Connectome (48 page)

[>]
   
subtler kinds of rewiring: In the Knudsen experiments, rewiring could be seen relative to the map in the inferior colliculus. A similar strategy could work in sensory and motor areas of the cortex, which generally contain analogous maps. Many other areas are not organized according to such simple maps, however, so rewiring is more difficult to detect.
In the Knudsen experiments, rewiring could be seen relative to the map in the inferior colliculus. A similar strategy could work in sensory and motor areas of the cortex, which generally contain analogous maps. Many other areas are not organized according to such simple maps, however, so rewiring is more difficult to detect.

[>]
   
No new neurons:
Rakic 1985 cemented the dogma.

[>]
   
Elizabeth Gould:
Gould et al. 1999.

[>]
   
“most startling”:
Blakeslee 2000.

[>]
   
champion at self-repair:
Taub 2004.

[>]
   
prevailed in the neocortex:
Most of the evidence comes from monkeys, but Bhardwaj et al. 2006 additionally studied the human brain.

[>]
   
hippocampus and the olfactory bulb:
Kornack and Rakic 1999, 2001. New neurons in these regions of the adult rat brain had previously been shown by Joseph Altman in the 1960s, but his pioneering discovery had been largely ignored by his colleagues.

[>]
   
“gateway” to memory:
Kempermann 2002.

[>]
   
memories of smells:
Lledo, Alonso, and Grubb 2006.

[>]
   
fingers fused together:
Flatt 2005.

[>]
   
died as survived:
Cowan et al. 1984.

[>]
   
wasteful to create:
Buss, Sun, and Oppenheim 2006.

[>]
   
I'll call
regeneration: When neuroscientists use the term
regeneration,
they are usually referring to the regrowth of axons after they are severed, but I call this
rewiring
. My usage of
regeneration
is typical of biology, and refers to the creation and elimination of cells.

[>]
   
since the 1960s:
Gross 2000 reviews the history of such reports and speculates about why they were ignored.

[>]
   
grain of truth:
Kornack and Rakic 2001 charged that Gould had erroneously identified non-neuronal cells as neurons. There are many types of brain cells that are not neurons.

[>]
   
foster learning and plasticity:
On a related note, some critics say that the Rosenzweig experiments reveal the effects of deprivation, not enrichment. The fancy cages with toys and companions should not be regarded as “enriched,” as they only relieve the deprivation of the ordinary laboratory cage. The latter is a highly impoverished environment compared with the rats' natural habitat.

[>]
   
migrate into the zone:
Carmichael 2006.

 

8. Seeing Is Believing

 

[>]
   
and Francis Crick:
Watson and Crick relied on the data of Rosalind Franklin, who was a crystallographer. She died prematurely and could not share their Nobel Prize.

[>]
   
didn't fully recognize its significance:
Leeuwenhoek reported his observations of sperm in a letter to the president of the Royal Society of London. Embarrassed by the subject, he stressed that the specimen was the natural product of his marriage bed, and asked the president to suppress the letter if he found it offensive (Ruestow 1983).

[>]
   
called them “animalcules”:
Actually animalcules seem like an afterthought in the letter, because they are mentioned only in the last paragraph (Leeuwenhoek 1674).

[>]
   
three clergyman, a lawyer, and a physician:
Dobell 1960 describes Leeuwenhoek's life and career, and collects many of his letters.

[>]
   
single, very powerful lens:
Ford 1985 describes the history of the single-lens microscope and argues that Leeuwenhoek made his best lenses by letting molten glass solidify into small globules. Ruestow 1996 notes that Leeuwenhoek also made some lenses by the more standard method of grinding glass, as he claimed in his writings.

[>]
   
individual neurons:
Figure 26 depicts Golgi-stained neurons from the cortex (superior temporal sulcus) of the adult rhesus monkey. The image extends from the white matter at the bottom to layer 3 of the cortex at the top, a distance of roughly 1.5 millimeters.

[>]
   
single dark strand:
Those of you who are observant may notice that the pasta shown in Figure 27 is actually bucatini, which is thicker than spaghetti and has a hole running down the center. (It has a wonderful chewy texture, and I recommend it highly.) If every strand of the bucatini were stained with a unique color, it might be possible to trace the paths of
all
the strands, even in a somewhat blurry image. Researchers have actually implemented this strategy by genetically engineering mouse neurons to fluoresce in random colors, a method that Jeff Lichtman wittily named “Brainbow” (Livet 2007; Lichtman 2008). However, the number of distinguishable colors is limited, so Brainbow may be insufficient for tracing a large number of densely entangled neurites. It may be possible to improve the situation by combining Brainbow with sharper images, like those produced by recently invented methods of light microscopy that beat the diffraction limit (Hell 2007). In another approach, Tony Zador has proposed genetically engineering each neuron to contain a random RNA or DNA sequence. The sequence could be unique for every neuron, because the number of possibilities is so large—much larger than the number of distinguishable colors. Other molecular tricks and genomic technologies would be used to find the sequences for every pair of connected neurons, yielding the connectome. We don't yet know whether these directions of research will provide alternatives to electron microscopy, the standard method of finding connectomes. I mention them only to make clear that connectomics is going through an exciting period of innovation.

[>]
   
why
Golgi's stain:
From a solution of potassium dichromate and silver nitrate, silver chromate precipitates in a small fraction of neurons, for some unknown reason.

[>]
   
Golgi looked in his microscope:
Guillery 2005. Cajal's view was called the “neuron doctrine” and Golgi's the “reticular theory.”

