Genetically-encoded intracellular EM stain for connectomics

The field of connectomics would greatly benefit from a genetically-encoded stain for specific proteins that works intracellularly.  This could be used to provide positive contrast for axon tracts or to identify if a synapse is excitatory, inhibitory, or neuromodulatory. Horseradish peroxidase (HRP) is a classic approach to EM staining but does not work in reducing environments, such as inside cells.  Several groups (Ting, Looger) have tried to engineer away the di-sulfide bonds of HRP, to allow it to work intracellularly, but all have failed to maintain activity after removing these bonds. An alternative approach, miniSOG, from the Tsien lab, is a genetically-encoded tag that can stain via singlet oxygen generation, but requires light to activate it. In Nature Biotechnology, Alice Ting’s group reports APEX, a perioxidase that works intracellularly, which they then engineered to monomerize and improve staining performance. This tool could find extensive use in connectomics of complex brain tissues.

 

UPDATE : DIADEM Final Results

The DIADEM automated neuronal reconstruction contest has finished.  Accurate, fast, and high-resolution automated neuron reconstruction is of vital importance to cracking the mystery of how neural circuits perform. Even with perfect knowledge of the firing patterns of every cell in a circuit, our understanding of how these patterns are produced and how the information is processed would be quite limited.  True understanding requires knowledge of the precise wiring diagram.  This prize is a good first step towards bringing awareness of this tricky problem to the world’s best computer scientists.

$75,000 in prize money was to go to the group that was able to produce high-quality reconstructions of neuronal structures at least 20x faster than by-hand reconstructions.  In the finals, the fastest speed achieved was 10X the by-hand method. Some groups were hindered by slight variances in the source data formatting, which normally isn’t a big deal unless you only have 20 minutes to produce as much reconstruction as possible…

Since no group was able to beat the hard floor, but substantial progress was made, the money was distributed amongst these finalists.

Badrinath Roysam Team, $25,000
“for the better overall generality of their program in producing robust reconstructions by integration of human and machines interactions.”

Armen Stepanyants Team, $25,000
“for the better overall biological results in the spirit of pure automation.”

Eugene Myers Team, $15,000
“for the excellent quality and strength of their algorithm.”

German Gonzalez Team, $10,000
“for their deeper potential, more original approach, and ultimate scalability of their proposed solution.”

Deniz Erdogmus Team
“for elevating themselves above the current state of automated reconstructions…with a deep understanding of the technical and scientific problems.”

Congrats to the placing teams.

Compressed Sensing in Neuroscience

Wired has a nice lay-person write-up of the rapidly developing field of compressed sensing. This is a technique that allows accurate reconstructions of highly undersampled sparse datasets. This field really took off in 2004 when Emmanuel J. Candès discovered that a tomography phantom image could be reconstructed exactly even with data deemed insufficient by the Nyquist-Shannon criterion. It is probably the hottest topic in imaging theory today.

Modified Shepp-Logan phantom with enhanced contrast for visual perception.

According to this review, Compressed Sensing MRI, its successful application requires three conditions to be met :

• Transform Sparsity: The desired image must have a sparse representation in a known transform domain (i.e., it must be compressible by transform coding),
• Incoherence of Undersampling Artifacts: The aliasing artifacts in a linear reconstruction caused by k-space undersampling must be incoherent (noise-like) in the sparsifying transform domain.
• Nonlinear Reconstruction: The image must be reconstructed by a non-linear method which enforces both sparsity of the image representation and consistency of the reconstruction with the acquired samples.

These conditions are well met by MRI imaging.  This decoding technique dramatically shortens the required sampling times in an MRI magnet, which reduces the impact of motion artifacts, the bane of high-resolution MRI.

Unfortunately, I don’t think it is very applicable to situations where signal/noise of the underlying source is poor, like counting action potentials in shot-noise limited in vivo calcium imaging. But it’s use is spreading into other related problems, such as mapping the functional connectivity of neural circuitry.  Tao Hu and Mitya Chklovvskii apply the compressed sensing algorithms in Reconstruction of Sparse Circuits Using Multi-neuronal Excitation (RESCUME) from the latest Advances in Neural Information Processing Systems journal. They measure a post-synaptic neuron’s voltage while stimulating sequentially random subsets of multiple potentially pre-synaptic neurons. The sparseness of connectivity allows them to map connectivity much faster than by a sequential method.

UPDATE: If you want to get a better sense of the breadth and depth of the applications of compressed sensing, check out Igor Carron’s comprehensive site, Compressive Sensing : The Big Picture.

CSHL Meeting – Session VII – Super-Resolution Optical Techniques

Jean-Louis Bessereau – Ultrastructural mapping of functional domains of synapse at the synapse using high pressure imaging
High pressure freezing instantaneously converts up to 0.3mm thick water into amorphous ice. C Elegans only .1mm thick at maximum. HPF entire worm to obtain EM of ‘living’ synapses. Vesicle priming occurs within 100nm of presynaptic density, directly across from post receptors. However, vesicle recycling occurs only at sites >150nm lateral from presynaptic release sites.

