. Rage, M.J. O’Connell, P.N,Taylor and H.L. Anderson 2002 NatureMaterials 1 160.

Cha, J.N., 1999 Proc. Nat. Acad. Sci. U.S.A. 96361.

Chen, J. and N.C. Seeman 1991 Nature 350 631.Cheng, Ji-Xin, L.D. Book and X.-S. Xie 2001

Optics Lett. 26 1341.Ciamician, G. 1912 Science 36 385.Clark, T.D., J. Tien, D.C. Duffy, K.E. Paul and

G.M. Whitesides 2001 J. Am. Chem. Soc. 1237677.

Colvin, V.L., N. Goldstein and A.P. Alivisatos1992 J. Am Chem. Soc. 144 5221.

Craig, D., A. Krammer, K. Schulten and V.Vogel 2001 Proc. Nat. Acad. Sci. 98 5590.

Curry, J.D., 1977 Proc. R. Soc. Lond. B. 196 443.Demers, L.M., D.S. Ginger, S.-J. Park, Z. Li, S.-

W. Chong and C.A. Mirkin 2002 Science 2961836.

Eigler, D.M. and E.K. Schweizer 1990 Nature344 524.

Elghanian, R., J.J. Storhoff, R.C. Mucic, R.L.Letsinger and C.A. Mirkin 1997 Science 2771078.

Falini, G., S. Albeck, S. Weiner and L. Addadi1996 Science 271 67.

Fernandez-Lopez, S., H.-S. Kim, E.C. Choi, M.Delgado, J.R. Granja, A. Khasanov, K.Kraehenbuehl, G.!Long, D. Weinberger,K.M. Wilcosen and M.R. Ghadiri 2001Nature 412 452.

Finnefrock, A. C., R. Ulrich, A. Du Chesne, C.C. Honeker, K. Schumacher, K. K. Unger, S.M. Gruner and U. Wiesner 2001 Angew.Chem. Int. Ed. 40 1208 113 1248.

Fréchet, J.M.J., 2002 Proc. Nat. Acad. Sci. 994782.

3/20/03

48

Ghadiri, M.R., J.R. Granja, R.A. Milligan, D.E.McRee and N. Khazanovich 1993 Nature366 324.

Gracias, D.H., J. Tien, T.L. Breen, C. Hsu andG.M. Whitesides 2000 Science 289 1170.

Groves, J. T., L.K. Mahal and C.R. Bertozzi2001 Langmuir 17 5129.

Gulbis, J.M., M. Zhou, S. Mann and R.MacKinnon 2000 Science 289 123.

Guo P., S. Grimes and D. Anderson 1986 Proc.Nat. Acad. Sci. 83 3505.

Gust, D., T. A. Moore and A. L. Moore 2001Accounts of Chemical Research 34 40.

Hao, B., W. Gong, T.K. Ferguson, C.M. James,J.A. Krzycki and M.K. Chan 2002 Science296 1462.

Hawker, C.J. and J.M.J. Fréchet 1992 Polymer33 1507.

Henry, C.M., 2001 Chemical & EngineeringNews 35 Cover Story.

Heslot, H., 1998 Biochemie (Paris) 80 19.Hess, H. and V. Vogel 2001 Reviews in

Molecular Biotechnology 82 67.Hess, H., J. Clemmens, D. Qin, J. Howard and

V. Vogel 2001 Nanoletters 1 235.Hess, H., J. Howard and V. Vogel 2002a

Nanoletters 2 1113.Hess, H., J., Clemmens, C.M. Matzke, G.D.

Bachand, B.C. Bunker and V. Vogel 2002bApplied Phys. A, 75 309.

Hinman, M., Z. Dong, M. Xu, and R.V. Lewis1993 in Biomolecular Materials (eds., Viney,C. et al) Mats. Res. Soc. 25.

Hiratsuka, Y., T. Tada, K. Oiwa, T. Kanayamaand T.Q.P. Uyeda 2001 Biophys. J. 81 1555and references therein.

