BioMolecular Vibrational Spectroscopy: Part 1: Principles of Infrared, Raman Spectra and Techniques Lectures for Warwick CD Workshop, Dec. 2011

Tim Keiderling University of Illinois at Chicago [email protected] T

Tentative Schedule — can vary with interests Part I: • Optical Spectroscopy (general)—low resolution, fast response • Vibrational Theory

– Biologically relevant Vibrational Modes – IR and Raman spectra - structure (qualitative) • IR Instrumentation; FTIR principles • Raman Instumentation • Practical Demonstrations (lab? Break?) – background material • Peptide methods—solution, solid • Protein Sampling Techniques (aqueous), ATR

Part II: • Application Examples T

Structural Biology Optical Spectroscopy is limited for determining structure – lacks site specificity but often fits important QUESTIONS

• often need to know just the conformation • structural determination of fold family may suffice, generally not after atomic structure

• In BioTech processes one must monitor effect of mutation and environmental changes   

need to get this information rapidly and in a cost effective manner Measure all phases/types of samples Look at fast time-scale events

T

Electro-Magnetic Spectrum Electron Excitation Wavenumber (cm-1)

107

Spectral Regions

X-ray

14,285

Near-IR

Electron Transition 106

105

Molecular Vibration 104

103

Ultraviolet

Infrared

4,000

400

Mid-IR

Molecular Rotation 102

10

1

Microwave 100

Far-IR

T

Vibrational Spectroscopy - Biological Applications There are many purposes for adapting IR or Raman vibrational spectroscopies to the biochemical, biophysical and bioanalytical laboratory • Prime role has been for determination of structure. We will focus on secondary structure of peptides and proteins, but there are more – especially DNA and lipids • Also used for following processes, such as enzyme-substrate interactions, protein folding, DNA unwinding • More recently for quality control, in pharma and biotech • New applications in imaging now developing, here sensitivity and discrimination among all tissue/cell components are vital

T

Optical Spectroscopy - Processes Monitored UV/ Fluorescence/ IR/ Raman/ Circular Dichroism Excited State (distorted geometry)

Ground State (equil. geom.)

Diatomic Model

Analytical Methods Absorption UV-vis absorp. hn = E - E & Fluorescence. grd

n0 nS

ex

Fluorescence hn = Eex - Egrd Raman: DE = hn0-hns = hnvib

Infrared: DE = hnvib Q molec. coord.

move e- (change electronic state) high freq., intense

CD – circ. polarized absorption, UV or IR

Raman –nuclei, inelastic scatter very low intensity

IR – move nuclei low freq. & inten. T

Optical Spectroscopy – Electronic, Example Absorption and Fluorescence Essentially a probe technique sensing changes in the local environment of fluorophores

What do you see? (typical protein)

eg. Trp, Tyr

Change with tertiary structure, compactness

 (M-1 cm-1)

Intrinsic fluorophores

Amide absorption broad, Intense, featureless, far UV ~200 nm and below T

ROA

1340 1300

935

R

I -I

L

Optical Spectroscopy - 0IR Spectroscopy

1665

1640

5

4.3 x 10

I + II + I

L

Protein and polypeptide secondary structural obtained from b) jack bean concanavalin A vibrational modes of amide (peptide bond) groups LR

a) human serum albumin Aside: Raman is similar, but different What do you see? – LOTS! amide I, little amide II, intense amide III 2.5 x 109 4

0

1340

-4

935 0.6

5

4.7 x 10

Absorbance

0.5

1220 1241

L

1658 1345

I

1640

II III

LR

0.1

8

R

6.3 x 10

0.0 2000

0

0

1800

1600

1400

1200

1000

1299 Wavenumbers 1342 (cm-11665 ) ROA 9 serum albumin 1554 a) human 1683 2.5 x 10 1462 1677 1295 1316 ROA 5 8

1426

2.6 x 10 1240 9.0 x 10 800 1200 1400 1340 5 1000 1220 ROA 4.7 x 10935 1300 1241 -1 1345 wavenumber / cm

Goal—try to give this meaning L

1665

1300

0.2 b) jack bean concanavalin A

LL

R

L RR

I - I II -+I I

Amide III (1300-1230 cm-1)