[>]
   
as Golgi envisioned:
Have you ever heard the joke “Economics is the only field in which two people can share a Nobel Prize for saying opposing things”? The quip probably dates from 1974, when the prize was shared by the economists Gunnar Myrdal and Friedrich Hayek, who were shocked to find themselves honored at the same event, given that their views were so diametrically opposed. At his banquet speech Hayek suggested that a prize for economics was a bit dangerous. Myrdal even wrote a paper calling for the abolition of the prize (Myrdal 1977). He argued that economics was a “soft” science, so its prize, established in 1968, did not belong with the “real” Nobel prizes in the “hard” sciences, which were originally established by the will of Alfred Nobel in 1895. According to Lindbeck 1985, this was ironic coming from Myrdal, who had lobbied strongly for the creation of the economics prize in the first place. Based on the 1906 Nobel Prize to Golgi and Cajal, should we also regard neuroscience as a “soft” science? Perhaps neuroscience is somewhere in between economics and physics. It's true that Golgi and Cajal had opposing views, but no one called for the abolition of the Nobel Prize for Physiology and Medicine, as far as I know. And they both turned out to be correct, so the Nobel committee did the right thing.

[>]
   
new stains:
These are based on big and heavy atoms like osmium, uranium, and lead, which reflect electrons well.

[>]
   
Figure 28:
This transmission electron microscope image comes from the rat hippocampus. It can be found along with many other interesting images of neurons and synapses at
synapse-web.org
.

[>]
   
the diffraction limit:
Recently physicists have realized that it's possible to beat the diffraction limit using fluorescence microscopy, which was not available to Golgi (Hell 2007).

[>]
   
in a light microscope:
The blurred version of the image is due to Winfried Denk, who simulated the point-spread function of a 1.4 numerical aperture (NA) microscope objective assuming a wavelength of 500 nanometers.

[>]
   
edge of a saw is blunt:
As a hybrid of saw and knife, serrated knives are one of those irritating intermediate cases that are the bane of the classifier. We will ignore them.

[>]
   
2 nanometers wide: More precisely, 2 nanometers is the edge radius of curvature claimed by several manufacturers of diamond knives on their websites. In the published literature, one can find reports of 4 nanometers (Matzelle et al. 2003).
More precisely, 2 nanometers is the edge radius of curvature claimed by several manufacturers of diamond knives on their websites. In the published literature, one can find reports of 4 nanometers (Matzelle et al. 2003).

[>]
   
Keith Porter and Joseph Blum:
Porter and Blum 1953. Bechtel 2006 recounts the history of biological electron microscopy.

[>]
   
ultramicrotome mounted inside:
Denk and Horstmann 2004.

[>]
   
“scanning electron microscopy”:
Earlier researchers had used transmission electron microscopy (TEM), which sends electrons through thin slices of tissue. (This is similar to viewing a photographic negative by holding it up to a light.) The scanning electron microscope instead bounces electrons off the surface of the object being imaged.

[>]
   
thin as 25 nanometers:
This number is important, because it sets the resolution of the 3D image stack in the vertical direction. Electron microscopy has much finer resolution (nanometers or less) in the two lateral directions. The vertical resolution is much coarser.

[>]
   
eventually achieved 30 nanometers:
Hayworth's original design, shown in Figure 30, was called ATLUM rather than ATUM. The
L
stood for “lathe,” a kind of rotary machine tool. The plastic block containing brain tissue was mounted on an axle. Each turn of the axle pushed the block past the diamond knife, shaving a thin slice off. Hayworth initially thought that the rotary motion would control slice thickness more precisely. Since then, he has returned to the traditional linear motion of a conventional ultramicrotome, like the back-and-forth of meat in a deli slicer.

[>]
   
eliminates the need for a diamond knife:
Knott et al. 2008 describes the method of focused ion beam (FIB) milling. Bock et al. 2011 describes a modification of the transmission electron microscope to produce images with larger field of view, speeding up the rate of data acquisition.

 

9. Following the Trail

 

[>]
   
walls of the axon go by:
The molecular car is kinesin, and the track is called a microtubule.

[>]
   
one billion collisions:
CMS Collaboration 2008.

[>]
   
had not been invented yet:
When Brenner gave the kickoff lecture for my 2007 class on connectomics, he expressed disdain for the term. He recommended that the field be christened “neuronomy” instead, quipping that “neuronomy is to neurology as astronomy is to astrology.”

[>]
   
Richard Goldschmidt:
White et al. 1986.

[>]
   
sausage is stuffed with spaghetti:
We are stretching our Italian food analogy. Perhaps Thai food would be better, as summer rolls typically do contain noodles.

[>]
   
just thin enough:
Ideally, the slice thickness would be the same as the spatial resolution of the 2D images produced by the electron microscope. Then the 3D image would have the same spatial resolution in all directions. But it's not possible to slice that thin, so the image inevitably has poorer resolution in the third dimension.

[>]
   
repeatedly wrote the same symbol:
They wrote with felt-tip pens on transparent acetate sheets, which were placed on top of the original photographic plates. To make the process even more complex, sometimes they would trace two neurites that started out separate but merged at a branch point. Once they realized the two neurites were part of the same neuron, they went back and changed all the letters of one neurite to match the other.

[>]
   
302 neurons of the worm: To be more precise, the 282 somatic neurons are described. There are also 20 pharyngeal neurons, which form an almost independent nervous system (Albertson and Thomson 1976). Errors were corrected, inconsistencies resolved, and gaps filled in by Chen, Hall, and Chklovskii 2006. The updated version was published at
wormatlas.org
.
To be more precise, the 282 somatic neurons are described. There are also 20 pharyngeal neurons, which form an almost independent nervous system (Albertson and Thomson 1976). Errors were corrected, inconsistencies resolved, and gaps filled in by Chen, Hall, and Chklovskii 2006. The updated version was published at
wormatlas.org
.

[>]
   
touch to the head:
Chalfie et al. 1985.

[>]
   
John Fiala and Kristen Harris:
Fiala 2005.