THE RLBL NEWSLETTERrlbl.chem.upenn.edu/pdf/rlbl28.pdfported news of the progress and new developments in our Core research areas. For the current Issue we take ... Following this article

THE RLBL NEWSLETTER 1

THE RLBL NEWSLETTERRegional Laser and Biotechnology Laboratories

At the University of PennsylvaniaD I R E C T O R : P R O F . R O B I N M . H O C H S T R A S S E R

Number 28 http://rlbl.chem.upenn.edu December 2003

I N S I D E T H I S I S S U E

1 Editorial

1 Research Article: Preassembly and Ligand-Induced

Changes of the Interferon γ Receptor Complex in Cells

6

9

Research Article: Direct Observation of Triplet Emission

of Single Molecules and Phosphorescence Quenching

by Molecular Oxygen

Research Article: Two-Dimensional Infrared Measure-

ments of the Coupling between Amide Modes of an α-

Helix

12 RLBL Resources and Recent Publications

15 Application Form

EditorialBy Thomas Troxler

The previous few Issues of our Newsletter mainly re-ported news of the progress and new developments inour Core research areas. For the current Issue we takethe opportunity to introduce some of our collaborativeresearch projects. These collaborative projects repre-sent one of the three pillars of our research along withthe Core and Service activities. They use the emergingtechnologies in major research applications. At thesame time they also increase the scope of the applica-tions of advanced laser technology in biology andwiden the range of scientific and technological chal-lenges to our laboratory. Due to the variety of experi-

Collaborative ResearchReports

P R E A S S E M B L Y A N D L I G A N D -I N D U C E D C H A N G E S O F T H E

I N T E R F E R O N γ R E C E P T O R

C O M P L E X I N C E L L S

By Ch. D. Krause, E. Mei, J. Xie, Y. Jia, M. A.Bopp, R. M. Hochstrasser and S. PestkaDepartment of Molecular Genetics, Microbiology andImmunology, Robert Wood Johnson Medical School,UMDNJ, Piscataway, NJ

Department of Chemistry, University of Pennsylvania

IntroductionInterferon-γ (IFN-γ) is a protein that plays a prominentrole in the activation of the immune system. In generalinterferons initiate signal transduction through specificcell surface receptors. (1-3). IFN-γ binds to the IFN-γreceptor binding subunit (IFN-γR1, receptor chain 1), aspecies-specific cell surface transmembrane receptorchain. A second transmembrane protein, IFN-γR2 isalso required for the signal transduction process.

Although a great deal of information has been accu-mulated about cell surface receptors from biochemicalexperiments, no direct measurements of the receptorstructure has been made in cells in the presence andabsence of binding ligands. In most cases, the modelsof multichain receptor complexes were deduced from acombination of cross-linking, immunoprecipitation of

continued on page 2, left column

continued on page 2, right column


mental expertise and experimental arrangements thatare set up in the RLBL, our environment is an idealplace for such collaborations and many of them havebeen exceptionally successful.

The first article by Krause et al. is a collaboration withthe University of Medicine and Dentistry of New Jerseyand describes recent results about the ligand-inducedchanges of an interferon-γ receptor complex. Thesestudies involved fluorescence microscopy on livingcells using single and two photon excitation and fluo-rescence resonance energy transfer techniques. Di-rect measurements of distances between interferon-greceptor chains before and after engagement of ligandsuggest a new paradigm for receptor structure andfunction. The second article by Mei et al., a collabora-tion with the Biophysics Department here at the Uni-versity of Pennsylvania, reports on findings about sin-gle molecule phosphorescence spectroscopy, espe-cially the measurement of a distribution function ofoxygen quenching rates. These studies help to ad-dress the blinking of fluorescence of single moleculesby studying directly the phosphorescence emission.Following this article is a third report by Fang et al. incollaboration with scientists at Mount Holyoke College,which applies our core research technique of two-dimensional infrared spectroscopy to the study of vi-brational interactions within an α-helix. Synthetic pep-tides were labeled with both 13C=16O and 13C=18O inthe secondary structure to establish common patternsin the two-dimensional IR spectra. We hope that theseresearch reports show the wide scope of projects thatare being carried out at the RLBL.

Visit our websiteWe would like to encourage you to visit our website athttp://rlbl.chem.upenn.edu. An extended descriptionof the resources is available to outside users, as isinformation about new instrumentation when it be-comes available for collaborative or service researchprojects. Please don’t hesitate to contact us if you areinterested in any of the instruments and approaches todiscuss possible collaborative and service projects, orif you have any other questions concerning our activi-ties. [email protected]

the constituent proteins, and delineation of the signaltransduction events. All these results were deducedfrom indirect measurements. These models impliedthat the receptor complex is assembled by ligand.However, more recent observations have begun tosuggest that receptor chains can be preassociated inthe cell membrane.

To provide a direct way to clarify the various models ofreceptor action, we established an approach to meas-ure interactions of receptor chains in living cells. Fo-cusing on the IFN-γ receptor we used fluorescenceresonance energy transfer (FRET) to assess thestructure of the receptor chains and the effects of lig-ands directly.

Indirect evidence exists about the structure of theIFN-γ receptor complex prior to its activation by IFN-γ.Accordingly, most models of the IFN-γ receptor con-sider it to consist of four chains: two IFN-γR1 and twoIFN-γR2 chains. The ligand IFN-γ is a dimer that thatbinds to the two IFN-γR1chains. Upon IFN-γR1 bindingthe receptor chains associate to form a condensedcomplex. But very little is directly known about thephysical association of the receptor chains of suchmultiprotein complexes and whether chains from onereceptor type directly interact with another.

FRET is a powerful spectroscopic technique that util-izes the strong dependence of the energy transfer effi-ciency on the distance between the donor and accep-

Research article continued from page 1Editorial continued from page 1

continued on the next page

Figure 1: Schematic of the structure and FRET proc-ess of the IFN-γ receptor complex. See text for details.


Feature article continued

tor moiety. For example, this technique has been usedto determine protein-protein interactions (4) or thedistance between to two parts of the ribosome (5). Be-cause it is a noninvasive technique, in-vivo measure-ments of acceptor/donor interactions are easily possi-ble under physiological conditions. Here we report ourresults of FRET in examining the IFN-γ receptorstructure directly and provide evidence for a newmodel of its structure and function.

