Beruflich Dokumente
Kultur Dokumente
arch.
Optimizing
your live-cell
microscopy:
Tricks and
trade-offs
Produced by the
Sponsored by Science/AAAS Custom
Publishing Office
TA B L E O F C O N T E N T S
Optimizing
Introductions
your live-cell
2 A life well lived in the age of microscopy
microscopy: Sean Sanders, Ph.D.
Science/AAAS
ScienceCareers.org
White paper: Superresolution microscopy
4 Bringing cellular dynamics to light with live-cell microscopy
Jeffrey M. Perkel
Research articles
8 Detryrosinated microtubules buckle and bear load in
contracting cardiomyocytes
Patrick Robison, Matthew A. Caporizzo, Hossein Ahmadzadeh et al.
Application note
1
O PT IM IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S
confocal imaging. S E C T I O N O N E | I N T RO D U C T I O N S
A life well one of the few direct ways of doing so is with some flavor of
microscopy. From its invention in the early 1600s (the origin of Shedding by the drawings and pictures created by generations
who sought to depict their world, and the accuracy and
light on life
frequency of the images mark how well we were and are
lived in
the first light microscope is somewhat in dispute), inventors and
researchers have been constantly improving on the original design able to describe and understand the phenomena of the world.
through hardware and software upgrades, pushing the boundaries Over the last two centuries the rapid progression of innovative mi-
the age of of resolution, speed, and depth of imaging. Seldom does a year pass
without a significant advance in this field, most especially in the realm
croscopy techniques has allowed unprecedented visualization and
description of microscopic phenomena, making a light microscope
microscopy of fluorescent microscopy (which garnered a Nobel Prize in Chemistry an indispensable tool in every biological or medical laboratory. As
in 2014). the light microscope has developed, it has given rise to hundreds of
For the vast majority of the time since the invention of the micro- different imaging techniques, from brightfield and differential interfer-
scope, in order to view a specimen, it has needed to be prepared ence contrast to modern fluorescence and superresolution microsco-
through a permanent fixation procedure. Also, samples are often spe- py, which all have one thing in common: They are based on light and
cially treated to reveal the location of specific cellular components, its interaction with the specimen. Nowadays light microscopes cannot
a process that requires the fixation step. However, this methodol- match the resolution of electron microscopes or the penetration of X-
ogy provides only a single instant in time for researchers to analyze. ray microscopes, but they remain unchallenged when it comes to the
Sometimes this is sufficient. But in order to study the kinetics of a imaging of dynamic processes in living cells, tissues, and organisms—
cellular event or track changes in a cell over time, the ability to image observing life as it happens. Traditionally, light microscopes were
a live cell is paramount. Additionally, the process of preparing fixed divided into two distinct groups: widefield microscopes, which can
cells has been found to cause artifacts that can confound the interpre- acquire an entire image in a single exposure, and laser scanning mi-
tation of the data. croscopes, that image a specimen sequentially, point-by-point. While
Caveat emptor. Together with the clear advantages of live-cell the latter produce crisper images of thick specimens, the former re-
imaging come new challenges. Dead cells can be treated relatively cord images much faster and with less damage to the specimen.
harshly without much concern. Live cells on the other hand, frequently Since Carl Zeiss, Ernst Abbe, and Otto Schott laid the foundations
require careful handling that includes a constant temperature, stable of modern microscopy 170 years ago, the company they founded
CO2 levels, and sufficient nutrients. Moreover, the lasers used in many delivered most of the radical microscope innovations, from the simple
fluorescent microscopes to create the beautiful and colorful images // INNOVATION achromatic and apochromatic lenses of the 19th century, to the com-
we have come to know are so powerful that they can easily damage MADE BY ZEISS plex imaging instruments of today. The latest innovation from ZEISS
cells even during short-term illumination. Live-cell imaging is, in its is a novel confocal detector named Airyscan. It blurs the traditional
most fundamental form, a constant balance between achieving the divide between widefield and laser scanning microscopes and offers
necessary data and not damaging or killing the cells. users the best of both worlds: high-speed, gentle imaging combined
A variety of clever solutions have been developed recently that with superb resolution and contrast, capabilities unmatched by stan-
surpass the methods and materials of yesteryear, including brighter dard confocal microscopes.
fluorescent dyes that require lower-energy activation light, more We hope you will find this collection of publications and peer-
sensitive detectors that can image both more quickly and in lower reviewed articles interesting and inspiring, an example of how in-
light, and techniques and technologies that use hardware and novative light microscopy continues to enable researchers to push
software tricks to minimize the amount of light bathing the sample. the boundaries of human knowledge. We believe that modern light
Your new standard for fast and gentle confocal imaging
microscopy can make a difference in your research as well. We are
None of these provides a complete solution, but these advances
and others have allowed researchers to better balance the best Discover ZEISS LSM 880 with Airyscan – the new excited
confocaltolaser
contribute to microscope
scanning your success.
resolution, fastest imaging speed, maximum imaging depth, and best that offers high sensitivity, improved resolution inBernhard
x, y and z, Zimmermann,
and high speed.Ph.D.
cell health to obtain the optimal results. The content provided in this
new supplement gives you an indication of how they’re succeeding All in one system. Find out more and book a hands-on demonstration in one
Senior Director Segment of
Marketing Life Sciences
in these efforts. We’ve provided three recent Science and Science our ZEISS Microscopy Labs now. Carl Zeiss Microscopy GmbH
Advances articles and an exclusive white paper in this timely update microscopy@zeiss.com
on live-cell imaging. We hope you enjoy reading it. www.zeiss.com/lsm880
R
esearchers have no shortage of options for livecell One alternative to confocal microscopy is pairing widefield
microscopy, from standard brightfield and differential microscopy with deconvolution strategies. Widefield images
interference contrast microscopy, to confocal and super- typically subject cells to lower light intensity than confocal. But,
resolution approaches. Here, we consider some of the more since the resulting pictures tend to be blurrier, especially with
popular fluorescence-based strategies. thick specimens, they can be cleaned up using computational
The most straightforward approach, perhaps, is basic wide- algorithms to produce a sharper reconstruction.
field fluorescence microscopy, in which an entire field of view is
illuminated with excitation energy, and total fluorescence from
the area is recorded at once using an area detector, such as a
CCD (charge-coupled device) or CMOS (complementary metal-
Airyscan image of oxide semiconductor) camera. The simplest microscopy meth- “Superresolution microscopy” is a
od available, widefield imaging also yields the blurriest images,
the central nervous
as there’s no mechanism to exclude out-of-focus light from
catch-all term describing a half-dozen
system of a Drosophila
melanogaster embryo.
reaching the detector. Thus, according to a review on the Carl or so different methods.
Zeiss Microscopy Online Campus, “widefield imaging achieves
the optimum results when the features of interest are either
large (such as an organelle) or highly punctate in nature” (4). Another alternative strategy, also compatible with live-cell
One strategy for overcoming those limitations is confocal imaging, is confocal microscopy using ZEISS’s Airyscan detec-
Bringing cellular movie to a single frame. And for many questions, the resulting
data are good enough to provide the answers. But some ques-
tions simply cannot be answered from single images. In those
microscopy. A bright light source (usually a laser) is focused
onto a particular point in the sample. The resulting fluores-
tor. In a typical confocal, one or more pinholes are used to re-
ject out-of-focus light; in-focus emission light is captured using
dynamics to light
cence is then captured after it passes through a small pinhole a point detector (such as a photomultiplier tube) or spectral
cases, researchers turn to live-cell imaging. aperture, which blocks out-of-focus light, producing a sharp array. As a result, the researcher must balance signal with clar-
image. The focal plane is then stepped perpendicular to the ity: By making the pinhole smaller, image sharpness increases,
with live-cell
I
n theory, live-cell microscopy is no different from any other slide (i.e., along the z axis), creating multiple optical planes but less light is captured, meaning more intense excitation light
microscopy method. But live cells do pose significant techni- that can be virtually assembled to create a 3D representation must be used. Airyscan circumvents that problem by eliminat-
microscopy
cal issues, and researchers pursuing live-cell methods have of the original sample. ing the pinhole and projecting the emission light onto a 32-el-
devised strategies for dealing with them. Confocal systems generally take one of two approaches ement array of hexagonal detectors. “Each detector element
The sine qua non of live-cell microscopy is, obviously, live to reconstruct each optical slice: Either the diffraction-limited functions as a single, very small pinhole,” according to compa-
cells, and both microscope vendors and third-party firms offer illumination point is raster-scanned across the sample (as in ny literature, leading to stronger signals, increased resolution,
By Jeffrey M. Perkel
tools (such as climate-controlled growth chambers) to facili- laser-scanning confocal microscopy), or multiple points are im- and decreased imaging speed.
tate such research. But it’s not enough that the cells be alive; aged simultaneously (as in spinning-disk confocal microscopy). “Superresolution microscopy” is a catch-all term describing
generally they also have to be healthy, and not all microscopy The difficulty is that whatever z position is being imaged, the a half-dozen or so different methods. In general, they also fall
L
ight microscopy has countless applications in today’s techniques are compatible with that state. Fluorescence-based entire sample volume is illuminated, increasing the likelihood of into two categories, the first of which includes localization-
IMAGE: COURTESY OF JULIA SELLIN, MICHAEL HOCH, LIMES INSTITUT, UNIVERSITY OF BONN
life sciences laboratory, from documenting cellular and approaches, especially, often require extended and intense phototoxicity and photobleaching. The spinning-disk approach based approaches such as PALM (photoactivated localization
organismal structure and mapping RNA localization, to sample irradiation with high-powered light sources, which can is generally considered faster and less phototoxic, but both can microscopy) and STORM (available from ZEISS and Nikon,
tracking cellular dynamics. First, though, researchers must an- induce phototoxicity. Fixed cells can easily stand this assault—so be applied to live cells, assuming the experiment is designed respectively). Normally, all fluorescent molecules in a diffrac-
swer a fundamental question: Are they going to image the cells long as the fluorescent dyes or proteins being imaged do not appropriately. Newer resonant scanning heads, such as that in tion-limited spot will fluoresce, making it impossible to dif-
while they’re alive, or after they’re dead? succumb to negative effects such as photobleaching. But live the Nikon A1R+ confocal system can accelerate scanning-based ferentiate fluorophores separated by less than about 200 nm.
For many projects, the answer is: both. Yet the question cells can become stressed and apoptotic, or at the very least microscopy to 30 frames per second or higher, depending on Localization-based approaches circumvent that problem by
F
or those specifically interested in membrane dynamics, Conversely, an imaging approach that is very fast must, by
TIRF (total internal reflection fluorescence) microscopy is necessity, not linger over any one position, meaning that fewer
a popular live-cell imaging technique. In TIRF, excitation photons will be collected, causing signal-to-noise to suffer—
energy is directed at the glass-sample interface at an oblique unless brighter excitation light is used.
angle such that it creates an evanescent wave parallel to the Microscopists are no stranger to this tug-of-war—it’s one of
slide surface, extending only about 100 nm into the sample. the reasons most research institutions contain microscopy core
As a result, only those fluorophores relatively near the slide sur-
facilities stocked with multiple instruments (economic consider-
face will fluoresce, producing a sharper image. In one recent
ations being another key factor, of course). And in many cases,
Dual-color PALM image of Chinese Hamster Ovary cells using example, Lynn Enquist and colleagues at Princeton Univer-
sity used live-cell TIRF microscopy to image the exit of newly researchers can obtain sufficient quality data out of whatever
fast sequential laser switching (Dronpa-Paxillin, red; tdEOS-
Vinculin, green). formed viral particles from cells (9). instrument they have available, even if it isn’t the top-of-the-line
A conceptually related idea is light-sheet microscopy. Here, hardware.
rather than illuminating the entire sample volume (as in bright- The good news for researchers is this: With so many tools
field and confocal microscopy), specimens are hit with a thin pla- in the toolbox, finding the right one for your imaging needs is
nar “sheet” of light from the side and imaged from an orthogonal easier than ever.
direction. By translating that sheet up and down, the system can
then capture 3D volumes repeatedly over time, producing a 4D
Jeffrey M. Perkel is a freelance science writer based in Pocatello, Idaho.
dataset. Commercial light-sheet implementations are available,
but researchers can build their own using the detailed instruc-
tions laid out by the OpenSPIM (selective plane illumination mi- References
croscopy) project (openspim.org). 1. K. Xu, G. Zhong, X. Zhuang, Actin, spectrin, and associated proteins
Superresolution SIM image of synapses at the neuromuscular Some researchers, though, need to up the ante. In late 2015, form a periodic cytoskeletal structure in axons. Science 339, 452–456
junction in Drosophila melanogaster. Triple staining for the for instance, Philipp Keller and colleagues of the Janelia Re- (2013).
presynaptic active zone marker Brp (green), postsynaptic search Campus described a custom microscopy system called 2. M. M. Frigault et al., Live-cell microscopy—tips and tools. J. Cell Sci.
glutamate receptors (red), and the presynaptic membrane (blue). “IsoView,” which exploits two perpendicular light sheets to cap- 122, 753–767 (2009).
ture four separate images of a sample simultaneously. That strate- 3. J. B. Grimm et al., A general method to improve fluorophores for live-
gy is particularly useful when imaging relatively opaque samples, cell and single-molecule microscopy. Nat. Methods 12, 244–250
such as Drosophila embryos, and the team used it to (among oth- (2015).
er things) track cellular motion during fruit fly gastrulation over a 4. M. E. Dailey et al., Live-cell imaging techniques. Carl Zeiss Microscopy
Yale University described a method for two-color live-cell STED
Online Campus, available at http://zeiss-campus.magnet.fsu.edu/
imaging using HaloTag- and SNAP-tag-based fluorophores, 3-hour period, collecting 3D volumes at 0.25 Hz (10).
IMAGES: (FROM TOP) COURTESY OF H. SHROFF, H. HESS, HHMI JANELIA FARM, ASHBURN, VIRGINIA, USA;
articles/livecellimaging/techniques.html.
which they used to differentiate the lumen and membrane of Also at Janelia, in 2014 Betzig developed a higher-resolution
Light-sheet image of hepatocyte spheroids showing nucleus 5. A. Chmyrov et al., Nanoscopy with more than 100,000 ‘doughnuts.’
the endoplasmic reticulum (6). And in 2012, Susan Cox and col- light-sheet-based strategy called “lattice light-sheet microscopy.”
(HNF4a/HNF3b) and cytoskeleton (CK18) staining. Nat. Methods 10, 737–740 (2013).
This approach differs from IsoView essentially in the thickness of
IMAGE: COURTESY OF JOHN TNG WEIQUAN, NG HUCK HUI, GENOME INSTITUTE OF SINGAPORE
leagues of King’s College London developed a computational
6. F. Bottanelli et al., Two-colour live-cell nanoscale imaging of
strategy called “Bayesian analysis of blinking and bleaching” the generated sheets and the resolution of the resulting images,
intracellular targets. Nat. Commun. 7, 10778 (2016).
(3B) that accelerated PALM to the point that the team could and Betzig used it to image subcellular events in 3D over extend-
W
podosome dynamics. Nat. Methods 9, 195–200 (2012).
4-second resolution (7). Those researchers interested in imaging cells relatively deep hen it comes to microscopy, there is no one perfect
8. D. Li et al., Extended-resolution structured illumination imaging
In general, however, the localization techniques are consid- within tissue sections and even live animals may benefit from two- instrument for every occasion. Confocal microscopy
of endocytic and cytoskeletal dynamics. Science 349, aab3500
ered too slow for many live-cell applications (as many thou- photon or multiphoton microscopy. Like confocal and light-sheet excels at removing out-of-focus light, but that may
(2015).
sands of “frames” must be collected to reconstruct a single microscopy, multiphoton microscopy creates thin optical sections not be absolutely necessary if you’re imaging only very thin
9. I. B. Hogue et al., Cellular mechanisms of alphaherpes virus egress:
image), and often require light of such high intensity as to be in tissue (12). But it does so via a different mechanism, requiring specimens. Superresolution microscopy yields magnificent
Live cell fluorescence microscopy of pseudorabies virus exocytosis.
phototoxic in many instances. The one exception is SIM, thanks the fluorophore to almost instantaneously absorb two low-energy detail, but that really only matters if the objects being studied
PLOS Pathog. 10, 1004535 (2014).
to its speed and relatively gentle illumination conditions. But photons to fluoresce rather than one high-energy photon. are small enough or close enough together to be otherwise
10. R. K. Chhetri et al., Whole-animal functional and developmental
SIM also yields the lowest resolution of any superresolution This strategy offers several potential benefits. First, because unresolvable.
imaging with isotropic spatial resolution. Nat. Methods 12, 1171–1178
method, about 100 nm. multiple absorption events occur only in the focal plane, out-of- Microscopy, in other words, is a balancing act. In a 2015 sup- (2015).
