Synchrotron IR Spectromicroscopy: Chemistry of
Hoi-Ying N. Holman, Hans A. Bechtel, Zhao Hao, and Michael C. Martin
Lawrence Berkeley National Laboratory
Advanced analytical capabilities of synchrotron IR spec-
tromicroscopy meet the demands of modern biological
research for studying molecular reactions in individual
living cells. (To listen to a podcast about this article, please
go to the Analytical Chemistry multimedia page at
The ability to modify cellular processes in response to changes
in internal or external environment is critical for maintaining an
efficient functional state in living cells. Much of our quantitative
understanding of cellular molecular reactions has come from
traditional biochemistry experiments that are either averaged over
large populations or performed in vitro with purified biomolecules.
Although these approaches have clarified many detailed mecha-
nisms, they are not sufficient to reveal the phenotypic differences
that exist even within a genetically homogeneous population.
These phenotypic variations are important in fields ranging from
ecology to pathogenesis1-3and may arise from population
heterogeneities in cell cycle phase, cell ageing, epigenetic regula-
tion,4or infrequently and transiently expressed maintenance
genes5or stochastic gene switching.6Furthermore, recent ad-
vances in sequencing and functional metagenomic and metapro-
teomic profiling have yielded cellular genetic “blueprints”, reveal-
ing complex networks of regulatory and metabolic processes. The
chemical reaction components of these networks in living cells
may exhibit both spatial and temporal separationsand are
therefore difficult to simulate in non-biological in vitro systems.
The challenge is to identify those cells of ecological or medical
importance within a large population and track their biochemical
reactions in situ in real time. In this article, we highlight advances
that now allow IR spectromicroscopy to address this challenge.
Central to these advances is the synchrotron light source, which
enables high-throughput noninvasive spectroscopic microanalysis.
This capability can precisely target subpopulations with diffraction-
limited spatial resolution and accuracy to track their chemical
reactions with high molecular specificity without the use of
labelssall at timescales that are associated with important biologi-
THE IR APPROACH
Mid-IR region spectroscopy (∼2.5-15.5 µm wavelength, or
∼4000-650 cm-1wavenumber) is powerful and nondestructive
and provides label-free fingerprint-like spectra originating from
the characteristic vibrational frequencies of various chemical
bonds and, therefore, functional groups.7IR spectroscopy
provides a wealth of chemical information about the sample
without a priori knowledge and has excellent sensitivity to
hydrogen bonding. Changes in hydrogen bond lengths and
Anal. Chem. 2010, 82, 8757–8765
10.1021/ac100991d 2010 American Chemical Society
Published on Web 09/14/2010
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
angles of as little as 0.01 Å or 1°, respectively, can provide clear
differences in a vibrational spectrum.8This specificity and
sensitivity makes the technique an excellent tool for studying
the structure and function of biological macromolecules, which
both affect and are affected by their immediate hydrogen
The application of IR spectroscopy to the study of biological
tissues and cells began >60 years ago when two research groups
reported using a reflecting microscope with mid-IR light to obtain
detailed molecular information in biological specimens.9,10They
demonstrated that even though an IR spectrum is a sum of the
contributions gathered from all biomolecules (i.e., proteins, amino
acids, lipids, and nucleic acids), distinct absorption bands exist
that can be related to known chemical groups in biomolecules.
This idea was soon applied by many others to investigate the
chemical composition of biological samples in the absence of
distinctive morphological features.11-13In spite of the apparent
simplicity of this IR spectroscopy-based microscopy method and
the extensive characterization of spectral fingerprints representa-
tive of biological macromolecules, the approach remained uncom-
mon, while fluorescence microscopy became the popular method
for studying live cellular processes such as the dynamics of gene
expressions. Following Digilab’s pioneer work in FTIR instru-
mentation in 1970s, IR spectroscopy-based microscopy rose again
in the 1990s with the improved speed and sensitivity of fast FTIR
spectromicroscopy (also called microspectroscopy). An excellent
summary of the theory and practice of FTIR spectromicroscopy
can be found in an article by Bhargava and Levin.14Subsequent
improvements in instrumentation and data evaluation methods
(aided by the availability of low-cost, high-speed computers) led
to the robust and reliable use of FTIR spectromicroscopy to study
the chemical composition and structure of complex biological
samples from a range of biological systems including microorgan-
isms, plants, animals, and humans.15
Improvements in two technical parameters have had significant
implications for FTIR spectromicroscopy studies of tissues, intact
cells, and microorganisms: 1) spatial resolution, which determines
the measurement area within the biological sample and therefore
the length scale of the heterogeneity that can be studied and 2)
detection sensitivity, which determines the data collection time
and therefore time resolution for investigating ongoing biological
processes in real time. Both these parameters can be significantly
improved by replacing the thermal emission source (e.g., a globar)
in the conventional FTIR spectromicroscope with a bright syn-
chrotron IR source.16,17
THE BRILLIANCE OF SYNCHROTRON IR
A synchrotron is a high-energy electron storage ring optimized
for the production and collection of light radiated by relativistic
electrons as they traverse a curved path through a magnetic field.
