Microscopy in Infectious Disease Research
—Imaging Across Scales
Vibor Laketa
Department of Infectious Diseases, Virology, University of Heidelberg, Germany
German Center for Infection Research, partner site Heidelberg, Germany
Correspondence to Vibor Laketa: Department of Infectious Diseases, Virology, University Hospital Heidelberg, Im
Neuenheimer Feld 344, D-69120 Heidelberg, Germany. vibor.laketa@med.uni-heidelberg.de
https://doi.org/10.1016/j.jmb.2018.06.018
Edited by P-Y Lozach
Abstract
A comprehensive understanding of host–pathogen interactions requires quantitative assessment of molecular
events across a wide range of spatiotemporal scales and organizational complexities. Due to recent technical
developments, this is currently only achievable with microscopy. This article is providing a general perspective
on the importance of microscopy in infectious disease research, with a focus on new imaging modalities that
promise to have a major impact in biomedical research in the years to come. Every major technological
breakthrough in light microscopy depends on, and is supported by, advancements in computing and
information technologies. Bioimage acquisition and analysis based on machine learning will pave the way
toward more robust, automated and objective implementation of new imaging modalities and in biomedical
research in general. The combination of novel imaging technologies with machine learning and nearphysiological
model systems promises to accelerate discoveries and breakthroughs in our understanding of
infectious diseases, from basic research all the way to clinical applications.
© 2018 Elsevier Ltd. All rights reserved.
Why microscopy?
Microscopy and infectious disease research have
been inseparable partners ever since in the 1670s
Anthonie van Leeuwenhoek used his newly invented
microscope to examine a sample of plaque he had
scraped from his own teeth, and observed—for the
first time—bacteria and other microorganisms that
share the world with us [1] (Fig. 1a). It took 200 years
until the work of Robert Koch and Louis Pasteur in
the second half of the 19th century demonstrated
that bacteria can cause the diseases. Microscopy
was used to examine the samples of infected people
and animals, and it quickly became an indispensable
research tool to investigate infectious diseases. In
contemporary photographs, both Koch and Pasteur
are often shown “as peeking through a microscope,”
and their microscopes are displayed in museums
around the world.
Due to their small size, viruses were not visible
with the contemporary light microscopes, so they
were not discovered until the late 1890s. However, it
was only after the first image of a virus (mouse
ectromelia virus) was acquired and published by
Helmut Ruska and colleagues in 1938, using an
electron microscope [2] (Fig. 1b), that the scientific
establishment finally accepted the existence of
viruses as particles and not as “toxins or elements
of ferment nature,” as in the view held by many at the
time. This was substantiated in the subsequent two
decades when, owing to improvements in electron
microscopy (EM) technology and sample preparation,
EM micrographs revealed the beautiful structure of T
bacteriophages [3] (Fig. 1c). These microscopybased
discoveries in the mid-20th century firmly
opened the door to the field of virology. Undoubtedly,
microscopy was fundamental in infectious disease
research, as it was necessary for the discovery of
infectious agents by direct observation and later on
also for their diagnosis. However, microscopy in those
early days of infectious disease research did not yet
reveal its full potential.
A big boost to microscopy came when computer
technology and digital detectors (such as charged-
0022-2836/© 2018 Elsevier Ltd. All rights reserved. J Mol Biol (2018) 430, 2612‐2625
Review
coupled device cameras) had sufficiently matured
to be used for microscopic imaging. By coupling
the microscope to a digital detector and a computer
[4, 5], the microscope was transformed from a tool
that “only” magnifies an image to an exact, quantitative,
analytical instrument. Another breakthrough was the
introduction of fluorescence microscopy, dramatically
increasing the contrast and allowing the observation of
sub-resolution structures specifically stained with
fluorescent labels (although they could still not be
spatially resolved until recently).
Based on development of new imaging modalities,
fluorescent probes and sensors, computer technology,
image analysis algorithms and new biological model
system, the field of biological microscopy is undergoing
a revolution. Only now we are starting to appreciate
and make use of the full potential of this microscopybased
approach. Modern microscopes have evolved
into complex robotic instruments capable of automatically
interrogating a wide range of biological processes
at vastly different spatiotemporal scales and organizational
complexities, from structural studies on macromolecular
scale all the way to whole organ/body
imaging in living animals (Fig. 2).
Microscopy is also an integral component of
modern medicine, where it provides a diagnostic
tool for many bacterial, fungal and parasitic infections
(light microscopy) [6] as well as for viral
infections (EM) [7]. In many developing regions,
light microscopy is the most important procedure for
the diagnosis of infections, facilitated by new lowcost,
smartphone-based microscopes to be used
in the field [8]. However, in this overview, I will
concentrate on the role light microscopy-based
imaging currently has in infectious disease research
specifically focusing on novel imaging modalities, as
well as highlight the challenges for the future.
The goal of infectious disease research is to obtain
a comprehensive understanding of replication,
spread and pathophysiology of infectious agents in
order to develop new and better medical treatments.
Classical biochemical, genetic and genomic approaches
have been employed over the years to
yield important insights in host–pathogen interactions.
Most of these experimental approaches are
population-based (“bulk”), end-point analyses where
obtained information represents an average across
the population and where important parameters can
be missed as they become “averaged out” in the bulk
measurement. It is important to realize that the
different states of biological systems, such as, for
example, the differences between the health and
Fig. 1. Examples of first microscopy
utilization in the discovery of
infectious agents. (a) Early illustrations
of dental plaque bacteria by
Antonie van Leeuwenhoek in 1683
(adapted from Ref. [1]). (b) Electron
micrograph of mouse ectromelia virus
by Helmut Ruska and colleagues in
1938, the first virus whose size and
shape were shown to the scientific
world (adapted from Ref. [2]).
(c) Electron micrograph showing
intricate structural organization of
the T bacteriophage (adapted from
Ref. [3]).
2613Review: Tackling Host‐Pathogen Interactions with Imaging
disease, are ultimately governed by individual,
stochastic and often rare molecular events. To truly
understand different physiological states, we need to
employ a reductionist experimental approach that is
able to capture and quantitatively examine these
individual molecular events. Due to recent innovations,
light microscopy is particularly suited to
sample complex spatio-temporal dynamics of living
systems at sufficient resolution, and thus able to
provide realistic representations of the biological
systems. It is a big technical challenge to observe
and quantify these stochastic molecular events. An
even bigger challenge is to understand how they
give rise to the different typical patho-physiological
states that we observe in “bulk” measurements. In
this respect, microscopy-based analysis yields
quantitative information that can be integrated into
predictive multiscale mathematical models of infection
that will hopefully be used to reconcile the
findings obtained by the reductionist approach with
the ones that are based on population
measurements.
Super-resolution light microscopy
For a long time, the observation of molecular and
cellular determinants relevant for host–pathogen
interactions was hampered by the inability of a
conventional microscope to spatially resolve any two
objects that are closer together than the Abbe
diffraction limit. In simple terms, if two objects are
closer than approx. 200 nm, they will not be seen as
two individual objects in a conventional microscope,
but as one, and no further details can be seen below
this limit. This fundamental limit of light microscopy
cannot be overcome by building better optics.
However, recently developed super-resolution imaging
techniques have found ways to circumvent
this limit, by only allowing a subset of molecules to
emit fluorescent light at any one time. This can be
done by using patterned light to spatially modulate
the fluorophore emission [e.g., in stimulated emission
depletion microscopy (STED) [9, 10]] or by
temporally modulating the emission via fluorophore
“blinking” [in single-molecule localization microscopy
(SMLM) [11, 12]]. Since spatial resolution in STED
microscopy scales inversely with the power of the
depletion laser and in SMLM it is determined by the
number of detected photons per blinking event [13],
there is no more fundamental spatial resolution limit
of fluorescence light microscopy. Indeed, a method
able to localize single molecules with 1-nm precision
and resolving emitters only 6 nm apart has been
reported [14]. However, due to technical challenges
connected mainly with fluorophore photostability and
labeling density, the resolution in biological imaging
typically achieved with super-resolution methods is
in the range of 20–40 nm.