In a 2015 report in Science detailing technical modifications focus fluorophores remain dark in this method, not to mention plement to Science, Betzig described that balance in terms of 11. B.-C. Chen et al., Lattice light-sheet microscopy: Imaging molecules to
that increase SIM resolution to between 84 nm and 45 nm, unaffected by the incident light, thereby increasing signal-to- a tetrahedron, with vertices representing spatial resolution, cell embryos at high spatiotemporal resolution. Science 346, 1257998
Janelia Research Campus investigator and Nobel Prize winner noise while minimizing photobleaching and phototoxicity. And, viability, imaging depth, and speed (14). Science writer Mike (2014).
Eric Betzig detailed the difficulty of applying superresolution because the technique uses lower-energy, higher-wavelength May, who interviewed Betzig, explains it thusly: 12. R. K. P. Benninger, D. W. Piston, Two-photon excitation microscopy
methods to live cells. Many such techniques, he wrote, “place light, that light can penetrate deeper into the tissue than most Imagine a point inside that tetrahedron as a representation of for the study of living cells and tissues. Curr. Protoc. Cell Biol. 59,
extraordinary demands on the photon budget, represented microscopy methods are capable of reaching—several hundred an imaging method’s combination of features. If it improves one 4.11.1–4.11.24 (2013).
by the product of the number of fluorescent molecules in the microns at least. That said, multiphoton microscopy traditionally feature, say spatial resolution, it moves closer to that vertex and 13. R. Amor et al., Widefield two-photon excitation without scanning: Live
specimen and the number of photons each can emit before suffers from a relatively low scan speed, slowing image capture. away from the others, thereby reducing its capabilities in the re- cell microscopy with high time resolution and low photo-bleaching.
bleaching irreversibly. They also require specialized photo- In one recent report, researchers at the University of Strathclyde maining three characteristics. To attain higher spatial resolution, PLOS ONE 11, e0147115 (2016).
switchable labels and excitation intensities of 103 to 108 W/cm2, in Scotland describe a widefield-based image-capture strategy for example, an imaging method requires more pixels, and that 14. M. May, Top tips from the superresolution microscopy pros, in
which are orders of magnitude greater than the 0.1 W/cm2 un- to circumvent that problem, which they used to image calcium requires more measurements, which takes more time and might Microscopy now: Getting the most from your imaging, pp. 13–16.
der which life evolved” (8). signaling events in rat neurons at up to 100 Hz (13). damage the cells by exposing them to more light. Sponsored supplement to Science, Oct. 2 (2015).
◥
◥ MTs not anchored to other cytoskeletal or sarco- Confocal active frame, rest frame 0ms 1.0
◥ MTs not anchored to other cytoskeletal or sarco-
RE
RE SE
RESEA
SEAA RC
RC HH AA RT
RT IC
IC LL EE meric proteins could rearrange or slide passively
meric proteins could rearrange or slide passively
with the surrounding medium; anchored MTs 0.8
with the surrounding medium; anchored MTs
Pearson’s
could directly experience contractile force and
MICROTUBULES could directly experience contractile force and
MICROTUBULES themselves deform under load; or the MTs could subsarcolemmal SiR tubulin
themselves deform under load; or the MTs could 0.6
break and/or disassemble and reform. These pos- 35ms
break and/or disassemble and reform. These pos-
Detyrosinated
Detyrosinated microtubules
microtubules
sibilities offer divergent mechanisms for the reg-
sibilities offer divergent mechanisms for the reg-
ulation of mechano-signaling and the overall
ulation of mechano-signaling and the overall
0.4
rest 70ms
0 1 2
Time (s)
3 4
to quantify.
contracting
contracting cardiomyocytes
cardiomyocytes
Standard confocal imaging, although capable
Standard confocal imaging, although capable
of resolving MTs in living cells (17), suffers from
of resolving MTs in living cells (17), suffers from
Airyscan rest
MT amplitude (µm)
MTs uncoupled from sarcomeres and buckled less during contraction, which allowed bulin, based on the fluorophore silicon rhodamine
MTs uncoupled from sarcomeres and buckled less during contraction, which allowed bulin, based on the fluorophore silicon rhodamine
sarcomeres to shorten and stretch with less resistance. Conversely, increased (18) (Fig. 1C and movie S2), we imaged internal 1µm
sarcomeres to shorten and stretch with less resistance. Conversely, increased (18) (Fig. 1C and movie S2), we imaged internal
detyrosination promoted MT buckling, stiffened the myocyte, and correlated with impaired MTs during contractions triggered by a 1 Hz elec-
detyrosination promoted MT buckling, stiffened the myocyte, and correlated with impaired MTs during contractions triggered by a 1 Hz elec- 20
function in cardiomyopathy. Thus, detyrosinated MTs represent tunable, compression- trical field stimulation in isolated mouse cardio-
function in cardiomyopathy. Thus, detyrosinated MTs represent tunable, compression-
Counts
trical field stimulation in isolated mouse cardio- 3.3µm
resistant elements that may impair cardiac function in disease. myocytes. We were able to capture MT behavior ~ 3.3µm
resistant elements that may impair cardiac function in disease. myocytes. We were able to capture MT behavior
during contraction and found that longitudinally
during contraction and found that longitudinally
oriented MTs frequently deformed and developed 10
oriented MTs frequently deformed and developed
long with its well-defined transport func- modify their mechanical properties and binding sinusoidal buckles. Because SiR tubulin may poly-
A
long with its well-defined transport func- modify their mechanical properties and binding sinusoidal buckles. Because SiR tubulin may poly-
tions, the microtubule (MT) network serves interactions. Detyrosination, a PTM of a-tubulin, merize MTs (18), we also generated adenovirus
tions, the microtubule (MT) network serves interactions. Detyrosination, a PTM of a-tubulin, merize MTs (18), we also generated adenovirus ~ 4.7µm
multiple mechanical roles in the beating has recently been shown to augment MT-dependent encoding a small fragment of the MT-binding pro-
multiple mechanical roles in the beating has recently been shown to augment MT-dependent encoding a small fragment of the MT-binding pro- 0
cardiomyocyte. MTs function as mechano- mechanotransduction in dystrophic cardiac and tein ensconsin fused to three copies of GFP (EMTB-
cardiomyocyte. MTs function as mechano- mechanotransduction in dystrophic cardiac and tein ensconsin fused to three copies of GFP (EMTB- 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
transducers, converting changing contract- skeletal muscle (12). This specific PTM is also in- 3xGFP) to decorate MTs. We achieved similar Buckle Wavelength, λ ( µm) rest contracted
transducers, converting changing contract- skeletal muscle (12). This specific PTM is also in- 3xGFP) to decorate MTs. We achieved similar 1.65µm
ile forces into intracellular signals (1, 2). MTs may creased in animal models of heart disease (1, 13, 14), results (table S1) but with improved signal-to-
ile forces into intracellular signals (1, 2). MTs may creased in animal models of heart disease (1, 13, 14), results (table S1) but with improved signal-to-
also act as compression-resistant elements, which which raises a mechanistic question: If mechano- noise ratios (Fig. 1D and movie S3).
also act as compression-resistant elements, which which raises a mechanistic question: If mechano- noise ratios (Fig. 1D and movie S3).
could provide a mechanical impediment to cardio- signaling is altered, have the mechanical properties We measured blindly selected MTs for deforma- Fig. 1. MTs reversibly buckle in contracting cardiomyocytes. (A) The sub- cycles. Initial drop to ~0.96 is due to imaging noise. (G) Quantification of
could provide a mechanical impediment to cardio- signaling is altered, have the mechanical properties We measured blindly selected MTs for deforma-
myocyte contraction (3–5). If so, they must bear of the myocyte changed? tion with two parameters—amplitude and wave- surface (top) and interior (bottom) cardiomyocyte MT network. (B) High- buckling amplitude (measured from centerline to edge) and l (wavelength
myocyte contraction (3–5). If so, they must bear of the myocyte changed? tion with two parameters—amplitude and wave-
some of the compressive and tensile load of a Although the idea that a proliferated (and per- length (Fig. 1G). Where possible, the same MT was speed confocal imaging of MTs at rest (top) and during contraction (bottom) measured as twice the distance between consecutive inflection points).
some of the compressive and tensile load of a Although the idea that a proliferated (and per- length (Fig. 1G). Where possible, the same MT was
working myocyte. Unfortunately, little is known haps modified) MT network may mechanically followed through contraction. Amplitude rose labeled with SiR tubulin with brightness increased for comparison with (C). (H) Amplitude of MTs labeled with EMTB-3xGFP in resting (black) and con-
working myocyte. Unfortunately, little is known haps modified) MT network may mechanically followed through contraction. Amplitude rose
about MT behavior during the contractile cycle. interfere with contraction is attractive, the “MT quickly from resting to contracted levels (Fig. 1H), (C) Airyscan imaging of the same MTs as in (B) at rest and during con- tracted (red) cardiomyocytes. The threshold to determine buckling occur-
about MT behavior during the contractile cycle. interfere with contraction is attractive, the “MT quickly from resting to contracted levels (Fig. 1H),
During this cycle, Ca2+ 2+
-mediated actin-myosin hypothesis” has remained controversial [for re- with clearly visible buckles. Using a threshold of traction. (D) Wider view of MTs labeled with EMTB-3xGFP at rest (top) and rence (blue line) was two standard deviations above the mean resting value.
During this cycle, Ca2+-mediated actin-myosin hypothesis” has remained controversial [for re- with clearly visible buckles. Using a threshold of
interaction first shortens repeating contractile views, see (15, 16)]. Two important limitations two standard deviations above resting amplitude, during contraction (bottom). (E) MTs imaged throughout a contractile (I) Distribution of buckling wavelengths in cardiomyocytes shows a dom-
interaction first shortens repeating contractile views, see (15, 16)]. Two important limitations two standard deviations above resting amplitude,
units called sarcomeres, which are then stretched have hindered our understanding: (i) a reliance we found that two-thirds of MTs buckle under cycle (cyan) were overlaid onto the network configuration from the initial inant population between 1.6 and 1.7 mm, and a second population at 3.3 mm.
units called sarcomeres, which are then stretched have hindered our understanding: (i) a reliance we found that two-thirds of MTs buckle under
as the heart fills with blood during diastole. on blunt pharmacological tools (colchicine/Taxol) control conditions (Fig. 1H). frame at rest (red). (F) Colocalization analysis of (E) shows that MTs re- (J) A representative MT demonstrating buckles with wavelengths that cor-
as the heart fills with blood during diastole. on blunt pharmacological tools (colchicine/Taxol) control conditions (Fig. 1H).
Although an isolated MT would present minimal that have off-target consequences; (ii) a lack of MT buckles quickly reversed during relaxation, peatedly return to the same position. Pearson’s coefficient is used to esti- respond to the distance between one (1.65 mm) or two (3.3 mm) adjacent
Although an isolated MT would present minimal that have off-target consequences; (ii) a lack of MT buckles quickly reversed during relaxation,
resistance to myocyte compression, the stiffness direct observation of MTs under the stress and and the configuration of the MT network between mate goodness of fit to original MT configuration over several contractile sarcomeres.
resistance to myocyte compression, the stiffness direct observation of MTs under the stress and and the configuration of the MT network between
of the network within a living cell, with MT- strain of the contractile cycle. Here, we charac- contractions tightly colocalized with the network
of the network within a living cell, with MT- strain of the contractile cycle. Here, we charac- contractions tightly colocalized with the network
associated proteins and other cytoskeletal ele- terized MTs under contractile loads using a configuration from previous cycles (Fig. 1, E and
associated proteins and other cytoskeletal ele- terized MTs under contractile loads using a configuration from previous cycles (Fig. 1, E and indicative of ordered
ordered geometric
geometric constraints
constraints on
on ily
ments, can change by orders of magnitude (6, 7). high-resolution imaging technique and directly F), with a minimal mean reduction in Pearson’s indicative of ily conserved
conserved across
across eukaryotes
eukaryotes (20)
(20) and
and appears
appears ing this
ing this construct
construct in in isolated
isolated cardiomyocytes
cardiomyocytes
ments, can change by orders of magnitude (6, 7). high-resolution imaging technique and directly F), with a minimal mean reduction in Pearson’s the buckling
buckling MT.
It is in this context that MTs are proposed to act tested how MT detyrosination may regulate colocalization of 0.01 per contractile cycle (n = the MT. This
This was
was observed
observed inin certain
certain required
required for for life
life (21),
(21), yet
yet its
its functional roles are
functional roles are could effectively
could effectively reduce
reduce the the level
level of
of detyrosina-
detyrosina-
It is in this context that MTs are proposed to act tested how MT detyrosination may regulate colocalization of 0.01 per contractile cycle (n = cells
as compression-resistant elements (6, 8) that may load-bearing and the mechanical properties of 18 runs). The rapid and precise reversibility of the cells where faint transverse
where faint transverse staining
staining at
at the
the Z-disc
Z-disc still
still poorly
poorly understood.
understood. Because
Because detyrosination
detyrosination tion as
tion as shown
shown by by both
both immunofluorescence
immunofluorescence (Fig. (Fig.
as compression-resistant elements (6, 8) that may load-bearing and the mechanical properties of 18 runs). The rapid and precise reversibility of the shows
impair sarcomere shortening and thus cardiac the myocyte. network deformations suggested tight coupling shows MTs
MTs buckling
buckling between
between sarcomeric con-
sarcomeric con- can
can protect
protect MTsMTs from
from disassembly
disassembly (22, 23) and
(22, 23) and 2, A
2, A and
and B)B) and
and immunoblot
immunoblot (Fig. (Fig. 2,
2, C
C and
and D),D),
impair sarcomere shortening and thus cardiac the myocyte. network deformations suggested tight coupling straints
function, particularly in disease states associated to the contractile apparatus, and it argues against straints (movie
(movie S4).
S4). can
can facilitate
facilitate their
their cross-linking
cross-linking with
with intermediate
intermediate which resulted
which resulted inin aa 71%
71% reduction
reduction inin the
the amount
amount
function, particularly in disease states associated to the contractile apparatus, and it argues against filaments
with MT proliferation (8–10). Posttranslational MTs buckle under contractile load
MTs buckle under contractile load MT breakage and regrowth contributing to me- Detyrosination filaments (IFs)(IFs) (24, 25), we
(24, 25), we hypothesized
hypothesized thatthat the
the of polymerized,
of polymerized, detyrosinated
detyrosinated MTs, MTs, with
with aa con-
con-
with MT proliferation (8–10). Posttranslational
modification (PTM) of MTs (11, 12) could also MT networks in cardiomyocytes have two major
MT breakage and regrowth contributing to me-
chanical properties and signaling over the time Detyrosination regulates
regulates MT
MT buckling
buckling in
in high
high proportion
proportion of of detyrosination
detyrosination maymay confer
confer the
the comitant up-regulation
comitant up-regulation of of tyrosinated
tyrosinated tubulin
tubulin
modification (PTM) of MTs (11, 12) could also MT networks in cardiomyocytes have two major chanical properties and signaling over the time the heart
the heart resilient
features (Fig. 1A): an orthogonal grid just beneath scale of myocyte contraction. resilient load-bearing
load-bearing capabilities
capabilities ofof the
the cardiac
cardiac (Fig. 2,
(Fig. 2, C
C and
and D,
D, and
and fig. S2). Overexpression
fig. S2). Overexpression of of
features (Fig. 1A): an orthogonal grid just beneath scale of myocyte contraction. This
the membrane that wraps the myofibrils and a A notable feature of the MT buckles was the This robust
robust buckling
buckling behavior of the
behavior of the MT
MT network
network cytoskeletal
cytoskeletal network.
network. TTL
TTL also resulted in
also resulted in aa modest
modest (10%)
(10%) reduction
reduction in in
the membrane that wraps the myofibrils and a A notable feature of the MT buckles was the may
11 deeper network composed primarily of longitu- emergence of subpopulations of buckle wavelength may be be aa result
result of
of aa particularly
particularly high
high abundance
abundance Using
Using antibodies
antibodies specific
specific to
to detyrosinated
detyrosinated a- a- the density
the density of of the
the polymerized
polymerized MT MT network
network (Fig.
