Stained beauty, naked neurons: visualizing the brain through history

i-d6132d9b40b57f4738ccf1d266d1ca0e-Schoonverp172.jpg

Hippocampus: Broad Overview
Tamily Weissman, Jeff Lichtman, and Joshua Sanes, 2005
from Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century by Carl Schoonover

The first time I created a transgenic neuron, it was in a worm, C. elegans -- a tiny, transparent cousin of the earthworm. I injected DNA into the embryonic worm, let it grow up, and voila: there was one eerie green blotch like a little Pac-Man ghost, its long green axon a lime racing stripe running along the worm's transparent body. The worm wiggled, but I was the one hooked: science is beautiful.

You wouldn't necessarily know that from the generic glassware and industrial equipment that populate university lab benches. You get little hints from research papers that sandwich lushly colored but largely inscrutable figures between cryptic captions and more cryptic methodologies. But generally, outsiders don't realize that the daily bread and butter of many (if not most) neuroscience labs consists of generating beautiful images -- not just the familiar rainbow-splashed fMRI plots, but more complex visions of individual neurons, synapses, and networks. The quotidian benchwork of a lab is punctuated, like a train trip through a thick forests, with sudden, stunning views that make the waiting worthwhile. Some vistas reveal things you expected to see, but had only imagined; some are surprising, puzzling, inexplicable.

This is the world Carl Schoonover documents in his new book, Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century. Much more than a mere coffee-table lookbook, Schoonover's collection couples neuroscience images with lucid explanations of the methodologies used to generate them, from the Golgi stain to GFP. Essays by scientists like Joshua Sanes, Javier deFelipe and Terrence Sejnowski situate the images in their historical, technical, and biological context -- exactly the type of context you can't get from a scientific paper. The result is a book that's fascinating to flip through and read and re-read.

The casual use of fMRI images to festoon science journalism has created a general impression that brains can be incredibly cool to take "pictures" of, especially when they are "lit up." But how you get a rainbow activity map, or any other useful figure, out of a 2.2 lb lump of pink paté -- that's a part of the story they rarely tell. The truth is, brain tissue looks blah. It's boring. Sliced brain is about as pretty as a thinly chopped white mushroom; you can see only a few obvious structures, and those structures only grossly correlate with brain function. Under a light microscope, you might glimpse neuronal circuitry, except that the barely opaque cortex is a featureless mishmash of pale neurons that are nearly impossible to distinguish. A few patient researchers, like the German scientist Otto Friedrich Karl Deiters (discussed on page 41) managed to tease out individual neurons using only a light microscope. But more substantive investigation of the brain had to await better tools, and after briefly touching in his first chapter on the historic luminaries of gross anatomy, like da Vinci, Vesalius, Willis and Wren, Schoonover focuses on the methodologies developed over the last century and a half to see individual neurons.

i-9d17a9feaf3429589f4c4dcaa26c839c-Schoonoverp47.jpg
Olfactory Bulb
Camillo Golgi, 1875
from Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century by Carl Schoonover

The big breakthrough was the Golgi stain, discovered by Italian physician Camillo Golgi in 1873. Golgi's stain blackened neurons more or less randomly, while the rest remained transparent. For the first time, researchers could confidently and easily copy the shapes of entire neurons -- and realize how many different shapes there were, and that they were organized into layers and regions with great complexity.

But it was not Golgi himself who fully exploited the stain to push cellular neuroscience forward: it was his future intellectual rival, Santiago Ramon y Cajal. An essay contributed by Javier deFelipe explains how Cajal discovered and exploited Golgi's technique, the "revelatory power" of which led Cajal to write,

What an unexpected spectacle! On the perfectly translucent yelow background, sparse black filaments appeared that were smooth and thin or thorny and thick, as well as black triangular, stellate, or fusiform bodies! One would have thought that they were designs in Chinese ink on transparent Japanese paper. The eye was disconcerted, accustomed as it was [to earlier, less effective staining methods] where the indecision of the mind has to be reinforced by its capacity to criticize and interpret. Here everything was simple, clear and unconfused.

