whitepaperforjaykeasling&jeffreydietrich,instructor&gsi,bioengc230compact,ultra-mobile,bioinfraredimagingarray(cumbia)universityof…

1、<p>　　White Paper for Jay Keasling & Jeffrey Dietrich, Instructor & GSI, BioEng C230</p><p>　　Compact, Ultra-Mobile, Bioinfrared Imaging Array (CUMBIA)</p><p>　　University of Califo

2、rnia, Berkeley</p><p>　　Sisi Chen1, Jim Cheng2, Josh Hug2, Gondica Nguyen3</p><p>　　1Department of Bioengineering, 2Department of Electrical Engineering & Computer Sciences, 3Department of C

3、hemical Engineering</p><p>　　Overview & Technical Rationale</p><p>　　Proposed is a program for work that will enable the creation of a near-infrared imaging technology (with an eventual goal

4、 of mid-to-long range IR with future advances in photoswitch molecules like azobenzene) using microorganisms to transduce infrared light between 700-950 nm to the visible region creating a monochromatic image based on in

5、tensity of absorbed radiation1. Such technology will enable a no battery, ultra low cost and highly sensitive infrared pixel technology in a very compact pack</p><p>　　A depiction of the final proposed syst

6、em is shown in Figure 1. This program will focus on 5 main tasks:</p><p>　　Synthesis/selection of infrared sensitive molecules; primarily azobenzene-based molecules which are photoisomerizable from trans to

7、 cis conformation due to near-IR.</p><p>　　Binding the photosensitive azobenzene molecule to an intermediate bridge molecule and ammonium to create the photoswitchable compound.</p><p>　　Tetheri

8、ng of the photoswitchable compound to the target protein with an amino acid; luciferase in our case.</p><p>　　Characterization and optimization of the protein’s transduction of IR into visible light.</p&g

9、t;<p>　　Development of IR imaging array by embedding the protein throughout a film (such as agar) and mounting it in the path of the focused, IR image so that it can directly render that image into the visible reg

10、ion.</p><p>　　Figure 1. System overview of bioinfrared imaging array. Current photoswitchable molecules are limited to ~900 nm which limits this imager to the NIR region (700 nm – 1 µm).</p><

11、;p>　　Near infrared (NIR) technologies often still employ semiconductor-based (ex. silicon, GaAs) photodetector technologies which often requires a NIR lamp making it an active technology and thus more power hungry. M

12、ore power conserving passive NIR technologies with photocathode and phosphor screen have longer operational lifetimes, however the high energy charge required for the photocathode, limited luminous efficiencies and large

13、r bulk of these devices puts clear boundaries on the portability of </p><p>　　The bioinfrared imaging array proposed in this research will solve many of the current downfalls of IR imagers – power efficiency

14、 while retaining a bright display, small pixel size (E. coli ~1–2 µm) and moderately high sensitivity (due to direct transduction mechanism, thermal insensitivity </p><p>　　Figure 2. Biochemical process

15、 for light production. The luciferin protein (I) reacts with ATP to form adenylated luciferin (II). After cleavage with O2 (III) and cleavage of the dioxetanone ring (IV), we get oxyluciferin (V) which is in an excited

16、state. The return to ground state releases a photon of a certain wavelength.</p><p>　　due to a direct optical absorption mechanism like rhodopsin and no electrical resistive losses). The direct coupling of

17、 the sensing to the display pixel results in a fairly high resolution sensor and display which utilizes minimal energy like the photocathode-based NIR imagers today. Luminous efficiency of luciferase-luciferin light gen

18、eration can be as high as 90%, much higher than the 30% efficiency from LEDs and 20% from fluorescent lighting2. Also, the E. coli can regenerate the oxylucife</p><p>　　Overall, such a system also proves th

19、e general concept of utilizing very highly efficient luciferase-based displays and provides a biomimetic, low-cost alternative to current NIR technologies available today.</p><p>　　General Background</p&g

20、t;<p>　　Our goal is to engineer a biological system that produces visible light in response to IR radiation. In order to respond dynamically to the input IR, the system must be capable of producing a low latency

21、pulse of light with short duration. Because the machinery associated with transcription and translation is inherently slow, we must turn to protein engineering to create photoswitchable compounds that can respond instan

22、taneously to IR input. </p><p>　　A. Luciferase-mediated light production</p><p>　　To produce visible light in a biological system, we take advantage of the natural light production process evol