(Fig.
1Department
Department ofof Physiology,
Physiology, Pennsylvania
Pennsylvania Muscle
Muscle Institute,
Institute, deeper network composed primarily of longitu- emergence of subpopulations of buckle wavelength
Department Pennsylvania
University
University of
of Physiology, Pennsylvania
of Pennsylvania Perelman
Perelman School
Muscle
School of
Institute,
of Medicine,
Medicine, dinal elements that interdigitate the myofibrils. centered at ~1.65 mm, 3.3 mm, and perhaps even of “detyrosinated”
of “detyrosinated” MTs MTs in in adult
adult cardiomyocytes
cardiomyocytes tubulin,
tubulin, we we found
found aa high
high abundance
abundance of of detyrosi-
detyrosi- 2B), consistent
2B), consistent with
with an an increased
increased disassembly
disassembly of of
dinal elements that interdigitate the myofibrils. centered at ~1.65 mm, 3.3 mm, and perhaps even (19).
University of PA
Philadelphia, Pennsylvania
19104, USA.Perelman
22 SchoolofofMaterials
Department
Philadelphia, PA 19104, USA. 2Department of Materials
Medicine, Longitudinal MTs often run many sarcomeres in 4.7 mm (Fig. 1I). These corresponded closely to the (19). Detyrosination
Detyrosination is is aa PTM
PTM of a-tubulin where
of a-tubulin where nation
nation in the a-tubulin
in the a-tubulin network
network of adult myocytes
of adult myocytes tyrosinated MTs
tyrosinated MTs (22, 23). We
(22, 23). We complemented
complemented this this
PA 19104, USA. Department
Philadelphia,Engineering, of Materials Longitudinal MTs often run many sarcomeres in 4.7 mm (Fig. 1I). These corresponded closely to the the
Science
Science and
and Engineering, University
Science and Engineering,
University of
of Pennsylvania
Pennsylvania School
University Philadelphia,
School
of Pennsylvania School length but do not span the full cell. Given that length of one, two, or three contracted sarcomeres, the C-terminal
C-terminal tyrosine residue has
tyrosine residue has been
been cleaved
cleaved (Fig.
(Fig. 2, A and B), as expected (12, 19). To test
2, A and B), as expected (12, 19). To test the
the genetic
genetic strategy with aa pharmacological
strategy with approach
pharmacological approach
length but do not span the full cell. Given that length of one, two, or three contracted sarcomeres, by
of
of Engineering
Engineering and
of Engineering
33
and Applied
Applied Science,
andofApplied
Science, Philadelphia, PA
Science, Philadelphia,
PA 19104,
19104,
PA 19104, cardiomyocytes change shape during contraction, respectively. Although MTs buckling outside of by aa tubulin
tubulin carboxypeptidase (TCP); this
carboxypeptidase (TCP); this process
process role
role ofof detyrosinated
detyrosinated MTs,MTs, wewe generated
generated adeno-
adeno- to
to inhibit TCP using
inhibit TCP using parthenolide
parthenolide (PTL)(PTL) (26).
(26). PTL
PTL
USA. Department Medicine, University of Pennsylvania
USA. 3Department of Medicine, University of Pennsylvania cardiomyocytes change shape during contraction, respectively. Although MTs buckling outside of can
USA. Department
Perelman ofMedicine,
Medicine,Philadelphia,
University ofPAPennsylvania
Perelman School of Medicine, Philadelphia, PA 19104, USA.
School of 19104, USA. the MT cytoskeleton must accommodate this these populations could be found in our data with- can be readily reversed by tubulin tyrosine ligase
be readily reversed by tubulin tyrosine ligase virus
virus encoding
encoding TTL (AdV-TTL) with
TTL (AdV-TTL) with aa Discosoma
Discosoma treatment also reduced the fraction
treatment also reduced the fraction of detyrosi-of detyrosi-
the MT cytoskeleton must accommodate this these populations could be found in our data with- (TTL)
Perelman School of Medicine, Philadelphia, PA 19104, USA.
*Corresponding author. Email: bpros@mail.med.upenn.edu
*Corresponding author. Email: bpros@mail.med.upenn.edu change. There are three apparent possibilities: out difficulty, these subpopulations were strongly (TTL) (11).
(11). This
This tyrosination
tyrosination cycle
cycle is
is evolutionar-
evolutionar- red fluorescent
red fluorescent protein
protein (DsRed)
(DsRed) reporter.
reporter. Express-
Express- nated MTs, albeit
nated MTs, albeit to
to aa lesser
lesser extent (42%) than
extent (42%) than
*Corresponding author. Email: bpros@mail.med.upenn.edu change. There are three apparent possibilities: out difficulty, these subpopulations were strongly
8
8 Originally
Originally published
published 22on
online April 2016 in
22 April SCIENCE
2016 in SCIENCE sciencemag.org SCIENCE
sciencemag.org
sciencemag.org SCIENCE
SCIENCE aaf0659-2
SCIENCE sciencemag.org
22 APRIL 2016 • VOL 352 ISSUE 6284 9
sciencemag.org SCIENCE
SCIENCE sciencemag.org 22 APRIL 2016 • VOL 352 ISSUE 6284 aaf0659-1 SCIENCE sciencemag.org 9
OIT
T PTLIM
E IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S RE S E ARCH | R E S E A R C H A R T I C L E R ES E A RC H | R E S EA R C H A R T I C LE S E C T IION
O N T WO | RESEARCH ARTICLES
∆SL (µm)
1.8 0.20
We next measured mechanical resistance directly
1.60
0.25 20 using atomic force microscopy (AFM). AFM mea-
1.7
0.10
surements of transverse stiffness were performed
0.20
1.50 1.6 10 across a range of indentation rates. Myocyte stiff-
0.15
ness changed substantially with indentation rate
1.40
1.5 0.00 0 and was well fit by a standard linear solid model
0.0 1.0 2.0 3.0 Resting SL ∆SL Shortening
0 1 2 3 4 Log TTL Fluorescence time (ms)
(SLSM) [methods in supplementary materials (SM)]
Time (s) (µm) (µm)
(fig. S4 and Fig. 4A). At low rates (100 nm/s), the
Ctrl ** DMSO **
stiffness of the cardiomyocyte was essentially elas-
4 4 TTL 4 PTL tic, reported as E1, and was reduced by PTL treat-
[Ca2+]i (F/F0 )
10
SCIENCE sciencemag.org sciencemag.org
22 APRIL 2016 • VOL 352 ISSUE 6284 SCIENCE
aaf0659-3 SCIENCE sciencemag.org
aaf0659-4 22 APRIL 2016 • VOL 352 ISSUE 6284 11
sciencemag.org SCIENCE
OIT
T PTLIM
E IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S RE S E ARCH | R E S E A R C H A R T I C L E R ES E A RC H | R E S EA R C H A R T I C LE S E C T IION
O N T WO | RESEARCH ARTICLES
Step (µm)
ctrl MT titin
MT
PTL (Force) (Length) 10 ε δ
TTL
rceM
Force
0
TTL medium z-disk
Elastic Modulus (Pa)
600 E2 2.3
SL (µm)
actin myosin actin
2.1 contractile element
E1 1.8 sarcomeric anchor
400
0 = 0.1 µm
0.4 Fpeak
titin
rest
Force (µN)
200 Fs.s.
0.2
medium
5 µm 0
0 anchor anchor
0.1 1 10 100 Resting Sarcomere Length - 1.84µm 0 10 20 30 40 50 60
Velocity ( µm/s) Time (s)
Stretch 0.5 microtubule
Ctrl - F peak
E1(Pa) E2 (Pa) Viscosity (Pa*s)
500 400
TTL - F peak γ/2 ξ γ/2
100 0.4
Ctrl - F s.s.
400 TTL - Fs.s. contracted
Force (µN)
300 80 0.3 0.0
0.15
∆ SL (µm)
200 ** *** TTL
0.10
TTL
40
0.1 ***
100
100 20
***
-0.2 0.05
0.0
0 0 0 Stretch Sarcomere Length - 2.07µm 1.8 1.9 2.0 2.1 2.2 2.3
Ctrl TTL DMSO PTL Ctrl TTL DMSO PTL Ctrl TTL DMSO PTL SL (µm) -0.3 0.00
Fig. 4. Detyrosinated MTs regulate the viscoelasticity of cardiomyocytes. controller and were subjected to stretch. MTs visualized by EMTB-3xGFP contracted, detyrosination
(A) Elastic modulus of cardiomyocytes measured by AFM at various inden- (C, blue and aqua) at rest (top) and at a stretched length (bottom). (D) Rep- 0.0 0.2 0.4 0.0 0.2 0.4
tation velocities and fit to SLSM (see methods in SM). (B) Quantification of resentative force versus length protocol. A series of stepwise stretches (red) in Time (s) Time (s)
velocity-independent (E1) and velocity-dependent (E2) components of the 4-mm increments are applied to an isolated myocyte, which increases sarco-
elastic modulus, and SLSM fit–derived viscosity. Both TTL overexpression and mere length (SL, black). Passive tension (blue) generated by the step relaxes null TTL
PTL treatment significantly reduced elasticity and viscosity. There were no quickly from a peak value to a new steady state. (E) Force measurements
significant differences in these parameters between dimethyl sulfoxide (DMSO) binned according to measured change in sarcomere length with a given step
(gray) and AdV-null (black) transduced cells (P = 0.28, 0.34, and 0.33, re- size. TTL-overexpressing cells exert reduced peak passive tension during step
spectively). Reductions in stiffness due to TTL overexpression are also apparent changes in length, with a more modest reduction in steady-state tension. Data
in cells under stretch along the longitudinal axis. (C) Myocytes were attached are presented as means ± SEM; *P < 0.05, **P < 0.01, ***P < 0.001. Further
via glass cell holders (C, top, and fig. S5) to a force transducer and length statistical details are available in tables S3 and S4.
Fig. 5. Modeling MTs in the contracting sarcomere. (A) Mechanical sche- sarcomere length at peak contraction and buckling amplitude. (C) and
taut when the cell was stretched and held at long sarcomeric periodicity of buckles (Fig. 1I) also increase the average distance between detyrosi- matic of the modeled sarcomere. A force-generating contractile arm (top) is (D) recapitulate experimental observations for TTL-overexpressing myocytes
sarcomere lengths (movie S8). suggests an underlying structural constraint that nated tubulins that could interact strongly with coupled in parallel at the Z-disc to a spring element representing titin (orange), after this change. (E and F) The cardiac sarcomere, shown with MTs with
changes in MT or medium stiffness alone cannot MT anchoring points. This disruption of MT- a viscoelastic medium (yellow spring and dashpot), and a MT (green) with putative stiff anchors to the sarcomere, here, at the Z-disc. Contraction
Model of MT contribution to explain. We thus chose to model MT buckling sarcomere coupling produced model outputs (Fig. anchors (fuschia pink) to the Z-disc (gray). The anchor to the Z-disc is only reduces the distance between anchor points, which requires the MTs either to
cardiac contractility within a contractile model that includes a MT 5, C and D) that closely recapitulated our experi- engaged at regions of MT detyrosination. (B) TTL overexpression is modeled buckle (G) if the anchors are engaged or to slide (H), if the anchors are not
We next sought to develop a mathematical mod- compression-resistive element that’s interaction mental contractility and buckling results. by allowing the anchors to slide for 50 nm at each end before engaging and engaged and force incident on the MT remains low. Mathematical model
el to recapitulate the experimentally measured with the sarcomere can be varied (see model in An alternate possibility to the sliding anchor is transmitting force to the MT at a detyrosinated subunit. (C) The change in parameters are available in table S5.
changes in MT buckling and contractility when SM for details). that the anchor is completely uncoupled by sup-
detyrosination is reduced. Previous work mod- Using the mechanical scheme detailed in pressing detyrosination and reverts to buckling Desmin forms structural bundles that form a The two populations of MTs show similar overall removal of desmin should both decrease cyto-
eling MT buckles (6) suggests that three critical Fig. 5A, we fitted the contraction resulting from a behavior governed by local viscoelasticity rather complex with the Z-disc (29), and intermediate patterning in WT myocytes, except for a specific skeletal stiffness and prevent tyrosination-
variables determine buckling behavior; MT stiff- log-normal force input to derive both contractile than underlying structure, as proposed for less filaments can form detyrosination-dependent accumulation of detyrosinated (and not tyrosi- dependent changes in viscoelasticity. Blind studies
ness, stiffness of the surrounding medium, and and buckling parameters. By modifying the in- rigidly organized cell types, including developing cross-links with MTs (30, 31). nated) tubulin in transverse bands at the Z-disc in WT and KO myocytes revealed that desmin
force incident on the long axis of the MT. How cident force applied to a MT for a given amount myocytes (6). In either case, the coupling of MTs We first sought to determine whether desmin that colocalized with desmin (Fig. 6, D and E, KO myocytes were significantly less stiff than WT
these three variables are predicted to alter MT of sarcomere shortening (Fig. 5A and model in to the sarcomere is reduced, which impairs their preferentially associates with detyrosinated MTs. and fig. S7). Note that KO animals lacked this counterparts (Fig. 6, F and G), and that treatment
behavior and myocyte mechanics is described in SM), we simulated the effect of a sarcomeric an- ability to resist contraction. Cosedimentation of cardiomyocyte lysates showed transverse pattern completely (Fig. 6D and fig. with PTL no longer reduced viscoelasticity (Fig. 6,
fig. S6. Of the three, only a decrease in incident chor sliding and then catching at detyrosinated that desmin forms pellets with polymerized MTs S7H), although the Z-disc itself remained intact F and G).
force can explain our experimental observations regions of the MT. Inclusion of a 100-nm slide Potential role for desmin as a (Fig. 6A) in direct proportion to their level of (fig. S7C). In addition, KO myocytes had a denser
after suppressing detyrosination. If MT anchor- (50 nm at each anchor, see model) before MTs sarcomeric MT anchor detyrosination (Fig. 6, B and C), which indicated (fig. S7F) and more disorganized MT network MT detyrosination is sufficient to impair
ing to the sarcomere is disrupted, the reduced engage with the rest of the sarcomere is rea- The putative characteristics of the anchor—a me- a specific and sensitive interaction. We also co- (Fig. 6D and fig. S7, B and E), which suggested cardiomyocyte contractility
incident force on the MT may drop below the sonable, given the ~80% reduction in detyrosi- chanically stiff protein capable of forming complexes stained cardiomyocytes from desmin-depleted that desmin is required for proper MT network Increasing detyrosination correlates with impaired
critical force required for buckling, which results nated area observed by immunofluorescence with MTs and restricted to a spatially defined [knockout (KO)] and wild-type (WT) mice for organization. function in animal models of heart disease (13, 14).
in simultaneous decreases in buckling amplitude with TTL overexpression (Fig. 2C), and reflects region of the sarcomere—suggested the interme- desmin and both tyrosinated and detyrosinated If desmin cross-links with detyrosinated MTs Thus, we next tested whether increasing detyr-
(Fig. 2) and viscoelasticity (Figs. 3 and 4). The the fact that reductions in detyrosination would diate filament desmin as an immediate candidate. tubulin to observe any preferential interaction. to structurally reinforce the network, then the osination could directly impair cardiac contractility.