Hiratsuka, Y., T. Tada, K. Oiwa, T. Kanayamaand T.Q.P. Uyeda 2000 In t e rnat iona lMicroprocesses and NanotechnologyConference, Digest of Papers Microprocessesand Nanotechnology, Japan Soc. Appl. Phys.296.

Jackson, A.P. 1988 Proc. R. Soc. Lond. B. 234415.

Jiménez, M.C., C. Dietrich-Buchecker and J.-P.Sauvage 2000 Angew. Chem. Intl. Ed. 393284.

Kiick, K. L., E. Saxon, D.A. Tirrell and C.R.Bertozzi 2002 Proc. Natl. Acad. Sci. U.S.A. 9919.

Kikkawa, M., E.P. Sablin, Y. Okada, H. Yajima,R.J. Fletterick and N. Hirokawa 2001 Nature411 439.

Klaus, T., R. Joerger, E. Olsson and C.-G.Granqvist 1999 Proc. Nat. Acad. Sci. 9 613611.

Koltover, I., K. Wagner and C. R. Safinya 2000Proc. Natl. Acad. Sci. 97 14046.

Koltover, I., T. Salditt, J.O. Raedler and C. R.Safinya 1998 Science 281 78.

Krammer, A., D. Craig, W. Thomas, K.Schulten and V. Vogel 2002 Matrix Biology21 139.

Krammer, A., H. Lu, B. Isralewitz, K. Schultenand V. Vogel 1999 Proc. Nat. Acad. Sci.U.S.A 96 1351.

Krauss, R., H.-G. Weinig, M. Seydack, J.Gendig and U. Koert 2000 Angew. Chem. Int.Ed. 39 1835.

Lazaris, A., S. Arcidiacono, Y. Huang, J.-F.Zhou, F. Duguay, N. Chretien, E.Z. Welsh,J.W. Soares and C.N. Karatzas 2002 Science295 472.

Lee, S.W., C. Mao, C. Flynn and A.M. Belcher2002 Science 296 892.

Li M., K.K.W. Wong and S. Mann 1999 Chem.Mater. 11 23.

Liphardt, J., B. Onoa, S.B. Smith, I. Tinoco, Jr.and C. Bustamante 2001 Science 295 733.

Lipowsky R., 2001 Phys. Rev. Lett. 87 pro8101.Liu, H., J. Schmidt, G. Bachand, S. Rizk, L.

Looger, H. Hellinga and C. Montemagno2002 Nature Materials 1 173.

Loweth, C.J., W.B. Caldwell, X.G. Peng, A.P.Alivisatos and P.G. Schultz 1999 Angew.Chem. Int. Ed. 38 1802.

Lu, H.P., L. Xun and X.S. Xie 1998 Science 2821887.

Maglieri, T.J., J.C. Anderson and P.G. Schultz2001 J. Mol. Biol. 307 755.

Mahadevan L. and P. Matsudaira 2000 fromMIT Science 288 95.

Mahal, L. K., K.J. Yarema and C.R. Bertozzi1997 Science 276 1125.

Mahtab, R., J.P. Rogers and C.J. Murphy 1995J. Am. Chem. Soc. 117 9099.

Mahtab, R., J.P. Rogers, C.P. Singleton and C.J.Murphy 1996 J. Am. Chem. Soc. 118 7028.

Malenfant, P.R.L., L. Groenendaal and J.M.J.Fréchet 1998 J. Am. Chem. Soc. 120 10990.

3/20/03

49

Mao, C., T.H. LaBean, J.H. Reif and N.C.Seeman 2000 Nature 407 493; Erratum:Nature 408 750.

Mao, C., W. Sun, Z. Shen and N.C. Seeman1999 Nature 397 144.

Martin, R.E. and F. Diederich 1999 Angew.Chem. Int. Ed. 38 1351.

Meldrum, F.C., S. Mann, B.R. Heywood, R.B.Frankel and D.A. Bazylinski 1993 Proc. R.Soc. London B 251 231.