Synchrotron radiation (SR) spans a large electromagnetic spectral
range, beginning in the far-IR and extending through much of
the X-ray region. The synchrotron IR is 100-1000× brighter than
a conventional thermal IR source because the SR is concentrated
in a very small opening angle such that the effective source size
can be very close to an ideal point source.16Interested readers
are directed to an informative overview of SR by Sham and
Rivers.18When coupled to an IR microscope (Figure 1a), the
greater brightness of synchrotron IR light means that the photons
can be focused to a diffraction-limited spot that is dependent on
the numerical aperture (NA) of the microscope objective and the
wavelength (λ) of light. Typical IR microscopes use Schwarzschild
objectives with NA in the range of 0.4-0.7, and spot sizes (and
hence spatial resolutions) of 0.5-1.2 λ are achievable,16,19depend-
ing on definition, optical coupling, and synchrotron facility
The size of the individual cells relative to the size of the SR IR
beam is important in biological investigations. In the mid-IR region,
these 0.5-1.2 λ (i.e., ∼2-10 µm) diffraction-limited spot sizes are
smaller than eukaryote cells, are larger than most of the prokary-
otes and archaea (except for a few mega-bacteria such as the sulfur
bacterium Thiomargarita namibiensis that can reach a diameter
of 700 µm or the thermophilic archaea Staphylothemus marinus
that can grow to 15 µm in diameter), and are comparable to a
small cluster of prokaryote or archaea species. Also, with a
synchrotron source there is no loss in S/N to obtain a diffraction-
limited spot, unlike with a thermal source that must use apertures
to achieve micrometer spatial precision and resolution. Although
many more IR photons are focused on the biological sample when
using a synchrotron source than when using a conventional globar
Figure 1. Basics of SR-FTIR. (a) Schematic of an SR-FTIR
spectromicroscopy experiment, in which the thermal source of a
conventional FTIR microscope is replaced by a synchrotron source.
Mid-IR radiation from a synchrotron is transported to an FTIR
interferometer bench. After modulation by the interferometer, an IR
microscope with reflective optics focuses the beam onto the sample.
The stage is computer controlled and rasters the sample in the x-y
plane with 0.1 µm precision to obtain spectral maps across the
sample. (b) Demonstration of how improved spatial resolution can
enhance detection sensitivity and spatial accuracy of heterogeneous
samples within a larger population. The left panel demonstrates that
with a larger spot size (from a poor brilliance source such as a
conventional thermal emission source), the abnormal (red) circle is
averaged with the normal (blue) circles and can be hidden. In the
right panel, the smaller spot size (from a high brilliance source such
as a synchrotron), the abnormal circle is clearly identified (adapted
with permission from ref. 17; copyright Elsevier). (c) Comparison of
FTIR spectra of a single cell using a 6 × 6 µm2aperture from a
synchrotron source (red trace) and a thermal (blue trace) source. The
collection time for the synchrotron source system was ∼16 s (32
scans) compared to ∼500 s (1000 scans) for the thermal source
system (courtesy of P. Dumas).
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
source, the low photon energy and low peak powers (relative to
the alternative IR laser sources) of <∼50 mW have shown no
detectable effects on living cells.20
This improved spatial resolution and S/N achievable with
synchrotron IR light are of great benefit to measurements
involving heterogeneous samples, such as tissues or communities
of microbes. As illustrated in Figure 1b, the measurement is
targeted to a small area, more accurately pinpointing which a part
of the sample gives rise to a feature of interest.17In Figure 1c,
the improvement in S/N and time resolution of SR-FTIR is
demonstrated by comparing FTIR spectra from a single cell, as
measured with a synchrotron and with a conventional thermal
source using a 10-µm aperture. In this example, the collection
time using the synchrotron source was only 16 s (32 scans)
whereas the collection time using the thermal source was 500 s
(1000 scans). Even after extensive averaging, the quality of the
thermal source measurement is not sufficient to reveal the fine
molecular features within the vibrational spectrum. Indeed, prior
to the incorporation of synchrotron light, the low S/N made it
exceptionally difficult to apply conventional IR spectroscopy to
follow the dynamics of cellular processes.
The high SR IR brightness coupled with its noninvasive nature
has already made a significant impact in the fields of biomedi-
cine,21environmental ecology, and geochemistry.22In an early
study from our group, time-lapse SR-FTIR reflectance measure-
ments showed the mechanism by which a small subset of
microbes on a basalt rock specimen survived exposure to a high
concentration of toxic chromium(VI) when most of the surround-
ing microbial cells were killed.23Although geological materials
inherently have very rough surfaces, the small spot size of the
SR IR source allowed us to measure a small region containing a
handful of microorganisms, revealing that they were reducing
chromium(VI) to chromium(III). This capability of SR-FTIR also
yielded new insights into how the soil bacteria Mycobacterium sp.
JLS solubilize large recalcitrant organic pollutants such as poly-
cyclic aromatic hydrocarbons and metabolize them into biomass.24
This understanding of the dynamic microbial processes in
simulated geological environments has contributed to the design
of environmental cleanup strategies.
WATER: THE GOLDILOCKS PROBLEM (GETTING
IT JUST RIGHT)
Water presents one of the primary challenges of using IR
spectroscopy in studying living cells, even with the bright
synchrotron source. Water strongly absorbs mid-IR light and even
the absorption due to a thin layer of water can completely
dominate the spectrum. However, water is necessary for life and
is the most common ingredient (>70%) in living cells. So it is
essential to get the optical thickness of water “just right”: enough
to support life and ensure the validity of model systems, but not
so much that it masks the molecular signatures of interest.
Typically, <10 µm of bulk water is preferred to exploit the
molecular information across the full mid-IR spectral range.