Materials and MethodsTo carry out our FRET studies, IFN-γR1 and IFN-γR2chains with EBFP (enhanced blue fluorescent protein,the donor) and GFP (green fluorescent protein, theacceptor) fused to the carboxyl termini of their intra-cellular domains were constructed (Figure 1). Becauseit was necessary to have cells express two proteinslabeled with EBFP and GFP at similar levels, we useda single vector expressing both proteins for transfec-tion rather than co-transfection with two vectors. Thus,tandem vectors, in which transcription of each cDNA iscontrolled by its own separate promoter and polyade-nylation signal on a single plasmid, were constructedas described previously (6).

To demonstrate GFP and EBFP fluorescence andFRET, a confocal microscope was modified to includea monochromator associated with a back-illuminatedliquid nitrogen-cooled charge-coupled device (CCD)camera so that fluorescence emission spectra andimages of whole cells could be obtained from the illu-minated cells. The S65T variant of GFP with an exci-tation maximum at 488 nm was used in all our studies(7). Single-photon excitation at 488 nm of the GFPwith an argon laser delivering 0.5 µW at the sampleyielded the signature GFP emission having a maxi-mum at 509 nm. The BFP and EBFP have excitationand emission maxima at 380 nm and 445 nm, respec-tively. Because we found that single-photon excitationof the cells at 380 nm produced very high backgroundfluorescence, we used two-photon excitation of at 760nm to substantially reduce the sample excitation vol-ume along the quartz cover slides. This resulted insignificantly decreased background fluorescence. The760 nm light was produced by a femtosecond mode-locked Ti:sapphire laser. Typical laser power in this

case was 2 mW. We found little or no cellular damagedue to the two-photon beam compared with ultravioletexcitation. If FRET occurs between EBFP and GFP,then the emission maximum of GFP at 509 nm will beobserved instead of the pumped EBFP emission at445 nm.

The constructs expressing IFN-γR2/GFP and IFN-γR2/EBFP were individually transfected into COS-1cells, and images were taken with a camera attachedto a confocal microscope. The images show that bothIFN-γR2/GFP (Fig. 2, left) and IFN-γR2/EBFP (Fig. 2,right) were visualized by epifluorescence after trans-fection. Similar images were obtained for the labeledIFN-γR1 samples (not shown). To further demonstratethat the constructs were functional when transfectedtogether into cells, a tandem vector coexpressing IFN-γR2/GFP and IFN-γR1/EBFP was transfected intoChinese hamster ovary cells expressing none of thehuman IFN-γ receptor chains (CHO q3). The MHCClass I surface antigen induction in response to IFN-γin CHO q3 cells expressing IFN-γR2/GFP and IFN-γR1/EBFP showed that both receptor chains are func-tional. Similar activity measurements with other re-ceptor chains such as IFN-γR2/EGFP, IFN-γR2/EBFP,and FL-IL-10R2/GFP demonstrated that all these re-ceptor chains with fluorescent proteins fused to theircarboxyl termini were fully functional.

Results and DiscussionTransient transfectants in COS-1 cells expressingmatched (Hu-IFN-γR1 and Hu-IFN-γR2) and mis-matched (Hu-IFN-γR1 and Hu-IL-10R2) receptor pairswere prepared as shown in Figure 3. The matchedpair, Hu-IFN-γR1/EBFP and Hu-IFN-γR2/GFP, repre-sents the two chains of the Hu-IFN-γ receptor com-

Figure 2: Fluorscence images of human IFN-γR2/GFPand IFN-γR1/EBFP transfected into cells.


plex. The mismatched pair represents two chains fromsimilar receptor complexes, Hu-IFN-γR1/EBFP andHu-IL-10R2/GFP, which were not expected to interact.

FRET was measured upon excitation at 760 nm incells expressing each of these pairs of receptor chainsin the absence of ligand, given in Figure 4. The spec-trum of cells expressing the mismatched pair Hu-IFN-γR1/EBFP and Hu-IL-10R2/GFP shows the EBFPemission spectrum with its peak around 450 nm to-gether with background fluorescence. The minusculeamount of GFP fluorescence due to energy transferdemonstrates little or no interaction of these receptorchains. In contrast, the matched pair, Hu-IFN-γR1/EBFP and Hu-IFN-γR2/GFP, excited again at 760nm, exhibits the typical fluorescence emission signa-ture of the GFP (peak around 510 nm), demonstratingclearly that energy transfer occurs after excitation be-tween the EBFP and GFP proteins. The calculateddistance between the intracellular regions of IFN-γR1/EBFP and IFN-γR2/GFP chains from these datawas (5) 36 Å. Furthermore, these spectra revealedthat IFN-γR1 and IFN-γR2 chains are preassociatedprior to ligand binding as no ligand (IFN-γ) was addedyet. Subsequent addition of the ligand IFN-γ to cellsexpressing the mismatched pair IFN-γR1/EBFP andIL-10R2/GFP did not affect the spectrum (not shown):the spectra were found to be basically identical in thepresence or absence of IFN-γ. In contrast, the effect ofadding IFN-γ to the matched receptor pair expressedin cells showed that IFN-γ produced a change in theemission spectrum, causing a major reduction in theFRET compared with the FRET in the absence of IFN-

γ (Figure 5). The distance between the intracellularregions of Hu-IFN-γR1/EBFP and Hu-IFN-γR2/GFPchains in the presence of the ligand IFN-γ was calcu-lated to be 63 Å, almost twice as large as the 36 Ådetermined in the case where no IFN-γ ligand was

present. Therefore, the results strongly indicate thatthe intracellular domains of the IFN-γR1/EBFP andIFN-γR2/GFP chains move apart on addition of theligand IFN-γ.

Our results further demonstrate directly that the re-ceptor chains of the IFN-γ receptor complex are pre-assembled on the cell membrane of intact cells. Fur-thermore, the data prove that the receptor complexconsists of four (or possibly more chains) becauseboth the IFN-γR1 and IFN-γR2 pair (Figures 4 & 5) andthe IFN-γR2 and IFN-γR2 pair (Figure 6) are preasso-

Figure 3: Illustration of matched (FL-IFN-γR2/GFP andIFN-γR1/EBFP) and mismatched (IFN-γR1/EBFP and FL-IL-IFN-10R2/GFP) pairs of receptors.

