Spectroscopies for Structure Andy Howard Biology 555 Created

Spectroscopies for Structure Andy Howard Biology 555: Created 21 April 2014 Presented 24 October 2016 and 13 November 2018

Spectroscopy can be a structural tool • Several types of spectroscopy provide structural information. We’ve already discussed NMR. Today we’ll examine four others.

Agenda • Circular dichroism • • Theory Measurements Application to proteins Helicity Melting Curves Unfolded -> Folded Electron paramagnetic resonance • Concept • Intrinsic signals • Spin labels • Fluorescence • • • Generalities Tyr and trp Ligands Exposures Quenching Mass spectrometry • • Protein MS Electrospray MALDI-TOF Application to proteomics • Is this structural?

Circular dichroism • • This is a form of spectroscopy that capitalizes on the fact that chiral molecules (including polypeptides and polynucleotides) differentially absorb left- and right-circularly polarized light The extent to which they do so is wavelengthdependent, and measurements of those wavelength -dependent differences can tell us something about structure

Polarization of light • • • Remember that, with EM radiation, E and B are always perpendicular to one another and perpendicular to the direction of propagation Linear polarization: E oscillates in only one plane Circular polarization: direction of E rotates about propagation direction but maintains its magnitude Cartoons courtesy Wikipedia

Left- and rightcircular polarization • • • Left-circularly polarized light: the electric vector rotates counter -clockwise Right-circular polarized light: E rotates clockwise It’s possible to filter light to separate out the left- and rightcircularly polarized components

What happens when polarized light interacts with chiral matter? • • • E causes a linear displacement of charge B causes a circulation of charge Together that moves the electron in a helical motion Achiral molecules: every electron that absorbs more left-circular than right-circular will be offset by its mirror image Chiral molecules: there may be a differential absorption, and that differential absorption will be wavelength-dependent

Quantitation • • • We measure difference in absorption ΔA between left and right By Beer’s law ΔA = ClΔε Clearly Δε is wavelength-dependent; that’s why these measurements involve spectroscopy.

Ellipticity as a rotation angle • • Express the difference as the angle associated with this depiction of the difference in transmission This enables us to express the effect either as a CD value Δε(l) or an ellipticity θ(l) in radians or degrees or centi-degrees

Relationship between Δε and [θ] • • Intensity is proportional to |E|2 So θ = (IR½ – IL½) / (IR½ + IL½) But remember that Beer’s law is really an exponential absorption: I = I 0 e-q = I 0 e-Aln 10 But that means IR½ = I 0 e-(AR/2)ln 10 and IL½ = I 0 e-(AL/2)ln 10 Taylor expansion of the θ equation gives θ (radians) = (ΔA/4)(ln 10) or θ (degrees) = (ΔA/4)(ln 10) (180/π) Converting to centidegrees and dividing by concentration gives us [θ] = 100Δε[(ln 10)/4] (180/π) Thus [θ] = 3298. 2Δε

If you have lots of components rotating things… • • If we’re studying a polymer, and many monomers are doing the same kind of differential absorption, we’ll get additive effects on [θ] or Δε Therefore we may want to measure: mean residue ellipticity = [θ]/m, where m is the number of monomers in the polymer

Where would you make these measurements? • • • The largest amount of data can be derived from ranges of wavelengths where the molecule is absorbing a lot anyway For biopolymers this is in the ultraviolet range, typically from 190 to 260 nm. It’s harder to make reliable measurements below 190 nm, largely because oxygen absorbs strongly; even the 190 -200 nm range is tricky

Using CD on proteins • • Proteins are polymers of chiral amino acids All proteins (except poly. Gly!) exhibit a CD spectrum Fingerprint analysis could tell you when the protein is changing conformation… but we should be able to do better than that It turns out that secondary structural motifs have specific CD spectra—especially α-helices

CD spectra for pure secondary structures • Spectra from real proteins: rarely this clean!

Structural information • • A complex protein structure could be decomposed into a linear combination of the spectra shown in the previous slide Software packages exist to do that In practice, separating strands from random coils and disordered structures has limited predictive value So the primary application is to helicity

Practical application I: estimating helicity • • • Comparing an observed spectrum to an idealized alpha-helical protein spectrum can yield an estimate of the percentage of the structure that is helical Results correlate well with helix measurements derived from crystal or NMR structures Doesn’t require crystals or high concentrations or ultra-high purity

Results for Vitreoscilla hemoglobin • These structures are 80 -90% helical

Melting curves • • CD spectrum will be temperature-dependent if the structure changes as a function of temperature CD can be followed to measure unfolding, particularly with helical proteins Vitreoscilla hemoglobin melting curves

Structural transitions • Proteins with minimal structure can sometimes be induced to fold into a more regular structure • Via interaction with a ligand • Via protein-protein interaction • These transitions can be readily followed via CD

Electron paramagnetic resonance spectroscopy • • • An unpaired electron will interact with an external magnetic field via Maxwell’s equations EM radiation will generate an energy difference between low-energy spin state and high-energy spin state for unpaired electron Microwave electromagnetic fields are the correct energy range for resonating with typical differences between low and high