1677

1295 1316

0.4 c) hen lysozyme 5 4.3 x 10 0.3

I + II + I

Amide II (1580-1480 cm-1)

0 ROA

0

-2

L R

L R

I -I I -I

Amide I (1700-1600 cm-1)

ROA 8 9.0 x 10

D x 105

R

2

1641

1658 1600 1665

T

Spectroscopic Process (covered) • Molecules contain distribution of charges (electrons and nuclei, charges from protons) which is dynamically changed when molecule is exposed to light

• In a spectroscopic experiment, light is used to probe a sample. What we seek to understand is: – the RATE at which the molecule responds to this perturbation (this is response or spectral intensity – probability of transition)

– why only certain wavelengths cause changes (this is spectrum, the wavelength dependence of the response – energy levels) – the process by which the molecule alters the radiation that emerges from the sample (absorption, scattering, fluorescence, photochemistry, etc.) so we can detect it

T

Spectroscopic Process (covered) • Molecules contain distribution of charges (electrons and nuclei, charges from protons) which is dynamically changed when molecule is exposed to light

• In a spectroscopic experiment, light is used to probe a sample. What we seek to understand is: – the RATE at which the molecule responds to this perturbation (this is response or spectral intensity – probability of transition)

– why only certain wavelengths cause changes (this is spectrum, the wavelength dependence of the response – energy levels) – the process by which the molecule alters the radiation that emerges from the sample (absorption, scattering, fluorescence, photochemistry, etc.) so we can detect it

T

Quantum mechanical picture Full Hamiltonian describes electron and nuclear motion H = - Sab [2/2Maa2 - 2/2mei2 - Zae2/ria + e2/rij + ZaZbe2/Rab ] i.e.

n-KE

e-KE

n-e attr.

e-e repul.

n-n repul

• Born-Oppenheimer approx. separate electron-nuclear w/f y (r,R) = cu (R) fel (r,R) -- product fct. solves sum H • Electronic Schrödinger Equation – issue for CD (done prev.) Hel fel (r,R) = Uel (R) fe (r,R) – electron sol’n – nucl. pot. Vn(R) = Sab [Uel(R) + ZaZbe2/Rab] – nuclear potential energy

• Nuclear Schrödinger Equation Hn cu(R) = -[Sa (ħ2/2Ma) a2 + Vn (R)] cu(R) = Eu cu(R) T

Solving Vibrational QM • Nuclear Hamiltonian is 3N dim. – N atom, move x,y,z – Simplify  Remove (a) Translation (b) Rotation – Result: (3N – 6) internal coordinates  vibration

• Harmonic Approximation – Taylor series expansion: V(R) = V(Re) + Sab V/RaRe(Ra-Re) + ½ Sab 2V/RaRbRe(Ra – Re)(Rb – Re) + …

– 3rd term –non-zero / non-const. - harmonic – ½ kx2 – Ra, Rb mixed  Solution  “Normal coordinates” Qi = Sjcij qj

 H = -Si [2/2 2/Qi2+½ kQiQi2] = Si hi (Qi)

hi ci(Qi) = Ei ci(Qi)  Ej = (uj + ½) hnj solve as if independent Diatomic: n = (1/2p) √k/m k – force const. m = MAMB/(MA + MB)

T

Harmonic Oscillator Model for vibrational spectroscopy E re

r e

q

v=3

5 hn 2 3 hn 2 1hn 2

v=2

v=1

Ev = (v+½)hn IR v=0 Dv =  1 DE = hn n = (1/2p)(k/m)½

Raman 9hn 2 7 hn 2

v=4 r

(virtual state)

hn

re T

Spectral Regions and Transitions • Infrared radiation induces stretching of bonds, and deformation of bond angles – • Couples like motions into molecular mode • (ignore rotations for biomolecules in solution)

symmetrical stretch H-O-H

symmetrical deformation (H-O-H bend)

asymmetrical stretch H-O-H T

Characteristic vibrations and structure • heavier molecules  bigger m - lower frequency • H2 ~4000 cm-1 C–H ~2900 cm-1 C–D ~2100 cm-1 • HF ~4141 cm-1 HCl ~2988 cm-1 • F2

892 cm-1 Cl2

564 cm-1 I–I ~214 cm-1

• stronger bonds – higher k - higher frequency • CC ~2200 cm-1 C=C ~1600 cm-1 C–C ~1000 cm-1 • O=O 1555 cm-1 N O 1876 cm-1 NN 2358 cm-1

• frequency depends mass + bond strength T

Frequency structure, small and large molec.