12
SCIENCE sciencemag.org sciencemag.org
22 APRIL 2016 • VOL 352 ISSUE 6284 SCIENCE
aaf0659-5 SCIENCE sciencemag.org
aaf0659-6 22 APRIL 2016 • VOL 352 ISSUE 6284 13
sciencemag.org SCIENCE
O PT IM IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S RE S E ARCH | R E S E A R C H A R T I C L E R ES E A RC H | R E S EA R C H A R T I C LE S E C T I O N T WO | RESEARCH ARTICLES
desmin in MT pellet
b 120
MT cosedimentation -tu 0.8 L shTTL 500
yr TT
shows the interaction dT *** *** 0.9 800 400
Modulus (Pa)
100
between polymerized b 400
tu H
Modulus (Pa)
0.6
Modulus (Pa)
MTs (pellet) and desmin. α- 0.8 PD
GA 600 80
Pa*s
slope = 1.21 300
(B) Quantification of the in 300
sm 0.4 0.7 r2 = 0.88 tub 60
amount of detyrosination
de yr- 400
and desmin (relative to dT 200
200
40
the total amount of P DH 0.2 0.6
tub
α-
tubulin) in the MT pellet GA 0.5
200 100 100
20
from cardiomyocyte 0.0 0.5 0.6 0.7 0.8 0.9 1.0
lysates with and without dTyr desmin dTyr in MT pellet 0 0 0 0
PTL treatment. Data
0.1 1 10 100 E1 E2 Viscosity
Des KO desmin Velocity (µm/s)
were normalized to WT desmin
D E 0.3 F
desmin
DMSO level. (C) The ctrl shTTL Donor Hypertrophy DCM HCM
amount of desmin asso-
ciated with the MTs after 5 ***
*** dTyr-tub
dTyr-tub
0.1 µm
5µm
nation across several 0
GAPDH
experiments in rat cardi- dTyr-tub dTyr-tub
omyocytes and C2C12 0.1
TTL
cells. (D) Immuno- ctrl
fluorescence of desmin, shTTL -5 tyr-tub
***
merge
dTyr, and Tyr-tubulin 1s
shows dTyr-specific 0.0 GAPDH
contraction relaxation SL (µm)
transverse pattern in WT,
but not desmin KO, myo-
G 3.5
α-tubulin
H
cytes. (E) Overlay of dTyr- tyr-tub tyr-tub
tubulin and desmin. (See D 0.3 dTyr:des
dTyr-tubulin *** 80
Mander’s coefficient
tyr:des
fig. S7 for more 3.0 tyr-tubulin
* Healty donor
LVEF (%)
0.1
in myocyte stiffness and
viscosity in WT, but not 2.0 * 50
800
SEM; *P < 0.05, **P < * p = 0.2 10
80 0.5
0.01, ***P < 0.001 with
### 600 ###
respect to DMSO treat- 1000 400 ###
60 n=9 n = 17 n = 10 n = 10 n = 15
ment; ###P < 0.001 with n = 17 0 1 2 3 4 5 6 7
0.0
respect to untreated WT 400 *** dTyr-tubulin (fold change)
40 Healthy Hypertrophy DCM Ischemic PostVad HCM
myocytes. Further statis- 500 200 Donors DCM
tical details are available 200 20
in table S6. Fig. 7. Increasing detyrosination impairs contraction and is associated human heart lysates. (G) Data from pooled analysis; n = 17 healthy donors,
0 0 0 0 with human heart failure. (A) Western blot shows that shRNA against 9 hypertrophy, 17 DCM, 11 ischemic, 10 with DCM following ventricular
0.1 1 10 100 WT WT KO KO WT WT KO KO WT WT KO KO
+PTL +PTL
TTL selectively increases dTyr-tubulin without changing overall levels of assist device support (VAD DCM), and 15 HCM hearts. (H) There was a
Velocity ( µm/s) +PTL +PTL +PTL +PTL
a-tubulin. (B) Elastic modulus of control and shTTL-expressing myocytes negative correlation between LVEF and dTyr-tubulin expression in control
at various indentation rates. (C) shTTL myocytes demonstrate increases and hypertrophic cardiomyopathy patients. Data are presented as means ±
in E1, E2, and viscosity. (D and E) TTL suppression significantly reduces SEM; *P < 0.05, **P < 0.01, ***P < 0.001. Further statistical details are
Using an adenoviral construct expressing short and suppressed contractile velocity and mag- S7). Detyrosinated tubulin was significantly in- contractile magnitude and velocity. (F) Representative Western blots from available in tables S7 to S9.
hairpin RNAs (shRNAs) against TTL (shTTL), we nitude (Fig. 7, D and E). creased in patients with clinically diagnosed
suppressed TTL expression, which enhanced We next examined whether this modification hypertrophic and dilated cardiomyopathies (HCM
detyrosination (Fig. 7A and fig. S8). shTTL- correlated with functional deficits in human heart and DCM, respectively), along with a modest in- indicator of cardiac contractility (Fig. 7H). gested a specific link between detyrosination TTL was unchanged in all patient populations,
transduced myocytes were then tested for their disease. To this end, we analyzed left ventricular crease in total tubulin content (Fig. 7, F and G). There was no such correlation detected be- and LVEF. Myocardium from patients with which showed that a decreased expression of the
viscoelastic and contractile properties. The ex- tissue samples from healthy patient donors and Blind analysis of HCM patient data showed that tween LVEF and total or tyrosinated tubu- DCM all demonstrated considerably depressed tyrosinating enzyme does not explain the in-
cess detyrosination alone was sufficient to in- from patients exhibiting varying degrees of heart detyrosination inversely correlated with left ven- lin levels, nor any correlation between heart LVEF and variable, but increased, detyrosi- crease in detyrosinated tubulin in patients with
crease viscosity and stiffness (Fig. 7, B and C) disease due to several underlying causes (table tricular ejection fraction (LVEF), a primary weight and detyrosination (fig. S9), which sug- nated tubulin. heart disease (Fig. 7G). Because the molecular
14
SCIENCE sciencemag.org sciencemag.org
22 APRIL 2016 • VOL 352 ISSUE 6284 SCIENCE
aaf0659-7 aaf0659-8
SCIENCE sciencemag.org
22 APRIL 2016 • VOL 352 ISSUE 6284 15
sciencemag.org SCIENCE
Growing Gap in Life Expectancy by Income: Implications for 22 December 2015; accepted 17 March 2016
Princeton Center for Translational Research on Aging
Federal Programs and Policy Responses” (2015). Published online 21 April 2016
grant 2P30AG024928. Data and code are available at
10.1126/science.aaf1437
5. J. Pijoan-Mas, J. V. Ríos-Rull, Demography 51, 2075–2102 http://dx.doi.org/10.7910/DVN/C2VYNM.
T IT
O PTLIM
E IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S R ES (2014).E A RC H | R E PO R TS S E C T I O N T WO | RESEARCH ARTICLES
6. H. Waldron, Soc. Secur. Bull. 67, 1–28 (2007).
7. H. Waldron, Soc. Secur. Bull. 73, 1–37 (2013).
identity of TCP is unknown, it is unclear if up- appears to return in the form of a restoring force 9. M. R. Zile et al., J. Am.8. Coll.
J. Wilmoth,33,
2. B. Cardiol.
Bosworth, K.250–260 (1999). inofInternational
C. Boe, M. Barbieri,
Zhang, “Evidence
Differences
Increasing Differential NEURODEVELOPMENT
ACKN OWLED GMEN TS SUPPLEMENTARY MATERIALS ing that elevated levels at 1 mo are likely
10. G. Cooper 4th, Am. J. Physiol.in Mortality
HeartACirc.at Older
Physiol.Ages:
291, Dimensions
H1003– and Sources,
regulation of the detyrosinating enzyme may un- that would quicken sarcomere extension. This Mortality: Comparison of the
E. M. Crimmins, S. H. Preston, B. Cohen, Eds.
HRS and SIPP,” Center for We thank M. Barbieri, A. Case, A. Deaton, J. Goldstein, www.sciencemag.org/content/352/6286/708/suppl/DC1 not a developmental artifact. C1q was also
H1014 (2006).
derlie this effect. implies that energy used to deform MTs under- Retirement Research at Boston College Working Paper 2015-13
(National Academies Press, Washington, DC, 2011), I. Kuziemko, R. Lee, and K. Wachter, as well as seminar Materials and Methods similarly increased in the hippocampus of
Discussion
goes substantial loss. Buckling of the MT exerts
compressive force on the surrounding matrix and
11. C. Janke, J. C. Bulinski, Nat.
(2011). pp.
9. G. K. 6,
12. J. P. Kerr et al., Nat. Commun.
Rev. Mol.
(2015).
337–372.
4. National8526
Singh, M.
Cell Biol. 12,
Research
773–786
3. R. Chetty et al., JAMA 10.1001/jama.2016.4226 (2016).
Siahpush, Int. J. Epidemiol. 35, 969–979
(2015). Council, Committee on the Long-Run
Complement and microglia mediate
participants at Berkeley, the Chicago Federal Reserve,
Fundação Getúlio Vargas São Paulo, Bonn University,
University of Munich, Princeton University, ETH Zurich,
Figs. S1 to S8
Tables S1 to S4
References (36–38)
another model of AD, the APP/PS1 (prese-
nilin 1) mice (15) (fig. S2C). Notably, SIM
13. H. Sato et al., J. Cell Biol.(2006).
Our findings demonstrate a regulatory pathway
for MT load–bearing and myocyte mechanics
deforms the cytoplasm, which, due to its intrinsic
viscosity, can act as an energy sink during each
10. M.
14. S. Belmadani, C. Poüs, R.Growing
Méry, Mol. Cell. Biochem.Med.
139, 963–973 (1997).
Macroeconomic
Ezzati,
237,5,
Federal
A.
Ventura-Clapier, B.
Gap in Life
e66 (2008).
39–46
Programs
Effects of the Aging U.S. Population, “The
Friedman,
(2002).
S. C. Kulkarni, C. J. Murray, PLOS
R. Fischmeister,
Expectancy P.-F. by Income: Implications for
and Policy Responses” (2015).
early synapse loss in Alzheimer
and the University of Zurich for comments. Supported by
Princeton Center for Translational Research on Aging
22 December 2015; accepted 17 March 2016
Published online 21 April 2016
demonstrated colocalization of C1q with
PSD95-positive puncta in 1 mo J20 hip-
through posttranslational detyrosination of tu- cycle. This is reflected in the large viscous com- grant 2P30AG024928. Data and code are available at pocampus (Fig. 1C). A higher percentage
5. 462,
mouse models
15. E. White, Pflugers Arch. 11. C. J. Murray et
177–184 (2011). al., PLOS Med. 3, e260 (2006). 10.1126/science.aaf1437
J. Pijoan-Mas, J. V. Ríos-Rull, Demography 51, 2075–2102 http://dx.doi.org/10.7910/DVN/C2VYNM.
bulin. Detyrosinated MTs buckle under load in ponent of the MT contributions to myocyte me- 16. G. Cooper 4th, Heart12. Fail.H.Rev.
Wang,
(2014). 5, 187–201 (2000).
A. E. Schumacher, C. E. Levitz, A. H. Mokdad, of PSD95 colocalized with C1q in the hip-
contracting cardiomyocytes, which confers mechan- chanical properties observed at deformation rates 17. L. Cassimeris, P. Tran, Methods
C. J. Murray,Cell Biol. 97, xvii–xviii
Popul. (2010).
Health Metr. 11, 8 (2013). pocampus of J20 mice than in that of WT
6. H. Waldron, Soc. Secur. Bull. 67, 1–28 (2007).
18. G. Lukinavičius et al.,13.Nat.J. Methods
S. Olshansky11, 731–733 (2014).Aff. 31, 1803–1813 (2011).
et al., Health
ical resistance to contraction and regulating the consistent with contractile velocities both in this 7. H. Waldron, Soc. Secur. Bull. 73, 1–37 (2013). littermates (Fig. 1D and fig. S3), suggesting
19. S. Belmadani, C. Poüs, 14. R.E.Fischmeister,
R. Meara, S.P.-F. Méry, Mol.
Richards, D. M.Cell.
Cutler, Health Aff. 27, 350–360 Soyon Hong,1 Victoria F. Beja-Glasser,1* Bianca M. Nfonoyim,1* Arnaud Frouin,1
viscoelastic properties of the myocyte. and previous work (34). However, we do note a 8. J. Wilmoth, C. Boe, M. Barbieri, in International Differences
Biochem. 258, 35–48 (2004). that the C1q-associated synapses may be
(2008).
in Mortality at Older Ages: Dimensions and Sources, Li,2 Saranya Ramakrishnan,1 Katherine M. Merry,1 Qiaoqiao Shi,2
NEURODEVELOPMENT
Shaomin
The observation that MTs normally buckle, slight prolongation of the late phase of relaxation 20. C. Janke et al., Science 15. 308,
D. M.1758–1762
Cutler, F.(2005).
Lange, E. Meara, S. Richards-Shubik, marked for elimination.
E. M. Crimmins, S. H. Preston, B. Cohen, Eds.
21. C. Erck et al., Proc. Natl. Acad. Sci. U.S.A. 102, 7853–7858 Arnon Rosenthal,3,4,5 Ben A. Barres,6 Cynthia A. Lemere,2
rather than break or slide, strongly indicates that in TTL-overexpressing myocytes, which may rep- C. J. Ruhm, J. Health Econ. 30, 1174–1187 (2011).
(National Academies Press, Washington, DC, 2011), Punctate Aβ was found deposited in J20
Dennis J. Selkoe,2,7 Beth Stevens1,8†
they bear load and store elastic bending energy
during the cardiac contractile cycle. This has im-
resent the loss of a MT contribution to restoring
force. We consider it probable that the restoring
(2005). 16. J. K. Montez, L. F. Berkman, Am. J. Public Health 104, e82–e90
pp. 337–372.
185, 1159–1166 (2009).
22. L. Peris et al., J. Cell Biol.(2014).
9. G. K. Singh, M. Siahpush, Int. J. Epidemiol. 35, 969–979
23. M. Sirajuddin, L. M. Rice, R. D. Vale,
17. Human Biol. 16, 335–344
Nat. CellDatabase;
Mortality www.mortality.org.
Complement and microglia mediate
Synapse loss in Alzheimer’s disease (AD) correlates with cognitive decline. Involvement
hippocampus at 1 mo (fig. S4), long before
Aβ plaques deposit (11, 12), raising the ques-
(2006). tion of whether C1q increase in these pre-
plications for MT-dependent mechano-signaling
in muscle and other tissues (1, 32, 33) but also
force of other internal elastic elements, such as
titin, are likely to play a more dominant role, at
(2014). 18. D. D. Reidpath, P. Allotey, J. Epidemiol. Community Health 57,
10. M.
24. G. Kreitzer, G. Liao, G. G.344–346
1118 (1999).
Ezzati,
Gundersen, A. B.
(2003).
Med. 5, e66 (2008).
Friedman, S.
Mol. Biol. Cell 10, 1105– C. Kulkarni, C. J. Murray, PLOS
early synapse loss in Alzheimer
of microglia and complement in AD has been attributed to neuroinflammation,
prominent late in disease. Here we show in mouse models that complement and microglia
plaque brains is dependent on soluble Aβ
has direct implications for contractility. Our mod- least in the initial return toward resting sarco- 19. A. Case, A. Deaton, Proc. Natl. Acad. Sci. U.S.A. 112, levels. To test this hypothesis, we injected
mouse models
11. C. Res.
25. R. A. Whipple et al., Cancer J. Murray et al., PLOS
68, 5678–5688 Med. 3, e260 (2006).
(2015). (2008). mediate synaptic loss early in AD. C1q, the initiating protein of the classical
el of myocyte contractility demonstrates how mere length (27). Thus, an increase in detyrosina- 26. X. Fonrose et al., Cancer
15078–15083
12. H. Wang, A. E. Schumacher,
67, 3371–3378 C. E. Levitz, A. H. Mokdad, the mice with compound E, a γ-secretase
20. J.Res. Bound, A. Geronimus, (2007). J. Rodriguez, T. Waidman, “The complement cascade, is increased and associated with synapses before overt plaque
changing MT load–bearing and force transfer tion may increase myocyte stiffness and impair C.H.J.Granzier,
27. M. Helmes, K. Trombitás,Implications Murray,Circ.Popul.Res.Health Metr. 11, 8 (2013).
79, 619–626
of Differential Trends in Mortality for Social inhibitor that rapidly decreases Aβ produc-
(1996). 13. J. S. Olshansky et al., Health Aff. 31, 1803–1813 (2011). deposition. Inhibition of C1q, C3, or the microglial complement receptor CR3 reduces the
with the sarcomeres can substantially alter con- contraction by acting as an energy sink, without Security Policy,” University of Michigan Retirement Research 1 1 tion (12). Compound E markedly reduced
28. R. E. Mahaffy, C. K. Shih,
14. E. F. C.
R.MacKintosh, J. Käs,2014-314
Meara, S. Richards, Phys.
