Ph.D's in Press (January-February 2013)

Part 4 in an occasional feature, highlighting recently published articles featuring an author (or authors) who is a current member of the Stanford Neuroscience Ph.D program. (Part 1, Part 2, Part 3)* Starting 2013 on a strong note, we have two papers with first authors from the Stanford Neuro PhD community. First, Logan Grosenick (Suppes and Deisseroth labs) presents new variants of the GraphNet fMRI analysis method.  Second, Jordan Nechvatal (Lyons lab) reviews brain imaging studies on the effect of stress exposure therapy for anxiety disorders (including phobias and PTSD).

Logan Grosenick

Grosenick, Klingenberg, Katovich, Knutson and Taylor. Interpretable whole-brain prediction analysis with GraphNet. Neuroimage, 2013 Jan 5. pii: S1053-8119(12)01248-7. doi: 10.1016/j.neuroimage.2012.12.062. [Epub ahead of print]. (Link)

Abstract: Multivariate machine learning methods are increasingly used to analyze neuroimaging data, often replacing more traditional "mass univariate" techniques that fit data one voxel at a time. In the functional magnetic resonance imaging (fMRI) literature, this has led to broad application of "off-the-shelf" classification and regression methods. These generic approaches allow investigators to use ready-made algorithms to accurately decode perceptual, cognitive, or behavioral states from distributed patterns of neural activity. However, when applied to correlated whole-brain fMRI data these methods suffer from coefficient instability, are sensitive to outliers, and yield dense solutions that are hard to interpret without arbitrary thresholding. Here, we develop variants of the Graph-constrained Elastic Net (GraphNet), a fast, whole-brain regression and classification method developed for spatially and temporally correlated data that automatically yields interpretable coefficient maps (Grosenick et al., 2009b). GraphNet methods yield sparse but structured solutions by combining structured graph constraints (based on knowledge about coefficient smoothness or connectivity) with a global sparsity-inducing prior that automatically selects important variables. Because GraphNet methods can efficiently fit regression or classification models to whole-brain, multiple time-point data sets and enhance classification accuracy relative to volume-of-interest (VOI) approaches, they eliminate the need for inherently biased VOI analyses and allow whole-brain fitting without the multiple comparison problems that plague mass univariate and roaming VOI ("searchlight") methods. As fMRI data are unlikely to be normally distributed, we (1) extend GraphNet to include robust loss functions that confer insensitivity to outliers, (2) equip them with "adaptive" penalties that asymptotically guarantee correct variable selection, and (3) develop a novel sparse structured Support Vector GraphNet classifier (SVGN). When applied to previously published data (Knutson et al., 2007), these efficient whole-brain methods significantly improved classification accuracy over previously reported VOI-based analyses on the same data (Grosenick et al., 2008; Knutson et al., 2007) while discovering task-related regions not documented in the original VOI approach. Critically, GraphNet estimates fit to the Knutson et al. (2007) data generalize well to out-of-sample data collected more than three years later on the same task but with different subjects and stimuli (Karmarkar et al., submitted for publication). By enabling robust and efficient selection of important voxels from whole-brain data taken over multiple time points (>100,000 "features"), these methods enable data-driven selection of brain areas that accurately predict single-trial behavior within and across individuals.

Jordan Nechvatal

Nechvatal and Lyons. Coping changes the brain. Front. Behav. Neurosci., 22 February 2013 | doi: 10.3389/fnbeh.2013.00013. (Link)

Abstract: One of the earliest and most consistent findings in behavioral neuroscience research is that learning changes the brain. Here we consider how learning as an aspect of coping in the context of stress exposure induces neuroadaptations that enhance emotion regulation and resilience. A systematic review of the literature identified 15 brain imaging studies in which humans with specific phobias or post-traumatic stress disorder (PTSD) were randomized to stress exposure therapies that diminished subsequent indications of anxiety. Most of these studies focused on functional changes in the amygdala and anterior corticolimbic brain circuits that control cognitive, motivational, and emotional aspects of physiology and behavior. Corresponding structural brain changes and the timing, frequency, and duration of stress exposure required to modify brain functions remain to be elucidated in future research. These studies will advance our understanding of coping as a learning process and provide mechanistic insights for the development of new interventions that promote stress coping skills.