Fig. 2. Imaging across scales and organizational complexities. Development of new imaging modalities have enabled
visualization and quantification of infectious diseases processes with high spatiotemporal resolution and at scales
spanning many orders of magnitude in size—from nm to mm. Much of the infectious disease research was done on the
cellular monolayers using conventional microscopy technologies. Now, using CLEM and super-resolution microscopy, it is
possible to “zoom in” further and get the relevant information on molecular and structural level. But it is also possible to
“zoom out” (using SPIM) and examine the same processes in more physiological setting, in the context of organoid, organ
or living animal. (Individual images courtesy of medical illustrations database—https://smart.servier.com/)
2614 Review: Tackling Host‐Pathogen Interactions with Imaging
Fig. 3 (legend on next page)
2615Review: Tackling Host‐Pathogen Interactions with Imaging
The development of super-resolution technologies
provided a way forward across the field of biomedical
sciences but had an especially significant impact on
the infectious disease research. Not only is it now
possible to examine the host–pathogen interactions
on a nano-scale, such as human immunodeficiency
virus (HIV)-induced microdomain organization on
the plasma membrane of host cells during virus
budding [15, 16] or intracellular membrane reorganization
during hepatitis C virus (HCV) infection [17],
it is also possible to examine molecular determinants
on the pathogens themselves. This was demonstrated
by showing Env receptor clustering during
maturation of HIV virions [18] (Fig. 3a), by localization
of the ESCORT machinery within HIV particle
during budding from producer cells [19] and by
structural organization analysis of herpex simplex
virus [20]. Furthermore, super-resolution microscopy
was used to determine 3D macromolecular structures
of large cellular complexes such as the nuclear
pore complex [21] and viruses [22, 23] by making
use of particle averaging techniques originally
developed for cryo-EM [24]. To facilitate the 3D
super-resolution microscopy-based structural analysis
of viruses, an open-source software package that
combines super-resolution imaging and singleparticle
analysis to generate unbiased molecular
models of intact viruses or viral substructures has
been reported [25]. Although still far from the
resolution limit of EM, light microscopy has been
transformed into a new tool for structural biology and
now can be used to tackle questions that were
previously reserved exclusively for EM and X-ray
crystallography studies with the benefit of higher
throughput, specific labeling and possibility to obtain
native structures “in situ”.
However, as the resolution of the light microscopy
is approaching that of the EM, some of the
considerations well established in the EM field are
becoming relevant for light microscopy-based approaches.
All super-resolution imaging modalities
require high density of labeling. Nyquist criterion
dictates that the fluorophores should not be separated
by a distance larger than half of the desired
resolution. One consideration is that the labeling
density necessary for extracting structural information
by super-resolution imaging might induce
unwanted, non-physiological macromolecular organization
due to sterical interference between exogenously
introduced dyes and tagged fluorescent
proteins or due to non-physiological effects connected
to high fluorescent protein overexpression. The
second consideration is that chemical fixation
typically employed in light microscopy is resulting
in structural artifacts [26, 27] and is incompatible with
high-resolution structural analysis. An interesting
solution to this is to immobilize the biological material
by vitrification process—a technique that allows
maintaining the water content of cells by freezing
them to a temperature around −190 °C within
milliseconds. The fast freezing process prevents
the water from crystalizing into hexagonal ice
(occupies a volume 10% larger than liquid water)
and forms a glass-like layer of amorphous vitreous
ice. Vitreous ice occupies the same volume as water
and therefore preserves cellular structures in nearnative
conditions without inducing structural changes
[28]. This sample immobilization approach is typically
employed in cryo-EM but not yet established for the
light microscopy approaches. Super-resolution light
microscopy under cryogenic conditions is a very
promising new imaging modality with various technical
challenges to overcome. Besides near-native
structure preservation, light microscopy of vitrified
samples will benefit from increased fluorophore
photostability under cryogenic conditions [29]. It
has been demonstrated that the single-molecule
localization-based super-resolution approach is feasible
under cryogenic conditions and able to yield
resolutions below the diffraction limit of the light [30].
Furthermore, when coupled to cryo-electron tomography,
the approach was able to identify different
conformations of macromolecular complexes involved
in a bacterial secretory system [31].
Another fundamental advantage of super-resolution
light microscopy over EM is the ability to examine
dynamic processes in living specimens. Despite pilot
experiments reporting successful super-resolution
Fig. 3. Examples of applications of new imaging modalities in infectious disease research. (a) HIV budding sites (Gag,
green; Env, red) on the surface membrane of HeLa cells shown in conventional microscopy (left) and super-resolution
microscopy (STORM) (right). Only using super-resolution microscopy is it possible to realize the structure of the viral
budding sites where Env molecules accumulate around the Gag cluster rather than directly co-localizing with it. The scale
bar represents 200nm (Images courtesy of Walter Muranyi and Hans-Georg Kräusslich, Department of Infectious
Diseases, Virology, University of Heidelberg, Germany). (b) Fluorescence image of GFP-tagged HCV subgenomic
replicon overlaid with EM image in a correlative approach. Areas of high fluorescence intensity correlate with the sites of
complex HCV-induced membranous compartments essential for HCV pathogenesis. (Image courtesy of Inés RomeroBrey
and Ralf Bartenschlager, Department of Infectious Diseases, Molecular Virology, University of Heidelberg,
Germany). (c) Organoid, derived from human small intestine cells, expressing innate immunity sensor IRF3-GFP shown in
bright field (right) and imaged using SPIM (left). Upon sensing of viral infection, the IRF3-GFP sensor will translocate from
the cytosol into the nucleus. With SPIM, it is possible to analyze this process within the context of complex close-tophysiological
model system. The scale bar represents 50 μm. (Images courtesy of Megan Stanifer and Steeve Boulant,
Department of Infectious Diseases, Virology, University of Heidelberg, Germany).
2616 Review: Tackling Host‐Pathogen Interactions with Imaging
live cell imaging in infectious disease research [32,
33], this application remains challenging. The necessary
high sampling frequency in space and time
required to obtain a super-resolution image invokes
huge demands on the photostability of fluorescent
reporters. There is a clear need for better, brighter,
more photostable, fluorophores especially in the far
red part of the visible light spectrum. Better fluorophores
have been continuously developed over the
past years and this will continue in the future [34].
However, another, more challenging, aspect to
consider is the effect of increased light irradiation
during live cell super-resolution imaging on normal
cellular physiology. Higher sampling density induces
more cellular photodamage and perturbs the biological
processes under examination [35]. Therefore,
even if we would have the hypothetical ideal
fluorescent probe, it would not necessarily directly
translate into successful live cell super-resolution
imaging. One way to prevent photo-induced cellular
damage is by using various scavengers of reactive
oxygen species (ROS) such as Trolox [36] and similar
compounds. Another way would be to design better
model systems for live cell imaging. The processes
behind light-induced cell death are not known. It is
assumed that ROS generated by fluorophore
photobleaching and triplet-state interactions are a
major source of photodamage, but it is not clear if that
is the only process. All the sources of ROS
species and the exact mechanisms are not clear yet.
There is evidence that some cell types are more
photodamage-resistant than others and therefore
more suitable for live cell imaging. For example,
HeLa cells can be challenged with order-of-magnitude
higher light doses than U2OS or COS-7 cells without
apparent photodamage [35]. It is perhaps the time to
investigate in detail what is the molecular and
physiological basis for this resistance. Once we
know the genetic basis for the resistance then, with
the help of CRISPR–Cas9 technology [37], relevant
biological model systems could be genetically
modulated to become more photodamage-resistant.
Lastly, the total light exposure could be reduced by
implementation of innovative illumination schemes. A
new illumination concept for STED microscopy has
recently been developed that reduces the sample
exposure to the STED beam by actively modulating
the STED power on per pixel basis depending on the
spatial distribution of the fluorophores in the sample so
that it only uses as much on/off-switching light as
needed to image at the desired resolution [38].
With this illumination scheme, it was possible to
reduce the sample exposure to light by up to 2 orders
of magnitude depending on the sample [38]. Another
novel, super-resolution live cell imaging concept has
been demonstrated that uses consecutive sequence
of reversible, on-stage cryo-arrests followed by superresolution
image acquisition within living, but arrested,
cells [39]. Many issues connected to live cell superresolution
approaches will be mitigated by optimizing
the microscopy hardware; however, physical limits will
be reached that cannot easily be overcome. Novel
computational procedures that improve the quality of
acquired microscopy images and enable biological
observations beyond the physical limitations of
microscopes are becoming increasingly important. A
purely analytical approach has been described that is
able to reconstruct super-resolved images from a
sequence of 100 images acquired in a standard wide
field or confocal microscope by taking into account
temporal fluorophore intensity fluctuations and
achieving resolutions in the range of 70 nm [40].
Using deep learning networks, it was possible to
restore microscopy images even when using 60 times
less photons during acquisition [41]. This remarkable
achievement will allow for acquisition of low signal-tonoise
super-resolved images using minimal light
exposure but without the compromise on the final
output and could become standard in biological image
restoration in the years to come. Hardware developments
together with the design of new photostable
fluorescent probes, image restoration by machine
learning and photodamage-resistant model systems
will pave the way to more robust super-resolution live
cell imaging.