D. M.Rev.Cutler, Health Aff. 27, 350–360 number of phagocytic
Soyon Hong, Victoria microglia, as well
F. Beja-Glasser, * as the extent
Bianca of early1synapse
M. Nfonoyim, * Arnaudloss. C1q1 is
Frouin,
tractile properties. Our experimental data show providing significant energetic return during Center Working Paper (2014). soluble Aβ levels in J20 mice; there was a
Lett. 85, 880–883 (2000). (2008). necessary
Shaomin Li, 2
for Saranya
the toxicRamakrishnan,
effects of soluble 1
Katherine
b-amyloid M.(Ab)
Merry, 1
oligomers
Qiaoqiaoon synapses
Shi, 2
and
that such changes in MT load–bearing can be relaxation. 21. J. B. Dowd, A. Hamoudi, Int. J. Epidemiol. 43, 983–988
15. Biol. 181, 667–681 (2008).
29. P. Konieczny et al., J. Cell(2014).
D. M. Cutler, F. Lange, E. Meara, S. Richards-Shubik, 3,4,5
corresponding reduction of C1q deposition
achieved by posttranslational modifications of Consistent with this, an increase in detyrosi-
hippocampal
Arnon Rosenthal, long-term Ben A. Barres,6Finally,
potentiation. Cynthiamicroglia
A. Lemere, 2
in adult brains engulf synaptic
(Fig. 1E), suggesting that Aβ upregulates
30. G. Liao, G. G. Gundersen,C. J. Biol. Chem.J.273,
J. Ruhm, 9797–9803
Health Econ. 30,(1998).
1174–1187 (2011).
22. T. Goldring, F. Lange, S. Richards-Shubik, “Testing for Changes material
Dennis J.inSelkoe, 2,7
a CR3-dependentBeth Stevens 1,8
process † when exposed to soluble Ab oligomers. Together,
the MTs themselves, particularly detyrosination. nation was associated with clinical contractile 31. G. Gurland, G. G. Gundersen, Cell Biol.L.131,
16. J. K.J.Montez, 1275–1290
F. Berkman, Am.(1995).
J. Public Health 104, e82–e90 C1q. To further address whether the in-
in the SES-Mortality Gradient When the Distribution of
Signal. 5, ra56 (2012).
32. R. J. Khairallah et al., Sci.(2014). these findings suggest that the complement-dependent pathway and microglia that
The measured reductions in buckling and visco- dysfunction in human hearts. Our cellular studies Education Changes Too,” National Bureau of Economic crease of C1q is dependent on soluble
33. M. Prager-Khoutorsky, 17.A.Human
Khoutorsky,MortalityC. W.Database;
Bourque, Neuron www.mortality.org. prune
Synapseexcess
loss synapses
in Alzheimer’sin development
disease (AD) arecorrelates
inappropriately activateddecline.
with cognitive and mediate
Involvement
elasticity and the increase in contractile speed of demonstrate that acute reduction of detyrosina- Research Working Paper 20993 (2015).
83, 866–878 (2014).18. D. D. Reidpath, P. Allotey, J. Epidemiol. Community Health 57, Aβ, and if so, which species, we injected
23. A. S. Hendi, Int. J. Epidemiol. 44, 946–955 (2015). synapse lossand
of microglia in AD.complement in AD has been attributed to neuroinflammation,
PTL-treated, TTL-overexpressing myocytes can tion with genetic or pharmacologic approaches 34. H. Tagawa et al., Circ. Res.344–34680, 281–289 (2003). (1997). soluble Aβ oligomers or monomers into
24. A. Aizer, J. Currie, Science 344, 856–861 (2014). prominent late in disease. Here we show in mouse models that complement and microglia
all be attributed to changing the way MTs in- can boost contractility and reduce mechanical 19. A. Case, A. Deaton, Proc. Natl. Acad. Sci. U.S.A. 112,
25. D. Brown, A. Kowalski, I. Lurie, “Medicaid as an Investment in
lateral ventricles of WT mice. Hippocam-
teract with the sarcomere and impairing their stiffness. Additionally, these approaches are 15078–15083 (2015). mediate synapticassociation
enome-wide
enome-wide loss early in
association AD. C1q,
studies
studies the initiating
impli- brain (1).protein
synapse During of thewe
loss occurs, classical
development, used
C1qsuperresolu-
and C3 local- pus contralateral to the injection site was
ACKNOWLEDGMENTS
G
Children: What Is the Long-Term Impact on Tax Receipts?”
ability to act as compression resistors. It is also able to induce large changes in detyrosination, 20. J. Bound, A. Geronimus, J. Rodriguez, T. Waidman, “The
The authors thank C. W. Ward National
and E. L. Bureau
Holzbauroffor kindly providing
Economic Research Working Paper 20835 complement
cate cascade,
catemicroglia
microglia andiscomplement-related
and increased and associated
complement-related ize to with
tion synapses
structured
synapses before
illumination
and mediate overt
synapse plaque
microscopy
elimination examined to avoid any surgery-related
Implications of Differential Trends in Mortality for Social deposition. Inhibition of C1q,disease
C3, or(AD)
the microglial complement receptor CR3 reduces
possible that detyrosinated MTs anchored to one while only slightly altering the overall MT cyto- TTL and EMTB-3xGFP plasmids, (2015).respectively, as well as for
Security Policy,” University of Michigan Retirement Research pathways
pathways in Alzheimer’s
Alzheimer’s disease (AD)
(1). (SIM) (10) tomicroglia
by phagocytic quantify synapse
(5–7). density the
We hypothesized in effects. Oligomeric Aβ (oAβ), which is
sarcomere form bundles with MTs anchored to skeleton, which minimizes off-target consequences. 26. S. Cahodes,
discussion; R. Bloch for providing CenterdesminWorking
S. Kleiner,
KO mice
Paper
M.littermate
F. Lovenhem, M. Grossman,
and2014-314 (2014).
number
(1). of phagocytic
Previous
Previous research
research microglia,
has has as well
demonstrated
demonstrated bothas the extent
hippocampal
that of early
this normal CA1 synapse
stratumloss.
developmental C1q is
radiatum
synaptic of fa-
pruning prefibrillar in nature and acts as a me-
controls; Carl Zeiss Microscopy of Child Health Insurance
“EffectAiryscan Access on Schooling,”
adjacent sarcomeres. If so, disrupting bundling Thus, interfering with detyrosination may repre- 21. J. for B. Dowd, A.instrumentation
Hamoudi, Int. J. use;Epidemiol. 43, 983–988 necessary
both for the
beneficial andtoxic
beneficial effectsroles
and detrimental
detrimental of soluble
roles
of com- ofb-amyloid
milial
pathway (Ab) oligomers
AD-mutant
is activated oninsynapses
human
early amyloid
the AD brain and
precur-
and diator of synapse loss and dysfunction in
IonOptix for technical supportNational
and
(2014). cell Bureau
holder of Economic
materials; and Research
Y. E. Working Paper
hippocampal long-term potentiation. Finally, microglia in (hAPP)
adultloss.
brains engulf synaptic
would also effectively uncouple MTs from force- sent an attractive and novel therapeutic strategy Goldman, E. L. Grishchuk,22. and20178 (2014). for discussion and
R.Goldring,
J. Composto complement
plement and
and microglia microglia in plaque-related
in plaque-related neuro- sor protein
mediates synapse (“J20”) transgenic mice AD (4), but not the relatively innocuous
T. F. Lange, S. Richards-Shubik, “Testing for Changes material in
generating structures. Regardless of the mecha- for increasing contractility. 27. S. Miller, L. R. Wherry, “The Long-Term Health
guidance. B.L.P., P.R., M.A.C., A.I.B.,
in the and K.B.M. wereGradient
SES-Mortality
Effects
responsible forthe Distribution of
of Early pathology (2,a3);CR3-dependent
neuropathology (2, 3); however,
however, process
their
their roles in roles when
synapse in exposed
(11). to soluble
TheQuantification
degree Abofoligomers.
colocalized
of region-specific Together,
synapsepre-loss
and is monomeric Aβ or vehicle, induced C1q
Coverage,” SocialWhen these findings a suggest that the of complement-dependent pathway ofand microglia thatinand
Life Medicaid Science Research Network
nism, disrupting coupling to sarcomeres would In conclusion, our data show that MTs exhibit experimental design; B.L.P., P.R., M.A.C., Changes
Education C.Y.C., andToo,”
Working Paper 2466691 (2015).
A.I.B.National
carried Bureau of Economic synapse
loss, loss,pathological
a major major pathological
correlate correlate
cognitive postsynaptic puncta
a stronger correlate [synaptophysin
cognitive decline AD deposition (Fig. 2A and fig. S5). A higher
reduce the incident force on the MT, and buckling divergent mechanical behavior because of the out experiments and analyzedResearch data; H.A.Working
and V.B.S. designed
Paper 20993Teens:
(2015).Using and Eligibility prune
of
decline excess
cognitive
in (4),synapses
AD decline remain toin
in AD development
be(4), to are
remainEmerg-
identified. be inappropriately
postsynaptic activated
than counts ofdensity and
plaques,95tangles, mediate
(PSD95) and(Fig. 1A);
neuronal percentage of PSD95 colocalized with C1q
28. L. R. Wherry, B. Meyer, “Saving
occurrence would drop. differences in how they couple to the rest of the and executed the mathematical A. model.
23. Discontinuity
S. Hendi, B.L.P.to and
Int. P.R. cowrote
J. Epidemiol.
Estimate 44, 946–955
the Effects (2015).
of Medicaid Eligibility,”
synapse
identified.
ing loss
research in AD. research
Emerging
implicates microgliaimplicates
and immune- mi- synaptotagmin and homer
loss (8, 9). To determine how early(fig.synapse
S1, Alossto in oAβ-injected versus monomer-injected
the manuscript. All authors 24.participated
National J.inCurrie,
A. Aizer, Bureau the critical
of Science review
Economic 344, and856–861
Research (2014).Paper 18309
Working
The striking periodicity of buckles in untreated cardiac cytoskeleton. The tyrosinated portions of revision of the manuscript. This
25. (2013).work was supported by funding
D. Brown, A. Kowalski, I. Lurie, “Medicaid as an Investment in
croglia and immune-related
related mechanisms in brain wiring mechanisms
in the healthy in D)]
occurs,revealed
we useda superresolution
significant lossstructured of synapses illu- mice (Fig. 2B), in a manner similar to this
myocytes lends further support to the idea of a the network, moving readily with the myocyte brain wiring
enome-wide in the association
healthy brain (1). During
studies impli- in J20(1).hippocampus
mination
brain microscopy
During at 3(10)
(SIM)
development, toC1q 4 months
to quantify old
syn-
and C3 local- colocalization in J20 mice. Together, these
G
from National Institute for29. Arthritis
L.Children:and What
R. Wherry, Musculoskeletal
S.IsMiller,
the Long-Termand SkinImpact
R. Kaestner, B. D. on Tax “Childhood
Meyer, Receipts?”
sarcomeric anchor. The preferential association of during contraction, provide little contractile re- Diseases (NIH) (T32AR053461-09 National
Medicaid to P.R.
Bureauand T32HL007954
Coverage of and
Economic to HealthWorking
Research
Later Life Paper 20835
Care Utilization,” development, C1q andand
cate microglia C3 localize to synapses
complement-related (mo),
apse an age
ize todensity
synapses that
in hippocampalprecedes
and mediate CA1 plaque
stratum
synapse deposi-
radiatum
elimination findings show an early and aberrant in-
desmin with detyrosinated tubulin and the insen- sistance. Conversely, the detyrosinated portions M.A.C.) and from National Heart, NationalLung,Bureau
(2015). and Blood Institute (NIH)
of Economic Research Working Paper 20929 and
1
mediate
F.M. Kirby synapse
Neurobiology
pathways elimination
Center, by (AD)
Boston Children’s
in Alzheimer’s disease phago-
Hospital(1). tion
of (11, 12).
byfamilial
phagocytic Synapse
AD-mutant
microglia loss
human inamyloid
(5–7). preplaque
We J20
precursor
hypothesized crease and synaptic localization of C1q in
sitivity of desmin KO animals to changes in de- of the MT network, forming complexes with des- (R00-HL114879 to B.L.P.).26. The(2015).
S.VIVA experiments
Cahodes, usedM.
S. Kleiner, shared
F. Lovenhem, M. Grossman, (BCH) and
cytic Harvard Medical
microglia
Previous (5–7).School
research has(HMS),
We Boston, MA that
hypothesized
demonstrated both CA1
proteinwas
that this confirmed
(hAPP)
normal (“J20”) by
developmentalelectron
transgenic micemicroscopy
synaptic (11).pruning
Quan- multiple AD model systems. Furthermore,
of Child Health Insurance 2
experimental facilities from 30.the
J.“Effect
Nano
Ludwig, Science
D. L. andMiller, J. Econ. Access
Engineering
Q. on Schooling,
122, 159–208 (2007).” 02115, USA. Ann Romney Center for Neurologic Diseases,
tyrosination strongly implicates desmin as at least min IFs, produce a cross-linked MT-IF network National Bureau of Economic Research Working Paper this normal
Departmentbeneficialdevelopmental
andBrigham
of Neurology, detrimental
andsynaptic pruning
rolesHospital
Women’s of com- (fig.
pathwayS1G).
tification Confocal
isofactivated
colocalized imaging
early pre- andalso
in the AD showed
postsynaptic
brain and fluorescent in situ hybridization (FISH)
Center on Molecular Function 31. H. at Hoynes,
the Nano-Bio Interface at the
D. Whitmore-Schanzanbach, D. Almond,
one component of a sarcomeric anchoring com- that confers robust resistance to contraction. This 20178 Run
“Long
University of Pennsylvania supported (2014).
Impacts
by NSF underofgrant
Childhood
DMR08- Access to the Safety Net,” pathway
plement
(BWH) is activated
and and
HMS, microglia
Boston, MA early inUSA.
the 3AD
in plaque-related
02115, brain
Alector and
neuro-
Inc., 953 synapse
puncta
mediates loss
synapsein CA1,
[synaptophysin loss.CA3,andand dentate gyrus
postsynaptic den- demonstrated up-regulated C1qa expres-
plex of detyrosinated MTs. Note that myocytes orthogonal MT-IF grid requires tightly periodic 32802. We thank R. Composto, 27. National
S.who
Miller, L. R. Wherry,
Bureau
acknowledges of Economic NSFLong-Term
the“The Health Effects
Research Working Paper of Early
18535 mediates
Indiana synapse
Street,
pathology San3);
(2, loss. CA their
Francisco,
however, USA. 4in
94107, roles Annexon
synapse of 3 95
sityThemo J20 hippocampus
(PSD95)
degree of (Fig.
region-specific butsynapse
1A); synaptotagminnot in loss stria-
and is sion in microglia (fig. S6), implicating mi-
Life Medicaid Coverage,” Social Science Research Network Biosciences, 280 pathological
Utah Avenue Suite 110, South
lacking desmin have decreased viscoelasticity, MT deformations to accommodate myocyte mor- Polymers Program under grant (2012).
DMR09-07493. Procurement of The
loss, degree
a major of region-specific
correlate ofSan
synapse loss
cognitive tum
homer (fig.
a stronger (fig.S1E).
S1, A Synapse
correlate levelsa significant
to D)]ofrevealed
cognitive were
decline not al-
inloss
AD croglia as a major source of C1q in these
32. A. Working
Isen, M. Paper 2466691 (2015).
Rossin-Slater, R. Walker, “Every Breath You Take Francisco, CA 94080, USA. 5Department of Anatomy,
human heart tissue was enabled by grants from NIH (HL089847 is
University of California San Francisco (UCSF), San decline
a stronger correlate of cognitive tered in 1 mo J20 brains versus
despite a denser MT network, which supports phology changes during contraction. These defor- 28. Every
L. R. Wherry,
Dollar You’llB. Meyer,
Make:“Saving Teens: Using
The Long-Term and Eligibility
Consequences of decline in AD (4), remain to be identified. Emerg-
Francisco,
of synapses
than counts inofJ20 hippocampus
plaques, tangles, 3 towild
at and type
4neuronal
months preplaque brains.
and HL105993) to K.B.M. H.A. and V.B.S. are supported by the in AD than
the idea that MT network organization and cross- mations require a considerable amount of energy Discontinuity
the Clean Air Act to Estimate
of 1970,”the EffectsBureau
National of Medicaid Eligibility,”
of Economic ing94143,
CA USA. 6counts
research implicates
Department ofofplaques,
microglia tangles,
Neurobiology,and immune-
Stanford and (WT)
old (8, littermates
loss(mo), 9).anToage determine (fig.how
that precedes S1F),
plaque
early suggesting
deposition
synapse loss To test whether C1q and oAβ act in a
National Institute of Biomedical Imaging and Bioengineering
linking is a stronger determinant of myocyte me- to form and dissipate a large fraction of that en-
(NIH) under award number
National Bureau
Research WorkingofPaper Economic 19858Research
(2014). Working Paper 18309 neuronal
University
related loss
School
mechanisms of(8, 9). To
Medicine,
in determine
Palo
brain Alto,
wiring how
CA 94305,
in the early
USA.
healthy that
(11,
occurs, the
12). we hippocampal
Synapse
used superresolution synaptic
loss in preplaque J20 loss
structured CA1atwas 3
illu- common pathway to eliminate synapses,
33.R01EB017753
A.(2013).