*Regarding the mechanics of this feature: This is purely through the magic of an ongoing My NCBI search for the names of Neuro PhD students. I wouldn't be surprised if there were some false negatives in the data set. Neuro students - let me know if I've missed your paper, and I'll gladly add it.

Airplane Brain Quiz!

Here's a tidbit I feel like sharing: Yesterday, I was flying from LA to SF - the last leg of a excruciating 22-hour journey back home from Israel. During that 15 minute window before landing, when my kindle had to be turned off just in case its electric presence flummoxed my Southwest airplane, I flipped through the inflight magazine.

And found this gem of sort-of neuroscience: a Brain Quiz (aka an advert for something in a pill bottle called "AlphaBrain". The website for AlphaBrain is so full of dubious neuro-technobabble that I'm categorically refusing to provide a link.)

 

With a befuddlement mostly provided by substantial amounts of jetlag (still feeling it. woohoo), I stared longest at question 3:

Ready for the answers? Curious which one of the rightmost boxes could possibly be the "most accurate association" with GABA (the major inhibitory neurotransmitter that is involved in just about everything)?

Here it goes.

Acetylcholine:

Mental speed, focus, memory. Commentary: Uh, I guess so. But maybe also muscle movements, seeing as how acetylcholine is THE transmitter at the neuromuscular junction. And I'm not too sure how what "mental speed" means, but acetylcholine is involved in attention, which I guess could work with the focus thing. And screwing with acetylcholine does affect learning/memory/plasticity, so I guess that's fine. Whatever.

Serotonin:

Positive mood. Commentary: Did you know that the vast majority of serotonin release is in the gastrointestinal tract, where it regulates intestinal movements? Mis-regulated intestinal movements sure leaves me in a bad mood. But sure, in the brain, release of serotonin does regulate mood. Drugs that increase serotonin levels in the brain are prescribed as antidepressants (e.g. selective serotonin re-uptake inhibitors, SSRI's), or used (and abused) as psychedelics (e.g. LSD, mescaline, MDMA).

Dopamine:

Coordination, pleasure, mental drive. Commentary: Pleasure? Ugh. Try "reward-driven learning". Does the coordination come from the loss of movement accompanying the death of dopaminergic neurons in Parkinson's disease? Not really a loss of coordination, so much as a categorical degeneration of motor control. "Mental drive" likely refers to the deficits in mental acuity, attention, and memory that accompany dopaminergic cell loss in Parkinson's. Also, reduced dopamine concentrations have been associated with ADHD, which could be characterized by less "mental drive". I guess. Maybe.

GABA:

Relaxation, sense of calm. Commentary: GABA, aka gamma-aminobutyric acid, aka the main inhibitory neurotransmitter in the CNS. It's diverse roles, reduced to the fact that many potent anesthetics are either GABA receptor agonists or positive modulators  (e.g. alcohol, valium). Oh well. Note: for those interested in the differences in GABAergic inhibition between awake and anesthetized states, I direct you to a great recent publication by Michael Hausser and Matteo Carandini. First author Bilal Halder shows that in the mouse visual system, synaptic inhibition was substantially stronger in awake animals, when compared with anesthetized animals. A fun finding, given the (radically oversimplified) hypothesis that anesthetics work by increasing inhibition within the CNS. Insert spirited discussion about the difference between general changes in GABAergic tone (produced by anesthetics) and temporally/spatially/neuron specific synaptic inhibition (observed in awake conditions, likely disrupted by anesthetics).

Citation: Halder, Hausser and Carandini (2013). "Inhibition dominates sensory responses in the awake cortex." Nature 492, 97-100. Link.

True Facts about the Elephant Seal

Recently, my lab decided to ditch the whole “doing science on a Friday” thing, and instead, go on a field trip. Roughly 1.5 hours from Stanford lies Ano Nuevo, a California State Park, and home of the largest mainland breeding colony of northern elephant seals in the entire world.(1)

In celebration of a fantastic afternoon filled with elephant seal babies, battles, and breeding, some photos of the Ano Nuevo colony, accompanied by some facts about the elephant seal.