23、ved by fireflies, bacteria, and many deep-sea organisms. In each of these organisms, the luciferase enzyme oxidizes a luciferin substrate to produce an intermediate product in an electronically excited state. The retur

24、n of the electron to ground state is accompanied by the emission of a photon of light in the visible range (Fig 2). Because the light production process was independent</p><p>　　Beetles, including the firef

25、ly, synthesize a variety of different luciferase enzymes that catalyze the same luciferin substrate to produce light with distinctly different wavelengths. Since this multiplicity is potentially useful for engineering m

26、any channels of light transduction, only the firefly luciferase system will be presented in detail (Fig 2). The enzyme consists of two domains: a large N-terminal domain and a smaller C-terminal domain connected by a di

27、sordered loop. The intersection</p><p>　　B. IR detection</p><p>　　The input detection in this system will take advantage of the chemical compound azobenzene, which has been characterized to cha

28、nge conformation upon absorption of light at specific wavelengths. An azobenzene molecule is composed of two phenyl groups linked by an N-N double bond. When the molecule absorbs light at a specific wavelength, it can

29、change from a trans to a cis conformation. Depending on how the molecule is functionalized, this absorption wavelength can reach near IR1 (Fig. 3). The</p><p>　　Figure 3. Structure of azobenzene with an a

30、bsorption peak between 825-950nm. </p><p>　　Researchers have been recently successful in conjugating this azobenzene molecule to potassium ion channels in neuronal cells to create a 'tethered photoswitch

31、’ (Fig. 4). Their photoswitch consists of three molecules: maleimide for tethering the entire compound to a cysteine residue on the ion channel, azobenzene, and a quaternary ammonium group to block the channel. In the

32、stable trans conformation, the tether is long enough for the ammonium group to block the pore of the channel. However, w</p><p>　　Figure 4. Azobenzene photoswitch conjugated to a potassium ion channel. At

33、rest, the azobenzene tether is extended and the tetraethylammonium (TEA) ligand blocks the pore. When the channel is exposed to UV light, the azobenzene bends into a cis conformation, opening the channel. </p>&l

34、t;p>　　Technical Challenges and Approach</p><p>　　A. Synthesis and selection of azobenzene-based molecules</p><p>　　Depending on what functional groups are attached to azobenzene molecule, th

35、e azobenzene will absorb different frequencies of light. The longest wavelength for which an azobenzene molecule has been designed is in the near IR range (Fig. 4)4. Because our project is mainly concerned with the bio

36、logical engineering of light transduction, we will not concentrate too much effort on developing chemical compounds for IR detection. For now, we will first attempt to develop the system using more conven</p><

37、;p>　　B. Creating a Photoswitch</p><p>　　Our photoswitch will be similar to the one used by Banghart et. al4. The photoswitch will contain a luciferase-azobenzene bridge, the azobenzene itself, and on the

38、 other side of the azobenzene a ligand that will block the binding pocket of luciferase. When the structure is excited by IR illumination, the azobenzene will switch configuration, freeing up the luciferase binding pocke

39、t to allow luciferin to bind, thus creating light.</p><p>　　By looking at the crystal structure of luciferase in Figure 5, we see that the active domain contains a highly conserved glutamic acid residue5. Th

40、us, it seems that using quaternary ammonium may be an acceptable choice of blocking ligand. Unfortunately, Banghart’s maleimide-azobenzeine-ammonium switch would not work in our case. Since the protein of interest in ou

41、r case, luciferase, is a cytosolic protein, the introduction of a maleimide based switch would probably cause cell death, since photo</p><p>　　Figure 5. "(Left Side) Representation of the concave molecu

42、lar surface of the large N-terminal domain of firefly luciferase, looking down onto the Yshaped system of valleys with b-sheet A on the left-hand side, b-sheet B on the right-hand side and the b-barrel at the top. The co

43、lours range from blue for the hypervariable residues, through white, to red for the three invariant residues Lys206, Glu344 and Asp422. (Right Side) Representation of the active site of firefly luciferase, viewed in a si

44、m</p><p>　　Instead, our photoswitch will be pre-bound to an amino acid and incorporated into the protein by the ribosome. For reasons to be explained later, we will not be able to use cysteine or any of the

45、standard 20 amino acids, and as a result, we will also probably need to use a molecule other than maleimide to serve as a bridge between the amino acid and the azobenzene. Thus, our photoswitch will be an amino acid-brid

46、ge-azobenzene-ammonium (AA-BR-AZO-QA) molecule. Construction of the AA-BR-AZO-QA pho</p><p>　　C. Redefining the UAG codon to code for the Photoswitch</p><p>　　In order to tell the ribosome where

