Single Molecule Biophysics

fret 1 fret 2

Figure 1.Left: Model of an 8 base-pair RNA with Cy3 and Cy5 dyes attached at the ends. Right: (A) Single-fluorophore-sensitive FRET data from the donor (Cy3, green and symbols) and the acceptor (Cy5, red) of a dye pair attached to the 8mer as shown on the left. (B) FRET ratio calculated from the data and background. (C) Proximity ratio calculated from the data and background

In the last two decades, the development of techniques to measure and manipulate single organic or biological molecules is facilitating new understandings of molecular interactions and dynamics in biophysics, cell and molecular biology, and polymer science. In biophysics and molecular biology, the ability to identify, track, and measure the conformation and interaction of e.g., single proteins or RNA molecules, is leading to new insights in protein folding and RNA interactions and functionality; it is helping to validate models of molecular dynamics; and it is leading to new insights in cell signaling and protein expression in living systems. In polymer science, single-molecule-sensitive techniques aid enormously in understanding transport, flow, and phase transitions.

Our group specializes in the use of optical techniques for physical and spectroscopic measurements on single molecules in biophysical and polymeric systems. Most of our recent work is biophysical in nature, and so we emphasize this in our name. Some examples from our current and past work are given on this page, with our more recent interests near the top of the page.

Single-molecule-sensitive measurements tend to fall into three categories; optical measurements (typically utilizing single molecule fluorescence or nonlinear spectroscopies), electronic measurements (e.g., single ion-channel measurements), and force measurements using optical or magnetic tweezers or microcantilevers. We work with both optical and physical techniques, and sometimes make use of the PALM/STORM microscope facility in our department for superresolution microscopy.

The Nanomechanics of Cellulose Synthesis


Figure 1. Fluorescence images of CESA complexes in living A. thaliana plants. Complexes are labeled with Green Fluorescent Protein to make them visible in these images. Top image is of a typical seedling: the total length of the plant is about 2 cm. (a) An image taken in the root. (b) An image taken in the hypocotyl. The small bright spots are individual cellulose synthase complexes. The larger bright spots that appear mostly out of focus are rapidly moving golgi bodies.

Cellulose, a polysaccharide, is the most abundant biopolymer on earth. This linear chain of D-glucose units was discovered in 1838, but its uses date back much farther: papermaking, for example, has been practiced for almost two thousand years. Cellulose is tough, renewable, biodegradable and biocompatible, all of which make it a desirable starting material for a host of modern inventions. Most commonly we may be familiar with textiles made from cellulose such as rayon and tencel, and the increasing interest in cellulose for the manufacture of biofuels.

Despite its historical significance and continuing industrial potential, the biological mechanism of cellulose synthesis is poorly understood. While it is known that cellulose is produced by large enzymatic complexes located in the plant cell membrane, the structure and mechanism of these complexes are largely unknown.

Towards developing an understanding cellulose synthesis, we study and characterize the mechanics and dynamics of cellulose synthase complexes. Using single-molecule-sensitive microscopy, we can image and track the motion of individual cellulose synthase complexes in living plants. In this way we gather information useful and necessary to understand, and develop a physical model of, cellulose synthesis.

This work is funded by NSF/PoLS 1205989 and is a collaborative effort with Tobias Baskin in the biology department at UMass.

RNA "Kissing" Complexes

R1R2 Rop

Figure 1.Left: Model of the R1inv-R2inv kissing complex. Right: Model of Rop protein, which binds kissing complexes.

Antisense interactions, or hairpin loop-loop “kissing” interactions, are ubiquitous in regulatory and enzymatic function of RNA. It is frequently the case that these complexes are transiently bound by a protein, with structural consequences for the antisense complex. Two examples are under study in our laboratory. Recently we have had success in visualizing a subtle conformational change in the R1inv-R2inv complex upon binding to Rop protein. The R1inv-R2inv system is derived from the RI-RII system from the ColE1 plasmid of E. coli, which is among the earliest models for the regulatory function of RNA (manuscript in preparation).

Our work on this system was done with the help of John Marino at NIST and Lynne Regan at Yale University, and was funded by NSF/MCB 0920139

Free in captivity: Single molecules nanodroplets

Single-molecule sensitive measurements are now commonly used both in vitro and in vivo to better understand the structural transformations and interactions between biological molecules. In vitro studies often involve tethering molecules to an inhomogeneous glass surface or confining them in a heterogeneous gel. The presence of the heterogeneous surface can perturb or inactivate the molecules and modify the photophysics of the attached fluorophores, problems that are widely acknowledged but often ignored.

Aqueous nanodroplets in oil provide a more homogeneous environment for observation of confined molecules (Reiner 2006). In contrast to a surface-attached molecule, droplet-confined molecules diffuse and rotate freely in their nanoenvironment (Tang 2008) . Nanodroplets are relatively easy to generate, manipulate, and mix (Goldner 2010), affording new opportunities for the observation of chemical reactions on a molecule-by-molecule basis (Tang 2009).

Most recently, we have been observing single-molecule-sensitive FRET inside of freely diffusing and optically tracked droplets, and we are developing droplet array devices for parallel measurement of confined molecules. Work on this project is funded by NSF/MCB 0920139 and NSF/DBI-1152386

Home-test Home-Test2

Figure 2. Left:Droplet injector in action, demonstrating droplet mixing. Middle: schematic of experiment involving optical detection and measurement of single droplet-confined molecules. Right: single droplet, approximately 1 micron diameter, in optical trap.