Figure 4: Comparison of fluorescence emission spectraof cells expressing the matched and mismatched pairs ofreceptor chains.

Figure 5: Comparison of fluorescence emission spectraof cells expressing the matched pair of receptor chains inthe presence and absence of IFN-γ.


ciated. It also became evident from our data that theIFN-γR2 and IFN-γR2 chains are preassociated evenin the absence of the IFN-γR1 chain, but that the pres-ence of the IFN-γR1 brings the two IFN-γR2 chainscloser together. However, IFN-γ shows no effect onthe separation distance of IFN-γR2 chains when IFN-γR1 is not present. This is in agreement with previ-ously published reports, where it was found that IFN-γdoes not bind to the IFN-γR2 chain and that IFN-γ ex-hibits no activity in the presence of only the IFN-γR2chain only (8, 9, 10). In the presence of IFN-γR1 how-ever the efficiency of FRET between two IFN-γR2chains decreases upon addition of IFN-γ to cells.

Thus the interaction of IFN-γ with its receptor complexalso results in the spreading of the intercellular do-mains of the two IFN-γR2 chains. We suggest that thisparadigm is likely to be applicable to other receptorchains and that receptor chains are preassociated incells ready for activation by ligand. This conclusionleads to the deduction that cell surface receptor chainshave specific sites, receptor association regions, thatenable the receptor chains to preassociate. Moreover,the increase in distance between the intracellular re-gions of the receptor chains after the addition of ligandto cells opens the receptor complexes for the attach-ment of the signal transduction components that areexcluded from the receptor complexes until ligand en-gagement. Thus, our direct measurements of the dis-tances between the IFN-γ receptor chains before andafter engagement of ligand suggest a new paradigmfor receptor structure and function.

General comments and conclusionsPrevious techniques used to measure protein-proteininteractions often removed the proteins from the con-text of their superstructure within membranes andother cellular components with which they are an intri-cately associated. Thus, high affinity interactions, butnot low affinity interactions, can be measured by theseprocedures. In contrast, as we showed in our work, theuse of FRET in combination with microscopy permitsthe direct visualization of all, low and high affinity, in-teractions in cells. Association constants that are tooweak to be observed in solution after disruption ofcells can be clearly seen by the FRET technique (Figs.4, 5, and 6) because the receptors remain anchored inthe membrane.

Because FRET is a real-time technique, changes inprotein-protein interactions or protein positions incomplexes within living cells can be determined overthe periods of time or under changing conditions of thesystem. Continued use of the FRET technique shouldprovide answers to questions about association ofspecific receptor chains of multichain complexes, andwhether chains from one receptor type directly interactwith those from another type. Furthermore, suchstudies can be extended to the downstream signaltransduction events by determining the order, interac-tions, and kinetics of these processes.

Literature1. Petska, S. et al. (1997) Cytokine Growth Factor

Review 8, 189-206.2. Langer, J. A. and Petska S. (1988) Immunol. To-

day 9, 393-400.3. Petska, S. et al. (1987) Annu. Rev. Biochem. 56,

727-777.4. Van der Meer, B. W. et al. (1994) Resonance En-

ergy Transfer: Theory and Data, J. Wiley.5. Langlois, R. et al. (1976) J. Mol. Biol. 106, 297-

313.6. Kotenko, S. V. et al. (1997) EMBO J. 16, 5894-

5903.7. Ormo, M. et al. (1996) Science 273, 1392-1395.8. Soh, J. et al. (1994) Cell 76, 793-802.9. Jung, V. et al. (1990) J. Biol. Chem. 265, 1827-

1830.10. Hemmi, S. et al. (1994) Cell 76 803-810.11. Krause, Ch. D. et al. (2002) Molecular & Cellular

Proteomics 1, 805-815.

Figure 6: Comparison of the fluorescence spectra ofcells expressing the FL-IFN-γR2/EBFP and Fl-IFN-γR2/GFP in the presence and absence of the IFN-γR1chain and the presence and absence of ligand.


probe

D I R E C T O B S E R V A T I O N O F

T R I P L E T E M I S S I O N O F S I N G L E

M O L E C U L E S A N D P H O S P H O -R E S C E N C E Q U E N C H I N G B Y

M O L E C U L A R O X Y G E N

By E. Mei, R. M. Hochstrasser and S. Vini-gradovBiophysics & Biochemistry and Chemistry Departments,University of Pennsylvania

IntroductionNewly emerged room temperature single moleculespectroscopy has opened a way for studying hetero-geneity present in materials and bio-macromolecules,as well as the dynamics of entities like proteins andenzymes. By studying molecules one at a time, onecan probe chemical dynamics and kinetics and obtaindirect distributions of physical and chemical properties.Such distributions are not directly accessible by con-ventional ensemble methods. So far a wide range ofsingle molecule detection technologies based upondifferent detection schemes has been explored.Among them, single molecule spectroscopy based onfluorescence detection has proven to be most powerfuland effective. Its advantages include high sensitivity,low noise level and easy applicability in bio-systems.

In the time-intensity records of these fluorescence ex-periments, so-called “blinking”, when the fluorescencesignal fluctuates between states of high and very lowemission intensity, is often observed. This blinking caninterfere with the data interpretation, especially if thereare other dynamics present that can cause a rapidchange in emission efficiency, which will cause similaron/off characteristics of the fluorescence signal (1,2).Although the blinking process has been studied exten-sively, no detailed and generally applicable mecha-nism has been established so far (3,4).

In this work, we address the problem of blinking bystudying not the fluorescence but the phosphores-cence emission. Since the triplet state is the lowestexcited state of the molecule, this study should enableus to gain insight into the underlying photo-physicaland photo-chemical dynamics, which are obscured in

the case of fluorescence detection. Furthermore, usingoxygen as a triplet quencher, we are able to obtainunprecedented information about the detailed mecha-nism of phosphorescence quenching at the singlemolecule level. Quenching of phosphorescence hasalso been applied to biological studies on a molecularscale. For example, oxygen quenching of tryptophan(5,6) and metalloporphyrin (7) phosphorescence hasbeen used to evaluate pathways of oxygen diffusioninside proteins and look into the dynamics of oxygenchannels.