Correlative light-electron microscopy
Despite the advances in super-resolution technologies
described above, EM still remains the method of
choice when ultrastructural analysis (below 10 nm) of
isolated viral particles or virus-induced intracellular
structures needs to be conducted. EM analysis is
usually performed on cellular volumes that are only
50–300 nm thick and a few microns wide. Finding a
molecule or a molecular event of interest in such
minute volume is a major challenge and a limitation of
an EM approach. This is especially relevant for
infectious disease research where one often needs
to look for very rare or very transient molecular events.
Examples are latent HIV infections, which only affect
one in a million CD4+ T cells in vivo [42], unmodified
HCV, which will only replicate in one out of hundred
thousand hepatocyte-derived cells (Huh7) [43], or
HIV-reporter HeLa cells, which internalize hundreds of
HIV particles, but only a few virions undergo reverse
transcription and eventually enter the nucleus [44].
Finding the proverbial needle in a haystack is a major
bottleneck in addressing questions in infectious
disease research. Another limitation of the EM is the
difficulty to distinguish a molecule of interest from the
surrounding cellular context which is also visualized.
Specifically labeling molecules in EM by immunogold
methods is possible but only available for a small
subset of molecules where we are lucky enough to
have a working antibody–antigen combination and is
usually only possible on the section surface. Correlative
light-electron microscopy (CLEM) has been
2617Review: Tackling Host‐Pathogen Interactions with Imaging
developed to overcome these limitations. In CLEM
approach, the sample is first examined by fluorescent
light microscopy, exploiting the light microscopy
advantages such as the possibility to specifically
label molecules of interest with fluorescent probes,
track dynamic and transient events as well as examine
large fields of view. Subsequently, the same sample
and the same object of interest are examined by EM to
provide the detailed structural information and cellular
context at ultrastructural resolution. This way CLEM
combines the strengths of fluorescent light microscopy
with those of the EM and is particularly suitable for
“needle in a haystack” type of experiments as well as to
complement the analysis of dynamic, transient molecular
events with ultrastructural analysis. For these
reasons, the CLEM approach proved to be extremely
powerful in addressing the open questions in the field
of infectious disease research. Although rudimentary
implementations of correlating light and electron
imaging modalities to address host–pathogen interactions
has been around since 1985 [45], the power of
the CLEM approach for infectious diseases research
was really demonstrated in proof of principle studies
where single, fluorescently labeled, HIV-1 particles
were observed on the cellular surface during the
internalization process using the light microscopy
followed by high-precision localization of the same
particles and ultrastructural analysis by 3D EM [46]. In
recent years, a similar approach has been employed to
analyze molecular composition, ultrastructural architecture
and spatiotemporal development of virusinduced
intracellular alterations such are “viral replication
factories” induced by HCV [47, 48] (Fig. 3b),
dengue virus [49] and semliki forest virus [50], nuclear
aggregates induced during herpesvirus varicellazoster
virus assembly [51] or virological synapse
during retrovirus murine leukemia virus spread [52].
CLEM can also be combined with super-resolution
imaging to derive high precision correlation between
light and electron imaging modalities as it was done in
the study of the molecular machinery and underlying
ultrastructural morphology involved in HIV-1 budding
[19]. Light microscopy can be incorporated in the EM
workflow at almost any stage; however, two main
CLEM approaches can be distinguished—when
light microscopy is performed before embedding
(pre-embedding CLEM) and after embedding (postembedding
CLEM). The main advantage of preembedding
CLEM is that the light microscopy and
EM can be performed using conventional light
microscopy and EM protocols resulting in generation
of high-quality light microscopy/EM data. Main disadvantage
is that the structures of interest may change
between light microscopy and EM steps resulting in
low-accuracy correlation. Post-embedding CLEM provides
excellent correlation but a compromise must be
found in the sample preparation procedure between
optimally preserving sample structure, and optimally
preserving the fluorescent signal essentially yielding
sub-optimal light microscopy and EM data quality.
Regardless of which CLEM approach is employed, the
use of embedding resins and heavy metal stains
prevents achieving nanometer to sub-nanometer
structural resolution. Furthermore, chemical fixation
of samples within the CLEM workflow could introduce
structural artifacts [26, 27]. This led to a development of
a special implementation of CLEM called cryo-CLEM
which performs the entire correlative workflow under
cryogenic conditions. The sample is first immobilized
by vitrification [28], followed by consecutive light and
electron microscopic analysis, while keeping the
sample in a vitrified state during the entire procedure.
In cryo-CLEM, contrast is generated directly by the
electron density differences within the molecules, while
objects of study and their history or specific state are
first identified by their fluorescence profile. By using
vitrification and avoiding resin embedding as well as
any heavy metal staining procedures, cryo-CLEM
permits the morphological and structural characterization
of defined biological objects near their native state
and at the highest possible resolution. Cryo-EM
combined with single-particle averaging/subtomogram
averaging is capable of determining macromolecular
structures at resolutions typically around 3–4 Å,
although resolutions below 2 Å (resolution where
individual carbonyl bonds and aromatic rings are
resolved) have been reported [53]. The same approach
has been used to solve the structure of
immature HIV-1 capsids [54] and mature Dengue
virus [55] at 3.9- and 3.5-Å resolutions, respectively.
Furthermore, cryo-EM has been used to determine the
structure of large macromolecular complexes such as
26S proteasome [56] and nuclear pore complex [57] at
below 30-Å resolution directly in vitrified cells (“in situ”)
by single-particle averaging/subtomogram averaging
of approximately 1000 particles. It is expected that
the combinations of cryo-CLEM and single-particle
averaging/subtomogram averaging approaches will be
able to yield 3D structures of relevant macromolecular
complexes for infectious disease research “in situ” and
at single-digit nanometer resolution, which will have a
major impact on the molecular understanding of host–
pathogen interactions.
Selective plane illumination microscopy
Molecular biology studies of host–pathogen interactions
are often performed in monolayer or suspension
cells of established transformed cell lines. It
is becoming increasingly clear that such systems are
fundamentally different and often only partially
recapitulate the relevant physiology of the complex
biological processes as they occur in infected
patients or animals. Relying solely on such model
systems creates a risk of fabricating an oversimplified
and potentially skewed view on these fundamentally
important patho-physiological processes.
Therefore, there has been a strong tendency in
2618 Review: Tackling Host‐Pathogen Interactions with Imaging
biomedical research in recent years to move toward
more physiologically relevant model systems of
increasing complexity (such as complex 3D cell
and organotypic cultures, organs, tissues, entire
animals or samples from biopsies of patients) [58].
For a long time, confocal and multi-photon microscopy
was the only method that allowed for
microscopy-based research on complex 3D model
systems. Confocal and multi-photon microscopy
technology continues to evolve with major improvements
introduced at a steady pace by commercial
providers and academia alike. Confocal microscopy
has entered the super-resolution regime with introductions
of new detectors and image restoration
routines (Airyscan by Zeiss, HyVolution by Leica and
OSR by Olympus). Developments in adaptive optics
ensure better signals and resolutions at increased
depth [59–61]. Confocal and multi-photon microscopy
is still the method of choice when examining nontransparent
model systems that extend up to 1–2 mm
along the optical axis or for microscopy performed in
living rodents (intravital microscopy) [62]; however,
these methods are fundamentally limited in speed and
employment of high-power lasers is incompatible with
high spatial/temporal sampling or long-term observation
of living samples due to phototoxicty considerations.