[gallery columns="1" type="slideshow" ids="2525,2535,2537,2539,2549,2551,2547,2559,2545,2543,2553,2555,2561,2557,2563"]

Some Facts(2), Including Facts from Peer-Reviewed Journal Articles(3)

There are two species of elephant seals, northern and southern. Northern elephant seals hang out in the North Pacific, ranging from Baja California to Alaska. Southern elephant seals, being aptly named, inhabit the sub-Antarctic and Artic waters.

Northern elephant seals are very large. The only seal bigger than a northern elephant seal is a southern elephant seal. They can be mistaken for very large logs (this, I have done). Adult males can grow to over 13 feet, 4,500 pounds; females generally weigh in at 10 feet, 1,500 pounds.

A “seal” can belong to one of three families of fin-footed mammals: Odobenidae (walruses), Otariidae (eared seals, e.g. sea lions), or Phocidae (earless, or true seals). Elephant seals are true seals - they don’t have external ear flaps, and they get around while on land by throwing themselves along the ground. It’s hysterical to watch, until you realize that 4,500 pounds of seal is throwing itself at you, at a rate of up 8 miles an hour.

For most of the year, elephant seals are solitary animals, spending most of their time migrating. Ano Nuevo female elephant seals, fitted with satellite tracking equipment, have ventured as far north as Alaska, and as far west as the International Date Line.(4)

The only natural predators of the elephant seals are great white sharks and orcas. Downside: you’re someone’s idea of a tasty snack. Upside: at least it’s an apex predator.

During the 19th century, humans hunted the elephant seal nearly to extinction, for their blubber (used for lamp oil, similarly to whale blubber). Massive conservation efforts, and the invention of electricity, have restored population numbers from less than 100 seals in 1910, to approximately 150,000 today.(5)

During the breeding season, elephant seals throw a beach party during which time males fight to establish dominance, females give birth and then mate with the dominant males. During this time, the elephant seals will abstain from both food and water.

The “elephant” part of the elephant seal’s name is not a comment about its size. Instead, it refers to the adult male elephant seals nose (or proboscis), which, if you want to be excessively polite about it, looks like an elephants trunk. Dominant males inflate their noses to produce a noise that sounds like a cross between a stalling chain saw and an elephant with irritable bowel syndrome.(6)

[quicktime]http://www.stanford.edu/group/neurostudents/cgi-bin/wordpress/wp-content/uploads/2013/02/MVI_4945.mov[/quicktime]

The proboscis isn’t merely the elephant seal’s attempt to win the animal kingdoms Ugliest Mammal Award, the enlarged nose contains highly convoluted nasal cavities (measuring up to 3140 cm2 in an adult male); this enlarged surface area allows elephant seals to reabsorb enough moisture from their exhalations to maintain water balance during the extended fast of the breeding season.(7)

Despite being mammals, and thus needing air to breath, elephant seals spend most of their time deep underwater – 91% of their time at sea is spent diving Our Ano Nuevo docent stated that the movement of an elephant seal descending underwater is be best described as “the same motion as a leaf on the wind”. A 4,500 pound leaf.(8) An “integrative hierarchical Bayesian state-space” model of Southern elephant seal movements, used to quantify how environmental factors influence an individual seal’s movement, is a thing that exists.(9)

Elephant seals hunt deep underwater, where light is scarce. Elephant seals are not equipped with echolocation, one very useful way to find stuff to eat when hunting in very dark water (see: whales). Instead, elephant seals have adapted their vision to be highly sensitive to low intensity light, with peak sensitivity at 485 nm. Coincidentally, 485 nm is the wavelength of bioluminescence produced by the southern elephant seal’s main prey: myctophid fish.(10)

Lastly, antibodies against the parasite Toxoplasma gondii(11) have been detected in Southern elephant seals.(12) Make of that what you will.