Materials and MethodsWe employed two molecules, PtOBCP (= Pt octabut-oxycarbony porphyrin) and Ru(dpp)3Cl2 ( = Ru(II)-tris-4,7-diphenyl-1,10-phenanthroline dichloride) for ourstudies. Absorption and emission spectra of PtOBCPin deoxygenated DMF are shown in Figure 1. Thephosphorescence lifetime is 36 µs in the absence of

Figure 1: Molecular structure of PtOBCP (a) andRu(ddp)3Cl2 (b).

NN

NN

N

NRu

2+

2Cl-

b

a

N

N

N

N

Pt

BuO2C

BuO2C

CO2Bu

CO2Bu

CO2Bu

CO2Bu

BuO2C

BuO2C


oxygen. Samples for single molecule imaging andspectroscopy where prepared by spin casting of a 1%PMMA solution onto a quartz cover slip. Typical sam-ple concentrations were 0.1 nM. Bulk phosphores-cence decay time measurements in solutions orPMMA films were performed using previously de-scribed methods (7). Single molecule experimentswere performed with an inverted confocal microscopeand Argon ion laser as an excitation source. A liquidnitrogen cooled CCD detector and an imaging spec-trograph were coupled to the microscope and used torecord the single molecule phosphorescence spectra.Oxygen concentration of the samples was adjustedusing an inverted glass funnel connected to gas tanksof nitrogen, oxygen and argon. A system of multipleinlets and flow meters allowed for gas mixtures to bevaried in the desired proportions.

Results and DiscussionFigure 2 compares single molecule phosphorescencespectra of PtOBCP and Ru(ddp)3Cl2 with the corre-sponding bulk spectra. Phosphorescence was ob-served in the same region and with the same wave-

length characteristics in both cases. Adding oxygen tothe sample clearly quenches the phosphorescenceemission as shown in Figure 3. In these experimentssingle molecules were first observed under a pure Ar-gon atmosphere. Oxygen was then added and addi-tional images of the same region were taken.

The relationship between oxygen concentration n(O2)and triplet lifetime t is usually described in bulk solu-tions by the Stern-Volmer equation

I0 / I = τ0 / τ = τ ( 1 / τ0 + k2*n(O2) ) = 1 + KQ*n(O2)

where I0 and τ0 are the intensity and phosphores-cence lifetime in the absence of oxygen, k2 is the bi-molecular rate constant for the quenching and KQ isthe related constant. This relationship was also foundto be valid in the case of our single molecule images.A linear relationship should be observed between theintensity ratio and the oxygen concentration. Thequenching constant itself will depend on the diffusionrate of O2 to the phosphorescent and on the mecha-nism of quenching within the encounter complex. Usu-ally it is assumed that quenching occurs via a shortrange energy exchange, resulting in the production ofsinglet oxygen.

Single molecules of Ru(ddp)3Cl2:

Individual quenching events could be measured in thecase of triplet quenching of Ru(ddp)3Cl2 by oxygen.This permitted an experimental evaluation of the dis-tribution of quenching constants that are hidden fromensemble average measurements in a typical bulkexperiment. Figure 8 shows such an experimentallydetermined distribution of quenching constants. Themean bimolecular rate constant k2 is equal to 2.2x108 lmol-1s-1, which corresponds to a mean equilibriumquenching constant KQ of 1120 l mol-1. The distributionfunction is not well fit by a gaussian function and itclearly stretches out to larger quenching constants.This suggests that there is a range of accessibilities tooxygen that is very likely related to the distance distri-bution of Ru(ddp)3Cl2 molecules from the surface ofthe film. Reported quenching rate constants for similarRu(II) complexes in solutions range from 2.2 - 4.2x109

l mol-1s-1 [41]. Thus it appears that on average thesolubility factor of Ru(ddp)3Cl2 in the surface layer isabout 0.1.

Figure 2: Absorption (a) and emission (b) spectra ofPtOBCP in dimethylformamide.

Wavelength (nm)

600500400

800700600

a

b

Abso

rban

ceEm

issi

on In

tens

ity


Single molecule of PtOBCP:

The signal to background ratio was considerably lowerin the experiments with PtOBCP than in the case ofthe Ru complex, so that our data are less extensive.Nevertheless, we were able to determine the quench-ing equilibrium constant KQ for a few molecules in Fig-ure 4 to be on the order of 100 l mol-1. This corre-sponds to a value of k2 = 2.8x106 l mol-1s-1. Similarbulk experiments yielded a value of k2 = 2.4x106 l mol-1s-1 in PMMA, in reasonable agreement with the singlemolecule data.

We could further show that PtOBCP could be em-ployed a single molecule oxygen sensor. This resultcould be useful for further studies of heme protein dy-namics because PtOBCP can be incorporated intovarious proteins. Such an ultra-small sensor could becrucial for non-invasive intracellular oxygen detectionand diffusion rate determinations. There is a need inbiochemical and medical research to measure oxygenconcentrations in living cells or tissues. In principle,our studies suggest that both PtOBCP andRu(ddp)3Cl2 offer this possibility by employing timegated methods to discriminate between the fast scat-tered signal and the slower phosphorescence.

Fluorescence Blinking

It is very common to observe large intensity jumps inthe fluorescence signal of single molecules on timescales that are long compared with ordinary photo-physical parameters. Sudden changes in the orienta-tion of the emitting dipole can affect the overall emis-sion signal strength, but in the two cases presentedhere, PtOBCP and Ru(ddp)3Cl2 are immobilized in aPMMA film. The observed dark states most likely cor-respond to reversible chemical transformationswhereby the chemistry might involve oxygen, thepolymer matrix or both. To provide trapping times onthe order of tens of seconds, escape barriers wouldrequire to be in excess of 10 kcal/mol. One possibilityinvolves the formation of long lived Pt porphyin cationradicals. Such involvement of cation radicals was pre-viously suggested in the case of single moleculestudies of light harvesting complexes LH2 (8). It alsohas been shown that ZnTPP cation radicals can sur-vive in some confined molecular environment for min-utes (9). Nevertheless, further experiments are re-quired to establish with certainty if the recovery of thephosphorescence signal from such dark states aftertens of seconds is a thermal or photo-induced proc-ess.