Visualization and quantification of pathogen
behavior in the complex 3D systems requires microscopy
methods that allow for high temporal and spatial
resolution with minimum phototoxicity. In recent years,
selective plane illumination microscopy (SPIM, also
known as lightsheet microscopy) has emerged as the
technical advance that provides the essential basis for
such studies. SPIM is a fluorescence microscopy
technique where only a thin slice (usually a few
hundred nanometers to a few micrometers) of the
sample is illuminated perpendicular to the direction of
observation [63]. As only the actually observed
section is illuminated, this method reduces photodamage
and stress induced by light on a living
sample. In fact, this is the most gentle fluorescence
microscopy technique to date. Because SPIM scans
samples by a plane of light instead of a point and
employs a high-sensitivity camera for detection, it can
acquire images with high signal-to-noise ratios at
speeds several orders of magnitude faster than those
offered by point-scanning methods such as confocal
and multi-photon microscopy. The SPIM approach
has revolutionized developmental biology research by
allowing for long-term, high-resolution observation of
developing embryos [64]. Together with advancement
of tissue clearing methods [65, 66], it led to the
generation of detailed, single-cell resolved, highly
multiplexed models of complex organs relevant for
infectious disease research such as the lymph node,
brain, liver and so on [66]. Despite the mentioned
benefits of SPIM technology and the fact that it
has been available for more than a decade, the
potential of SPIM technology has not been fully
exploited in infectious disease research. SPIM has
been employed in proof-of-principle studies to analyses
HIV-1 assembly sites dynamics in host cells [33]
or parasite Toxoplasma gondii behavior in 3D
epithelial model systems [67]. It has also been used
to study zika virus (ZIKV) tropism in cerebral
organoids [68] and Trypanosoma brucei life cycle in
tsetse fly [69]. However, there are several potential
reasons for low application of the SPIM in infectious
disease research. The SPIM approaches are difficult
to implement in general due to high demands on IT
and computational infrastructure that are necessary to
support the storage, transfer, visualization and analysis
of data sets ranging from 100s of GB to multiple
TB in size that are produced in a typical experiment.
As high-end IT infrastructure is getting less expensive
and both commercial (Arivis, Imaris, etc.) and open
source software for SPIM data visualization and
analysis [70–72] are becoming more robust and
readily available, it is expected that SPIM will have
wider use in the field of infectious disease research
and others. More specifically, model systems purposefully
developed for infectious disease research
that fill the gap between simple cellular monolayer
cultures and living animals (such as spheroid and
organoid 3D cultures) amenable for SPIM are largely
missing (Fig. 3c). That is certainly going to change in
the next years as development of near-physiological
organoid model systems derived from human stem
cells has become an area of significant interest [73,
74] (Fig. 3c). Furthermore, objectives, detectors and
lightsheet thickness in SPIM systems are not fully
optimized for sub-cellular resolution which have
prevented addressing many open questions in infectious
disease research. Recent implementations of
new illumination strategies such as Bessel beams [75]
and lattice lightsheet [61, 76] achieve lightsheets that
are thinner than the depth of focus of the detection
objective and uniform across a large field of view. With
such illumination strategies, nearly isotropic 3D
resolution (230 nm laterally and 350 nm axially) has
been demonstrated, thus allowing for analysis of
single-molecule dynamics across multiple cells simultaneously
within a 3D stem cell spheroid or formation
of a T-cell immunological synapse in a 3D matrix
at diffraction-limited resolutions and at speeds of
1 volume/s [76]. Recently, by combining structured
illumination microscopy and SPIM, a 100-nm lateral
resolution was demonstrated [77]. Although lattice
lightsheet-based SPIM system is commercially
available, instrumentation utilizing new SPIM modalities
is still very rare worldwide and not widely used by
the scientific community. Specific advantages and
disadvantages of these approaches will have to be
evaluated by the scientific community in the years to
come. Being able to analyze physiology of host–
pathogen interactions in 3D model systems at
spatiotemporal scales offered by SPIM approaches
in combination with innovative illumination techniques
2619Review: Tackling Host‐Pathogen Interactions with Imaging
promises major advances in the field of infectious
disease research.
Microscopy by machine learning
One major limitation of all mentioned imaging
modalities is their low experimental throughput.
Super-resolution microscopy (especially SMLM techniques),
(cryo)CLEM and SPIM tend to require manual
interventions during the entire experimental workflow,
from laborious adjustments of image acquisition
parameters, over subsequent quality control to—
often non-intuitive—tweaking of the data processing
parameters, for image acquisition, processing and
analysis. This severely limits the data that can be
obtained. A snapshot of one or two objects of interest
clearly does not constitute a data set from which
conclusions can be derived. In order to derive
statistically significant information and turn the bioimage
data into scientific discovery, one often needs to
collect and analyze a large number of images.
Furthermore, this needs to be done over many
experimental conditions and preferentially in an unbiased
manner. This issue is further aggravated by the
trend of moving the biomedical research toward more
physiologically relevant 3D model system (such as
stem cell-derived organoids, spheroids, entire organs/
organisms and other 3D model systems) where
suddenly large volumes of biological material need to
be swept in order to find objects of interest and derive
relevant quantitative information. One approach to
tackle this issue is to develop fully automated
microscopy workflows which would automatically
acquire and process the data. There has been a
substantial progress in the last decade in automating
microscopy acquisition and analysis using standard
imaging modalities (such as confocal and widefield
microscopy) [78, 79]. This approach, referred to as
high-throughput high-content screening (HT-HCS),
has been extensively used to study the effects of
systematic downregulation (via siRNA technology) or
upregulation (overexpression) of each gene in the
genome on biological processes across biomedical
disciplines including infectious disease research
[80–82]. Extending a decade of HT-HCS know-how
to new imaging modalities proved to be more
challenging than anticipated. It is difficult to extend
some of the established HT-HCS pipelines to new
imaging modalities due to more complex acquisition
and analysis procedures that still require substantial
human input. However, the need for HT-HCS approach
in new imaging modalities has been recognized and
automated acquisition and analysis pipelines are being
developed for super-resolution [83, 84], CLEM [85] and
SPIM [86] approaches.
A particularly powerful approach, applicable to any
imaging modality, would be combining the automated
image acquisition with machine learning to perform
automated morphological profiling and image analysis
[87]. Machine-learning technology gives computers
ability to learn patterns in the data without being
explicitly programmed and then predict those patterns
in the new data. Spearheaded by large IT companies
such as Google, Microsoft, IBM and others, machinelearning
techniques are already present in many
aspects of modern society from web searches and
speech recognition in smartphones to self-driving
vehicles [88]. They have proved to be very powerful at
discovering the intricate structure of high-dimensional
data such as bioimages [89] which typically contain
multiple phenotypes whose discriminating morphologies
are not easily described by a few parameters. If
a car can already segment, classify and understand its
complex surrounding environment and make autonomous
decisions by harnessing the power of deep
neural networks, we can ask ourselves, why microscopes
would not be able to do the same within the
field of biomedical research. Automated acquisition
and analysis approaches using machine learning are
expected to have a major impact on how microscopy
is performed in biomedical research in the future. This
would increase the objectivity, reproducibility and
sensitivity of analysis of very heterogeneous pathophysiological
processes, as well as be able to identify
new and rare phenotypes that might escape human
attention. However, the restrictions and limitation of
machine learning approach as well as the extent of its
applicability in biomedical research still need to be
carefully evaluated by the scientific community before
adopting it for a wider use.
High-throughput imaging where the entire specimen
is indiscriminately recorded at maximal spatial and
temporal resolution may not be a feasible approach
due to both phototoxicity and computational restrictions
(such as for example for the SPIM-based imaging
approach). For this, but applicable to literally any
microscopy-based approach, we need an automated
image acquisition scheme, probably based on machine
learning approaches (e.g., by using deep neural
networks) to interrogate the biological specimen in its
entirety quickly at low resolution and decide what
needs to be acquired at higher resolution or from which
areas it is worth to extract a time lapse information in
case of live cell imaging. This way, those data that do
not contain the interesting information are not part of
the acquired data set significantly relieving the
pressure on IT infrastructure for storage and image
analysis. Such approaches would have a significant
impact on infectious disease studies where already
the search for individual transient and rare molecular
events provides a significant bottleneck in experimental
throughput. Such software has been under
development for classical imaging modalities [90,
91], also by implementing aspects of machine
learning [92], but this needs to be made easy to
use, compatible with various commercial hardware
implementations and extended to new imaging
modalities.
2620 Review: Tackling Host‐Pathogen Interactions with Imaging
Conclusion
In infectious disease research, discoveries by direct
visual observation have always been the crucial step
toward understanding. With major breakthroughs in
the last decade, microscopy has advanced to the point
where we can literally visualize molecular mechanisms
in model systems of increased complexity and
even in vivo. New imaging modalities such as superresolution
microscopy, CLEM and SPIM have opened
the possibility to interrogate our model systems at vast
ranges of spatio-temporal scales—from nanometers
to millimeters and from microseconds to days (Fig. 2).
These technologies are essential for achieving a
comprehensive understanding of the infection process,
as they are the only ones in our experimental
arsenal that allow sufficient spatio-temporal resolution
to capture and analyze individual events that govern
pathophysiological states in living systems. Technologies
such as super-resolution microscopy and CLEM
have already been established in infectious disease
research and are providing major progress in the field.
However, technical developments are still needed in
the years to come to enable good structural preservation
of fixed specimens for super-resolution and
analysis of infection processes in living systems.