47、 to place our photoswitch inside the luciferase protein, we’ll need to change the genetic code of E. Coli. By introducing amber suppressor tRNAs loaded with AA-BR-AZO-QA we will effectively transform UAG into a photoswit

48、ch codon instead of a stop codon6. In order to load our amber suppressor tRNA with AA-BR-AZO-QA we will need a tRNA/synthetase pair for the amino acid we choose. tRNA/synthetase pairs have been successfully implemented i

49、n E. Coli for at least 30</p><p>　　D. Selection of Amino Acid and Incorporating the Photoswitch into the Cell</p><p>　　We’d like to select an amino acid which is stable inside E. Coli, is taken

50、up by cellular machinery, which is capable of being linked to azobenzene by a bridge molecule , and for which a working synthetase/tRNA pair exists7. Several novel amino acids have been identified which meet our first tw

51、o criteria. The third criterion is probably an easy problem for someone with solid chemistry knowledge, and at least 30 amino acids have been identified which satisfy the fourth. </p><p>　　It is possible tha

52、t the addition of the BR-AZO-QA to the amino acid may cause these properties to change. In that case, it may be necessary to try several different amino acids for our photoswitch before we find one that meets all of our

53、criteria. In the event that we find an amino acid that is acceptable except that we cannot get it into the cell, we can proceed by using electroporation, though this will reduce the lifetime of our cells to a few hours.&

54、lt;/p><p>　　E. Selection of Photoswitch Location</p><p>　　It will be very difficult to predict ahead of time where exactly to place the UAG codon. Most likely, we will have to try a number of diffe

55、rent positions near negatively charged residues on the active site. From the crystal structure, it appears that the glutamic acid residue at position 344 is a good initial target, as it is very highly conserved, negative

56、ly charged and looks like it should be easy to access, though it would probably be a good idea to have someone with more expertise do a liter</p><p>　　F. Construction and Testing of Plasmids</p><p

57、>　　One plasmid will contain the code for a UAG-containing luciferase, as well as the luciferin regenerating enzyme (LRE) gene needed to cycle oxyluciferin back to luciferin. The other plasmid will contain the tRNA/syn

58、thetase pair. By keeping the tRNA/synthetase pair separate, we make it easier to try out new amino acid photoswitches without having to change the luciferase plasmid. Likewise, when we have to explore where to place the

59、UAG by creating a library of luciferase genes, this prevents us fr</p><p>　　Milestones & Deliverables</p><p>　　Phase 1 (June 2007-December 2008)</p><p>　　Synthesis/selection of

60、infrared sensitive molecules; primarily azobenzene-based molecules which are photoisomerizable from trans to cis conformation due to near-IR.</p><p>　　Begin research on binding the photosensitive azobenzene

61、molecule to an intermediate bridge molecule and ammonium to create the photoswitchable compound.</p><p>　　Phase 2 (January 2009-June 2010)</p><p>　　Complete photoswitchable compound with possibl

62、e alternative structures.</p><p>　　Tethering of the photoswitchable compound to the target protein with an amino acid; luciferase in our case.</p><p>　　Phase 3 (June 2010-December 2011)</p>

63、;<p>　　Characterization and optimization of the protein’s transduction of IR into visible light.</p><p>　　Development of IR imaging array by embedding the protein throughout a film (such as agar) and

64、mounting it in the path of the focused, IR image so that it can directly render that image into the visible region.</p><p><b>　　Impact</b></p><p>　　Near infrared (NIR) technology is

65、used in a variety of fields. Typical applications of NIR are used in the pharmaceutical, food, beverage and agrochemical industries for quality control and composition testing. NIR is also used as a medical diagnostics

66、 tool to test blood sugar levels and oximetry. </p><p>　　In the pharmaceutical industry, NIR can be used to monitor the degradation of the gelatin capsule multiple individual tablets. Currently, a NIR cam

67、era can analyze approximately 1300 tablets simultaneously8. Although pharmaceutical manufacturing facilities do have NIR cameras in use, the instruments do not have the sensitivity to scan the entire tablet leading to s

68、ome inaccurate readings. Increasing the resolution of the image will increase the accuracy of the testing. </p><p>　　In the food and agrochemical industries, chemical testing is used for food quality and c

69、omposition testing. Chemical methods tend to be expensive and time consuming especially for products that contain multiple components. NIR is used as a non-invasive method to analyze products quickly and reliably. <