ConclusionsIn summary, we have demonstrated the successfulobservation and characterization of single moleculephosphorescence and the experimental determinationof oxygen quenching constants. Based on our experi-ments, a single molecular oxygen sensor employingeither Ru(ddp)3Cl2 or PtOBCP seems to be feasible.This would allow for the non-invasive measurement of

Figure 3: Spectra of a single molecules of (a) PtOBCP inPMMA and (b) Ru(dpp)3Cl2 on quartz surface. The spec-tra of bulk PtOBCP in PMMA and bulk Ru(dpp)3Cl2 onquartz surface are shown in (c) and (d)

Figure 4: Distribution of quenching constants as definedin the text, of 377 single molecules of Ru(dpp)3Cl2 on aquartz surface.

Wavelength (nm)

a b

c d

700600 700600

Rel

ativ

e In

tens

ity

Quenching Constants (l mol-1)

Occ

urre

nce 40

20

8000600040002000


oxygen concentration in cell and tissues.

Acknowledgement

This work was supported by NIH GM 48130 and NIHRR 001348 grants.

Literature1. Lu, H. P. et al., (1998) Science 282, 1877-1882.2. Jia, Y. et al., (1999) Chem. Phys. 247, 69-83.3. Weston, K. D. et al., (1998) J. Chem. Phys. 109,

7474-7485.4. Yip, W.-T. et al., (1998) J. Phys. Chem. 102, 7564-

7575.5. Lakowicz, J. et al., (1973) Biochemistry 12, 4161-

4179.6. Calhoun, D. et al., (1988) Biochemistry 27, 8466-

8484.7. Khajehpour, M. et al., (2003) Proteins: Structre,

function and genetics 53, 656-666.8. Bopp, M. A. et al., (1996) Proc. Natl. Acad. Sci.

USA 94, 10630-10635.9. Morishima, Y. et al., (1996) Macromolecules 29,

6505-6509.10. P. George, (1953) Biochem. J. 55, 220.11. Mei, E. et al, (2003) J. Am. Chem. Soc. 125,

13198-13204.

T W O D I M E N S I O N A L I N F R A R E D

M E A S U R E M E N T S O F T H E

C O U P L I N G B E T W E E N A M I D E

M O D E S O F A N α- H E L I X

By C. Fang, W. Barber-Armstrong, R. M.Hochstrasser and S. DecaturDepartment of Chemistry, Holyoke College, MA

Chemistry Departments, University of Pennsylvania

IntroductionThe development of methods to obtain structuralchanges in proteins is an important experimentalchallenge to which two-dimensional infrared spectros-copy (2D IR), developed at the RLBL, can contributesignificantly. The development of 2D IR started withthe observation of infrared photon echo in the earlyseventies followed by two pulse echoes in the visible.As optical lasers improved, investigations into faster

responding liquids and glasses became feasible andnew techniques like heterodyning, gating and threepulse echoes were successfully introduced. With thecurrent availability of femtosecond laser technologyheterodyning of phase-locked IR echoes and dual-frequency IR pump-probe experiments are suitable tomeasure 2D IR spectra of peptides and small proteins.All these techniques open up the exciting possibility ofstudying protein dynamics in real time over severalorders of magnitudes through a detailed analysis ofvibrational mode coupling in the multidimensional IRspectrum.

IR spectroscopy provides the fingerprints of nuclearmotion in any molecule. Within proteins, motions ofeach peptide unit give rise to characteristic vibrationalfrequencies of which the amide I band around 1600-1650 cm-1 is the most intense. According to normalmode calculations, its main contribution comes fromthe C=O stretch. The amide I band also shows struc-ture sensitivity due to frequency shifts from exter-

Fig. 1: Example structures of the peptide understudy, Ac-(AAAAK)3AAAAY-NH2 for (a) aperfect helix and (b) a random coil. In practice,a distribution of structures is expected whoseaverage structure lies nearer to (a) at 0oC andcloser to (b) at 50oC.


nal/internal hydrogen bonding and electrostatic cou-pling to neighboring units. Utilization of linear IR spec-tra to deduce structures of proteins and peptides hasbeen limited because of the congestion from overlap-ping amide I bands associated with different structuraldomains. Methods based on nonlinear IR spectros-copy overcome these shortcomings by spreading thespectrum into more frequency dimensions. 2D IRspectroscopy has emerged as a powerful tool to in-vestigate structural dynamics not accessible in a linearapproach.

Materials and Spectroscopic Results25-residue helical peptides based on Ac-(A)4K(A)4 -K(A)4K(A)4Y-NH2 [6] containing the isotopic amidecarbonyl labels 13C=16O and 13C=18O were obtained bystandard peptide synthesis, purified by reverse-phaseHPLC and characterized by electrospray mass spec-trometry. The 18O amino acids were synthesized by18OH2 exchange. The isotopomers contain none, oneor one of each of the isotopic amide carbonyl labelsand are referred to as [0,0], [0,11], [12,13], [11,13] and[11,14]. The numbers refer to the isotopically labeledamino acids in the chain. With this set of isotopomerswe were able to distinguish between direct neighborand non-direct neighbor interactions.

Figure 2 shows linear IR spectra of the four isoto-pomers [0,11], [12,13], [11,13] and [11,14] in the re-gion of the amide I band obtained with a conventionalFTIR instrument at room temperature. A baselinespectra consisting of a sample cell with solvent but nopeptide was subtracted in all cases to obtain thesespectra. The thick spectral line shape was fitted to fourgaussian peaks for each case corresponding to themain helical 12C=16O peak at 1632 cm-1, a weaker non-helical 12C=16O contribution at 1655 cm-1, the 13C=16Oisotopomer helical peak at 1594 cm-1 and the 13C=18Oisotopomer helical peak at 1571 cm-1. The peak sepa-rations between the two isotopic labels in the [12,13],[11,13] and [11,14] isotopomers were 24.7, 22.4 and28 cm-1. Diagonal anharmonicities, i.e. the change infrequency of a single vibration as its excitation numberincreases, for both isotopically substituted amideswere determined by broad-band pump/probe experi-ments [1] and were in the range of 12-15 cm-1 for allisotopomers.