Historically, most applications of real-time synchrotron IR spec-
tromicroscopy to the study of cellular processes in living cells
were limited to hydrated microbial systems maintained in moist
microchambers23-25or experiments using isotope labeled
A traditional approach to minimize water absorption during
living cell experiments is attenuated total reflectance (ATR) FTIR
equipped with a flow chamber. In ATR FTIR, the IR beam
internally reflects off a high index crystal (e.g., diamond or
germanium), and the attenuation of the IR beam caused by the
absorption of the evanescent wave is measured and analyzed.28
For living cell experiments, cells are cultured on ATR crystals in
a specially designed flow chamber that can carefully control the
culturing/experimental conditions (pH, temperature, ionic strength,
flow shear stress, and delivery of materials such as nutrients and
drugs) for hours.29-31The evanescent field formed at the cell/
ATR crystal interface has a typical penetration depth of <1 µm,
thereby reducing the optical path length in water. However, this
small penetration depth, though important in reducing the path
length through water, prevents the ATR FTIR technique from
examining chemical and biological processes in cells within
biological systems with an extensive extracellular matrix. Another
important consideration is the effect of growing the cells on the
ATR crystal. Many cells require a substratum for normal growth,
and the use of an ATR crystal may affect extracellular matrix
properties that are important in cellular phenotype. This penetra-
tion-depth limitation makes it less than ideal for use in studies of
the resistance of bacterial biofilms to antibiotics, the resistance
of lung cancer cells to chemotherapy, or the directional movement
and proliferation of colon cancer during metastasis.
We used a more straightforward microfluidic approach in-
stead.32Figure 2 shows an early version of our open-channel
microfluidic device. It was fabricated on a silicon chip using deep
reactive ion etching to create hydrophilic microstructures 10-15
Figure 2. Microfluidic platform to minimize water absorption. (a)
Microfluidic SR-FTIR microscopy platform design and setup. (b) Plane
view depiction of an example chip with several parallel etched
microstructures for multiple simultaneous experiments. (c) Flow maps
of the microstructures, simulated from experimentally measured paths
of near-neutral density polystyrene beads (yellow arrows) and
superimposed on a snapshot image. Top: flow in a microchannel.
Bottom: flow in a microwell. Scale bars ) 10 µm. Velocity ) ∼60 µm
s-1(adapted from ref. 32).
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
µm deep and at least 40 µm wide. A continuous thin-film (<10
µm) laminar flow is maintained by balancing the hydrostatic
pressure in a microliter-sized feeder droplet at the microchannel
inlet with the capillary forces present at the outlet. In this setup,
the bulk flow rate in the open microchannel is typically ∼60 µm
s-1and is controlled by adjusting the elevation of the off-chip
reservoir that supplies the feeder droplet. Mid-IR photons
emitted from the synchrotron are focused onto the targeted
areas within the microstructures, and SR-FTIR measurements
are made by the transflection or transmission modes. To image
the biochemical properties and the distribution of bacterial
activity across a biofilm, it is raster scanned with equal
micrometer-sized steps, collecting a full SR-FTIR spectrum at
each position. The ability of the platform to provide a sustain-
able environment for SR-FTIR measurements of living bacteria
has been confirmed by using the reporter bacterium E. coli
Several other groups have reported uncomplicated flow cham-
ber techniques.33,34They commonly placed living cells between
two parallel IR transparent windows (e.g., CaF2) that are sepa-
rated by a spacer of <12 µm to minimize the aqueous media
absorption. Most recently, a more sophisticated closed-chamber
microfluidic device has been fabricated by etching channels
and wells into two CaF2crystals.35Although effective in its
simplicity, we found that the spectra from closed-chamber or
closed-channel devices are often marred by complicated sinu-
soidal patterns of interference fringes that probably arise from
interference between multiple reflections from the parallel
surfaces. In our device, the sloping free interface between water
and air, which is produced by balancing the hydrostatic
pressure in a microliter-size feeder droplet at the microchannel
inlet with the capillary forces present at the outlet, has allowed
us to completely eliminate the interference fringes during FTIR
THE RIDDLES OF BACTERIAL BIOFILMS
How do bacteria in biofilms survive antibiotics that should
kill them? A bacterial biofilm is a population of cells growing on
a surface and enclosed in a self-produced polymeric matrix in an
aqueous environment. Bacteria within a biofilm can become up
to 1000× more resistant to antibiotics than free-floating bacteria
of the same species,36yet only ∼1% of their genes show differential
expression between these two populations.37
Figures 3a-c show a dramatic example of using open-channel
microfluidics with SR-FTIR spectromicroscopy to study why the
antimicrobial agent mitomycin-C (MMC) does not kill some E.
coli in biofilms.32MMC is activated only after it has entered a
cell, where it is reduced to a hydroquinone form that covalently
cross-links to guanine residues in DNA to form DNA-MMC
adducts.38The presence of DNA-MMC adduct signals at ∼986
cm-1reflects cellular uptake and action of MMC.32Figure 3b
shows chemical images before and after the introduction of MMC
into the microfluidic system. Despite an uninterrupted supply of
MMC-laden water, the SR-FTIR image plot shows that MMC
uptake is highly localized. DNA-MMC adduct signals are
especially strong in areas that were either closer to the MMC
source (circles 1 and 2) or in areas formerly rich in protein amide
III (∼1310 cm-1). In areas with little DNA-MMC adduct, the
protein amide III signal increased with MMC exposure. Shown
in Figure 3c are analyses of vector-normalized SR-FTIR spectra
(over the 900-1800 cm-1region) at different locations; they also
indicate localized small (∼10-20 µm) spatial-scale changes in
biochemical contents consequent to MMC exposure. We
speculate that the localization of MMC may be caused by flow
heterogeneities inside the biofilms, perhaps related to internal
flow diversions around regions with higher protein amide III
signal. The heterogeneous spectral features and behavior,
however, may also reflect localized cellular diversification
processes in response to MMC toxicity, such as metabolic
modification and/or migration to more favorable living areas.
Bacterial real estate: sustainable or a bubble? Our group
also used the open-channel SR-FTIR platform to improve the
understanding of the growth and development of bacteria on
surfaces, a more subtle but growing topic of research interest.