EPR physics

EPR signals from proteins • • • Rate of rotation of a spin influences the resonance behavior of the spin Tryptophan and tyrosine can harbor unpaired electrons under appropriate conditions Creation of these states can be monitored Technique is sensitive (low background) Generally independent of protein size

Spin labels • • If you covalently or noncovalently attach a ligand that contains an unpaired electron to a protein it will produce a signal This can be monitored to enable an analysis of the neighborhood of the spin label

Nature of the labels • • • Often the unpaired electron is on a nitrogen atom Nitroxide (-N=O) is a common instance Nitroxide often incorporated into a ring (e. g. pyrrolidine) to enable residue-specific reactivity

Applications to structure Next several slides from Vanderbilt Univ. structural biology

Mobility correlates with position within the protein

Multiple labels • • If you’ve labeled a protein in more than one place per monomer, you can estimate the distance between the labels from analysis of dipolar coupling between the labels This works over a narrow range of distances— 1 - 2. 5 nm; but that’s a range that’s considerably wider than NMR or XAFS

Distance determination in proteins

Fluorescence spectroscopy • • • Fluorescence is emission of a higherwavelength photon after absorption Signal tends to be sharp and lowbackground Wide applications, some non-structural

Fluorescence for structure • • Trp, tyr, and phe absorb soft-UV reasonably well They re-emit at characteristic wavelengths Those wavelengths and the fluorescent yield are environment dependent, which is what makes this a useful structural tool Tryptophan is the most important Amino acid Lifetime, nsec Absorption λ, nm Absorptivity Fluores. Quantum cence λ, nm yield Trp 2. 6 280 5600 348 0. 20 Tyr 3. 6 274 1400 303 0. 14 Phe 6. 4 257 200 282 0. 04

Using trp fluorescence • • • Fluorescence peak for buried trp is blueshifted (10 -20 nm) and usually more intense compared to exposed trp: less dielectric Quenched by nearby acidic amino acids Unfolding can be monitored by looking at the fluorescence λmax and (for a specific system) at the intensity

Extrinsic fluors • Covalent ligands • Enable probes of micro-environment • Sophisticated chemistries available for attachments to particular groups (primary amines, thiols) • Noncovalent ligands work too • • These will be in (bound unbound equilibrium) Often chosen so that they only fluoresce when bound Often anionic Natural and human-produced

Green fluorescent protein • • • Protein from jellyfish Protein contains ser-tyr-gly post-translationally modified to 4 -(p-hydroxybenzylidine)-imidazolidin-5 -one Intense absorption @395, 475 nm; emits @509 Can be covalently attached at N or C ends to other proteins to use as a fluorescent tag Also used as a cell-component tag Mutants with different spectral properties are available

Mass spectrometry • • • Direct measurement of mass-to-charge ratios is potentially useful We can do this either with intact proteins or with proteolytic digests of proteins Once m/z resolution had reached 0. 1% or better, people started yearning to use MS on proteins

Built-in problem • • MS wasn’t used much on macromolecules until the 1980’s because the macromolecule would fragment when ionized Two techniques for protecting the protein from fragmentation were developed: • Electrospray Ionization (ESI) • Matrix-Assisted Laser Desorption Ionization (MALDI)

Electrospray Ionization • • • Liquid containing the analyte is dispersed by electrospray into a fine aerosol You want a lot of the solvent to disappear, so you mix water with volatile organics like methanol Conductive cosolutes like acetic acid are added to increase conductivity and provide source of ions

ESI-MS, continued • • The tiny droplets produced in ESI can then be subjected to MS analysis As originally developed, it was coupled to a single MS instrument; more often it’s now coupled to tandem (MS – MS) equipment

MALDI • • Sample is mixed with a matrix (organic crystals in acetonitrile or ethanol), applied to a metal plate Pulsed laser irradiates sample, pulling it off the plate Analyte is ionized in a hot plume of gases Ions are accelerated into the MS instrument

MALDI-TOF • • • Typical implementation is a time-of-flight spectrometer, where the instrument records the arrival time of the ions to the detector Often combined with HPLC so that the individual samples are at least somewhat separated by molecular weight or charge Can even be used in identifying bacteria

Coupling MS with proteolysis • • • Standard technique in proteomics: Fragment the protein with a protease Do ESI-MS or MALDI-MS on the fragments Consult a library of existing fragment mass spectra and look for correspondence Can be used on massively complex mixtures Is this really a structural technique? Sort of…

MS protocols

Hydrogen exchange • • As with NMR and neutron diffraction, we can substitute deuterium for hydrogen in our proteins and try to identify which amide protons have been exchanged This method has become reasonably practical as a structural technique, partly because software for identifying the exchanged amide protons is getting better

Other structural approaches • • Chemically cross-link the protein at residues that are close together in space but not in sequence before fragmenting the protein Laser-induced covalent binding to protein will probe how accessible particular parts of the protein are

MS in proteomics