Footnotes:

1. For more on Ano Nuevo, including park and colony history and visitor information, go their excellent website. Back to text

2.  Source: Marine Mammal Center; National Geographic. Back to text

3. Methods: A Pubmed search for Mirounga generated an extensive list of journal articles relating to elephant seals. Journal articles were selected from said list on the basis the level of awesomeness evident in the abstract. Back to text

4.  Source: Robinson et al (2012). Foraging behavior and success of a mesopelagic predator in the northeast Pacific Ocean: insights from a data-rich species, the norther elephant seal. PLoS One. 7(5):e36728. Back to text

5 Source: Marine Mammal Center. Back to text

6. Go home evolution, you are drunk. Back to text

7. Source: Huntley et al (1984). The contribution of nasal countercurrent heat exchange to water balance in the northern elephant seal, Mirounga angustirostris. J Exp Biol 113:447-54. Back to text

8. Which is less like a leaf on the wind, an elephant seal, or Walsh, piloting Serenity? Thinking about which option just made you sadder? (This joke is dedicated to K.Bryant) Back to text

9. Source: Bestley et al (2013). Integrative modeling of animal movement: incorporating in situ habitat and behavioural information for a migratory marine predator. Proc Biol Sci. 280(1750):20122262. Back to text

10. Nicely done, evolution. Source: Vacquie-Garcia et al (2012). Foraging in the darkness of the Southern Ocean: Influence of bioluminescence of a deep diving predator. PLoS One: 7(8):e43565. Back to text

11. Let Neuro Ph.D candidate Patrick House remind you all about Toxoplasma gondii. Back to text

12. Source: Rengifo-Herrera et al (2012). Detection of Toxoplasma gondii antibodies in Antarctic pinnipeds. Vet Parasitol: 190(1-2):259-62. Back to text

Day in the (Grad School) Life: White eggs, brown eggs [updated with results]

[Updated 2/7/13. Click here to skip to results of the hatch] So, I work in a chicken lab.*

What this involves:

• a weekly delivery of fertilized eggs from a farm located in California's Central Valley;
• storing the egg delivery in a wine fridge set to 55 degrees celsius
• two times a week, placing a set of fertilized eggs in an industrial incubator for 3 weeks
• waiting for a cheeping flock to chicks to hatch

When I first joined my lab, we exclusively used white eggs - ones hatched by white leghorns.

In recent months however, the farm has been sending us eggs that look like this:

Clearly, these are not eggs hatched by a white leghorn (leghorn eggs are described as "pearl white" Source: Henderson's Chicken Breed Chart).

Now, I guess I could go to the internet, and carefully research the egg coloration/patterns of common chicken breeds. If I did that, I would probably be able to narrow down the potential breed of chicken currently growing inside the eggs pictured above. But that would be boring.**

Instead, I'm waiting until Wednesday, when the first batch of mystery brown eggs is scheduled to hatch. I'm betting the chicks in the uniformly brown eggs are either Rhode Islands, the most common layer of brown eggs. As for the speckled eggs, they may be eggs laid by the same breed as the uniformly brown eggs, or they could represent an additional breed.

Will all the chicks be the same breed, despite the range in egg coloration? What will that breed be? I'll be finding out (hopefully) on Wednesday.

In conclusion, to quote a post-doc with whom I've been discussing our inability to acquire white eggs: "Aw yeah science!"

------------------------------------------------------------------------------------------

*Technically, I'm a senior graduate student in the lab of Dr. Eric Knudsen, studying neural mechanisms underlying visual attention in the avian optic tectum. For the historical minded: Eric's lab has a long history of working with another avian model, the barn owl. In recent years, the focus in his lab has begun to shift to work in chickens. Yes, the barn owl is a much more majestic bird than the chicken. **For the scientific aspects of my research, it doesn't really matter what breed of chicken I'm using. Yes, it would be more elegant to use only one breed of bird, but doing experiments with brown eggs is better than not doing experiments with non-existant white eggs.

[Update 2/7/13 - the results of the Brown Egg Hatch]

[gallery type="slideshow" ids="2709,2711,2713,2715,2717,2719,2721,2723,2727"]

What I'm Reading: You patched how many cells now?