Further technical developments are necessary to
achieve automated image acquisition with minimal
human intervention which will dramatically increase
the experimental throughput. SPIM approaches have
not gained a strong foothold in the infectious disease
field yet. However, this is expected to change as new
technical developments enable infectious disease
research on close-to-physiological model systems
(such as human stem cell-derived organoids and
other complex 3D model systems) and at subcellular
resolution. SPIM, together with human stem cellderived
organoid model systems, allow for comprehensive
dissection of the structure–function relationships
between diverse cell types, pathogens and their
surrounding microenvironments. This will have a
major impact in infectious disease research as shifting
basic research toward close-to-physiological biological
systems will yield results with higher predictive
value for estimating pathogenicity or pathogen control
(immunological or drug-mediated) in the infected host
and therefore higher translation potential.
However, the complexity of living systems and
host–pathogen interactions in health and disease
require more than visual inspection in order to derive
comprehensive understanding of these processes.
Modern microscopy-based research requires quantitative
measurements and statistics of a larger data set
to identify rare events and judge the variability and
significance of the findings which can only be
achieved by an approach that utilizes fully automated
acquisition and analysis protocols. Automated largevolume
sampling combined with computational systems
analysis using machine learning approaches
(support-vector machine or deep neural networks) will
provide a way forward. Taken together, modern
microscopy technology is currently uniquely positioned,
like at the time of the field's pioneers, van
Leeuwenhoek, Koch and Pasteur, to propel the
infectious disease research to a new frontier.
Acknowledgments
I thank Holger Lorenz [Center of Molecular Biology
(ZMBH), University of Heidelberg, Germany], Timo
Zimmermann [Centre for Genomic Regulation (CRG),
Barcelona, Spain], Martin Spitaler (Max Planck Institute
for Biochemistry, Munich, Germany), Christian
Conrad [German Cancer Research Center (DKFZ),
BioQuant Center, University of Heidelberg, Germany]
and Anna Kreshuk [Interdisciplinary Center for Scientific
Computing (IWR), Heidelberg, Germany] for
critical reading of the manuscript, comments and
discussions. Walter Muranyi, Hans-Georg Kräusslich,
Inés-Romero Bray, Ralf Bartenschlager, Megan
Stanifer and Steeve Boulant (Department of Infectious
Disease, University of Heidelberg, Germany) for
providing images for the Fig. 3. V.L. is supported by
TTU HIV of the German Center for Infection Research
(DZIF).
Received 10 May 2018;
Received in revised form 3 June 2018;
Accepted 8 June 2018
Available online 24 June 2018
Keywords:
super-resolution light microscopy;
CLEM;
SPIM;
machine learning;
virus
Abbreviations used:
EM, electron microscopy; STED, stimulated emission
depletion microscopy; SMLM, single-molecule localization
microscopy; HIV, human immunodeficiency virus; HCV,
hepatitis C virus; ROS, reactive oxygen species; CLEM,
correlative light-electron microscopy; SPIM, selective
plane illumination microscopy; HT-HCS, high-throughput
high content screening.
References
[1] A. Leeuwenhoek, Microscopical observations about animals
in the scurf of the teeth, Philos. Trans. 14 (1684) 568–574.
[2] B. Von Borries, E. Ruska, H. Ruska, Bakterien und Virus in
übermikroskopischer Aufnahme, Klin. Wochenschr. 17
(1938) 921–925.
2621Review: Tackling Host‐Pathogen Interactions with Imaging
[3] S. Brenner, G. Streisinger, R.W. Horne, S.P. Champe, L.
Barnett, S. Benzer, M.W. Rees, Structural components of
bacteriophage, J. Mol. Biol. 1 (1959) IN9–IN21, https://doi.org/
10.1016/S0022-2836(59)80035-8.
[4] S. Inoue, Video image processing greatly enhances contrast,
quality, and speed in polarization-based microscopy, J. Cell
Biol. 89 (1981) 346–356, https://doi.org/10.1083/jcb.89.2.346.
[5] Y. Hiraoka, J. Sedat, D. Agard, The use of a charge-coupled
device for quantitative optical microscopy of biological
structures, Science 238 (1987).
[6] V.T. Anchinmane, R.T. Shedge, A review of malaria
diagnostic tools: microscopy and rapid diagnostic test,
Asian J. Med. Sci. 1 (2011) 119–127, https://doi.org/10.
3126/ajms.v1i2.2965.
[7] C.S. Goldsmith, S.E. Miller, Modern uses of electron
microscopy for detection of viruses, Clin. Microbiol. Rev. 22
(2009) 552–563, https://doi.org/10.1128/CMR.00027-09.
[8] M.A. Saeed, A. Jabbar, “Smart diagnosis” of parasitic
diseases by use of smartphones, J. Clin. Microbiol. 56
(2017), e01469-17. https://doi.org/10.1128/JCM.01469-17.
[9] S.W. Hell, J. Wichmann, Stimulated-emission-depletion
fluorescence microscopy, Opt. Lett. 19 (1994) 780–782,
https://doi.org/10.1364/OL.19.000780.
[10] T.A. Klar, S. Jakobs, M. Dyba, A. Egner, S.W. Hell,
Fluorescence microscopy with diffraction resolution barrier
broken by stimulated emission, Proc. Natl. Acad. Sci. 97
(2000) 8206–8210, https://doi.org/10.1073/pnas.97.15.8206.
[11] E. Betzig, G.H. Patterson, R. Sougrat, O.W. Lindwasser, S.
Olenych,J.S.Bonifacino,M.W.Davidson,J.Lippincott-schwartz,
H.F. Hess, Imaging intracellular fluorescent proteins at
nanometer resolution, Science (80-.) 313 (2006) 1642–1646,
https://doi.org/10.1126/science.1127344.
[12] M.J. Rust, M. Bates, X. Zhuang, Sub-diffraction-limit imaging
by stochastic optical reconstruction microscopy (STORM),
Nat. Methods 3 (2006) 793–795, https://doi.org/10.1038/
nmeth929.
[13] C. Eggeling, K.I. Willig, S.J. Sahl, S.W. Hell, Lens-based
fluorescence nanoscopy, Q. Rev. Biophys. 48 (2015)
178–243, https://doi.org/10.1017/S0033583514000146.
[14] F. Balzarotti, Y. Eilers, K.C. Gwosch, A.H. Gynnå, V.
Westphal, F.D. Stefani, J. Elf, S.W. Hell, Nanometer
resolution imaging and tracking of fluorescent molecules
with minimal photon fluxes, Science (80-.) 355 (2017)
606–612, https://doi.org/10.1126/science.aak9913.
[15] M. Lehmann, S. Rocha, B. Mangeat, F. Blanchet, H. Uji-i, J.
Hofkens, V. Piguet, Quantitative multicolor super-resolution
microscopy reveals tetherin HIV-1 interaction, PLoS Pathog.
7 (2011) https://doi.org/10.1371/journal.ppat.1002456.
[16] W. Muranyi, S. Malkusch, B. Müller, M. Heilemann, H.G.
Kräusslich, Super-resolution microscopy reveals specific
recruitment of HIV-1 envelope proteins to viral assembly
sites dependent on the envelope C-terminal tail, PLoS
Pathog. 9 (2013) https://doi.org/10.1371/journal.ppat.
1003198.
[17] D. Eggert, K. Rösch, R. Reimer, E. Herker, Visualization and
analysis of hepatitis C virus structural proteins at lipid
droplets by super-resolution microscopy, PLoS One 9
(2014) https://doi.org/10.1371/journal.pone.0102511.
[18] J. Chojnacki, T. Staudt, B. Glass, P. Bingen, J. Engelhardt,
Maturation-Dependent HIV-1 Surface Protein Redistribution
Revealed by Fluorescence Nanoscopy, 2012 524–529.
[19] S.B. Van Engelenburg, G. Shtengel, P. Sengupta, K. Waki, M.
Jarnik,S.D.Ablan,E.O.Freed,H.F.Hess,J.Lippincott-schwartz,
Distribution of ESCRT machinery at HIV assembly sites
reveals virus scaffolding of ESCRT subunits, 343 (2014)
653–656, https://doi.org/10.1126/science.1247786.
[20] R.F. Laine, A. Albecka, S. Van De Linde, E.J. Rees, C.M.
Crump, C.F. Kaminski, Structural analysis of herpes simplex
virus by optical super-resolution imaging, Nat. Commun. 6
(2015) 1–10, https://doi.org/10.1038/ncomms6980.
[21] A. Szymborska, A. De Marco, N. Daigle, V.C. Cordes, J.A.G.