Nonlinear 2D IR spectra of the same isotopomerswere obtained precisely as in previous heterodynedecho experiments at 6 µm [2-5]. Three infrared pulseswith wave vectors k1, k2 and k3 were incident on thesample. Either k1 (rephasing spectrum) or k2 (non-rephasing spectrum) arrives first, followed with a timedelay τ simultaneously by k2/k3 or k1/k3. The echo sig-nal at wave vector –k1+k2+k3 was observed with afourth heterodyning laser with wave vector k4 at a de-lay time t after the second excitation pulse sequence.Scanning the two delays τ and t and Fourier trans-forming the resulting two-dimensional sample re-sponse yields the 2D IR spectrum along the frequencyaxes ωτ and ωt. For example, Figure 3a shows the(absolute value) 2D IR rephasing spectrum of the[11,13] isotopomer and Figure 3b corresponding ab-sorptive spectrum. Absolute value thereby means thatthe real and imaginary part of the Fourier transformwere squared and added, and the absorptive spectrumconsists of an addition of the real parts of therephasing and non-rephasing spectra.

Like the linear FTIR spectrum, the 2D IR spectrum inFigure 3a reveals three main peaks along the diago-nal. The intensities are different and more pro-nounced, because the 2D IR spectrum scales with thefourth power of the transition dipole moment, and con-trary to the linear spectrum, the 2D IR spectrum isbackground free. In addition to the diagonal peaks,

�

��

��

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

� � ��

��

��

��

��

��

��

��

��

��

��

��

� ��

��

Figure 2: Linear FTIR spectra of the isotopomers [0,11],[12,13], [11,13] and [11,14]. The thick line is the experi-mental curve and the thin lines represent gaussian fits.


there are several off-diagonal features clearly visible inthe 2D spectrum. These off-diagonal peaks are cou-pling peaks between the two corresponding diagonalpeaks in analogy to a 2D NMR spectrum. The crosspeaks, also observed in the case of the other isoto-pomers not shown here, reveal that the isotopomerresidues are coupled in all cases. No such couplingcould be extracted from the linear spectrum.

In the absorptive 2D IR spectrum shown in Figure 3bagain for the [11,13] isotopomer, the anharmonicitiesof all vibrations are easily revealed. Along the diago-nal, “double” peaks are due to the 0→1 and 1→2 tran-sitions. The displacement of these transitions revealsthe diagonal anharmonicity of the corresponding vi-brations. Off-diagonal anharmonicities between differ-ent vibrations within the peptide molecule are againrevealed in the many cross peaks identifiable in thisspectrum. They are on the order of 3 - 9 cm-1 depend-ing on the isotopomer.

In order to estimate the coupling constants from thesemeasured off-diagonal and diagonal anharmonicities,a model is needed which will be able to reproduceboth the linear and nonlinear spectra. Such a model isdescribed in more detail in reference 1. Briefly, cou-pling constants βij needed to generate the observedset of two fundamentals, two overtones and one com-bination band were βij[12,13] = 9.5 cm-1 for the isoto-

pomer with nearest neighbor isotopic substitution andβij[11,13] ≈ βij[11,14] = -0.5 cm-1 for the isotopomerswith one or two residues between the substituted sites.

In summary, the insertion of 13C=16O and 13C=18O la-bels at known residues on the α-helix enabled themeasurement of the coupling between the selectedamide-I vibrational modes using a 2D IR approach.The observed signal strength indicates that one in 35residues could have been detected, meeting the re-quirement for application of the 2D IR method to singleresidue specificity in small proteins.

AcknowledgementThis work was supported by grants: NIH GM 12592and RR 001348 to RMH, and NSF CHE9984844 andNIH R15GM54334 to SMD.

Literature1. Hamm, P.; Lim. M.; Hochstrasser, R. M.; (1998) J.

Phys. Chem. B 102, 6123.2. Asplund, M. C., Zanni, M. T., Hochstrasser, R. M.

(2000) Proc. Natl. Acad. Sci. USA 97, 8219.3. Zanni, M. T., Gnanakaran, S., Stenger, J.,

Hochstrasser, R. M. (2001) J. Phys. Chem. B 105,6520.

4. Ge, N.-H., Zanni, M. T., Hochstrasser, R. M.(2002) J. Phys. Chem. A 106, 962.

5. Ge, N.-H., Hochstrasser, R. M. (2002) Chem.Commun. 5 17.

6. Silva, R.A.G.D., Nguyen, J.Y., Decatur, S.M.(2002) Biochemistry 41, 15296.

1540 1580 1620 1660 1700

1540

1580

1620

1660

1700

-ωτ/

2πc

ωt /2πc

Figure 3a: Absolute value two-dimensional IR spectra ofthe [11,13] isotopomer. See text for further details.

1540 1560 1580 16001540

1560

1580

1600

ωt /2πc

ωτ/

2πc

Figure 3b: Absorptive two-dimensional IR spectra of theisotopomer [11,13]. See text for further details.


C U R R E N T T E C H N O L O G I C A L

D E V E L O P M E N T A N D R E S E A R C H A T

T H E R L B L

The main subjects under investigation at RLBL areshown below. If your research may be interfaced withany of these approaches we urge you to contact us. Afull description of each of these topics is also availableat our Web site http://rlbl.chem.upenn.edu.

• Two-dimensional infrared spectroscopy (2D-IR)and infrared analogues of NMR: Heterodynedphoton echo spectroscopy and spectrally resolvedthree pulse IR photon echoes are employed to in-vestigate the amide I, amide II and amide A regionand other transitions of peptides and small pro-teins. One and two color experiments are feasibleto obtain quantitative information on the anhar-monic potential surfaces, the vibrational modecoupling, the frequency distributions and vibra-tional frequency correlation functions with thehighest time resolution. Efforts are under way tocouple the 2D-IR technique with temperature jumpmethods.

• Methodologies to investigate protein folding andmacromolecular conformational dynamics: Detec-tion and characterization of intermediate states inconformational dynamics and unfolding is anotherdeveloping technology at RLBL. Laser-based tem-perature-jump instruments are available for theseinvestigations in combination with transient ab-sorption techniques in the visible and infraredspectral regions.