Same for vibrational modes of amide (peptide bond) groups a

Amide I (1700-1600 cm-1)

b

Amide II (1580-1480 cm-1)

rc Amide III (1300-1230 cm-1)

I

II

For polymer -- repeated structural elements have overlap/coupled spectra T

Vibrational Transition Selection Rules Harmonic oscillator: only one quantum can change

D vi = ± 1, D vj = 0; i  j . These are fundamental vibrations Anharmonicity permits overtones and combinations

Normally transitions will be seen from only vi = 0, since most excited states have little population. Population, ni, is determined by thermal equilibrium, from the Boltzman relationship:

ni = n0 exp[-(Ei-E0)/kT], where T is the temperature (ºK) – (note: kT at room temp ~200 cm-1) T

Anharmonic Transitions Real molecules are anharmonic to some degree so other transitions do occur but are weak. These are termed overtones (D vi = ± 2,± 3, . .) or combination bands (D vi = ± 1, D vj = ± 1, . .). [Diatomic model] E/De

D0—dissociation energy DE02 = 2hnanhrm - overtone

DE01 = hnanh--fundamental ( r - re )/re

T

Vibrational Selection Rules • Interaction of light with matter can be described as the induction of dipoles, mind , by the light electric field, E:

mind = a . E

where a is the polarizability

• IR absorption strength is proportional to  ~ ||2, transition moment between Yi

Yf

• To be observed in the IR, the molecule must change its electric dipole moment, µ , in the transition—leads to selection rules

dµ / dQi  0 relatively easy, ex. C=O str. intense • Raman intensity is related to the polarizability, I ~ 2, where da / dQi  0 for Raman trans. T

Complementarity: IR and Raman

If molecule is centrosymmetric, no overlap of IR and Raman

T

Peak Heights • Beer-Lambert Law: • A = lc –

A = Absorbance

–

 = Absorptivity

–

l = Pathlength

–

c = Concentration

An overlay of 5 spectra of Isopropanol (IPA) in water. IPA Conc. varies from 70% to 9%. Note how the absorbance changes with concentration.

• The size (intensity) of absorbance bands depend upon molecular concentration and sample thickness (pathlength)

• The Absorptivity () is a measure of a molecule’s absorbance at a given wavenumber normalized to correct for concentration and pathlength – but as shown can be concentration dependent if molecules interact T

Peak Widths Water Water

Benzene

• Peak Width is Molecule Dependent

• Strong Molecular Interactions = Broad Bands • Weak Molecular Interactions = Narrow Bands T

Atomic resolution

Ca chain

Structural Biology Level of structure determination needed depends on the problem

Secondary structure

Segment 23 fold (tertiary)

T

Chain conformation depends on f, y angles If (f,y) repeat, they determine secondary structure

Polymer analysis Study the repeat units

Detection requires method sensitive to amide coupling Far UV absorbance broad, little fluorescence—coupling impact small

T

Physical method of detection must sense secondary structure — e.g. couple amides IR/Raman— coupling comparable to band width, intensity maximum is characteristic of structure – frequency basis

Circular dichroism --dipole and through-bond chiral coupling of local modes (excitations)  circularly polarized transitions, DA = AL-AR - Develops characteristic band shapes (intensity)

Theoretically try to understand spectra/structure relation IR ~ D=m.m~|dm/dQ|2 (Raman ~ |da/dQ|2) Major activity, ECD, VCD ~ R = Im(m.m) for analysis!