Many microbial processes important in pathogenesis and ecology
are initiated in microscopic spaces, and recent reports indicate
that the colonizing bacterial cells actively seek out confined spaces
where biofilm initiation, formation, and evolution could be
influenced by fluid dynamics, nutrient supplies, or waste removal.
We compared spectroscopically the dynamics of biofilm forma-
tion under two different conditions: in microchannels with higher
rates of nutrient supplies and waste product removal and in
microwells with lower rates of nutrient supplies and more waste
accumulation.32Figure 4 shows the remarkable differences. For
microchannels of high nutrient supplies and waste removal, four
different markers of biofilm formation increase asymptotically:
Figure 3. Interactions of antibiotics with E. coli. (a) Bright-field optical
images of the 1 day old E. coli biofilm before (0 h) and during (8 h)
0.15 µg mL-1concentration MMC exposure. (b) SR-FTIR chemical
image plots of the intensity before (upper panels) and during (lower
panels) MMC exposure at ∼985 cm-1(DNA-MMC adducts), ∼1310
cm-1(protein amide III), and ∼1080 cm-1(polysaccharides). Three
distinct biochemical regions formed during MMC treatment: a high
MMC-uptake and a low MMC-uptake region separated by an area of
high polysaccharide content. Arrows indicate MMC stream direction.
Circles in panels mark selected locations. Scale bars ) 10 µm. (c)
Preprocessed SR-FTIR spectra at selected locations before (blue)
and during (red) MMC exposure (adapted from ref. 32).
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
polysaccharide (∼1080 cm-1), glycocalyx (∼1130 cm-1), DNA/
RNA (∼1240 cm-1), and protein amide III (∼1310 cm-1)
(Figure 4a). This result provides chemical evidence that biofilm
formation proceeds via multiple convergent genetic pathways. The
abundance of glycocalyx carbohydrates in sites with large biofilm
growth confirms the belief that glycocalyx synthesis is crucial to
the formation of bacterial biofilms.39In contrast, biofilm formation
in microwells exhibits cyclic growth patterns as indicated by the
sequential rise and fall of the different biomolecule signals (Figure
4b). The lack of spectroscopic evidence of glycocalyx suggests
that biofilm attachment was weaker in the microwell. A very recent
discovery is that some bacteria can produce a factor (mixtures of
D-amino acids) that could prevent biofilm formation and cause
the breakdown of existing biofilms.40Future applications of the
SR-FTIR approach to this biological consideration may add a
dimension of understanding to the question: why do some bacteria
prosper in biofilms whereas others “move away”?
BIOLOGICAL CHEMISTRY AS SEEN BY
HYDROGEN BONDS IN CELLULAR WATER
IR sensitivity to structured water molecules constitutes a major
advantage of IR spectromicroscopy over other vibrational methods
such as Raman spectroscopy methods and enables the detection
to the hydrogen bond structure that links cellular water molecules
to ions or other small molecules is the key to this new approach,
which works well for several reasons. First, >70% of a cell’s contents
of water is a sensitive reflection of the cellular chemical environment
as radicals, small organic acids, and hydrogen gas, which are
expected to be present during the functional metabolism of stress
adaptive response.41,42Furthermore, the IR spectrum of OH stretch
vibrations in the hydride-OH dominated stretch region (1900-3800
cm-1) has already been widely used to characterize the dynamics
of hydrogen-bonding structures in both water clusters and
condensed phases such as aqueous liquids. These studies can
be used to guide the interpretation of IR spectra of OH stretch
vibrations in the hydride-OH region and to link the variations to
the presence of ions and other small molecules in liquid or other
condensed phases (for discussion and references, see ref. 44).
Unraveling how descendants of ancient bacteria cope with
the stresses of the modern world. We have used SR-FTIR to
study how descendents of some of the oldest bacteria can survive
oxygen stress from the modern Earth atmosphere. Microbes, such
as the ubiquitous sulfate reducing bacteria (SRB), are the oldest
(>3.5 billion years) and the smallest (about 1/8000th the volume
of a human cell) living organisms on Earth. Strong selection
pressure over time and their high surface-to-volume ratio may have
enabled these species to cope with many environmental fluctua-
tions. Genome sequencing shows that some SRB descendents
have acquired the genetic “blueprints” to survive transiently in
an oxygenated atmosphere. This includes the potential capabilities
to reduce O2 and cope with the subsequent toxic reactive
oxygen species (ROS) in the modern atmosphere, such as
superoxide anion radicals (O2•-), peroxides (H2O2), and hy-
droxyl radicals (HO•-).
In the case of our model SRB, Desulfovibrio vulgaris Hilden-
borough, only a small number of cells in a population can survive
transient exposure to atmospheric oxygen (Figure 5a). These few
“survivors” often exist individually or in small groups of several
cells to form localized microscopic communities. They must be
interrogated without spectroscopic influence from neighboring
cells in the population that are dying. We also know from
transcriptomics analysis that the time scales of the oxygen
adaptive response processes can potentially range from minutes
to hours.43In order to avoid long data acquisition times and thus
capture the response dynamics at shorter time scales, a brilliant
IR source providing high S/N is required.