With the structure of the nervous system clarified, it was possible to make inferences about brain function, like which direction information flows among the cells of the retina. The small arrows in the drawing below represent Cajal's deductions -- from logic and structure alone -- about the transmission of visual information to the brain.

i-51fa3df2c69b49a35bdb2bef2e9b383d-p68detail.jpg
Retina (detail)
Santiago Ramon y Cajal
from Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century by Carl Schoonover

Cajal's lucid sketches and prose portray the situation as clearer than it was; Schoonover includes a photograph of one of Cajal's slides, observing that even after Golgi staining, "there remains nonetheless a confusing tangle of wires in the background." Many scientists doubted Cajal's findings, until, in a story related by deFelipe, Cajal secured the confidence and vocal approval of an influential German histologist. Even Golgi would not accept Cajal's drawings as evidence that neurons were not a continuous network, but separate cells (this made things rather awkward when they accepted the Nobel Prize together in 1906).

It may make me a bad empiricist, but the thing I love most about Cajal's work is his style. Cajal's hand is readily recognizable, from the heavily inked neurons to the lowercase letters and arrows scattered through the drawing. Cajal was not faithfully copying his data; he was extracting patterns and using design to depict concepts about brain function. As Schoonover puts it,

he apparently did not sketch with pen in one hand and the microscope's focus knob in the other. Instead, he is said to have drawn from memory, in the afternoon, after a morning of observation. Though the details were of course of great importance, what mattered most to Cajal was the general form, the common properties, the essence of the specimen's overall architecture. . . His amply documented interest in the visual arts suggests why his renditions of biological samples are so exquisite and hints, perhaps, at why he was able to identify underlying forms where others saw only lines.

Indeed: biology is all about patterns, and finding hidden patterns takes more than a new staining technique. It takes a certain kind of big-picture approach to the data. One of the more common failings I have seen among graduate students, myself included, is an inability to look up from the weeds long enough to see big patterns.

That said, there was only so much the Golgi stain, and other stains applied to dead tissue, could do to reveal patterns of structure and function in the brain. The next step in Schoonover's book is GFP (green fluorescent protein, originally cloned from Pacific jellyfish) which is inserted transgenically into the animal and turned on in its neurons as it grows. The neurons "light up" (quite literally, not in the metaphoric sense used too often to describe fMRI) when exposed to blue light. The result is a living animal with one or more labeled neurons -- just like the racing-stripe worm I made over a decade ago.

While I agree that GFP is where the action is, methodologically and aesthetically, my main quibble with Schoonover's book is that he elides the century between Cajal and GFP. That century was full of other visualization techniques which here receive short shrift. I rarely used GFP to visualize cells in my own graduate work; mainly I used antibody staining and in situ hybridization, which label proteins and RNA, respectively. Fluorescent tags can be coupled to either technique, producing a result that looks much like GFP expression, though lacking many of its benefits. But one can also visualize the results with more "old-fashioned" (relative to GFP, not to the Golgi stain) chemical techniques, like horseradish peroxidase-DAB staining (which looks like (p. 182-83) and lacZ/beta-galactosidase staining (represented on 178-79). These techniques aren't as sexy and sharp as fluorescence; the results they give aren't as clean. But they're cheap, robust, and you can see them with your naked eye (you don't need a specialized fluorescent light microscope). As a result they've been a big part of day-to-day visualization in neuroscience labs, and the foundation for important discoveries over the last several decades.

That said, Schoonover only has so much room in his book, and I wouldn't give a page less to images like the fluorescent panorama of the hippocampus at the top of the post (detail below).

i-4b254b933f5dfd74504108e82629db64-p172detail.png
Hippocampus: Broad Overview - detail
Tamily Weissman, Jeff Lichtman, and Joshua Sanes, 2005
from Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century by Carl Schoonover

This image of the hippocampus was created by breeding mice carrying different variants of the GFP gene - not only green fluorescent protein, but also its red and blue counterparts. The result? Simultaneous visualization of several different classes of neurons in various layers of the hippocampus, laced through each other like multicolored fingers. (Take a deep breath and ask yourself, where would you even start to decipher this?)

It was in Lichtman's and Sanes's labs that the Brainbow mouse -- GFP to the max -- was conceived by postdoc Jean Livet. The Brainbow technique (explained by Sanes in his essay on p. 79-80) takes the genes for green, red and blue fluorescent proteins, introduces all three into neurons in multiple copies, then randomly flips each copy on and off. What you get is different levels of red, blue and green fluorescence expressed in each individual neuron -- much like what happens in the pixels of your color television screen. Voila, a technicolor mouse brain, in which it is even easier to tell where one neuron ends and another begins.