Briggs, J. Ellenberg, Nuclear pore scaffold structure analyzed
by super-resolution microscopy and particle averaging,
Science (80-.) 341 (2013) 655–658, https://doi.org/10.1126/
science.1240672.
[22] F. Huang, G. Sirinakis, E.S. Allgeyer, L.K. Schroeder, W.C.
Duim, E.B. Kromann, T. Phan, F.E. Rivera-Molina, J.R.
Myers, I. Irnov, M. Lessard, Y. Zhang, M.A. Handel, C. JacobsWagner,
C.P. Lusk, J.E. Rothman, D. Toomre, M.J. Booth, J.
Bewersdorf, Ultra-high resolution 3D imaging of whole cells,
Cell 166 (2016) 1028–1040, https://doi.org/10.1016/j.cell.2016.
06.016.
[23] D. Salas, A. Le Gall, J.-B. Fiche, A. Valeri, Y. Ke, P. Bron, G.
Bellot, M. Nollmann, Angular reconstitution-based 3D reconstructions
of nanomolecular structures from superresolution
light-microscopy images, Proc. Natl. Acad. Sci. 114 (2017)
9273–9278, https://doi.org/10.1073/pnas.1704908114.
[24] J.A.G. Briggs, Structural biology in situ-the potential of
subtomogram averaging, Curr. Opin. Struct. Biol. 23 (2013)
261–267, https://doi.org/10.1016/j.sbi.2013.02.003.
[25] R.D.M. Gray, C. Beerli, P.M. Pereira, K.M. Scherer, J.
Samolej, C.K.E. Bleck, J. Mercer, R. Henriques, VirusMapper:
open-source nanoscale mapping of viral architecture
through super-resolution microscopy, Sci. Rep. 6 (2016) 1–8,
https://doi.org/10.1038/srep29132.
[26] U. Schnell, F. Dijk, K.A. Sjollema, B.N.G. Giepmans,
Immunolabeling artifacts and the need for live-cell imaging,
Nat. Methods 9 (2012) 152–158, https://doi.org/10.1038/
nmeth.1855.
[27] D.R. Whelan, T.D.M. Bell, Image artifacts in single molecule
localization microscopy: why optimization of sample preparation
protocols matters, Sci. Rep. 5 (2015) 1–10, https://doi.
org/10.1038/srep07924.
[28] J. Dubochet, M. Adrian, J.J. Chang, J.C. Homo, J. Lepault, A.W.
McDowall, P. Schultz, Cryo-electron microsopcy of vitrified
specimens, Q. Rev. Biophys. 21 (1988) 129–228, https://doi.org/
10.1017/S0033583500004297.
[29] C.L. Schwartz, V.I. Sarbash, F.I. Ataullakhanov, J.R.
Mcintosh, D. Nicastro, Cryo-fluorescence microscopy facilitates
correlations between light and cryo-electron microscopy
and reduces the rate of photobleaching, J. Microsc. 227
(2007) 98–109, https://doi.org/10.1111/j.1365-2818.2007.
01794.x.
[30] R. Kaufmann, P. Schellenberger, E. Seiradake, I.M. Dobbie,
E.Y. Jones, I. Davis, C. Hagen, K. Gruenewald, Superresolution
microscopy using standard fluorescent proteins in
intact cells under cryo-conditions, Nano Lett. 14 (2014)
4171–4175, https://doi.org/10.1021/nl501870p.
[31] Y.W. Chang, S. Chen, E.I. Tocheva, A. Treuner-Lange, S.
Löbach, L. Søgaard-Andersen, G.J. Jensen, Correlated
cryogenic photoactivated localization microscopy and cryoelectron
tomography, Nat. Methods 11 (2014) 737–739,
https://doi.org/10.1038/nmeth.2961.
[32] J. Schneider, J. Zahn, M. Maglione, S.J. Sigrist, J. Marquard,
J. Chojnacki, H.G. Kräusslich, S.J. Sahl, J. Engelhardt, S.W.
Hell, Ultrafast, temporally stochastic STED nanoscopy of
millisecond dynamics, Nat. Methods 12 (2015) 827–830,
https://doi.org/10.1038/nmeth.3481.
2622 Review: Tackling Host‐Pathogen Interactions with Imaging
[33] P. Hoyer, G. de Medeiros, B. Balázs, N. Norlin, C. Besir, J.
Hanne, H.-G. Kräusslich, J. Engelhardt, S.J. Sahl, S.W. Hell, L.
Hufnagel, Breaking the diffraction limit of light-sheet fluorescence
microscopy by RESOLFT, Proc. Natl. Acad. Sci. 113
(2016) 3442–3446, https://doi.org/10.1073/pnas.1522292113.
[34] V. Sakin, G. Paci, E.A. Lemke, B. Müller, Labeling of virus
components for advanced, quantitative imaging analyses,
FEBS Lett. 590 (2016) 1896–1914, https://doi.org/10.1002/
1873-3468.12131.
[35] S. Wäldchen, J. Lehmann, T. Klein, S. van de Linde, M.
Sauer, Light-induced cell damage in live-cell super-resolution
microscopy, Sci. Rep. 5 (2015), 15348. https://doi.org/10.
1038/srep15348.
[36] I. Rasnik, S.A. McKinney, T. Ha, Nonblinking and long-lasting
single-molecule fluorescence imaging, Nat. Methods 3
(2006) 891–893, https://doi.org/10.1038/nmeth934.
[37] M. Jinek, K. Chylinski, I. Fonfara, M. Hauer, J.A. Doudna, E.
Charpentier, A programmable dual-RNA-guided DNA endonuclease
in adaptive bacterial immunity, Science (80-.) 337
(2012) 816–821, https://doi.org/10.1126/science.1225829.
[38] J. Heine, M. Reuss, B. Harke, E. D’Este, S.J. Sahl, S.W. Hell,
Adaptive-illumination STED nanoscopy, Proc. Natl. Acad.
Sci. (2017) 201708304, https://doi.org/10.1073/pnas.
1708304114.
[39] M.E. Masip, J. Huebinger, J. Christmann, O. Sabet, F.
Wehner, A. Konitsiotis, G.R. Fuhr, P.I.H. Bastiaens, Reversible
cryo-arrest for imaging molecules in living cells at
high spatial resolution, Nat. Methods 13 (2016) 665–672,
https://doi.org/10.1038/nmeth.3921.
[40] N. Gustafsson, S. Culley, G. Ashdown, D.M. Owen, P.M.
Pereira, R. Henriques, Fast live-cell conventional fluorophore
nanoscopy with ImageJ through super-resolution radial
fluctuations, Nat. Commun. 7 (2016), 12471. https://doi.org/
10.1038/ncomms12471.
[41] M. Weigert, U. Schmidt, T. Boothe, A. Müller, A. Dibrov, A.
Jain, B. Wilhelm, D. Schmidt, C. Broaddus, S. Culley, M.
Rocha-Martins, F. Segovia-Miranda, C. Norden, R.
Henriques, M. Zerial, M. Solimena, J. Rink, P. Tomancak,
L. Royer, F. Jug, E.W. Myers, Content-aware image
restoration: pushing the limits of fluorescence microscopy,
BioRxiv Prepr. (2017) 1–16, https://doi.org/10.1101/236463.
[42] T.W. Chun, L. Carruth, D. Finzi, X. Shen, J.A. DiGiuseppe, H.
Taylor, M. Hermankova, K. Chadwick, J. Margolick, T.C.
Quinn, Y.H. Kuo, R. Brookmeyer, M.A. Zeiger, P. BarditchCrovo,
R.F. Siliciano, Quantification of latent tissue reservoirs
and total body viral load in HIV-1 infection, Nature 387
(1997) 183–188, https://doi.org/10.1038/387183a0.
[43] K.J. Blight, A.A. Kolykhalov, C.M. Rice, Efficient initiation of
HCV RNA replication in cell culture, Science (80-.) 290 (2000)
1972–1974, https://doi.org/10.1126/science.290.5498.1972.
[44] K. Peng, W. Muranyi, B. Glass, V. Laketa, S.R. Yant, L. Tsai,
T. Cihlar, B. Müller, H.-G. Kräusslich, Quantitative microscopy
of functional HIV post-entry complexes reveals association
of replication with the viral capsid, elife 3 (2014) 1–21,
https://doi.org/10.7554/eLife.04114.
[45] C.L. Rieder, S.S. Bowser, Correlative immunofluorescence
and electron microscopy on the same section of Eponembedded
material, J. Histochem. Cytochem. 33 (1985)
165–171, https://doi.org/10.1177/33.2.3881520.