Mark Ellisman – Multiscale light and electron microscopic imaging of the nervous system
Two-color correlated light and EM microscopy using FlAsH and ReAsH. Quantum dot immunohistochemistry for multicolor correlated light and EM. QDs of different wavelength are differently sized and can be distinguished on EM.

Eric Betzig – Superresolution optical location of single proteins.
Sparsely photoactivate PA-GFP tagged proteins with very weak laser pulse. Then image with high laser power to collect light from many point sources. Determine point source centers, then repeat process many times to find protein locations to 2-25nm resolution. See the science 2006 paper.

Winfried Denk – Automated circuit reconstruction
EM :
When doing automated serial EM, individual sections can get lost or crumpled before being imaged. Instead image the block face then shave off a section. Resolution is significantly degraded, because lower power 3keV to limit z-penetration. Penetration goes as E^(5/3). Monte Carlo simulation of backscatter at 3 vs 10keV to determine point spreads and depth. Do 30nm sections, which is very [pun alert] cutting edge. Circuit reconstruction fidelity is limited by the dimension of least resolution, so usually Z section thickness. Block face EM actually used in the 1940s. Can still identify synaptic densities. Use backscatter to see heavy things. Constructing whole c. elegans. Demonstrate by hand reconstruction of some axons, dendrites and synapses as proof of principal. Using ion gun to acquire without need of a vacuum.

Staining Technology :
Colloidal lanthanum greatly enhances contrast of extracellular space but poor tissue penetration. Kevin Briggman and Denk like using extracellular HRP currently. Good membrane contrast. Moritz Helmstadter working on automated segmentation, collaborating with Sebastian Seung @ MIT for neural network based methods. Neural network method looks pretty good.

Lukyanov – New fluorescent proteins
Cloned the first RFP from coral, DsRed. DsRed tends to aggregate due to tetramer nature. Comparison of DsRed2, TurboRFP, TagRFP and mCherry. Brightness by e*QY. 100,172,134,44 respectively. TagRFP has shorter emission wavelength (<600nm) than mCherry (618nm). Can distinguish TagRFP from mCherry by the significantly difference in lifetime. [What about mCherry2?]

Redshifted FPs.
Katusha excitation 588, em 635. QY 0.34 e 45,000. Not quite as redshifted as mPlum but significantly brighter. But not monomeric. Made mKate which has similar brightness but is monomeric.

Cyan to green photoswitchable PS-CFP. Also made Dendra, monomeric green to red photoconvertible FP. [How much bleaching to red occurs to get 90% conversion? In our hands no photoswitchable protein allows total conversion without significant bleaching.]

Visualization of targent protein degredation in real time at single cell level using Dendra2. Zhang Biotechniques Apr 2007. IkBa-Dendra2 degredation down to 20% at 5hr with cycloheximide treatment. Fluoresence stays fixed with proteasome inhibitor. 20min protein halflife following PMA treatment. Hydrogen peroxide sensor HyPer of cpYFP inserted into OxyR-RD. OxyR transcription factor forms reversible S-S bonds in bacteria. Around 2-fold ratio change 490/410 between 0 and 250nM H2O2. Showed some small responses in cells to EGF stimulation. [Is this response reversible? Never showed a recovery trace.] cpCitrine145 with m13 on N, calmodulin on C term makes a GCaMP like sensor, 12-16x maximum ratio change. [How EXACTLY is this sensor different from GCaMP2?]

Killer Red, genetically encoded photosensitizer. Screened different natural FPs for phototoxic proteins that kill bacterial colonies under light. Most FPs non-toxic, 2-5fold increase in bacterial phototoxicity, but 1 FP had 1000x increase. In mammlian cells, expression in cytosol is not enough to cause sufficient oxidative stress to kill cell with light. But target to mitochondria and can kill cells with light. Targeted to cell membrane, blebs occur in 3 minutes. Use to kill off specific muscle cell with light in zebrafish. CALI of phospholipase C1-d PH domain by fusion to killer red. However it is still dimeric, so doesn’t fuse well to some things.

Karel Svoboda – Meeting Summary
Many spatial and timescales of neural questions require development of variety techniques. Interface of new imaging techs and genetically targeted probes making lots of progress on addressing these questions. The developmental talks were some of the best applications of the new technologies. These meetings will be measured by the crystallization of new, unexpected directions in research. Of course GECIs have been very exciting, but in the last 2 years also seen big progress on…
EM
-sectioning and data collection, segmentation and reconstruction, EM on targeted neurons.
Light based approaches
– PALM, STED –EM type resolution in far field optical microscope, spectral multi-plexing, synapse specific tracing.
Optical remote control with light.
ChR2 – The silver bullet
HR – hyperpolarlize
Rhodopsin – seconds time-scale modulation of plasticity
Optical switches without transgenes
Applications

See you in 2009!