Fenelon, S. andH.NSF grant CMMI-
Preston, Demography 49, 797–818 7 8
Prothena Biosciences, Dublin, Ireland. Stanley Center for mo is likely not a result of
chanical properties than network density per se. ergy due to viscous interactions. This has im- 1312392. The data are included in
29. (2012). the main manuscript and in the
L. R. Wherry, S. Miller, R. Kaestner, B. D. Meyer, “Childhood
confirmed
mination by electron
microscopy microscopy
(SIM) (10)abnormal
to(fig. S1G). syn-
quantify Con-
syn- we injected oAβ into lateral ventricles of
Psychiatric Research, Broad Institute of MIT and Harvard,
Both the desmin and MT networks have ele- portant implications for MT load–bearing across supplementary materials34. online.
D.Medicaid
de Walque, CoverageJ. Hum.andResour.
Later Life45, Health
682–717Care Utilization,”
(2010). Cambridge, MA 02142, USA. aptic
focal development.
imaging
apse density in also showed synapse
hippocampal CA1 stratum lossradiatum
in CA1, C1qa knockout (KO) mice (16). Soluble
ments perpendicular to their typical orientation, cell biology, as well as for the altered mechanical 35. C. National
E. Finch,Bureau E. M.ofCrimmins,
Economic ScienceResearch305, Working Paper 20929
1736–1739
1
F.M. Kirby
*These Neurobiology
authors contributed Center,
equally Boston Children’s
to this work. Hospital
†Corresponding of We
CA3, asked
and
familial whether
dentate
AD-mutant gyrus the
human of classical
3amyloid
mo J20 comple-
hippo-
precursor oAβ induced a significant loss of colocal-
particularly near the sarcolemma, which may stiffness and mechano-signaling in cardiac disease. SUPPLEMENTARY MATERIALS
(2015).
(2004). (BCH) and
author. Harvard
Email:
2
Medical School (HMS), Boston, MA
beth.stevens@childrens.harvard.edu ment
proteincascade
campus but not
(hAPP) isinup-regulated
(“J20”)striatum
transgenic in
(fig.mice preplaque
S1E). Synapse
(11). Quan- ized synapsin- and PSD95-immunoreactive
30. J. Ludwig, D. L. Miller, Q. J. Econ. 122, 159–208 (2007).
www.sciencemag.org/content/352/6284/aaf0659/suppl/DC1 02115, USA. Ann Romney Center for Neurologic Diseases,
alter how those elements interact with the cyto-
Department of Neurology, Brigham and Women’s Hospital
brains
tification when synapses pre-
of colocalized are and
already vulner-
postsynaptic puncta in WT mice within 72 hours (Fig.
Materials and Methods 31. H. Hoynes, D. Whitmore-Schanzanbach, D. Almond,
skeleton and plasma membrane. However, we REFERENCES AND NOTES (BWH) and HMS, Boston, MA 02115, USA. 3Alector Inc., 953 able.
puncta C1q immunoreactivity
[synaptophysin (13) (antibody
and postsynaptic den- 3A, left panel) (17). In contrast, oAβ failed
REFERENCES
1. B. L. Prosser, C.AND NOTES
W. Ward, W. J. Lederer, Science 333, 1440–1445
Supplementary Text 7 12“Long6Run MAY Impacts
2016 of Childhood
• VOL Access to
352 ISSUE the Safety Net,”
6286 sciencemag.org SCIENCE
believe that the preponderance of the contractile
Figs. S1 to S11
National Bureau of Economic Research Working Paper 18535 Indiana Street, San Francisco, CA 94107, USA. 4Annexon now
sity 95 available
(PSD95) at Abcam)
(Fig. was elevatedand
1A); synaptotagmin in to induce synapse loss in C1qa KO mice (Fig.
1. B.(2011).
L. Prosser, C. W. Ward, W. J. Lederer, X-ROS signaling: Rapid
resistance that results from detyrosination is due Tables S1 to S9
(2012). Biosciences, 280 Utah Avenue Suite 110, South San J20
homer brains
(fig. S1, asAearly
to D)]as 1 mo and
revealed preceding
a significant loss 3A, right panel), suggesting that C1q is re-
2. mechano-chemo
B. L. Prosser, C. W. Ward, W. J. Lederer,
transduction in heart.Cardiovasc. Res. 98,
Science 333, 32. A. Isen, M. Rossin-Slater, R. Walker, “Every Breath You Take Francisco, CA 94080, USA. 5Department of Anatomy,
to longitudinal MTs in an orthogonal grid with 307–314 (2013).
1440–1445 (2011). doi: 10.1126/science.1202768; References (35–42)
Every Dollar You’ll Make: The Long-Term Consequences of University of California San Francisco (UCSF), San Francisco,
synapse
of synapses lossin (Fig. 1B and fig. S1).
J20 hippocampus at 3 to C1q eleva-
4 months quired for oAβ-induced synapse loss in vivo.
transverse desmin IFs because of the simple nu- 3. pmid:
M. R. Zile et al., Circulation 98, 567–579 (1998).
21903813 Movies S1 to S8 the Clean Air Act of 1970,” National Bureau of Economic CA 94143, USA. 6Department of Neurobiology, Stanford tion was an
old (mo), region-specific,
age that precedes particularly
plaque depositionin the To determine whether local, acute inhibi-
merical majority of cytoskeletal elements in this 2.4. B.S.L.
Nishimura
Prosser, etC. al.,
W.Circ.
Ward, W.98,
Res. J. 81–87 (2006).
Lederer, X-ROS signaling is Research Working Paper 19858 (2014). University School of Medicine, Palo Alto, CA 94305, USA. hippocampus
(11, 12). Synapseand loss frontal
in preplaque cortex, J20 twoCA1 wasre- tion of C1 activation could similarly blunt
configuration. 5. enhanced
H. L. Granzier, T. C. Irving,
and graded Biophys.
by cyclic J. 68, 1027–1044
cardiomyocyte (1995).
stretch. 12 December 2015; accepted 33. 29 January 2016 7
Prothena Biosciences, Dublin, Ireland. 8Stanley Center for
A. Fenelon, S. H. Preston, Demography 49, 797–818 gions
confirmed vulnerable
by electron to microscopy
synapse loss (fig.(14)
S1G).(Fig.
Con- the synaptotoxic effects of oAβ, we used an
6. Cardiovasc.
C. P. Brangwynne
Res. 98,et al., J. Cell Biol.
307–314 173,doi:
(2013). 733–741 (2006).
10.1093/cvr/cvt066 10.1126/science.aaf0659 (2012). Psychiatric Research, Broad Institute of MIT and Harvard,
Despite the fact that detyrosinated MTs store 3.7. M.
M.R.Mehrbod, M. R.
Zile et al., K. Mofrad, PLOS
Constitutive ONE 6,
properties ofe25627 (2011).
adult mammalian 1B and
focal fig. S2A).
imaging also C1q
showedimmunoreactivity
synapse loss in was CA1, antibody against C1q (anti-C1q) (ANX-M1,
34. D. de Walque, J. Hum. Resour. 45, 682–717 (2010). Cambridge, MA 02142, USA.
energy during sarcomere contraction, which pro- 8. cardiac
H. Tsutsui, K. Ishihara,
muscle G. Cooper 4th,
cells. Circulation Science 260,
98, 567–579 682–687
(1998). 35. C. E. Finch, E. M. Crimmins, Science 305, 1736–1739 *These authors contributed equally to this work. †Corresponding comparable
CA3, and dentate between gyrus J20of and3 mo WT J20mice hippo-at Annexon Biosciences), which blocks the
vides compression resistance, little of this energy (1993).
doi: 10.1161/01.CIR.98.6.567; pmid: 9714115 (2004). author. Email: beth.stevens@childrens.harvard.edu postnatal
campus butday not21 in (P21)
striatum (fig.(fig.
S2B),
S1E).suggest-
Synapse classical complement cascade (see fig. S7
16 7 12 sciencemag.org SCIENCE
6 MAY 2016 • VOL 352 ISSUE 6286 SCIENCE sciencemag.org Originally published sciencemag.org
6 May 2016 in SCIENCE
SCIENCE 17
O PT IM IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S RE S EAR CH | R E P O R T S R ES E A RC H | R E PO R TS S E C T I O N T WO | RESEARCH ARTICLES
Fig. 2. Oligomeric Ab increases C1q and microglial phagocytic activity. (A and B) Soluble Ab oligomers in WT mice led to elevation of C1q (green) (A)
and a higher percentage of PSD95 (red) colocalization with C1q versus monomers (B). (C and D) oAb induced high levels of CD68 (green) immunoreactivity
in Iba1-positive (red) microglia in WTmice (C), but not in those of C1qa KO mice (D). Both had negligible changes in morphology. See fig. S10. Scale bar, 10 mm (A),
5 mm (B), or 20 mm (C). Means ± SEM; n = 3 to 5 mice per treatment group per genotype. *P < 0.05 using two-tailed t test (B) or *P < 0.05, **P < 0.01 versus
control-treated or ##P < 0.01 versus Ab monomer–treated using two-way ANOVA followed by Bonferroni posttest (C).
18
SCIENCE sciencemag.org 6 MAY 2016 • sciencemag.org 7 13
SCIENCE
VOL 352 ISSUE 6286 7 14 6 MAY
SCIENCE 2016 • VOL 352 ISSUE 6286
sciencemag.org sciencemag.org SCIENCE
19
RE S EAR
O PT IM IZING YOUR CH
LI V E -CE|L LR E
MPIOCRO
R T S S CO P Y: TR I C K S A N D T R A D E - O F F S RE S EAR CH | R E P O R T S S E C T I O N T WO | RESEARCH ARTICLES
20 6 MAY 2016 • VOL 352 ISSUE 6286 7 15 SCIENCE sciencemag.org sciencemag.org SCIENCE 6 MAY 2016 • VOL 352 ISSUE 6286
SCIENCE sciencemag.org 7 15 21
O PT IM IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S S E C T I O N T WO | RESEARCH ARTICLES
the regulation of complex patterning signals per well of a 96-well plate), iPS cells formed Bioengineered 3D IOS in iPS cell– sors in DP cells (34, 35). The follicle variable
BIOMEDICAL ENGINEERING derived explants generated using the
during embryogenesis and self-formation of embryoid bodies (EBs) in 3 days, and the region mediates the hair cycle, which depends
CDB transplantation method
pluripotent stem cells in three-dimensional EBs were allowed to grow for 1 week (Fig. on the activation of Lgr5-positive follicle epi-
Bioengineering a 3D (3D) stem cell culture (2, 3, 18–22). However,
ectodermal organs, such as the skin, hair folli-
1B). The relative mRNA expression levels of
undifferentiated iPS markers (Nanog) were
In the iPS cell–derived explants generated
via the CDB transplantation method, we
thelial stem cells during the telogen-to-anagen
transition. In addition, the follicle epithelial
transplantation model
a complex system that plays important roles of EBs after 7 days in culture, we observed licles compared with the not-stimulated con- neered hair follicles of the CDB transplants,
in waterproofing, cushioning, protecting that the iPS cell–derived EBs had differenti- dition (Fig. 2A). The morphology of EBs and which were produced using the EGFP-labeled
deeper tissues, excreting waste, and thermo- ated into both epithelial and mesenchymal the expression of Sox2, p63, Sox17, Pax6, Snail, iPS cell clone. All cells composed of bioengi-
regulation. The integumentary organs include cells (Fig. 1D). EBs expressed Sox2 and p63 and Twist were not different with or without neered 3D IOS showed green fluorescence
Ryoji Takagi,1∗ Junko Ishimaru,1∗ Ayaka Sugawara,1* Koh-ei Toyoshima,2,3,4
the skin and its appendages (hair, sebaceous as integumentary markers and expressed Wnt10b signaling (Fig. 1D and fig. S3). The (Fig. 2G, left panels). Sox9 and CD34 dou-
Kentaro Ishida,5 Miho Ogawa,2,3,4 Kei Sakakibara,1 Kyosuke Asakawa,2 glands, sweat glands, feathers, and nails) Sox17 as an endodermal epithelial marker. hair shaft in the explant is black (Fig. 2A), al- ble-positive epithelial stem cells were also
Akitoshi Kashiwakura,1 Masamitsu Oshima,5 Ryohei Minamide,2 Akio Sato,4 (23). The integumentary organs arise from EBs also expressed proteins for the neural though the hair shaft of the SCID mouse is observed in the bulge region of the normal iPS-
Toshihiro Yoshitake,4 Akira Takeda,4 Hiroshi Egusa,6 Takashi Tsuji2,3,5† organ germs through reciprocal epithelial- progenitor marker Pax6 and for neural crest white, indicating that hair follicles contain derived bioengineered hair follicles (Fig. 2G).
mesenchymal interactions in the skin field (4). markers such as Snail and Twist (Fig. 1D). melanocytes that originate from iPS cell–de- Lrig1-positive epithelial stem cells and Lgr5-
The integumentary organ system is a complex system that plays important roles During embryogenesis, the skin field forms These results indicated that both epithelial rived neural crest cells. Although the frequen- positive epithelial cells were also observed in
in waterproofing, cushioning, protecting deeper tissues, excreting waste, and through a regulated process of pattern for- and mesenchymal cells were generated in cies in the explants with or without Wnt10b the proper anatomical regions, which were de-
thermoregulation. We developed a novel in vivo transplantation model designated mation, and its appendage organs are then iPS cell–derived EBs after 7 days in culture. signaling were not different (Fig. 2B), the tected in the upper and lower portions of the
as a clustering-dependent embryoid body transplantation method and generated induced through epithelial-mesenchymal in- We next transplanted these EBs under number of bioengineered hair follicles in the CK15-positive bulge region, respectively (Fig.
a bioengineered three-dimensional (3D) integumentary organ system, including teractions according to a typical Turing model various conditions into the subrenal capsule CDB explants treated with Wnt10b was drasti- 2G). The calponin-positive arrector pili muscle
appendage organs such as hair follicles and sebaceous glands, from induced pluripotent of activator and inhibitor signals (23). Regen- of severe combined immunodeficient (SCID) cally higher than that of the explants without was connected to the bulge region epithelial
stem cells. This bioengineered 3D integumentary organ system was fully functional eration of the 3D IOS could contribute to re- mice in vivo. Both single iPS cells and single Wnt10b (Fig. 2C). In the explants treated with cells of both the natural hair follicle and the
following transplantation into nude mice and could be properly connected to surrounding generative therapies for patients with burns, EB transplants formed teratoma-like tissues, Wnt10b, the bioengineered hair shaft, which bioengineered hair follicle (Fig. 2H). These
host tissues, such as the epidermis, arrector pili muscles, and nerve fibers, without scars, and alopecia, and could be used as a which contained three germ layers, including was measured from the tip of the hair shaft to results indicate that various follicle stem cells
tumorigenesis. The bioengineered hair follicles in the 3D integumentary organ system novel assay system for nonanimal safety test- neural tissue, muscle, cartilage, and bronchial the DP, was longer than that in the explants and the arrector pili muscles were correctly
also showed proper hair eruption and hair cycles, including the rearrangement of ing of cosmetics and quasi-drugs (5, 24). How- epithelia, as reported previously (Fig. 1E, top without Wnt10b (Fig. 2D). Moreover, the hair arranged in the bioengineered hair follicles
follicular stem cells and their niches. Potential applications of the 3D integumentary ever, it is difficult to generate the complex 3D and middle) (26). By contrast, in multiple EBs follicles in explants treated with Wnt10b were constructed using the bioengineered 3D IOS,
organ system include an in vitro assay system, an animal model alternative, and a IOS using in vitro stem cell culture or in vivo transplanted using collagen gel, which con- abundant, and mature hair follicle structures which was derived from iPS cells.