[46] W. Kukulski, M. Schorb, S. Welsch, A. Picco, M. Kaksonen,
J.A.G. Briggs, Correlated fluorescence and 3D electron
microscopy with high sensitivity and spatial precision,
J. Cell Biol. 192 (2011) 111–119, https://doi.org/10.1083/
jcb.201009037.
[47] I. Romero-Brey, A. Merz, A. Chiramel, J.-Y. Lee, P. Chlanda,
U. Haselman, R. Santarella-Mellwig, A. Habermann, S.
Hoppe, S. Kallis, P. Walther, C. Antony, J. Krijnse-Locker,
R. Bartenschlager, Three-dimensional architecture and
biogenesis of membrane structures associated with hepatitis
C virus replication, PLoS Pathog. 8 (2012), e1003056.
https://doi.org/10.1371/journal.ppat.1003056.
[48] I. Romero-Brey, C. Berger, S. Kallis, A. Kolovou, D. Paul, V.
Lohmann, R. Bartenschlager, NS5A domain 1 and polyprotein
cleavage kinetics are critical for induction of double-membrane
vesicles associated with hepatitis c virus replication, MBio 6
(2015) 1–12, https://doi.org/10.1128/mBio.00759-15.
[49] L. Chatel-Chaix, M. Cortese, I. Romero-Brey, S. Bender, C.J.
Neufeldt, W. Fischl, P. Scaturro, N. Schieber, Y. Schwab, B.
Fischer, A. Ruggieri, R. Bartenschlager, Dengue virus
perturbs mitochondrial morphodynamics to dampen innate
immune responses, Cell Host Microbe 20 (2016) 342–356,
https://doi.org/10.1016/j.chom.2016.07.008.
[50] P. Spuul, G. Balistreri, L. Kaariainen, T. Ahola, Phosphatidylinositol
3-kinase-, actin-, and microtubuledependent
transport of semliki forest virus replication
complexes from the plasma membrane to modified lysosomes,
J. Virol. 84 (2010) 7543–7557, https://doi.org/10.
1128/JVI.00477-10.
[51] M. Lebrun, N. Thelen, M. Thiry, L. Riva, I. Ote, C. Condé, P.
Vandevenne, E. Di Valentin, S. Bontems, C. SadzotDelvaux,
Varicella-zoster virus induces the formation of
dynamic nuclear capsid aggregates, Virology 454–455
(2014) 311–327, https://doi.org/10.1016/j.virol.2014.02.023.
[52] F. Li, X. Sewald, J. Jin, N.M. Sherer, W. Mothes, Murine
leukemia virus Gag localizes to the uropod of migrating
primary lymphocytes, J. Virol. 88 (2014) 10541–10555,
https://doi.org/10.1128/JVI.01104-14.
[53] A. Merk, A. Bartesaghi, S. Banerjee, V. Falconieri, P. Rao, M.I.
Davis, R. Pragani, M.B. Boxer, L.A. Earl, J.L.S. Milne, S.
Subramaniam, Breaking cryo-EM resolution barriers to facilitate
drug discovery, Cell 165 (2016) 1698–1707, https://doi.org/
10.1016/j.cell.2016.05.040.
[54] F.K.M. Schur, M. Obr, W.J.H. Hagen, W. Wan, A.J. Jakobi, J.M.
Kirkpatrick, C. Sachse, H. Kräusslich, J.A.G. Briggs, An atomic
model of HIV-1 capsid-SP1 reveals structures regulating
assembly and maturation, Science 353 (2016) 506–508,
https://doi.org/10.1126/science.aaf9620.
[55] X. Zhang, P. Ge, X. Yu, J.M. Brannan, G. Bi, Q. Zhang, S.
Schein, Z.H. Zhou, Cryo-EM structure of the mature dengue
virus at 3.5-Å resolution, Nat. Struct. Mol. Biol. 20 (2013)
105–110, https://doi.org/10.1038/nsmb.2463.
[56] S. Asano, Y. Fukuda, F. Beck, A. Aufderheide, F. Forster, R.
Danev, W. Baumeister, A molecular census of 26S proteasomes
in intact neurons, Science (80-.) 347 (2015) 439–442,
https://doi.org/10.1126/science.1261197.
[57] A. Von Appen, J. Kosinski, L. Sparks, A. Ori, A.L. DiGuilio, B.
Vollmer, M.T. Mackmull, N. Banterle, L. Parca, P. Kastritis, K.
Buczak, S. Mosalaganti, W. Hagen, A. Andres-Pons, E.A.
Lemke, P. Bork, W. Antonin, J.S. Glavy, K.H. Bui, M. Beck, In situ
structural analysis of the human nuclear pore complex, Nature
526 (2015) 140–143, https://doi.org/10.1038/nature15381.
[58] P.J. Schweiger, K.B. Jensen, Modeling human disease using
organotypic cultures, Curr. Opin. Cell Biol. 43 (2016) 22–29,
https://doi.org/10.1016/j.ceb.2016.07.003.
[59] N. Ji, D.E. Milkie, E. Betzig, Adaptive optics via pupil
segmentation for high-resolution imaging in biological tissues,
Nat. Methods 7 (2010) 141–147, https://doi.org/10.1038/
nmeth.1411.
2623Review: Tackling Host‐Pathogen Interactions with Imaging
[60] C. Wang, R. Liu, D.E. Milkie, W. Sun, Z. Tan, A. Kerlin, T.W.
Chen, D.S. Kim, N. Ji, Multiplexed aberration measurement
for deep tissue imaging in vivo, Nat. Methods 11 (2014)
1037–1040, https://doi.org/10.1038/nmeth.3068.
[61] T. Liu, S. Upadhyayula, D.E. Milkie, V. Singh, K. Wang, I.A.
Swinburne, K.R. Mosaliganti, Z.M. Collins, T.W. Hiscock, J.
Shea, A.Q. Kohrman, T.N. Medwig, D. Dambournet, R.
Forster, B. Cunniff, Y. Ruan, H. Yashiro, S. Scholpp, E.M.
Meyerowitz, D. Hockemeyer, D.G. Drubin, B.L. Martin, D.Q.
Matus, M. Koyama, S.G. Megason, T. Kirchhausen, E.
Betzig, Observing the cell in its native state: imaging
subcellular dynamics in multicellular organisms, Science
(80-.) 1392 (2018) https://doi.org/10.1126/science.aaq1392.
[62] B. Stolp, K. Melican, Microbial pathogenesis revealed by
intravital microscopy: pros, cons and cautions, FEBS Lett.
590 (2016) 2014–2026, https://doi.org/10.1002/1873-3468.
12122.
[63] E.H.K. Stelzer, Light-sheet fluorescence microscopy
for quantitative biology, Nat. Methods 12 (2015) 23–26,
https://doi.org/10.1038/nmeth.3219.
[64] J. Huisken, D.Y.R. Stainier, Selective plane illumination
microscopy techniques in developmental biology, Development
136 (2009) 1963–1975, https://doi.org/10.1242/dev.
022426.
[65] A. Klingberg, A. Hasenberg, I. Ludwig-Portugall, A.
Medyukhina, L. Männ, A. Brenzel, D.R. Engel, M.T. Figge,
C. Kurts, M. Gunzer, Fully automated evaluation of total
glomerular number and capillary tuft size in nephritic kidneys
using lightsheet microscopy, J. Am. Soc. Nephrol. 28 (2017)
452–459, https://doi.org/10.1681/ASN.2016020232.
[66] W. Li, R.N. Germain, M.Y. Gerner, Multiplex, quantitative
cellular analysis in large tissue volumes with clearingenhanced
3D microscopy (C e 3D), Proc. Natl. Acad. Sci.
(2017) 201708981, https://doi.org/10.1073/pnas.1708981114.
[67] P.H. Jakob, J. Kehrer, P. Flood, C. Wiegel, U. Haselmann, M.
Meissner, E.H.K. Stelzer, E.G. Reynaud, A 3-D cell culture
system to study epithelia functions using microcarriers,
Cytotechnology 68 (2016) 1813–1825, https://doi.org/10.
1007/s10616-015-9935-0.
[68] M.F. Wells, M.R. Salick, O. Wiskow, D.J. Ho, K.A. Worringer,
R.J. Ihry, S. Kommineni, B. Bilican, J.R. Klim, E.J. Hill, L.T.
Kane, C. Ye, A. Kaykas, K. Eggan, Genetic ablation of AXL
does not protect human neural progenitor cells and cerebral
organoids from Zika virus infection, Cell Stem Cell 19 (2016)
703–708, https://doi.org/10.1016/j.stem.2016.11.011.