The first step to study the chemistry of adaptive survival
response is to identify a priori the few survivor cells in the D.
vulgaris population. Our previous experience shows that only cells
that are entering the early stationary phase of the cell cycle and
have accumulated an internal reserve of polyglucose and elemental
of polyglucose-accumulated D. vulgaris (Figure 5b and c) show the
distinct spectral features of the nonglycosidic polyglucose vibration
(υC-OH) band between 1055 and 1045 cm-1and the glycosidic
linkage vibration (υC-O-C) at ∼1175 cm-1(Figure 5b). This
spectral information is used to look for the potential survivors in a
thin layer of D. vulgaris clones that are maintained in an oxygen-
free moist atmosphere inside the microscope-stage environmental
chamber. By rastering the diffraction-limited synchrotron IR beam
across the thin film, one can identify spots where the IR spectrum
Figure 4. Biofilm dynamics. SR-FTIR time course analyses and
chemical images of biofilms (a) in a microchannel and (b) in a
microwell (as shown by four molecular markers at ∼1080 cm-1
(polysaccharides), ∼1130 cm-1(glycocalyx), ∼1240 cm-1(DNA/RNA
polysaccharides), and ∼1310 cm-1(protein amide III). Unlike the
microchannel data, in which signal intensity of key biomolecules
appeared to approach an asymptotic state, the microwell SR-FTIR
data were cyclic (cell growth and release). The chemical image plots
obtained after the second cycle show locally higher signal intensities
of protein amide III and DNA/RNA polysaccharides near the microwell
center, whereas the polysaccharide matrix accumulated near the
microwell edge. There is little spectroscopic evidence of glycocalyx
facilitating strong adherence to the microwell substrate. Scale bars
) 10 µm (adapted from ref. 32).
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
exhibits spectral features similar to the typical IR spectrum of
polyglucose-bearing living D. vulgaris.
In order to survive oxygen exposure, these polyglucose-bearing
bacteria must be metabolically active before exposure. Metabolic
activity can be verified by immediately making a short, sequential,
uninterrupted series of SR-FTIR measurements. The cells are
metabolically active if the analysis of time-difference spectra
(derived from the series of real-time spectra) reveals positive time-
difference absorption features (bands at ∼3190 cm-1and ∼3645
cm-1and a shoulder feature at ∼3745 cm-1) of water molecules
H-bonded with hydrogen gas.44Hydrogen gas production is
selected as an indicator because it is consistent with the central
metabolism of D. vulgaris under anaerobic conditions.
The graph in Figure 6a is an overview of the striking molecular
changes in D. vulgaris survivors during their exposure to
atmospheric oxygen. The trends seen in Figure 6b and 6c, which
were derived from spectrally integrated absorption intensities of
polyglucose and water bands,44reveal an unanticipated multiphasic
pattern. These two molecules are of immediate interest because
the reduction of dioxygen to water through aerobic respiration
of internal polyglucose reserves was previously believed to be the
primary step in air-tolerant SRB. Comparing trends in these two
figures shows that from 0 < t < ∼50 min, there is a substantial
decrease in the polyglucose band intensity (green inverted
triangles in Figure 6b), but little change in the water band intensity
(green inverted triangles in Figure 6c). At t > ∼50 min, the water
band intensity increases abruptly, whereas the rate of polyglucose
disappearance continues unabated until slowing down distinctly
later (t > ∼100 min). We elucidate the mechanism(s) underlying
this puzzling multiphasic behavior by analyzing the time-difference
spectra in the hydride-OH region.44
Figure 6d is a 2D time-frequency contour plot of the time-
difference spectra (negative values are shown in dark blue), with
difference spectrum snapshots beneath (Figure 6e). The contour
map reveals the remarkable changing absorption features from
υOH as a function of time, which reflects a cascade of chemical
reactions (Figure 6f) that sheds light on genetically controlled
pathways.44A summary of these time-dependent chemical events
and their consistency with some known putative events of oxygen-
stress adaptive response is highlighted in Figure 7. A comparison
indicates that during the initial phase (t < ∼50 min), the cells
consume polyglucose (green inverted triangles in Figure 6b) to
produce acetate (a two-carbon carboxylate; green circles in Figure
6e) that initially accumulates. Chemical reactions diversify during
the intermediate phase (∼50 < t < ∼150 min). At t > ∼50 min, the
spectra show two dominant types of chemical reactions. (i)
Carboxylates (acetate) are converted to CO2(compare the green
circles for acetate and the blue triangles for CO2in Figure 6e),
which coincides with increasing water content (see the inverted
green triangles in Figure 6c). This coincidence suggests an onset
of an ATP generating pathway. (ii) ROS begin to accumulate,
signifying that their rate of formation exceeds their removal by
protective enzymes and other mechanisms in D. vulgaris (red
squares in Figure 6e). Then, at t ≈ 70 min, a striking new spectral
feature indicates the formation of sulfate anions, which surprisingly
coincides with both the disappearance of ROS (compare the pink
Figure 5. Survivor SRB. (a) Viability of SRB after exposure to moist
air for 8 h. Live (green) and dead (red) were assessed using reagents
in LIVE/DEAD BacLight Bacterial Viability kit. (b) Typical IR absorption
spectra of stationary-phase (red) and exponential-phase (blue) D.
vulgaris (c) Transmission electron microscopy images of thin sections
poststained by the periodic acid thiosemicarbazide-osmium method
show intracellular polyglucose granules in stationary-phase but not
exponential-phase D. vulgaris (adapted with permission from ref. 44,
copyright 2009 National Academy of Sciences, U.S.A.).
Figure 6. Oxidative stress and adaptation. (a) Typical real-time SR-
FTIR spectra of D. vulgaris during the transition from an anaerobic
to aerobic environment. Sequential spectra are offset upward for
clarity. Because all spectra are derived using air as a reference, the
abrupt change in the spectral feature at ∼2348 cm-1is associated
with the presence of atmospheric CO2. Typical time-course of IR
intensity (normalized by the maximum value) of (b) polyglucose
content and (c) water. (d) Time versus frequency contour plot of SR-
FTIR time-difference spectra in the hydride-OH dominated stretch
region. (e) Snapshots of time-difference spectra for selected different
time points. The dashed line marks zero difference absorbance. (f)
Typical time-course of IR intensity (normalized by the maximum value)
of H-bonded species. (Bars: ( 10% error.) (Adapted with permission
from ref. 44, copyright 2009 National Academy of Sciences, U.S.A.)