• Investigations of single molecular assemblies us-ing confocal and atomic force microscopes: It isnow possible to examine the properties of singlemolecules using fluorescence and phosphores-cence in association with confocal microscopy.Studies using resonance fluorescence energytransfer, dual wavelength detection, polarizationresolution, two photon absorption are feasible forsingle molecule investigations of biomolecules andimaging techniques of single cells. The RLBL alsohas coupled single molecule detection methodswith mature time correlated photon counting tech-

nology for fluorescence lifetime imaging with po-larization resolution. Excitation wavelengthsavailable cover the whole visible range from 400– 850 nm and include continuous wave andpulsed picosecond and femtosecond lasers. Ad-ditionally, an objective-type total internal reflec-tion microscope (TIRFM) that images singlemolecules by means of a CCD camera has re-cently been set up and coupled to various lasersources. Typical penetration depth of the eva-nescent field is ~150 nm, ideally suited to studysingle molecules located in cell membranes.

• Dynamics of photoactivatable proteins and otherbiological structures: Methods are being usedand further developed to examine the responsesof biological systems to light by pump/probe andnonlinear spectroscopic techniques encompass-ing spectral regimes from the UV to the far IRand covering femtosecond to second timescales.Techniques include: single and multiple wave-length transient spectroscopy (UV/Vis, vibrationalIR, Terahertz), photon echoes, two photon ab-sorption and time-correlated single photoncounting.

• Energy transfer and fluorescence monitoring ofbiological dynamics: Monitoring fluorescencelifetimes and anisotropies reveals details ofstructural dynamics of peptides and proteins.Techniques have been developed at the RLBL tomonitor these properties of fluorescing specieson the femtosecond to nanosecond timescale.Vibrational energy dynamics, electron transferreactions, and UV cross-linking are open for in-vestigation.

• Development of time resolved far-IR (terahertz)probes for protein dynamical changes: Newpowerful sources of THz and far-IR radiation aredeveloped and used as laboratory THz source toaccess spectral information in the far IR region.


A S E L E C T I O N O F R E C E N T

P U B L I C A T I O N S

Direct Observation of Triplet State Emission of SingleMolecules: Single Molecule PhosphorescenceQuenching of Metalloporphyrin and OrganometallicComplexes by Molecular Oxygen and Their Quench-ing Rate Distributions. E. Mei, S. Vinogradov and R.M. Hochstrasser, J. Am. Chem. Soc. 125, 13198(2003).

Anthracene - BODIPY cassettes: Syntheses and en-ergy transfer. C.-W. Wan, A. Burghart, J. Chen, F.Bergstroem, L. B.-A. Johansson, M. Wolford, T. G.Kim, M. R. Topp, and R. M. Hochstrasser, Chemistry--A European Journal 9, 4430 (2003).

Nature of structural inhomogeneities during folding ahelix: Manifestation in IR. S. Gnanakaran, R. M.Hochstrasser, and A. E. Garcia, Abstr. Pap. Am.Chem. Soc. 226, PHYS-433 (2003).

Infrared study of the stability and folding kinetics of a15-residue β-hairpin. Y. Xu, R. Oyola, and F. Gai, J.Am. Chem. Soc. ACS ASAP (2003).

Temperature dependence of the CN stretching vibra-tion of a nitrile-derivated phenylaniline in water. C.-Y.Huang, T. Wang, and F. Gai, Chem. Phys. Lett. 371,731 (2003).

Smaller but slower: Folding kinetics of a β-hairpin anda three-helix bundle. Abstr. Pap. Am. Chem. Soc. 226PHYS-431 (2003).

Dual Frequency 2D-IR heterodyned photon-echo ofthe peptide bond. I. V. Rubtsov, J. Wang, and R. M.Hochstrasser, Proc Natl Acad. Sci. USA, 100, 5601(2003).

Dual frequency 2D-IR of peptide amide-A and amide-Imodes. I. V. Rubtsov, J. Wang, and R. M.Hochstrasser, J. Chem. Phys. 118, 7733 (2003).

Effect of Protein Dynamics upon reactions that occurin the heme pocket of horseradish peroxidase. M.Khajehpour, T. Troxler, and J. M. Vanderkooi, Bio-chemistry 42, 2672 (2003).

Motions of single molecules and proteins in trealoseGlass. E. Mei, J. Tang, Jane M. Vanderkooi, and R. M.

Hochstrasser, J. Am. Chem. Soc. 125, 2730 (2003).

Vibrational coupling between amide-I and amide-Amodes revealed by femtosecond two-color infraredspectroscopy. I. V. Rubtsov, J. Wang, and R. M.Hochstrasser, J. Phys. Chem. B ASAP (2003).

Dual frequency 2D-IR spectroscopy in peptides. I. V.Rubtsov, J. Wang, and R. M. Hochstrasser, Abstr.Pap. Am. Chem. Soc. 225 462-PHYS (2003).

Ultrafast singlet excited-state polarization in electroni-cally asymmetric ethyne-bridgedbis[(porphinato)zinc(II)] complexes. I. V. Rubtsov, K.Susumu, I. G. Rubtsov, and M. J. Therien, J. Am.Chem. Soc. 125, 2687 (2003).

Helix-coil kinetics of two 14-residue peptides. T.Wang, D. G. Du, and F. Gai, Chem. Phys. Lett. 370,842 (2003).

Using nitrile-derivated amino acids as infrared probesof local environment. Z. Getahun, C. Y. Huang, T.Wang, B. De Leon, W. F. DeGrado, and F. Gai, J. Am.Chem. Soc. 125, 405 (2003).

Ultrashort-lived excited states of aminophthalimides influid solution. T. G. Kim, M. F. Wolford, and M. R.Topp, Photochem. Photobiol. Sci. 2, 576 (2003).

Fluorescence quenching and lifetime distributions ofsingle molecules on glass surfaces. M. Lee, J. Kim, J.Tang, and R. M. Hochstrasser, Chem. Phys. Lett. 359,412 (2002).

Vibrational dynamics, mode coupling and structuralconstraints for acetylproline-NH2. I. V. Rubtsov and R.M. Hochstrasser, J. Chem. Phys. B 106, 9165(2002).

Single molecule fluorescence of cytochrome c. E. Mei,J. M. Vanderkooi, and R. M. Hochstrasser, Biophys. J.82, 230 Part 2 (2002).

Seeing the light: Preassembly and ligand-inducedchanges of the interferon γ receptor complex in cells.C. D. Krause, E. Mei, J. Xie, Y. Jia, M. A. Bopp, R. M.Hochstrasser, and S. Pestka, Mol. Cell. Proteomics 1,805 (2002).