This week, the list of journal articles that I simply must manage tofinish reading includes a technical tour de force from the lab of J Julius Zhu (UVA), entitled “The organization of two new cortical interneuronal circuits”. The co-first authors are Xiaolong Jiang (now a Research Assistant Professor at Baylor College of Medicine) and Guangfu Wang (currently a Research Associate at UVA). Now, I’ve not yet finished reading the paper, so I’m not going to be able to provide the kind of in-depth description of its findings/protocols/implication. (It’s only Wednesday, proper article-reading protocol requires at least 2 days of sitting half-read on my desk.) That being said, folks interested in cortical microcircuits, inhibitory networks, and astonishing efforts in patch-clamp electrophysiology would do well to pick up this article.

The general gist of the study: it is a painstaking examination of the connection patterns of two distinct subtypes of cortical layer 1 inhibitory interneurons. The authors provide convincing evidence for two never-before-described circuits, involving these layer 1 interneurons: one circuit serves to enhance a particular form of activation, known as complex spikes, in downstream layer 5 pyramidal neurons; the other circuit suppresses this same form of activation. The two subtypes of layer 1 interneuron generate their opposing effects on layer 5 pyramidal neuron activity via distinct connections with additional interneurons located in layer 2/3.

As I said, I won’t go in depth into the findings of the paper. Instead, I’d like to focus on the methods, which as I mentioned, could be described as a technical tour de force. But really, tour de force doesn’t quite capture the ludicrous amount of work that must have gone into this study. One the surface, the techniques used are fairly straightforward, patch clamp electrophysiology, a technique first developed by Burt Sakmann and Erwin Neher, and used daily by scientists in research labs across the world (including myself). However, it is in the application of the technique, that this study sets itself apart.

Some background for the non-patch-clamp physiologists in the room. Most of the time, patch-clamp electrophysiologists record from individual neurons within a brain slice; if the experimental question requires, sometimes two neurons will be patched simultaneously. Simultaneous recordings can be technically tricky, requiring a stable recording setup and a skilled patch-clamper; so although dual recordings are pretty common these days, they still represent a technical achievement. Simultaneously recording more than 2 neurons (say 3-4)? A rare event.

To demonstrate the patterns of connectivity between the various subtypes of neurons (those located in layer 1, 2/3 and 5), Jiang, Wang and company patch clamped up to 8 neurons simultaneously. In total, they report testing “14,832 connections between 1,703 L1 neurons, 3,310 L2/3 interneurons and/or 3,394 L5 pyramidal neurons in the cortical slices.”

To best illustrate my reaction to reading this particular sentence, I present my marginalia:

As a patch-clamp electrophysiology, I am absolutely staggered by the prospect of patching that many neurons. And to collect those numbers while attempting (and succeeding) in simultaneously patching up to 8 neurons… These folks are crazy.

Take a look at Figure 3, in which the authors characterize the connectivity between 7 simultaneously recorded neurons. Oh, and did I mention that the authors recovered the anatomy of the vast majority of their interneurons and pyramidal cells – 85% and 99%, respectively. (I’ve got a hit rate of 50-60%, on good day. Damn people.)

Technically, this paper is ridiculously impressive, in a “damn-how-long-did-that-take-you-no-wait-how-did-you-do-that-um-can-I-buy-you-a-patch-clamp-machine-please” way. And the science is pretty cool too (although again, I’ve not yet finished the paper, so “pretty cool too” should be seen as an initial impression only. And an understatement.).

But seriously, Xiaolong Jiang, Guangfu Wang, Alice J Lee, Ruth L Stornetta, and lab head J Julius Zhang have produced a paper of breathtaking technical mastery of patch clamp physiology. We all should read it and appreciate their hard, hard work. And then get them one of these: http://autopatcher.org/

Jiang, Wang, Lee, Stornetta and Zhu (2013). The organization of two new cortical interneuronal circuits. Nature Neuroscience, 16(2): 210-218. doi:10.1038/nn.3305. Available online here.