Mirkin, C.A., R.L. Letsinger, R.C. Mucic andJ.J. Storhoff 1996 Nature 382 607.

Montemagno, C.D. and G.D. Bachand 1999Nanotechnology 10 225.

Moore, J.S., 1997 Acc. Chem. Res. 30, 402 andreferences therein.

Moriyama, Y., H. Okamoto and H. Asai 1999Biophys. J. 76 993.

Morse, D.E., 2001 In: The Chemistry ofOrganic Silicon Compounds Rappoportand Apeloig, eds., John Wiley & Sons 3 805.

Naik, R.R., S.J. Stringer, G. Agarwal, S.E. Jonesand M.O. Stone 2002 Nature Materials 1 169.

Nêdêlec, F.J., T. Surrey, A.C. Maggs and S.Leibler 1997 Nature 389 305.

Nelson, D.R. 2002 Nanoletters 2 1125.Noji, H., 1998 Science 282 1844.Noji, H., R. Yasuda, M. Yoshida, K. Kinosita,

Jr. 1997 Nature 386 299.Oliver, S.R.J., T.D. Clark, N. Bowden and G.J.

Whitesides 2001 J. Am. Chem. Soc. 123 8119.Perkins, T.T., D.E. Smith and S. Chu 1994

Science 6 819.Pfohl, T., J.H. Kim, M. Yasa, H.P. Miller,

G.C.L. Wong, F. Bringezu, Z. Wen, L.Wilson, Y. Li, M.W. Kim and C.R. Safinya2001 Langmuir 17 5343.

Raedler, J.O., I. Koltover, T. Salditt and C.R.Safinya 1997 Science 275 810.

Reif, J.H., 1997 Proceedings of a DIMACSWorkshop, University of Pennsylvania,Rubin and Wood eds., AmericanMathematical Society 3 217.

Robinson, B.H. and N.C. Seeman 1987 Prot.Eng. 1 295.

Rodacy, P. 2002 Sandia NationalLaboratory/New Mexico, personalcommunication

Sambles, R., 2001 Nature 412 783.Saxon, E., and C.R. Bertozzi 2000 Science 287

2007.Seeman, N.C., 1982 J. Theor. Biol. 99 237.

Seeman, N.C., 1991 DNA & Cell Biol. 10 475.Seeman, N.C., 1999 Nature 397 144.Seeman, N.C., and A.M. Belcher 2002 Proc.

Nat. Acad. Sci. 10 1073.Shortreed, M.R., S.F. Swallen, Z.-Y. Shi, W.

Tan, Z. Xu, C. Devadoss, J.S. Moore and R.Kopelman 1997 Phys. Chem. B. 101 6318.

Simon, P.F.W., R. Ulrich, H.W. Spiess, U.Wiesner 2001 Chem. Mater. 13 3464.

Smith, B.L., T.E. Schäffer, M. Viani, J.B.Thompson, N.A. Frederick, J. Kindt, A.Belcher, G.D. Stucky, D.E. Morse and P.K.Hansma 1999 Nature 399 761.

Smith, D.E., H.P. Babcock and S. Chu 1999Science 283 1724.

Smith, D.E., S.J. Tans, S.B. Smith, S. Grimes,D.L. Anderson and C. Bustamante 2001Nature 413 748.

Song, L., M.R. Hobaugh, C. Shustak, S.Cheley, H. Bayley and J.E. Gouaux 1996Science 274 1859.

Soong, R.K., G.D. Bachand, H.P. Neves, A.G.Olkhovets, H.G. Craighead and C.D.Montemagno 2000 Science 290 1555.

Srinivasan, G., C.M. James and J.A. Krzycki2002 Science 296 1459.

Stadler, R., C. Auschra, J. Beckmann, U.Krappe, I. Voigt-Martin and L. Leibler 1995Macromolecules, 28 3080.

Steinberg-Yfrach, G., J.-L. Rigaud, E.N.Durantini, A.L. Moore, D. Gust and T.A.Moore 1998 Nature 392 479.