Light induced helix formation. C. Y. Huang, S. He, W.F. DeGrado, D. G. McCafferty, and F. Gai, J. Am.Chem. Soc. 124, 12674 (2002).


Shining a light on molecules: Photochemistry andphotobiological sciences. R. M. Hochstrasser, Nature420, 6911 (2002).

Vibrational dynamics, mode coupling and structure ofacetylproline-NH2 dipeptide. I. V. Rubtsov and R. M.Hochstrasser, in Trends in Optics and Photonics, 72,386 (2002).

Local structure and dynamics of liquid acetone by het-erodyned 2D IR spectroscopy. N.-H. Ge and R. M.Hochstrasser, in Trends in Optics and Photonics, 72,255 (2002).

Two dimensional infrared spectroscopy: Studies of thedynamics of structures with femtosecond pulse Fouriertransform correlation spectroscopy. R. M.Hochstrasser, N. H. Ge, S. Gnanakaran, and M. T.Zanni, Bull. Chem. Soc. Jap. 75, 1103 (2002).

Effects of vibrational frequency correlations on two-dimensional infrared spectra. N.-H. Ge, M. T. Zanni,and R. M. Hochstrasser, J. Phys. Chem. A 106, 962(2002).

The temperature dependent structure distribution of ahelical peptide studied with 2D IR spectroscopy. M. T.Zanni, N.-H. Ge, Y. S. Kim, and R. M. Hochstrasser,Biophys. J. 82, 66 Part 2 (2002).

Femtosecond two-dimensional infrared spectroscopy:IR-COSY and THIRSTY. N.-H. Ge and R. M.Hochstrasser, Physchemcomm. 3, U1-U23 (2002).

R150A mutant of F TraI relaxase domain: Reducedaffinity and specificityfor ssDNA and altered fluores-cence anisotropy of a bound labeled oligonucliotide.M. J. Harley, D. Toptygin, T. Troxler, and J. F. Schild-bach, Biochemistry 41, 6460 (2002).

Distance dependence of electron transfer in rigid, co-facially compressed, pi-stacked porphyrin-bridge-quinone systems. Y. K. Kang, I. V. Rubtsov, P. M. Io-vine, and M. J. Therien, J. Am. Chem. Soc. 124, 8275(2002).

Synthesis, excited-state dynamics, and reactivity of adirectly-linked pyromellitimide –(porphinato)zinc(II)complex. N. P. Redmore, I. V. Rubtsov, and M. R.Therien, Chem. Phys. Lett. 352, 357 (2002).

Energy transfer enhanced quenching in conjugatedpolymer chemosensors: theory and applications. C. B.

Murphy, Y. Zhang, T. Troxler, V. Ferry, and W. E.Jones Jr., Polym. Mat. Sci. Eng. 87, 134 (2002).

Energy transfer enhanced quenching in conjugatedpolymer chemosenso; Theory and applications. C. B.Murphy, Y. Zhang, T. Troxler, V. Ferry, and W. E.Jones Jr., Abstr. Pap. Am. Chem. Soc. 224 279-PMSEPart 2 (2002).

Helix formation via conformation diffusion search. C.Y. Huang, Z. Getahun, Y. J. Zhu, J. W. Klemke, W. F.DeGrado, and F. Gai, Proc. Natl. Acad. Sci. USA 99,2788 (2002).

Conformational preferences and vibrational frequencydistributions of short peptides in relation to multi-dimensional infrared spectroscopy. S. Gnanakaranand R. M. Hochstrasser, J. Am. Chem. Soc. 123,12886 (2001).

Two-dimensional infrared spectroscopy: A promisingnew method for the time resolution of structures.M. T. Zanni and R. M. Hochstrasser, Curr. Opin. Struc.Biol. 11, 516 (2001).

Two-dimensional IR spectroscopy can be designed toeliminate the diagonal peaks and expose thecrosspeaks needed for structure determination.M. T. Zanni, N. H. Ge, Y. S. Kim and R. M.Hochstrasser, Proc. Natl. Acad. Sci. USA 98, 11265(2001).

Heterodyned two-dimensional infrared spectroscopy ofsolvent-dependent conformations of acetylproline-NH2.M. Zanni, S. Gnanakaran, J. Stenger, and R. M.Hochstrasser, J. Phys. Chem. B 105, 6520 (2001).

Dynamic infrared band-band spectroscopy of periph-eral light-harvesting complexes from R. acidophila. R.Kumble, T. D. Howard, J. Cogdell, and R. M.Hochstrasser, J. Photochem. Photobiol. A 142, 121(2001).

Fluorescence lifetime distribution of single moleculesundergoing Forster energy transfer. M. Lee, J. Tang,and R. M. Hochstrasser, Chem. Phys. Lett. 344, 501(2001).

Temperature-dependent helix-coil transition of an ala-nine based peptide. C. Y. Huang, J. W. Klemke, Z.Getahun, W. F. DeGrado, and F. Gai, J. Am. Chem.Soc. 123, 9235 (2001).


Application for Use of the RLBL

Title:

Keywords (optional)

NIH Axis Numbers (optional)Axis IAxis II

Investigators (PI first) Degree Department / Institution / Address

1.

2.

3.

NIH Support Sources NIH Start/End Date Other Support Sources Grant Number(s) (MM/DD/YY-MM/DD/YY) Agency and Grant Number(s)

1.

2.

Abstract: Describe briefly (200-250 words) the scientific goals and methods.

Logistics: Equipment to be supplied by applicant, needed from RLBL, and anticipated time.

Telephone Number:FAX Number:E-Mail address:Date:Mailing Address:

Send to: Professor R.M. Hochstrasser Director, RLBL Dept. of Chemistry University of Pennsylvania Philadelphia, PA 19104-6323


ADDRESS CORRECTION REQUESTED

Mailing AddressStreet Number and NameCity, State 98765-4321

R L B Lc/o T. TroxlerDepartment of Chemistry, UPenn231S 34th StreetPhiladelphia, PA 19104-6323

NON-PROFIT ORGAN.U. S. POSTAGE

PAID

PERMIT # 2563PHIILADLEPHIA, 4 PA

Documents

THE RLBL NEWSLETTERrlbl.chem.upenn.edu/pdf/rlbl28.pdfported news of the progress and new developments in our Core research areas. For the current Issue we take ... Following this article