I never tire, and do not imagine who could tire, of looking at Brainbow images, like this Lite-Brite-esque slice of brainstem by Jean Livet, winner of the 2007 Olympus Bioscapes competition. "Neurons or Pollack?" asked Wired, and they had a good point: how could something so beautiful be scientific data? (I like to think we at BioE expect no less). Brainbow images have an undeniably modern aesthetic. They're crisp and bold, like neon signs in a night cityscape, or satellite photos of the illuminated Earth. And while Brainbow photos lack the style Cajal brought to his sketches, they make up for it in sheer complexity and depth: you're seeing the brain itself, albeit through a technicolor lens. Without Cajal's carefully labeled diagrams interposed between you and messy, squiggly reality, the experience is closer to actually being in the lab.

Golgi staining, GFP and Brainbow are only a few of the techniques Schoonover collects in his book. PET, MRI and fMRI are represented, of course; diffusion MRI, which labels axon pathways; the old technique of cerebral angiography; an entire chapter on computer techniques for mapping and modeling connections between neurons; another chapter on electrophysiology. Confocal microscopy, two-photon microscopy, SEM and TEM are represented, including a lovely image of rope-like actin studded with myosin heads by John Heuser (1985), and a luminous false-color view of a growth cone by Knoll and Berger (2008). There's even stimulated emission depletion microscopy, a relatively new technique which tries to break the visualization limits set by the physics of light itself. Most of these techniques are described very briefly, so don't expect a research protocol; Schoonover's curation errs on the side of simplicity and accessibility. Mainly, he lets the images speak for themselves.

On some level, neuroscience is dizzyingly, recursively weird. Every pretty new press release picture of a brain is the cumulative product of centuries of creative and persistent minds trying to devise ways to "see" inside themselves. The fact that we've seen as much as we have is astonishing; the amount we haven't seen, or have seen but have yet to understand, makes me doubt we'll understand ourselves fully in my lifetime. But I'm grateful to Carl Schoonover for compiling a book that makes the snapshots from this scientific journey accessible outside the neuroscience lab community. As he says in his preface,

If the images are extraordinarily beautiful, I would argue that the principles underlying the techniques that created them are in some instances even more exquisite. The manifestations of beauty that we have been licensed to encounter in every nook and cranny of the brain depend, in every single case, on the cleverness and elegance of the strategies used to illuminate them. My goal is for the reader's enjoyment of the images to be accompanied by equal appreciation for the craft that underlies their genesis.

As Western art developed over the past century, the concept behind the physical art object acquired a status on a par with -- or, some might argue, superseding -- the thrilling of our senses by the object itself. Sol LeWitt's perspective -- "the idea becomes a machine that makes the art" - urges us to enjoy the aesthetic value of a concept independent of the actual item it engenders. By analogy, I often find the idea behind a technique to be irresistibly beautiful in and of itself. I propose then, that in science as in art we should delight not only in the physical manifestations of the data -- examples of which fill these pages -- but also in the ideas which produced them.

Bravo, Mr. (soon to be Dr.) Schoonover. I think that may be what I've been trying to say with Bioephemera all along.

i-fe799569d10f9c8949b1bbb88d4f2d3e-PortraitsMind_cover.jpg

all images and quotes from Carl Schoonover's Portraits of the Mind: Visualizing the Brain from Antiquity to the 21st Century, Foreword by Jonah Lehrer, Published by Abrams.

Related:

Atlantic Monthly's slideshow of images from Portraits of the Mind

Ira Flatow's 2007 interview with Jeff Lichtman about Brainbow

Carl Schoonover's website, where you can view more images from the book

An old riff on Cajal and scientific art from BioE

More like this

This article is reposted from the old Wordpress incarnation of Not Exactly Rocket Science. The blog is on holiday until the start of October, when I'll return with fresh material. At Harvard University, a group of creative scientists have turned the brains of mice into beautiful tangles of colour…
Jennifer Ouellette reports from a month-long program on the anatomy, development and evolution of the brain, at the Kavli Institute for Theoretical Physics, with a fantastic post called Tools of the brain trade. Inspired by a talk given by Winifred Denk, about reconstructing brain circuitry using…
THE latest issue of Technology Review contains a photo essay by yours truly, called Time Travel Through the Brain, in which I look at how techniques used to investigate the brain have evolved during the 100 year history of modern neuroscience. The essay begins with a drawing by the great Spanish…
For most of the nineteenth century, there was an on-going debate among researchers about the organization of the nervous system. One group of researchers, the so-called reticularists, believed that the nervous system consisted of a large network of tissue, or reticulum, formed by the fused…