}

Computable with ab initio QM techniques, ECD needs excited states IR & VCD relatively easy, Raman more basis set sensitive

T

Characteristic Amide Vibrations ~3300 cm-1

A – often obscured by solvent I - Most useful;

~1650 cm-1 IR intense, less interference (by solvent, other modes,etc) Less mix (with other modes)

1500-50 cm-1 II - IR intense mix 1300-1250 cm-1

700 cm-1

III - Raman Intense

IV – VII – difficult to detect, discriminate T

absorbance spectra of selected model peptides

Model polypeptide IR spectra -- Amide I and II

Absorbance

3

Helix—small frequency dispersion, central ones most intense, amide I, higher ones for amide II

helix

I

2

II

b-structure

Sheet—large frequency dispersion, characteristic split amide I, broad amide II

1 random coil

Coil—less well-defined broad amide I and II

0 1750 1700 1650 1600 1550 1500 1450 -1

Frequency based

Wavenumbers (cm )

Differentiation of conformations mostly due to coupling of amides not to H-bonds or other factors, although they contribute

T

Monitoring structural change - temperature Temperature dependent IR spectra of the helical peptide

Temperature dependence of amide I’ frequency unfolded

folded

IR frequency shift shows a sigmoidal curve and spectra have an isobestic point for thermal unfolding However, frequency shift is ~1635  ~1645 cm-1 – solvated helix

T

Tyr92 Tyr115

Tyr97

Tyr73 H1 H2

Ribonuclease A combined uv-CD and FTIR study

H3 Tyr25

Tyr76

Simona Stelea, Prot Sci 2001

• 124 amino acid residues, 1 domain, MW= 13.7 KDa • 3 a-helices  • 6 b-strands in6an  b AP b-sheet b sheet • 6 Tyr residues (no Trp), 4 Pro residues, (2 2 cis,) 2 trans) 29 Optical spectra senses dynamic equilibrium - unstructured systems

T

0.06

FTIR

Absorbance

0.05

Ribonuclease A

0.04 0.03 0.02

FTIR—amide I

0.01

Loss of b-sheet

0.00

1720

1700

1680

1660

1640

1620

1600

Wavenumber (cm-1) 0

Ellipticity (mdeg)

-2

Near –uv CD

-4 -6 -8

Loss of tertiary struct.

-10 -12 -14

Near-UV CD

-16 260

280

300

320

Wavelength (nm)

Far-uv CD

Ellipticity (mdeg)

5

Loss of a-helix

0

-5

-10

Far-UV CD -15 190

200

210

220

Wavelength (nm)

230

240

250

Spectral Change Temperature 30 10-70oC

T

Ribonuclease A

-6.4

1.0

Ci1 (x102)

0.5

-7.2

0.0

-7.6

-0.5

-8.0

-1.0

-5

PC/FA loadings Temp. variation Ci2 (x10)

FTIR

-6.8

FTIR (a,b)

10

-7 5

-5

-13 -15

-10

-17

-15

-10

5 0

-11

-5

Far-UV CD

Ci1

Near-uv CD (tertiary)

Ci2

0

Near-UV CD

-11

-10

-12

-15

Far-uv CD (a-helix)

Ci2

Ci1

-9

-20 -13

-25

Temp. Pre-transition evident in far-uv CD and FTIR, 31 not near-uv CD 0

20

40

60

80

-30 100

Temperature (oC)

T

Nucleic acid IR Nucleic Acids – less variation —helicity all about the same a) – monitor ribose conformation b) – single / duplex / triplex / quad – H-bond link bases

T

Other biopolymers Sugars – little done, spectra broad, some branch appl. Lipids – monitor order – self assemble – polarization Example is CH2 wag, but also stretch and scissor bend are characteristic Self assemble to lipid bilayer – membrane Polarization can tell orientation of lipid or protein in membrane

T

Combining Techniques: Vibrational CD “CD” in the infrared region Vibrational chirality Many transitions / Spectrally resolved / Local Technology in place  DA ~10-5 - limits S/N / Difficult < 700 cm-1 Same transitions as IR same frequencies, same resolution Band Shape from spatial relationships neighboring amides in peptides/proteins Relatively short length dependence AAn oligomers VCD have DA/A ~ const with n vibrational (Force Field) coupling plus dipole coupling Development -- structure-spectra relationships Small molecules – theory / Biomolecules -- empirical, Recent—peptide VCD can be simulated theoretically

T

Poly Lysine in D2O – Amide I’–Secondary structure VCD 2

10

(b)

(a)

0 VCD

0 -5

VCD

-4 -15

1.0

1.0

Absorbance

IR Absorbance

(c)

-10

-10

0.5

0.0 1750

VCD

1.0

IR

IR Absorbance

-5

-2

DA x 105

0

DA x 105

DA x 105

5

5

0.5

0.0 1700

1650

1600

Wavenumber (cm-1)