bioengineered organ replacement therapy. transplantation models and to recapitulate tained more than 30 EBs cultured for 7 days included the dermal sheath, DP, sebaceous
experience organ dysfunction as a result of the physiological functions of the skin using a under nonadhesive conditions (Fig. 1F), cys- glands, and subcutaneous adipose tissue (Fig. Orthotopic transplantation of the iPS
INTRODUCTION disease, injury, or aging (8). Our recent stud- bioengineered 3D IOS. tic epithelia were observed in the CDB trans- 2E). The expression of Shh, Msx2, Wnt10b, cell–derived bioengineered 3D IOS
ies provided the proof of concept that fully Here, we generated a bioengineered 3D plants (Fig. 1E, bottom). The tissue weight of β-catenin, and Lef1 in the epithelium, and We next investigated whether the iPS cell–
O
rganogenesis is a complex process in- functional regeneration of ectodermal or- IOS from iPS cells, which includes appendage CDB transplants was heavy compared with of Bmp4 and Notch1 in the mesenchyme of derived IOS could be transplanted into the
volving tissue self-organization, cell-cell gans, such as teeth, hair follicles, and salivary organs such as hair follicles and sebaceous single EB transplants, and organ formation the bioengineered hair follicles in the CDB cutaneous environment of adult nude mice
interactions, regulation of signaling mol- and lachrymal glands, could be achieved with glands, using a novel in vivo transplantation was observed in 4 to 9 g of explants (Fig. explants cultured with Wnt10b, was similar (Fig. 3A, upper panel). We isolated male cys-
ecules, and cell movement (1–3). Almost the transplantation of bioengineered organ model designated as the clustering-depen- 1G). The ratio of cystic tissue area in CDB to that observed in natural skin on embry- tic tissues with hair follicles in the explants,
all organs arise from organ germs, which germs using the organ germ method (9–13). dent embryoid body (CDB) transplantation transplants was significantly larger than that onic day 18.5 (fig. S4) (11). Thus, in the iPS cut them into small specimens containing 10
are induced by reciprocal epithelial-mesen- However, organ-inductive stem cells in a wide method. We first developed the conditions for in explants of single iPS cells and single EB cell–derived explants generated via the CDB to 20 follicular units, and transplanted them
chymal interactions in each organ-forming variety of organs, except hair follicles, exist the induction of various epithelial germ layers transplants (Fig. 1H and fig. S1). transplantation method, we observed the onto the backs of female nude mice using a
field (4). Current regenerative therapy uses only during embryonic organogenesis (2). in in vivo explants of iPS cells. Wnt signaling We next characterized the types of epithe- bioengineered 3D IOS, including the skin, follicular unit transplantation method (Fig.
tissue stem cell transplantation to restore Thus, we must develop techniques to recon- controlled the frequency of the 3D IOS, calcu- lium generated in the in vivo explants using hair follicles, dermis, sebaceous glands, and 3A, lower panels). At 14 days and thereafter,
damaged tissues and organs in a wide vari- stitute a wide variety of bioengineered organ lated by the number of hair follicles in the ex- the CDB transplantation method, such as subcutaneous adipose tissue. We confirmed we observed the eruption and growth of black
ety of diseases (5–7). Next-generation regen- germs using pluripotent stem cells, such as plants. After orthotopic transplantation of the ectodermal epithelium, nerve tissue, integu- these structures using other iPS cell clones, hair shafts (n = 216; Fig. 3B). After 3 months,
erative therapy consists of organ replacement embryonic stem (ES) cells and induced plu- bioengineered IOS into nude mice, the bioen- ment, and endodermal epithelium, includ- including a gingiva-derived iPS cell clone we did not observe tumorigenesis in the trans-
regenerative therapy, which aims to repro- ripotent stem (iPS) cells (2, 14). gineered hair follicles showed full functional- ing the respiratory tract and gastrointestinal (ectodermal origin), the stomach-derived plantation areas (n = 171). To confirm that the
duce the reciprocal epithelial-mesenchymal Pluripotent stem cells, such as ES and iPS ity, including the ability to undergo repeated tube, by histochemical analyses with specific iPS-Stm-FB/gfp-99-1 clone (endodermal ori- transplanted 3D IOS was of iPS cell origin,
interactions during embryogenesis. This or- cells, can be induced to differentiate into spe- hair cycles through the rearrangement of vari- antibodies for CK5, CK10, Muc2, Cdx2, Pax6, gin), and the hepatocyte-derived iPS-Hep- we performed Y-chromosome fluorescence in
gan regeneration therapy represents a fun- cific somatic cell lineages using cytokines that ous stem cell niches. Our current study thus villin, Tuj1, CC10, and E-cadherin (Fig. 1I and FB/Ng/gfp-103C-1 clone (endodermal origin) situ hybridization (Y-FISH) with male mouse–
damental approach to treating patients who mimic the patterning and positioning signals demonstrates the development of a functional fig. S2). In the explants, we identified various (32). The gingiva- and stomach-derived iPS specific DNA probes (Fig. 3C and fig. S5).
during embryogenesis (15). In the embryo, bioengineered 3D IOS and the realization of types of epithelium arising from the three cell clones, but not the iPS-Hep-FB/Ng/gfp- Y-FISH–positive cells were observed in the
patterning signals indicating body axis and organ replacement therapy using iPS cells. germ layers, including ectodermal epithelium 103C-1 clone, generated the bioengineered 3D transplantation area (including the skin epi-
1
Department of Biological Science and Technology, Graduate organ-forming fields are strictly controlled by (CK5-positive), integument (CK10-positive), IOS, including hair follicles. These results in- thelium, dermis, sebaceous glands, intracuta-
School of Industrial Science and Technology, Tokyo University of signaling centers according to the embryonic and endodermal epithelium, including the re- dicate that the generation of a bioengineered neous adipose tissue, and hair follicles) of the
Science, Noda, Chiba 278-8510, Japan. 2Laboratory for Organ
RESULTS
body plan (16). These complex pattern forma- spiratory tract (Muc2-positive, CC10-positive, 3D IOS is not dependent on the origin of iPS cervical skin of recipient female mice. These
Regeneration, RIKEN Center for Developmental Biology, Kobe, In vivo induction of epithelial tissue
Hyogo 650-0047, Japan. 3Organ Technologies Inc., Minato-ku, tion signals in local areas of the body may and CK10-negative) and gastrointestinal tube cell clones from different germ layers because results indicated that the bioengineered IOS
Tokyo 105-0001, Japan. 4Department of Regenerative Medicine, lack centralized organizing signals, as ob- using a CDB transplantation method (villin-positive and Cdx2-positive). Although heterogeneity exists among iPS cell clones. could engraft in the cervical skin and revealed
Plastic and Reconstructive Surgery, Kitasato University of Medicine, served in the generation of teratomas, which To generate various epithelia from iPS cells, the transplantation of single ES or iPS cells In hair follicles, various stem cell types that it is of iPS cell origin. The engrafted hair
Sagamihara, Kanagawa 252-0374, Japan. 5Research Institute include disorganized neural tissues, cartilage, we first developed a CDB transplantation induced epithelium at low efficiency (27), the are maintained in particular regions, such as follicle must establish proper connections
for Science and Technology, Tokyo University of Science, Noda, muscle, and bronchial epithelia (17). Although method as an in vivo transplantation model. incidence ratio of epithelium was statistically CD34-, CK15-, and Sox9-positive follicle epi- with the surrounding tissues, such as the ar-
Chiba 278-8510, Japan. 6Division of Molecular and Regenerative
Prosthodontics, Tohoku University Graduate School of Dentistry,
it is difficult to control the development of To form the complete three-germ-layer epi- constant in explants generated via the CDB thelial cells in the follicle stem cell niche of the rector pili muscles and nerve fibers. The hair
Sendai, Miyagi 980-8575, Japan. specific types of compartmentalized tissues, thelial tissue from iPS cells, we established transplantation method (Fig. 1J). These re- bulge region (33, 34), Lrig1-positive epithelial follicles in the bioengineered 3D IOS had
*These authors contributed equally to this work. several groups recently reported neuroecto- iPS cells from murine gingiva (Fig. 1A) (25). sults indicate that our CDB transplantation stem cells in the upper bulge outer root sheath connected to calponin-positive arrector pili
†Corresponding author. Email: t-tsuji@cdb.riken.jp dermal and endodermal organs generated via In an in vitro culture of iPS cells (3000 cells method reproduced various epithelia. region, and multipotent mesenchymal precur- muscles in the correct polarization to iPS cell–
22 Originally published 1 April 2016 in SCIENCE ADVANCES sciencemag.org SCIENCE SCIENCE sciencemag.org 23
O PT IM IZING YOUR LI V E -CE L L M I CRO S CO P Y: TR I C K S A N D T R A D E - O F F S S E C T I O N T WO | RESEARCH ARTICLES
RESEARCH ARTICLE RESEARCH ARTICLE
A Day 0 7 30 C 4.0
Nanog
1.6
Nestin
1.0
Pax3
in vitro in vivo
Relative expression
0.8
3.0 1.2
Sorting Transplantation In vivo 0.6
iPS cells by CDB method organogenesis 2.0 0.8
0.4
1.0 0.4 0.2
D0
D3
D5
D7
D0
D3
D5
D7
D0
D3
D5
D7
Cells EBs Cells EBs Cells EBs
B iPS cells EB formation
Day 0 Day 3 Day 5 Day 7
D H&E IHC F
Sox2/p63 Sox17 Pax6 Snail/ Twist Top Side
Integumental
iPS cells
CL
CL
CL
Muc2/E-cad Cdx2/E-cad
Single EB
Gastrointestinal
CL CL
CL
Villin/Tuj1 CC10/E-cad
CDB
Respiratory
CL
CL
G H J
14 30
60
*
*
12
25
Area occupancy of epithelial cyst
* 50
(% of each epithelial type)
Frequency of epithelium
10
20
Tissue weight (g)
40
(% of specimen)
8
15 30
6
10 20
4
2 5 10
0 0 0
Single CDB iPS Single CDB Neural Skin/ Upper Lower Unclassified
EB cells EB rosette mucosa endodermal endodermal
tissues tissues
Fig. 1. Induction of epithelial tissues via the CDB transplantation method. (A) Schematic representation of EB cultures and the CDB transplantation
method. (B) Phase-contrast
Fig. 1. Induction images
of epithelial of iPSvia
tissues cellstheandCDB
the transplantation
formation of EBs, which method. were cultured
(A) Schematicin nonadherent
representationplastic wells
of EB for 3, 5,and
cultures or 7thedays.
CDB Scale bars, 100 mm.
transplantation Fig. 2. Analysis of the bioengineered hair follicle induced from iPS cells via the CDB transplantation method. (A) Macroscopic (left panels) and
(C) mRNA(B)
method. expression levels ofimages
Phase-contrast undifferentiated
of iPS cellsiPS andcell
the markers
formation(Nanog)
of EBs,andwhich
neural crest
were cell markers
cultured (Nestin andplastic
in nonadherent during
Pax3)wells forEB 3, formation.
5, or 7 days. *PScale
< 0.001 by
bars, microscopic (H&E staining; right panels) examination of the hair follicles and shafts in the CDB transplants. EBs were stimulated without Wnt10b (upper
Fig. 2. Analysis of the bioengineered hair follicle induced from iPS cells via the CDB transplantation method. Macroscopic (left panels) and
Student’s
100 µm. (C) t test.
mRNA(D) Hematoxylin and eosin
expression levels (H&E) staining iPS
of undifferentiated andcellimmunostaining
markers (Nanog) of EBs,
andusing
neural antibodies
crest cellof epithelial
markers (Sox2/p63
(Nestin and Sox17),
and Pax3) during neural progenitor
EB formation. *P panels) or with Wnt10b (lower panel) on day 7. Scale bars, 100 mm. (B) Frequency of the epithelial tissue, including hair follicle, in CDB transplants. The data
microscopic (H&E staining; right panels) examination of the hair follicles and shafts in the CDB transplants. EBs were stimulated without Wnt10b (upper
(Pax6),
< 0.001and neural crest
by Student’s markers
t test. (Snail and Twist)
(D) Hematoxylin after 7(H&E)
and eosin days. staining
The nuclei andwere stained usingofHoechst
immunostaining EBs, using33258 (white).ofScale
antibodies bars, 50
epithelial mm. (E) Macroscopic
(Sox2/p63 and Sox17), are presented as the means ± SEM from individual experiments; n = 13 (single iPS injection), n = 49 (CDB transplants without Wnt10b), and n = 15 (CDB
panels) or with Wnt10b (lower panel) on day 7. Scale bars, 100 µm. (B) Frequency of the epithelial tissue, including hair follicle, in CDB transplants. The
photographs
neural progenitor (left panels)
(Pax6),andand microscopy
neural crest(H&Emarkersstaining,
(Snailcenter and right
and Twist) afterpanels)
7 days.ofThein nuclei
vivo transplants
were stained under various
using transplantation
Hoechst 33258 (white). conditions.
Scale bars,The50in µm.
vivo transplants with Wnt10b). (C) Number of hair follicles in the CDB transplants. The data are presented as the means ± SEM from individual experiments; n = 13
data are presented as the means ± SEM from individual experiments; n = 13 (single iPS injection), n = 49 (CDB transplants without Wnt10b), and n =
transplants
(E) Macroscopic of 3000photographs
dissociated iPS
(leftcells (upper),
panels) andsingle EBs (middle),
microscopy and morecenter
(H&E staining, than 30 andEBs (lower)
right wereofplaced
panels) in vivointransplants
the subrenal capsule
under for 30transplantation
various days and then (single iPS injection), n = 74 (CDB transplants without Wnt10b), and n = 4 (CDB transplants with Wnt10b). *P < 0.001 by Student’s t test. (D) Comparative
15 (CDB transplants with Wnt10b). (C) Number of hair follicles in the CDB transplants. The data are presented as the means ± SEM from individual
analyzed.
conditions. (F)The
Macroscopic photographs
in vivo transplants of multiple
of 3000 EB in aiPS
dissociated collagen gel before
cells (upper), transplantation.
single EBs (middle), Scale
andbars,
more 1 mm.
than(G)30Weight of thewere
EBs (lower) in vivo transplants.
placed The data
in the subrenal analysis of the length of hair shafts from the hair tip to the DP in CDB explants treated with or without Wnt10b. *P < 0.05 by Student’s t test. (E) Histological
experiments; n = 13 (single iPS injection), n = 74 (CDB transplants without Wnt10b), and n = 4 (CDB transplants with Wnt10b). *P < 0.001 by Student’s t test.
are presented
capsule for 30as the and
days median
then±analyzed.
maximum(F) orMacroscopic
minimum from individual experiments;
photographs of multiple EBnin = a8 collagen
and n = 48 gelper experiment.
before Red circles
transplantation. indicate
Scale bars, the
1 mm.cyst, including
(G) Weight analysis of the hair follicles and their surrounding tissues in iPS cell–derived bioengineered 3D IOSs. The isolated cystic structures with hair follicles were
(D) Comparative analysis of the length of hair shafts from the hair tip to the DP in CDB explants treated with or without Wnt10b. *P < 0.05 by Student’s t
hair follicles,
of the in vivointransplants.
the explants. *Pdata
The < 0.001
are by Student’s
presented test.median
ast the (H) The±area occupancy
maximum of the cystic
or minimum fromlumen in the
individual whole specimens
experiments; n = 8 andof innvivo
= 48transplants of the
per experiment. observed macroscopically (left panels). In the H&E analyses, the boxed areas in the low-magnification macroscopic views are shown at a higher magnification
test. (E) Histological analysis of the hair follicles and their surrounding tissues in iPS cell–derived bioengineered 3D IOSs. The isolated cystic structures
three types of
Red circles conditions
indicate wasincluding
the cyst, compared. *Pfollicles,
hair < 0.001 in bytheStudent’s
explants. t test.
*P <(I)0.001
Histochemical
by Student’sandtimmunohistochemical
test. (H) The area occupancy analyses ofofthethe cystic
cystic epithelia
lumen inwhole
in the the in in other panels. CL, cystic lumen; EL, epidermal layer; DL, dermal layer; SD, subdermal tissue; SG, sebaceous gland; HO, hair opening; v, vessel. Scale bars,
with hair follicles were observed macroscopically (left panels). In the H&E analyses, the boxed areas in the low-magnification macroscopic views are shown
vivo explants
specimens ofof
in multiple iPS cell–derived
vivo transplants EBs.types
of the three Boxedof areas in thewas
conditions left compared.
panels show *PH&E staining.