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
hexagons to the red squares in Figure 6e) and an increase of water
content (green inverted triangles in Figure 6c). Others have
reported that D. vulgaris can potentially oxidize its accumulated
elemental sulfur and other reduced sulfur compounds and that
the oxidation is by means of an ATP/adenyl sulfate pathway that
couples the sulfate ion formation with oxygen reduction to water.
For longer times (t > ∼150 min), the intensity of the υOH band
of the water· · ·ROS bond declines (red squares in Figure 6e),
which suggests an improved ROS removal rate in the survivor
The mechanisms by which SRB survive transient exposure to
atmospheric oxygen have been controversial, partly because of
the approach of averaging over a large population. Our SR-FTIR
technique provides the first direct molecular observation of the
reduction of dioxygen to water by a small number of SRB through
aerobic respiration of internal polyglucose reserves as one crucial
step and identifies the transient chemical reactions that allow the
survivor subpopulation to adapt to extreme changes in its
INTERROGATING SUB-CELLULAR CHEMISTRY
IN LARGE LIVING CELLS
As described above, SR-FTIR is an excellent tool for interrogating
chemical reactions in living bacterial cells. Many of the techniques
described here for studying bacteria may be extended to study
subcellular chemical processes in larger cells and tissues. For
example, other research groups have used SR-FTIR to probe
subcellular chemistry in living cells that are several tens to several
hundreds micrometers in size. Jamin et al.25first demonstrated
the advantage of a synchrotron source by imaging the subcellular
components of single intact mammalian cells. After a time, Moss
et al. combined SR-FTIR with a flow-through system to examine
the growth of single live colorectal cancer cells over several
hours.45Metabolite formation in living unicellular algae (Chlamy-
domonas reinhardtii) was also recorded over time using recently
improved instrumentation technology.27In addition, several
groups reported the observation of subcellular chemical changes
in living cells as the cells responded to environmental changes.
For example, Heraud et al. mapped nutrient-induced biochemical
changes in living algal cells (Micrasterias hardyl) and reported
that the largest changes occurred in the chloroplast compart-
ments.46The Gough group reported significant chemical changes
in fungal hyphae committed to spore development.47More
recently, researchers in the Dillion group observed arsenic-
induced changes in membrane and secondary protein structures
in living HL60 cells.48Although these examples are frontier
breaking, they have only begun to scratch the surface of the
potential of SR-FTIR for studying cellular chemistry inside sub-
compartments of living mammalian cells, plant cells, and many
other eukaryotic cells.
Synchrotron IR spectromicroscopy shows great potential for
studying biochemical reactions in living microbial cells when
conventional methods of averaging over a large population are
inadequate. The diffraction-limited spot size of synchrotron IR
enables us to investigate chemistry in bacteria and archaea and
at a subcellular level in cells tens to hundreds of micrometers in
size. Examination of the riddles of biofilms and investigation of a
stress-adaptive response in live cells of different sizes illustrate
that SR-FTIR spectromicroscopy is particularly powerful for
systems that cannot be well understood by conventional ap-
proaches. From a careful analysis of the real-time multiple-
dimension spectra with time resolutions from seconds to days,
SR-FTIR spectromicroscopy studies can identify key chemical
reactions underpinning phenotypic diversity and individuality in
cells within a population, with applications to ecology, pathology,
and molecular medicine.
Recent advances in sequencing and functional metagenomic
and metaproteomic profiling have elucidated the “blueprints” for
cellular genetic, regulatory, and metabolic processes. These
analyses comprise an emerging base that allows research to focus
on a genome-based understanding and, through that understand-
ing, predict how cellular systems and their metabolic processes
shape the foundation of a biological system. Whereas previous
efforts in biochemistry and structural biology have focused on
determining the structure and function of individual cellular parts,
we now need to develop methods to probe the cell’s response to
chemical and physical perturbations, to identify cell-cell differ-
ences, and to detect the presence of rare cells or rare cellular
events in a population of many cellssall with sufficient spatial,
temporal, and chemical resolution. This requires a noninvasive
chemical imaging technology with spatial resolution at the scale
of individual whole cells and subcells and temporal resolutions
compatible with the rates of important chemical reactions in
One logical next step in the development of IR spectromicros-
copy is the improvement of imaging speed. This is required for
Figure 7. A summary of the evolving cellular chemical environment and possible survival mechanisms inside the same living D. vulgaris as in
Figure 6 during its transient oxygen-stress and adaptive response, as revealed by SR-FTIR measurements and analyses. Polyglucose is labeled
as PolyG; scale bar ) 0.5 µm. (Adapted with permission from ref. 44, copyright 2009 National Academy of Sciences, U.S.A.).
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
studying multiple subpopulations simultaneously within an iso-
genetic population and for developing a fundamental understand-
ing of population-level diversity. Currently, even with the huge
improvements in S/N from a synchrotron source, mapping large
areas at high spatial resolution with the raster scanning technique
can be extraordinarily time intensive. For example, imaging a 100
× 100 µm area with a 3-µm step size can take nearly 20 min. Using
longer average measurement times to improve the S/N can easily
increase the acquisition time to hours or days! Focal plane array
(FPA) technology has increased the speed of image acquisition
by multiplexing the acquisition of data, but low S/Ns have limited
its applicability to living cell measurements. One possible solution
is to couple the high brightness of synchrotron radiation with
these new detectors, but achieving the expected advantage
remains a challenge.49Presently, many of the over thirty syn-
chrotron IR facilities worldwide are investigating this prospect.