Steinberg-Yfrach, G., P.A. Liddell, S.-C. Hung,A.L. Moore, D. Gust and T.A. Moore 1997Nature 385 239.

Storhoff, J.J. and C. A. Mirkin 1999 Chem. Rev.99 1849.

Thomas, W.E., E. Trintchina, M. Forero, V.Vogel and E.V. Sokurenko 2002 Cell 109913.

Tour, J.M., 1996 Chem. Rev. 96, 537 andreferences therein.

Tully, D.C., K. Wilder, J.M.J. Fréchet, A.Trimble and C.F. Quate 1999 Adv. Mater. 11314.

Van Hest, J.C.M., K.L. Kiick and D.A. Tirrell2000 J. Am. Chem. Soc. 122 1282.

Vogel, V., W.E. Thomas, D.W. Craig, A.Krammer and G. Baneyx 2001 Trends inBiotech. 19 416.

Volkmer, A., Ji-xin Cheng and X. Sunney Xie2001 Phys. Rev. Lett. 87 023901

3/20/03

50

Waite, J.H. X.-X. Qin and K.J. Coyne 1998Matrix Biol. 17 93.

Wang, L., A. Brock, B. Herbrich and P.G.Schultz 2001 Science 292 498.

Whaley, S.R. and A.M. Belcher 2000 M R SProceedings 599 189.

Whaley, S.R., D.S. English, E.L. Hu, P.F.Barbara and A.M. Belcher 2000 Nature 405665.

Winfree, E., 1995 in: DNA Based Computers,Lipton and Baum eds., Am. Math. Soc., 199.

Winfree, E., F. Liu, L.A. Wenzler and N.C.Seeman 1998 Nature 394 539.

Winfree, W., 2000 J. Biomol. Struct. and Dyns.Conversation 11 263.

Winter, J.O., T.Y. Liu, B.A. Korgell and C.E.Schmidt 2001 Adv. Mat. 13 1673.

Wong, G. C. L., J.X. Tang, A. Lin, Y. Li, P.A.Janmey and C.R. Safinya 2000 Science 2882035.

Yan, H., X. Zhang, Z. Shen and N.C. Seeman2002 Nature 415 62.

Yasuda, R., H. Noji, J. Kinosita and M. Yoshida1998 Cell 93 1117.

Yin, J., E.C. Mundorff, P.L. Yang, K.U. Wendt,D. Hanway, R.C. Stevens and P.G. Schultz2001 Biochemistry 40 10764.

Yu, Q., J.M. Bauer and J.S. Moore 2001 AppliedPhysics Letters 78 2589.

Zeck, G. and P. Fromherz 2001 Proc. Nat. Acad.Sci. USA 98 10457.

Zhou, M., J.H. Morais-Cabral, S. Mann and R.MacKinnon 2001 Nature 411 657.

Zumbusch, A., G.R. Holtom and X. S. Xie 1999Phys. Rev. Lett. 82 4142.

1 Portions of Section 2 appear in a similar form in a report of another Federal agency workshop, as yet unpublished, inwhich an author (MDA) of this report was invited to participate.

2 Text similar to portions of Section 3 first appeared in “The Molecular Foundry” published by the Materials SciencesDivision of the Lawrence Berkeley National Laboratory, March 2001, written in part by MDA and APA who were alsoparticipants in this workshop.

DISCLAIMERThis document was prepared as an account of work sponsored by the United States Government. While this documentis believed to contain correct information, neither the United States Government nor any agency thereof, nor TheRegents of the University of California, nor any of their employees, makes any warranty, express or implied, or assumesany legal responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or processdisclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specificcommercial product, process, or service by its trade name, trademark, manufacturer, or otherwise, does not necessarilyconstitute or imply its endorsement, recommendation, or favoring by the United States Government or any agencythereof, or The Regents of the University of California. The views and opinions of authors expressed herein do notnecessarily state or reflect those of the United States Government or any agency thereof, or The Regents of theUniversity of California.