< 0.001 To identify
by Student’s epithelial
t test. types, suchand
(I) Histochemical as ectodermal epithelium,
immunohistochemical 500 mm (H&E; upper left) and 100 mm (others). (F) Schematic representation of stem/progenitor cells in the hair follicle and surrounding tissues on the skin at
at a higher magnification in other panels. CL, cystic lumen; EL, epidermal layer; DL, dermal layer; SD, subdermal tissue; SG, sebaceous gland; HO, hair
integument
analyses of (top panels),epithelia
the cystic and endodermal
in the in epithelium,
vivo explants including
of multiple the gastrointestinal
iPS cell–derived tube
EBs.(middle
Boxedpanels)
areas inandtherespiratory
left panels tract
show(bottom
H&E panels),
staining.we Toanalyzed
identify the anagen and telogen phases. APM, arrector pili muscle. (G) Immunohistochemical analyses of stem/progenitor cells in the follicles of natural pelage and
opening; v, vessel. Scale bars, 500 µm (H&E; upper left) and 100 µm (others). (F) Schematic representation of stem/progenitor cells in the hair follicle
CDB transplants
epithelial by immunostaining
types, such with specificintegument
as ectodermal epithelium, antibodies(top for CK5, CK10,
panels), and Muc2, Cdx2, villin,
endodermal CC10, Tuj1,
epithelium, and E-cadherin.
including The nuclei
the gastrointestinal were
tube stained
(middle using
panels) enhanced green fluorescent protein (EGFP)–labeled iPS cell–derived bioengineered hair follicles. These samples were immunostained with anti-Sox9 (red),
and surrounding tissues on the skin at the anagen and telogen phases. APM, arrector pili muscle. (G) Immunohistochemical analyses of stem/progenitor
Hoechst 33258 (blue).
and respiratory To identify
tract (bottom the nonspecific
panels), we analyzedfluorescence
CDB transplants signals in these immunohistochemical
by immunostaining analyses,for
with specific antibodies weCK5,
performed
CK10, Muc2, the experiments under
Cdx2, villin, CC10, anti-CD34 (green), anti-CK15 (red), anti-Lrig1 (white), and anti-Lgr5 (red) antibodies. Arrows, epithelial cells of bulge regions; arrowheads, Lrig1-positive
cells in the follicles of natural pelage and enhanced green fluorescent protein (EGFP)–labeled iPS cell–derived bioengineered hair follicles. These samples
the conditions
Tuj1, withoutThe
and E-cadherin. specific antibodies
nuclei against
were stained antigens
using Hoechst [negative
33258control
(blue).(NC)]. Scale bars,
To identify 1 mm (low-magnification
the nonspecific fluorescence images)
signals inand mm (high-magnification
100 immunohistochemical
these cells; *, background fluorescence of hair shafts. Scale bars, 200 mm (upper panels) and 50 mm (lower panels). (H) Histological and immunohistochemical
were immunostained with anti-Sox9 (red), anti-CD34 (green), anti-CK15 (red), anti-Lrig1 (white), and anti-Lgr5 (red) antibodies. Arrows, epithelial cells of
images).
analyses, (J)weThe frequencythe
performed of experiments
epithelial typesunderin CDB
thetransplants. Epithelialspecific
conditions without types inantibodies
CDB transplants
againstwere classified
antigens basedcontrol
[negative on the(NC)].
cell morphology
Scale bars,and 1 mmnumber
(low- analyses of the iPS cell–derived hair follicles and their surrounding tissues. Natural pelage and iPS cell–derived hair follicles were stained with calponin.
bulge regions; arrowheads, Lrig1-positive cells; *, background fluorescence of hair shafts. Scale bars, 200 µm (upper panels) and 50 µm (lower panels).
count. The dataimages)
magnification are presented
and 100asµm the means ± SEM fromimages).
(high-magnification individual (J)experiments;
The frequency n= of5.
epithelial types in CDB transplants. Epithelial types in CDB transplants Arrowheads indicate calponin-positive arrector pili muscles. Scale bars, 200 mm.
(H) Histological and immunohistochemical analyses of the iPS cell–derived hair follicles and their surrounding tissues. Natural pelage and iPS cell–derived
were classified based on the cell morphology and number count. The data are presented as the means ± SEM from individual experiments; n = 5. hair follicles were stained with calponin. Arrowheads indicate calponin-positive arrector pili muscles. Scale bars, 200 µm.
Takagi et al. Sci. Adv. 2016; 2 : e1500887 1 April 2016 3 of 10 Takagi et al. Sci. Adv. 2016; 2 : e1500887 1 April 2016 4 of 10
derived follicles, which were identical to natu- mentalized organized tissues derived from the replacement of skin physiological func- Fig. 3. Transplantation of the bioengineered 3D IOS.
ral pelage follicles (Fig. 3D). Nerve fibers had three germ layers, such as neural tissue (ec- tions, such as the protection of deeper tis- (A) Schematic representation of the methods used for the
also formed proper connections to the bulge todermal), cartilage (mesenchymal), and sues, waterproofing, and thermoregulation generation and transplantation of iPS cell–derived hair follicles.
region of the bioengineered hair follicles (Fig. gut/bronchial epithelia (endodermal), by an (5). However, it is well recognized that current Cystic tissue with hair follicles was isolated and divided into small
3D, lower panels). These results indicate that in vivo injection assay (37). Several cutane- therapies using cultured epithelial tissues suf- pieces containing 10 to 20 hair follicles (lower left panels). The
the iPS cell–derived hair follicles in the bio- ous types of mature teratomas, such as cystic fer from critical issues, including aesthetics small pieces were transplanted into the back skin of nude mice
engineered IOS were correct in structure and dermoids in the ovary and orbital regions, and the inability to excrete sweat and lipids using a follicular unit transplantation (FUT) method developed
established the proper connections to the sur- generate ectodermal organs, such as teeth from exocrine organs (44). In addition, artifi- in humans (lower right panels). Scale bars, 500 µm (lower left
rounding tissues. and hair follicles (17). However, the molecu- cial skin consisting of cultured epidermis and panels) and 1 mm (lower right panels). (B) Macromorphological
lar mechanisms underlying the development dermis without appendage organs and pores observations of two independent engraftments into the
Analysis of iPS cell–derived hair species of these dermoids are still unknown. In an is often used for testing during the develop- dorsoventral skin of nude mice showing the eruption and growth
and distribution in vivo injection assay of ES and iPS cells, ment of cosmetics and quasi-drugs, although of iPS cell–derived hair follicles. Scale bars, 1 mm. (C) Y-FISH
We next investigated the hair species and ectodermal organs could not be detected at this method has critical limitations related to analysis of the CDB transplants using male mouse–specific DNA
their distribution in the bioengineered IOS a considerable frequency in pluripotent stem the permeability of cosmetics and drugs and probes. H&E and differential interference contrast (DIC) images
in the cutaneous environment of recipient cell–derived teratomas. In 3D stem cell cul- the reliability of physiological responses (45). are shown in the left panels. FISH images are shown in the upper
adult nude mice. The iPS cell–derived IOS ture, bioengineered 3D tissues are limited to By contrast, it is expected that bioengineered left panel. Green and blue signals indicate Y-chromosome–
produced all types of pelage hairs, including neuroectodermal and endodermal tissues (2). 3D IOS can overcome these issues (46). Here, positive cells and nuclei, respectively. Boxed areas in the FISH
zigzag and awl/auchene and guard, in a simi- Thus, to generate a 3D IOS, we developed a we provide evidence for the generation of a image are shown at a higher magnification in the right panels.
CC10 (T-18) (1:100, goat; Santa Cruz Biotech- explants, were intracutaneously transplanted 35. C. A. B. Jahoda, C. J. Whitehouse, A. J. Reynolds, N. Hole,
nology), Tuj1 (TU-20) (1:200, mouse; Chemi- into the back skin of 6-week-old Balb/c nu/nu Exp. Dermatol. 12, 849–859 (2003).
36. R. R. Driskell, A. Giangreco, K. B. Jensen, K. W. Mulder, F. M.
con), and E-cadherin (36/E-Cadherin) (1:200, mice, as described above. The iPS cell–derived Watt, Development 136, 2815–2823 (2009).
mouse; BD Biosciences) in blocking solution. skins after transplantation were dissected at 37. K. Takahashi, S. Yamanaka, Cell 126, 663–676 (2006).
The primary antibodies were detected using the first or second anagen phase. The natural 38. C. M. Metallo, L. Ji, J. J. de Pablo, S. P. Palecek, Stem Cells
highly cross-absorbed Alexa Fluor 488 don- skins and iPS cell–derived skin were fixed and 26, 372–380 (2008).
39. T. Kadoshima et al., Proc. Natl. Acad. Sci. U.S.A. 110, 20284–
key anti-mouse immunoglobulin G (IgG; H placed horizontally on a plastic dish before 20289 (2013).
+ L) (1:500; Life Technologies), Alexa Fluor macroscopic photographs were taken. To ex- 40. A. Stathopoulos, M. Levine, Dev. Cell 9, 449–462 (2005).
488 goat anti-rat IgG (H + L) (1:500; Life amine the hair types and distribution of the 41. L. Jiménez-Rojo, Z. Granchi, D. Graf, T. A. Mitsiadis, Front.
Technologies), Alexa Fluor 488 donkey anti- species of hair shafts, we observed all samples Physiol. 3, 107 (2012).
42. M.-X. Lei, C.-M. Chuong, R. B. Widelitz, J. Invest. Dermatol.
rabbit IgG (H + L) (1:500; Life Technologies), using the SteREO Lumar.V12 and AxioCam 133, 7–9 (2013).
Alexa Fluor 594 donkey anti-mouse IgG (H microscope system (Carl Zeiss). The distance 43. X. Lim, R. Nusse, Cold Spring Harb. Perspect. Biol. 5,
+ L) (1:500; Life Technologies), Alexa Fluor between hair pores was measured with Axio- a008029 (2013).
594 donkey anti-goat IgG (H + L) (1:500; In- Vision Vs40 Image processing software (Carl 44. C. Niemann, F. M. Watt, Trends Cell Biol. 12, 185–192 (2002).
45. E. E. Kandyba, M. B. Hodgins, P. E. Martin, J. Invest.
vitrogen), Alexa Fluor 594 donkey anti-rabbit Zeiss). The position of natural and iPS cell– Dermatol. 128, 1039–1049 (2008).
IgG (1:500; Invitrogen), Alexa Fluor 594 goat derived hair pores, hair growth direction, and 46. E. Bellas, M. Seiberg, J. Garlick, D. L. Kaplan, Macromol.
The Fast module for ZEISS LSM 880 with Airyscan: Improving Resolution and SNR on a Confocal
The Airyscan detector design simultaneously improves SNR
Confocal superresolution imaging with four times the speed and resolution with a hexagonally packed 32 channel GaAsP
PMT array that is positioned in the pinhole plane (Fig. 1) and
and improved signal-to-noise ratio removing the traditional pinhole-PMT scheme. This concept
combined two preferable but opposed / conflictive settings
of the pinhole in the traditional LSM; an open pinhole would
Author: Joseph Huff, PhD
Carl Zeiss Microscopy, LLC, Thornwood, NY, USA allow a lot of light to pass and thus increasing signal to
noise, while a small pinhole opening will improve the resolu-
Dr. Annette Bergter tion of the system. Further, each of the 32 detector elements
Carl Zeiss Microscopy GmbH, Germany
behaves as a small independent pinhole where not only the
Date: April 2016 optical resolution is improved but also spatial distribution of
the light is recoded to additionally improve the spatial fre- Figure 1 Schematic of Airyscan detector for LSM 880
1. Mirror 2. Emission filters 3. Zoom optics 4. Airy disk 5. Airyscan detector
quency contrast of the LSM system beyond what a tradition-
First introduced in August 2014, the Airyscan detector from ZEISS represents a new detector concept for laser- al system provides. To use Airyscan for super resolution im-
scanning microscopy (LSM) that enables a simultaneous resolution and signal-to-noise (SNR) increase over aging, the adjustable zoom optics can be arranged to project
traditional LSM imaging. The Airyscan detector design substitutes the conventional LSM detector and pinhole 1.25 Airy units onto the detector where each detector ele-
scheme for an array of 32 sensitive GaAsP detector elements, arranged in a compound eye fashion that resides ment represents 0.2 Airy units yielding a 1.4 x optical resolu-
in the pinhole-plane while still generating an optical section. The new detection geometry allows for the tion improvement and collect up to 50 % more light than an
collection of the spatial distribution of light originating from every point of a microscopic fluorescent object at traditional LSM system with a 1 AU pinhole. Once combined
the pinhole allowing access to higher frequency information and while additionally collecting more light for with a linear deconvolution, a 1.7 x resolution improvement
ultra-efficient imaging. Based on the Airyscan detection concept, the next innovation from ZEISS has been in all three dimensions is achieved and the additional collec-
developed with the introduction of the Fast mode for Airyscan. The Fast mode concept utilizes the Airyscan tion efficiency of the detector provides a 4x – 8x SNR improve-
detector technology in combination with an illumination shaping approach to enhance acquisition speeds by ment affording researchers the ability to better quantify and
four times while simultaneously increasing SNR and resolution overcoming the traditional compromises of LSM study biological processes like microtubule buckling as a mea-
imaging. sure of measure of mechanical stress on a cell. (Fig. 2, [3]).
Improving Speed on a Point Scanning Confocal Figure 2 Application example of standard confocal compared to Airyscan
detection using cardiomyocyte Cells with SiR-tubulin to measure microtubule
Introduction Traditionally in order to improve acquisition speeds on a
buckling. (Left Panel) Traditional confocal still images from a timeseries
The Confocal Laser Scanning Microscope (LSM) has become 25 years, all major developments of commercially available point scanning confocal system, the speed at which the laser acquired at 30 fps with low SNR. (Right Panel) Airyscan still images from a
one of the most popular instruments in basic biomedical systems focused on novel methods to increase image contrast, spot rasters the field of view is increased resulting in a time series acquired at 30 frames/sec with high SNR and resolution.
Each series of stills represents the resting tubules, contracted tubules, and
research for fluorescence imaging. The main reason LSM has instrument versatility and acquisition speed. These innova- decrease the amount of time the laser spot spends on one resting tubules respectively. Images courtesy of Ben Prosser, University of
become so popular is that the technique affords researchers tions have provided the LSM with the flexibility to address pixel. For the maximum speed increase, resonance scanning Pennsylvania [3]
images with high contrast and a versatile optical sectioning research applications in many areas, providing data of cell approaches are utilized in LSM systems allowing frame rates
capability to investigate three dimensional biological struc- differentiation, cell tracking, live cell kinetics, protein expres- of around 30 Hz for a full field of view. However, to achieve
tures [1]. The optical sectioning ability of an LSM is a product sion, protein localization, neural network connectivity, tissue the higher acquisition rates a researcher will have to sacrifice
of scanning a diffraction limited spot, produced by a focused structure, gene expression and protein / gene function (list is the resulting SNR of an image. The reason for the image SNR
laser spot, across a sample to create an image one point at a not meant to be inclusive). However, until the introduction decrease is that as the acquisition speed is increased, the
time. Traditionally, the generated fluorescence for each point of the Airyscan detection concept, the LSM system flexibility laser spot will spend less time per pixel and thus fewer
is collected and focused back through an aperture (pinhole) has always forced a researcher into making a compromise on photons will be available resulting in a deterioration of Figure 3 Application example of a resonance scanning time series
compared to a time series from the Airyscan Fast acquisition mode using
onto a unitary detector (typically a PMT). The traditional a performance metric of a system in order to increase another image SNR (Fig. 3, upper panel). At some point the SNR can
cardiomyocyte Cells with tubulin-EMTB to measure microtubule buckling at
detection scheme allows the pinhole to reject the out of (i.e. speed for resolution, resolution for SNR etc) to meet get so low, that it becomes impossible to discriminate high frame rates. (Top panel) Still images from a resonance scanning time
focus light and the PMT to turn the remaining light from the application requirements. The introduction of both the individual features of the sample and structural information series acquired at 80 frames/sec. (Bottom panel) Still images from an Airyscan
Fast mode time series acquired at 96 frames/sec. Each series of stills
focal plane into a digital signal to form an optical sectioned Airyscan detection concept in 2014 and now the Airyscan is lost in the process to reach the necessary temporal represents the resting tubules, contracted tubules, and resting tubules
image [2]. As confocal systems developed over the past Fast mode represents a fundamental change where tradi- resolution, rendering the method non-usable for the aspired respectively. Images courtesy of Ben Prosser, University of Pennsylvania [3]
tional imaging compromises no longer exist and all metrics scientific goal.
can be improved simultaneously.
30 31
2 3
Application Note
Application Note
// INNOVATION
Go to poster.sciencemag.org/confocal MADE BY ZEISS
www.zeiss.com/lsm880
Finding answers in
demanding research.
ZEISS LSM 800
// INNOVATION
MADE BY ZEISS
www.zeiss.com/lsm800