From Photons to FRET: Quantitative Modeling of Fluorescence Resonance Energy Transfer from Cyanine Dyes on RNA

fret 1 fret 2

Figure 3. Left: Output from the MD simulation of a 16 base-pair RNA molecules with Cy3 and Cy5 dyes attached at the 5' ends. Notably, the dyes spend a significant fraction of their time stacked on the RNA, with intermittent brief excursions to unstacked and freely rotating (wandering) states. Right: The result of MD-MC calculations of FRET. (a) and (c) show the histogram of instantaneous FRET values obtained from the MD simulation for 5'(a) and 3'(c) attached dyes. "Forward" and "Reverse" refers to swapping the position of the dyes. In (b) and (d) are shown the FRET histogram predicted from the output of the MC model.

Flouresence resonance energy transfer (FRET) involves the use of two dyes. A "donor" dye is typically excited with a green or blue laser. Once excited the donor might fluorescence, or it might transfer its energy to a sufficiently close "acceptor" dye. The efficiency of this energy transfer, which can be read out as a change in color of the fluorescence (from the bluer donor to the redder acceptor), is referred to as "FRET." To the extent that dyes can be approximated as dipoles, FRET is sensitive to both the distance between dye molecules, and the angle between their transition dipole moments. To a large extent, the angular sensitivity FRET is often ignored by making an assumption that the dyes are freely rotating and so angular dependence can be averaged out.

Using a combination of molecular dynamics (MD) simulations and Monte-Carlo (MC) modeling of the relevant dye photophysics, we recently described how FRET can be predicted (Milas 2013). Using indocyanine dye-labeled RNA molecules for our models, we found that dyes attached to the 3' end of the RNA do, in fact, explore a large range of angles, while 5' attached dyes tend to stick, stacked on the end of the RNA. However, even the 3' dyes are "sticky" in that the "free rotation" of the dyes is better described as intermittent sticking at various positions on the RNA. In either case, the "free rotation" approximation, often used to interpret FRET, fails. Instead, we demonstrate that a "slow rotation" model works better in all cases.

One very recent consequence of this work is that it is now possible to observe subtle angular changes of an RNA "kissing" complex using FRET.

Characterization of a guanosine analogue, 3-MI, for use as a single molecule probe

g-fig1 g-fig2

Figure 2. Left Side: Top: Guanosine. Bottom: 3-MI. Right Side: The count rate per molecule, η, is plotted as a function of the input intensity. Diamonds: stilbene 3. Squares: 3-MI monomer. Asterisks: 36 mer oligo containing 3-MI.

Using Fluorescence Correlation Spectroscopy (FCS), We characterized the guanosine-analog 3-MI [3-Methyl-8-(2-deoxy-β-Dibofuranosyl)isoxanthopterin], a pteridine widely used in studies of DNA binding and dynamics Sanabia 2004 . The photon count rate per molecule, for both monomeric 3-MI and a 3-MI-containing oligonucleotide, and a comparison with a bright laser dye, are shown in Fig. 2, right side. For the monomer, we find a photon count rate per molecule above 4 kHz and a signal to background ratio of 5. For incorporated 3-MI, both parameters are a factor of 4 smaller. We investigated triplet and photobleaching behavior of 3-MI and the possibilities of using this analog in single molecule studies of DNA dynamics. The conclusion was that this is a good dye for use in burst experiments but unless a way can be found to minimize photobleaching, it will probably not work in a single molecule tracking experiment since it photobleaches too fast. Comparisons are made to the behavior of stilbene 3, a brilliant laser dye.

Studies of rotational dynamics of single molecules in Polymer thin films

Using polarization modulation confocal microscopy we measure the absorption dipole orientation and reorientation dynamics of individual dye molecules physisorbed to glass and embedded in thin, spin-cast polymer films under ambient conditions (Weston and Goldner, 2001) . Surprisingly, discrete jumps in absorption dipole orientation were observed for a significant fraction of dye molecules in all samples tested. A sub-population of dye molecules that is stationary on the time scale of these experiments (32 s) is observed and persists even at high excitation power. Fig. 4 shows an image acquired while modulating the polarization of the scanned excitation light, giving rise to apparent stripes that run through stationary molecules and making it very easy to identify molecules that change their orientation. A dependence of the reorientation dynamics on film thickness was identified: DiIC18 molecules reorient with higher frequency and a broader distribution of jump rates in progressively thinner polymer films.

film 1

Figure 4. Demonstration of the technique used to quickly assess the location and fraction of rotationally mobile molecules in a sample. The beginning of each scan line (horizontal) is synchronized with the initiation of polarization rotation. The bright and dark stripes correspond to excitation light parallel to or perpendicular to the absorption dipole orientation of a molecule. Discontinuity in the stripes indicates a molecule has reoriented during imaging. In (a) a 300 pixel × 600 pixel (10 µm × 20 µm) image is shown. In (b)-(g), expanded views of 6 different molecules taken from the image of (a) are shown. The molecule in (b) is stationary while the molecules in (c)-(g) are rotationally active.