The IR ENvironmental Imaging (IRENI) Facility at the Synchro-
tron Radiation Center (SRC) at the University of Wisconsin
Madison is the first FPA-based SR-FTIR system available to users
(http://www.src.wisc.edu/ireni/). The technology is still in its
infancy, but it offers exciting possibilities for rapidly unraveling
biochemistry in living cells and tissues.
Another logical direction is to integrate the SR-FTIR technology
with high-density microfluidic chips with plumbing networks.50
This advance will enable the automatic and precise manipulation
of fluids to provide the “just-right” aqueous dimensions for
circumventing obfuscation by water absorption and to provide the
accurate “on-demand” management of environmental conditions
to study the chemistry of living cells. This will permit investiga-
tions of many important cellular systems in aqueous environments
over long time periods, including harmful processes such as those
underlying chronic bacterial infections or beneficial processes
such as those enabling biofuel energy production by microbes.
This work was supported by the U.S. Department of Energy
Office of Biological and Environmental Research’s Structural
Biology Program through contracts DE-AC02-05CH11231 and
KP1501021 with Lawrence Berkeley National Laboratory. The
Advanced Light Source is supported by the Director, Office of
Science, Office of Basic Energy Sciences, of the U.S. Department
of Energy under contract DE-AC02-05CH11231.
Hoi-Ying N. Holman is the Director of the Berkeley Synchrotron Infrared
Structural Biology (BSISB) Program at the Advanced Light Source
(ALS), the Head of Chemical Ecology in the Earth Sciences Division
(ESD), and a Virtual Institute for Microbial Stress and Survival (VIMSS)
Investigator. Holman’s research focuses on the development and applica-
tion of technology to study the chemistry of living cells by integrating
spectroscopy, microscopy, and microfluidics. She received her Ph.D. in
Environmental Chemistry and Chemical Engineering from University of
California at Berkeley. Hans Bechtel is a Senior Scientific Engineering
Associate at the ALS IR beamlines and has over 13 years of experience
with spectroscopic methods and analysis. Bechtel received his Ph.D. in
Physical Chemistry from Stanford University and was a postdoctoral
researcher in the Spectroscopy Laboratory at the Massachusetts Institute
of Technology. Zhao Hao is a Senior Scientific Engineering Associate at
the Chemical Ecology and the ALS infrared beamlines. Hao’s current
research focuses on multimode imaging, plasmonic metamaterials, and
spectroscopic analysis and modeling. Hao received his Ph.D. from the
Institute of Physics, Chinese Academy of Sciences, was the recipient of
the 2000-2001 STA Fellowship from Japan’s Science and Technology
Agency, and taught general physics at Florida State University while
working at the National High Magnetic Field Laboratory before joining
ALS. Michael Martin is the Deputy Leader of the Scientific Support Group
at the ALS. Martin received his Ph.D. in Physics from the State University
of New York at Stony Brook. All are currently at Lawrence Berkeley
National Laboratory. Address correspondence to H.-Y. N. Holman at MS.
70A-3317L, Lawrence Berkeley National Laboratory, One Cyclotron Road,
Berkeley, CA 94720; email@example.com.
(1) Harrison, J. J.; Ceri, H.; Roper, N. J.; Badry, E. A.; Sproule, K. M.; Turner,
R. J. Microbiology 2005, 151, 3181–3195.
(2) Lewis, K. Nat. Rev. Microbiol. 2007, 5, 48–56.
(3) Losick, R.; Desplan, C. Science 2008, 320, 65–68.
(4) Avery, S. V. Nat. Rev. Microbiol. 2006, 4, 577–587.
(5) Zalzman, M.; Falco, G.; Sharova, L. V.; Nishiyama, A.; Thomas, M.; Lee,
S. L.; Stagg, C. A.; Hoang, H. G.; Yang, H. T.; Indig, F. E.; Wersto, R. P.;
Ko, M. S. H. Nature 2010, 464, 858–866.
(6) Acar, M.; Mettetal, J. T.; van Oudenaarden, A. Nat. Genet. 2008, 40, 471–
(7) Parker, F. S. Applications of Infrared, Raman, and Resonance Raman
Spectroscopy in Biochemistry; Plenum Press: New York & London, 1983.
(8) Jeffrey, G. A. An introduction to hydrogen bonding; Oxford University Press:
New York, 1997.
(9) Blout, E. R.; Mellors, R. C. Science 1949, 110, 137–138.
(10) Barer, R.; Cole, A. R. H.; Thompson, H. W. Nature 1949, 163, 198–201.
(11) Wood, D. L. Science 1951, 114, 36–38.
(12) Woernley, D. L. Cancer Res. 1952, 12, 516–523.
(13) Stevenson, H. J.; Bolduan, O. E. Science 1952, 116, 111–113.
(14) Bhargava, R.; Levin, I. W. Anal. Chem. 2001, 73, 5157–5167.
(15) Salzer, R.; Siesler, H. W., Eds.; Infrared and Raman Spectroscopic Imaging;
Wiley-VCH: Weinheim, 2009.
(16) Carr, G. L.; Reffner, J. A.; Williams, G. P. Rev. Sci. Instrum. 1995, 66,
(17) Dumas, P.; Sockalingum, G. D.; Sule-Suso, J. Trends Biotechnol. 2007, 25,
(18) Sham, T. K.; Rivers, M. L. Applications of Synchrotron Radiation in Low-
Temperature Geochemistry and Environmental Sciences 2002, 49, 117–147.
(19) Levenson, E.; Lerch, P.; Martin, M. C. Infrared Phys. Techn. 2008, 51, 413–
(20) Holman, H. Y. N.; Bjornstad, K. A.; McNamara, M. P.; Martin, M. C.;
McKinney, W. R.; Blakely, E. A. J. Biomed. Opt. 2002, 7, 417–424.