High pH – helix

1750

1700

1650

1600

0.5

0.0 1750

1700

1650

1600

Wavenubmer (cm-1)

Wavenumber (cm-1)

High pH, heating – sheet

Neutral pH - coil

T

VCD of DNA, vary A-T to G-C ratio base deformations

sym PO2- stretches

-1

big variation

little effect

All B-DNA forms

T

DNA VCD of PO2- modes in B- to Z-form transition B, A

B

Z

Z A

Experimental

B

Theoretical T

Protein RAMAN & ROA spectra III

I

hSA

R

I +I

L

a) human serum albumin

8

9.0 x 10

1340 935

1665

1300

R

I -I

L

ROA 0

1640

5

4.3 x 10

Con A

R

I +I

L

b) jack bean concanavalin A

9

2.5 x 10

1295

1677

1316

R

I -I

L

ROA 0

1220 1241

5

4.7 x 10

1658 1345

c) hen lysozyme

R

I +I

L

HEWL

8

6.3 x 10

1299

ROA

1462

1665 1683

0

R

I -I

1342 1554

L

ROA sign patterns stable but frequencies shift. Chirality selects out amide modes but Raman spectra dominated by aromatics Barron data

1426

5

2.6 x 10 800

1641

1240

1000

1200

T

1400

1600

IR & Raman Instrumentation - Outline • Principles of infrared spectroscopy • FT advantages • Elements of FTIR spectrometer • Acquisition of a spectrum • Useful Terminology • Mid-IR sampling techniques – Transmission – Solids • Raman instrumentation comparison • (Note—more on sampling variations later)

T

Techniques of Infrared Spectroscopy Infrared spectroscopy deals with absorption of radiation-detect attenuation of beam by sample at detector radiation source

Frequency selector

transmitted radiation

detector

Sample

Dispersive spectrometers (old) measure transmission as a function of frequency (wavelength) - sequentially--same as typical UV-vis Interferometric spectrometers measure intensity as a function of mirror position, all frequencies simultaneously--Multiplex advantage T

Comparison of IR Methods – Dispersive & Fourier Transform

But add to this now many laser-based technologies! Nicolet/Thermo drawings

T

New specialized experiments still use dispersive IR T/jump IR with diode laser

Dispersive VCD for Bio Apps

2-D IR setup with 4-wave mixing

T

Major Fourier Transform Advantages • Multiplex Advantage – All spectral elements are measured at the same time, simultaneous data aquisition. Felgett’s advantage. • Throughput Advantage – Circular aperture typically large area compared to dispersive spectrometer slit for same resolution, increases throughput. Jacquinot advantage • Wavenumber Precision – The wavenumber scale is locked to the frequency of an internal He-Ne reference laser, +/- 0.1 cm-1. Conne’s advantage

T

Typical Elements of FT-IR IR Source (with input collimator) – Mid-IR: Silicon Carbide glowbar element, Tc > 1000oC; 200 - 5000 cm-1 – Near IR: Tungsten Quartz Halogen lamp, Tc > 2400oC; 2500 - 12000 cm-1

IR Detectors: – DTGS: deuterated triglycine sulfate - pyroelectric bolometer (thermal) • Slow response, broad wavenumber detection – MCT: mercury cadmium telluride - photo conducting diode (quantum) • must be cooled to liquid N2 temperatures (77 K) • mirror velocity (scan speed) should be high (20Khz)

Sample Compartment – IR beam focused (< 6 mm), permits measurement of small samples. – Enclosed with space in compartment for sampling accessories T

Interference - Moving Mirror Encodes Wavenumber

Source

Detector

Paths equal all n in phase Paths vary interfere vary for different n T

Interferograms for different light sources

T

Dispersive Raman - Single or Multi-channel Eliminate the intense Rayleigh scattered & reflected light -use filter or double monochromator

–Typically 108 stronger than the Raman light

•Disperse the light onto a detector to generate a spectrum

Polarizer Sample

Lens

Filter

Scattered Raman - ns Laser – n0

Detector: PMT or CCD for multiplex Single, double or triple monochromator T

Synchrotron Light Sources – the next big thing Brookhaven National Light Source

Broad band, polarized well-collimated and very intense (and fixed in space!) Light beam output Where e-beam turns

T