[69] S. Schuster, T. Krüger, I. Subota, S. Thusek, B. Rotureau, A.
Beilhack, M. Engstler, Developmental adaptations of trypanosome
motility to the tsetse fly host environments unravel
a multifaceted in vivo microswimmer system, elife 6 (2017)
1–28, https://doi.org/10.7554/eLife.27656.
[70] H. Peng, Z. Ruan, F. Long, J.H. Simpson, E.W. Myers, V3D
enables real-time 3D visualization and quantitative analysis
of large-scale biological image data sets, Nat. Biotechnol. 28
(2010) 348–353, https://doi.org/10.1038/nbt.1612.
[71] T. Pietzsch, S. Saalfeld, S. Preibisch, P. Tomancak, BigDataViewer:
visualization and processing for large image data sets,
Nat. Methods 12 (2015) 481–483, https://doi.org/10.1038/
nmeth.3392.
[72] L.A. Royer, M. Weigert, U. Günther, N. Maghelli, F. Jug, I.F.
Sbalzarini, E.W. Myers, ClearVolume: open-source live 3D
visualization for light-sheet microscopy, Nat. Methods 12
(2015) 480–481, https://doi.org/10.1038/nmeth.3372.
[73] M.J. Workman, M.M. Mahe, S. Trisno, H.M. Poling, C.L.
Watson, N. Sundaram, C.F. Chang, J. Schiesser, P. Aubert,
E.G. Stanley, A.G. Elefanty, Y. Miyaoka, M.A. Mandegar, B.R.
Conklin, M. Neunlist, S.A. Brugmann, M.A. Helmrath, J.M.
Wells, Engineered human pluripotent-stem-cell-derived intestinal
tissues with a functional enteric nervous system, Nat. Med.
23 (2017) 49–59, https://doi.org/10.1038/nm.4233.
[74] J.A. Bagley, D. Reumann, S. Bian, J. Lévi-Strauss, J.A.
Knoblich, Fused cerebral organoids model interactions
between brain regions, Nat. Methods 14 (2017) 743–751,
https://doi.org/10.1038/nmeth.4304.
[75] T.A. Planchon, L. Gao, D.E. Milkie, M.W. Davidson, J.A.
Galbraith, C.G. Galbraith, E. Betzig, Rapid three-dimensional
isotropic imaging of living cells using Bessel beam plane
illumination, Nat. Methods 8 (2011) 417–423, https://doi.org/
10.1038/nmeth.1586.
[76] B.-C. Chen, W.R. Legant, K. Wang, L. Shao, D.E. Milkie, M.W.
Davidson, C. Janetopoulos, X.S. Wu, J.a. Hammer, Z. Liu, B.P.
English, Y. Mimori-Kiyosue, D.P. Romero, A.T. Ritter, J.
Lippincott-Schwartz, L. Fritz-Laylin, R.D. Mullins, D.M.
Mitchell, A.-C. Reymann, J.N. Bembenek, R. Bohme, S.W.
Grill, J.T. Wang, G. Seydoux, U.S. Tulu, D.P. Kiehart, E. Betzig,
Lattice light-sheet microscopy: Imaging molecules to
embryos at high spatiotemporal resolution, Science (80-.)
(2014) https://doi.org/10.1126/science.1257998.
[77] B.-J. Chang, V.D. Perez Meza, E.H.K. Stelzer, csiLSFM
combines light-sheet fluorescence microscopy and coherent
structured illumination for a lateral resolution below 100 nm,
Proc. Natl. Acad. Sci. 114 (2017) 4869–4874, https://doi.org/
10.1073/pnas.1609278114.
[78] R. Pepperkok, J. Ellenberg, High-throughput fluorescence
microscopy for systems biology, Nat. Rev. Mol. Cell Biol. 7
(2006) 690–696, https://doi.org/10.1038/nrm1979.
[79] M. Mattiazzi Usaj, E.B. Styles, A.J. Verster, H. Friesen, C.
Boone, B.J. Andrews, High-content screening for quantitative
cell biology, Trends Cell Biol. 26 (2016) 598–611, https://doi.
org/10.1016/j.tcb.2016.03.008.
[80] H. Zhou, M. Xu, Q. Huang, A.T. Gates, X.D. Zhang, J.C.
Castle, E. Stec, M. Ferrer, B. Strulovici, D.J. Hazuda, A.S.
Espeseth, Genome-scale RNAi screen for host factors
required for HIV replication, Cell Host Microbe 4 (2008)
495–504, https://doi.org/10.1016/j.chom.2008.10.004.
[81] A. Karlas, N. MacHuy, Y. Shin, K.P. Pleissner, A. Artarini, D.
Heuer, D. Becker, H. Khalil, L.A. Ogilvie, S. Hess, A.P.
Mäurer, E. Müller, T. Wolff, T. Rudel, T.F. Meyer, Genomewide
RNAi screen identifies human host factors crucial for
influenza virus replication, Nature 463 (2010) 818–822,
https://doi.org/10.1038/nature08760.
[82] M. Poenisch, P. Metz, H. Blankenburg, A. Ruggieri, J.Y. Lee,
D. Rupp, I. Rebhan, K. Diederich, L. Kaderali, F.S.
Domingues, M. Albrecht, V. Lohmann, H. Erfle, R.
Bartenschlager, Identification of HNRNPK as regulator of
hepatitis C virus particle production, PLoS Pathog. 11 (2015)
1–21, https://doi.org/10.1371/journal.ppat.1004573.
[83] A. Beghin, A. Kechkar, C. Butler, F. Levet, M. Cabillic, O.
Rossier, G. Giannone, R. Galland, D. Choquet, J.-B. Sibarita,
Localization-based super-resolution imaging meets highcontent
screening, Nat. Methods (2017) https://doi.org/10.
1038/nmeth.4486.
[84] J.P. Eberle, W. Muranyi, H. Erfle, M. Gunkel, Fully automated
targeted confocal and single-molecule localization microscopy,
Methods Mol. Biol. (2017) 139–152, https://doi.org/10.1007/
978-1-4939-7265-4_12.
[85] M. Schorb, L. Gaechter, O. Avinoam, F. Sieckmann, M. Clarke,
C. Bebeacua, Y.S. Bykov, A.F.P. Sonnen, R. Lihl, J.A.G.
Briggs, New hardware and workflows for semi-automated
2624 Review: Tackling Host‐Pathogen Interactions with Imaging
correlative cryo-fluorescence and cryo-electron microscopy/
tomography, J. Struct. Biol. 197 (2017) 83–93, https://doi.org/
10.1016/j.jsb.2016.06.020.
[86] C. Schmied, P. Steinbach, T. Pietzsch, S. Preibisch, P.
Tomancak, An automated workflow for parallel processing of
large multiview SPIM recordings, Bioinformatics 32 (2016)
1112–1114, https://doi.org/10.1093/bioinformatics/btv706.
[87] O.Z. Kraus, B.T. Grys, J. Ba, Y. Chong, B.J. Frey, C. Boone,
B.J. Andrews, Automated analysis of high-content microscopy
data with deep learning, Mol. Syst. Biol. 13 (2017) 924,
https://doi.org/10.15252/msb.20177551.
[88] Y. LeCun, Y. Bengio, G. Hinton, Deep learning, Nature 521
(2015) 436–444, https://doi.org/10.1038/nature14539.
[89] A. Kreshuk, U. Koethe, E. Pax, D.D. Bock, F.A. Hamprecht,
Automated detection of synapses in serial section transmission
electron microscopy image stacks, PLoS One 9 (2014)
https://doi.org/10.1371/journal.pone.0087351.
[90] J. Henriksson, J. Hench, Y.G. Tong, A. Johansson, D.
Johansson, T.R. Bürglin, Endrov: an integrated platform
for image analysis, Nat. Methods 10 (2013) 454–456,
https://doi.org/10.1038/nmeth.2478.
[91] A. Carro, M. Perez-Martinez, J. Soriano, D.G. Pisano, D.
Megias, iMSRC: converting a standard automated microscope
into an intelligent screening platform, Sci. Rep. 5
(2015), 10502. https://doi.org/10.1038/srep10502.
[92] C. Conrad, A. Wünsche, T.H. Tan, J. Bulkescher, F.
Sieckmann, F. Verissimo, A. Edelstein, T. Walter, U. Liebel,
R. Pepperkok, J. Ellenberg, Micropilot: automation of fluorescence
microscopy-based imaging for systems biology, Nat.
Methods 8 (2011) 246–249, https://doi.org/10.1038/nmeth.
1558.
2625Review: Tackling Host‐Pathogen Interactions with Imaging