(21) Miller, L. M. In Infrared and Raman Spectroscopic Imaging; Salzer, R.,
Siesler, H. W., Eds.; Wiley-VCH: Weinheim, 2009, pp 510.
(22) Holman, H.-Y. N.; Martin, M. C. In Adv. in Agronomy; Sparks, D., Ed.;
Elsevier: New York, 2006; Vol. 90, pp 79-127.
(23) Holman, H. Y. N.; Perry, D. L.; Martin, M. C.; Lamble, G. M.; McKinney,
W. R.; Hunter-Cevera, J. C. Geomicrobiol. J. 1999, 16, 307–324.
(24) Holman, H. Y. N.; Nieman, K.; Sorensen, D. L.; Miller, C. D.; Martin, M. C.;
Borch, T.; McKinney, W. R.; Sims, R. C. Environ. Sci. Technol. 2002, 36,
(25) Jamin, N.; Dumas, P.; Moncuit, J.; Fridman, W. H.; Teillaud, J. L.; Carr,
G. L.; Williams, G. P. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 4837–4840.
(26) Cameron, D. G.; Martin, A.; Mantsch, H. H. Science 1983, 219, 180–182.
(27) Goff, K. L.; Quaroni, L.; Wilson, K. E. Analyst 2009, 134, 2216–2219.
(28) Iwamoto, R.; Ohta, K. Appl. Spectrosc. 1984, 38, 359–365.
(29) Hutson, T. B.; Mitchell, M. L.; Keller, J. T.; Long, D. J.; Chang, M. J. W.
Anal. Biochem. 1988, 174, 415–422.
(30) Suci, P. A.; Vrany, J. D.; Mittelman, M. W. Biomaterials 1998, 19, 327–
(31) Kuimova, M. K.; Chan, K. L. A.; Kazarian, S. G. Appl. Spectrosc. 2009, 63,
(32) Holman, H.-Y. N.; Miles, R.; Hao, Z.; Wozei, E.; Anderson, L. M.; Yang, H.
Anal. Chem. 2009, 81, 8564–8570.
(33) Nasse, M. J.; Ratti, S.; Giordano, M.; Hirschmugl, C. J. Appl. Spectrosc.
2009, 63, 1181–1186.
(34) Tobin, M. J.; Puskar, L.; Barber, R. L.; Harvey, E. C.; Heraud, P.; Wood,
B. R.; Bambery, K. R.; Dillon, C. T.; Munro, L. K. Vib. Spectrosc. 2010, 53,
(35) Birarda, G.; Grenci, G.; Businaro, L.; Marmiroli, B.; Pacor, S.; Vaccari, L.
Microelectron. Eng. 2010, 87, 806–809.
(36) Stewart, P. S.; Franklin, M. J. Nat. Rev. Microbiol. 2008, 6, 199–210.
(37) Whiteley, M.; Bangera, M. G.; Bumgarner, R. E.; Parsek, M. R.; Teitzel,
G. M.; Lory, S.; Greenberg, E. P. Nature 2001, 413, 860–864.
(38) Portugal, J.; Sanchezbaeza, F. J. Biochem. J. 1995, 306, 185–190.
(39) Costerton, J. W.; Irvin, R. T.; Cheng, K. J. Annu. Rev. Microbiol. 1981, 35,
(40) Kolodkin-Gal, I.; Romero, D.; Cao, S.; Clardy, J.; Kolter, R.; Losick, R. Science
2010, 328, 627–629.
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010
(41) Woutersen, S.; Bakker, H. J. Nature 1999, 402, 507–509.
(42) Cowan, M. L.; Bruner, B. D.; Huse, N.; Dwyer, J. R.; Chugh, B.; Nibbering,
E. T. J.; Elsaesser, T.; Miller, R. J. D. Nature 2005, 434, 199–202.
(43) Mukhopadhyay, A.; Redding, A. M.; Joachimiak, M. P.; Arkin, A. P.; Borglin,
S. E.; Dehal, P. S.; Chakraborty, R.; Geller, J. T.; Hazen, T. C.; He, Q.; Joyner,
D. C.; Martin, V. J. J.; Wall, J. D.; Yang, Z. K.; Zhou, J.; Keasling, J. D. J.
Bacteriol. 2007, 189, 5996–6010.
(44) Holman, H. Y. N.; Wozei, E.; Lin, Z.; Comolli, L. R.; Ball, D. A.; Borglin, S.;
Fields, M. W.; Hazen, T. C.; Downing, K. H. Proc. Natl. Acad. Sci. U.S.A.
2009, 106, 12599–12604.
(45) Moss, D. A.; Keese, M.; Pepperkok, R. Vib. Spectrosc. 2005, 38, 185–191.
(46) Heraud, P.; Wood, B. R.; Tobin, M. J.; Beardall, J.; McNaughton, D. Fems.
Microbiol. Lett. 2005, 249, 219–225.
(47) Jilkine, K.; Gough, K. M.; Julian, R.; Kaminskyj, S. G. W. J. Inorg. Biochem.
2008, 102, 540–546.
(48) Munro, K. L.; Bambery, K. R.; Carter, E. A.; Puskar, L.; Tobin, M. J.; Wood,
B. R.; Dillon, C. T. Vib. Spectrosc. 2010, 53, 39–44.
(49) Petibois, C.; Piccinini, M.; Guidi, M. C.; Marcelli, A. J. Synchrotron Radiat.
2010, 17, 1–11.
(50) Thorsen, T.; Maerkl, S. J.; Quake, S. R. Science 2002, 298, 580–584.
Analytical Chemistry, Vol. 82, No. 21, November 1, 2010