70、;/p><p>　　More recently, NIR spectrometry in conjunction with magnetic resonance imaging (MRI) or computerized tomography (CT) scans for non-invasive scans to detect changes in blood hemoglobin. The use of the

71、se modalities combined is useful since less data points are required to create a clear image. By improving on NIR technology, the cost of building scanning devices will be reduced as well as the time required to perform

72、 the scan. NIR scans alone are not feasible at this time as these scans require</p><p>　　Given the flexibility of NIR applications, research in this field will lead to greater advances. Although NIR device

73、s currently are restricted to monochromatic images unless coupled with another device, multi-colored images are also possible. The click beetle is known for synthesizing many different types of luciferases that produce

74、different colors of light, ranging from green(546 nm) to yellow(578 nm) to orange (593 nm)9. The drastic change in emission peak can be achieved with just a singl</p><p>　　NIR devices that use azobenzene an

75、d luciferase pose very few safety concerns. Azobenzene can be converted to benzidine, a known human carcinogen, in the presence of an acid. According to the Environmental Protection Agency, azobenzene induced invasive

76、sarcomas in the abdominal cavity of rats following dietary administration10. As this is the case, care in handling and storing azobenzene should be considered. More recently, azobenzene has been used in many industrial

77、applications with little o</p><p>　　References: </p><p>　　1.Ahmed, S. A. M. Photochromism of dihydroindolizines. Part II - Synthesis and photophysical properties of new photochromic IR-sensitiv

78、e photoswitchable substituted fluorene-9 '-styrylquinolinedihydroindolizines. Journal of Physical Organic Chemistry 15, 392-402 (2002).</p><p>　　2.in Pink Tentacle (2006).</p><p>　　3.Baldw

79、in, T. O. Firefly luciferase: The structure is known, but the mystery remains. Structure 4, 223-228 (1996).</p><p>　　4.Banghart, M., Borges, K., Isacoff, E., Trauner, D. & Kramer, R. H. Light-activated

80、ion channels for remote control of neuronal firing. Nature Neuroscience 7, 1381-1386 (2004).</p><p>　　5.Conti, E., Franks, N. P. & Brick, P. Crystal structure of firefly luciferase throws light on a sup

81、erfamily of adenylate-forming enzymes. Structure 4, 287-298 (1996).</p><p>　　6.Mehl, R. A. et al. Generation of a Bacterium with a 21 Amino Acid Genetic Code. J. Am. Chem. Soc. 125, 935-939 (2003).</p>

82、;<p>　　7.Wang, L., Xie, J. & Schultz, P. G. Expanding the genetic code. Annu Rev Biophys Biomol Struct 35, 225-49 (2006).</p><p>　　8.Hamilton, S. J. & Lodder, R. A. in Biomedical Nanotechnol

83、ogy Architectures and Applications (eds. Bornhop, D. J. et al.) (Society of Photo-Instrumentation Engineers, 2002).</p><p>　　9.Wilson, T. & Hastings, J. W. Bioluminescence. Annual Review of Cell and Dev

84、elopmental Biology 14, 197-230 (1998).</p><p>　　10.IRIS. (ed. EPA, U.).</p><p>　　Budget Summary</p><p>　　PROJECT PERIOD: 1 June 2007 – 31 December 2011 (54 months)</p><p

85、>　　BUDGET SUMMARY BY MAJOR COST ITEM BY YEAR</p><p>　　Explanation of costs</p><p>　　Faculty: 1 Professor summer salary @ 50% (educated guess based on salaries publically available from univer

86、sity)</p><p>　　Research Support: 1 Grant administrator salary @ 10% (educated guess based on industry numbers on internet)</p><p>　　GSR: 3 grad students @ $3,793 EECS rate x 50% academic year an

87、d 100% summer</p><p>　　Supplies Expenses: reagents, computer networking charges, mailing, printing, faxing, publication costs, general office supplies (educated guess based on costs from assorted internet we

88、bsites)</p><p>　　Equipment: $300,000 for IR test set-up and biochem equipment for synthesis; $4,800 for laptop and computer accessories for grad students</p><p>　　Travel: 3 x $4000 for conferenc

89、es and meetings with possible collaborators, $3000 DARPA meetings</p><p>　　Employee Benefits: general tuition + health insurance costs x 3 grad students + educated guess of $3000 for professor and administra

90、tive support (based on benefits costs tables from university)</p><p>　　Overhead: 52%</p><p>　　Note: General increase in cost over years due to inflation. Accounted for by increases between 2-9%