Where's Bregma?

OR: A fun game for every surgery!

If you're interested in any of the important parts of the brain (i.e. not cortex), you're going to need do to stereotaxic surgery to target the area you're interested in. And to do stereotaxic surgery, you need fiducial coordinates on the skull, which canonically are bregma and lambda.

You would think that the definitions of bregma and lambda are well known, but I found out last year that I had a completely wrong definition of lambda. So in the interest of clarity, here's a diagram of a mouse skull with four possible locations for bregma and lambda (nose to the bottom). Without looking at the caption, can you, dear reader, identify the correct locations?

Diagram of mouse skull (not to scale). The bregma, for mice, is where the coronal suture intersects the midline, point B. Lambda is defined as the intersection of the midpoint and a curve fitting the "lamboid structure" (whatever that is). Basically, it's where the rear sinus would intersect the midline if the rear sinus didn't curve, point D. It is NOT point C, as any reasonable person would guess.

Now that you know the actual definitions of bregma and lambda, you can impress your professors at happy hour, and do accurate surgeries! Except bregma, is almost NEVER that simple in practice. The coronal suture curves this way and that; the left and right sides almost never meet at the same place on the midline. Sometimes the midline curves too!

Just for fun, I've diagrammed some common bregmas you might encounter during surgery. Which of the points listed below do you think are bregma? My opinion is in the caption.

Realistic bregmas (not to scale).
Top left: bregma where the two sides do not meet at the midline. The bregma here is at point B.
Top right: bregma where the mouse's right coronal suture veers. Here I'd say Bregmas is at C.
Bottom left: bregma where the midline curves. This is a bit trickier, but I'd say point B.
Bottom right: bregma where the midline shifts. Here there is actually not enough information! I would follow the midline all the way to lambda to see if it shifts back.

Disagree? Tell me in the comments!

The (Near) Future of Cell Type Specificity

I have been growing more interested in genomics, so this quarter I took a class on new techniques in genomics.* What I learned is that the most important aspect of modern genomics is cost. Sequencing gets exponentially cheaper every year, which makes answering old questions less expensive, and opens up new experimental possibilities. For example for cancer, in the past doctors might have genotyped a tumor once, while doctors now can sequence tumors in real time to identify drug resistance.

* If you miss the days of scientific tables, I suggest you check out the supplement of any genomics paper.

So how can genomics be useful for neuroscience? Sequencing is most famously used on DNA, but can be used on RNA too (RNA-seq). This makes it possible to characterize transcriptomes using sequencing rather than use microarrays. Recently, with cost reductions and new techniques, this has enabled single neuron transcriptomics. If you believe than a neuron’s mRNA and proteins define it, this means we can characterize single neurons with RNA-seq, then look at populations of neurons to figure out how neurons can be grouped into cell types.

So today I'd like to write about what cell specificity is, review two papers on single neuron transcriptomics, and the implications that this has on how systems neuroscience looks at cell-specific circuits.

What is a cell type?

The brain is composed of billions of neurons, each of which do something different, but many of which do something similar. Given that protein expression determines what a cell does, I would define a theoretical "cell type" as a group of cells that express a similar set of proteins and encodes a similar type of information. (I say "similar type of information" since neurons of the same type could encode different versions of the same information. For example, I would say all direction selective retinal ganglion cells are the same cell type whether they encode up or down movement. If different directions express different proteins, then that would change things (which I believe may actually be true)).

While that is a simple theoretical definition of a cell type, there is also the issue of how we deal with cell types in practice. Right now, the most popular way to approach cell-type specificity is transgenic mice (while I poked fun at Nature titles last month, a majority of them were about specific circuits controlling behaviour). To understand how cell-type specificity is done, it would be useful to have an example, so let's look at how Haubensak et. al. in David Anderson's lab did it.

Haubensak et al were interested in the central amygdala, which is involved in conditioned fear. To find genes specific to the central amygdala, they performed a microarray analysis on the transcripts from the brain region (today, people would probably scour the Allen Brain Atlas). This analysis yielded a few candidate genes like PKC-delta, CRH, and enkaphalin. Once they had their candidate genes, they performed immunohistochemistry and patching to characterize the cells' expression patterns and electrical properties. When they were satisfied that PKC-delta neurons were interesting, they injected viruses containing ChR2 into the central amygdala of PKC-delta-Cre mice, and performed their main fear conditioning experiments.

This illustrates the functional definition of neuron types today: Cre-driver lines with viruses injected in a specific location.

This functional definition of a cell type is incomplete a number of ways. It ignores that multiple types of cells in a given region could express the same gene; that is, there could be two types of cells that express PKC-delta. It also ignores that gene expression changes during development (this is partially overcome by the temporal precision of viruses). And it ignores the importance of that gene for the function of the cell.

As a general rule, I would guess that the farther an "identifying gene" gets from the synapse, the less specific it is as a cell marker. The closest genes to the synapse would be neurotransmitters and receptors. For example, in the arcuate hypothalamus, there are cell types identified for the neurotransmitters AgRP, POMC, and KNDy, each of which segregate. Acknowledging that there might be subsets of these cells (e.g. those that project to different places), these cell types are probably well defined. The major exceptions to this neurotransmitter heuristic are probably the neurotransmitters glutamate and GABA, which are expressed so ubiquitously that they are non-specific. (This has not, of course, stopped people from publishing well with VGat or Vglut2 cell lines.)

Examples of cell types in arcuate nucelus as determined by neurotransmitter. The set of neurons expressing GABA includes three different subsets of neurons, an example of how GABA and glutamate are poor cell markers. 

Moving farther from the synapse, many interneurons can be identified by the types of calcium binding proteins they express (e.g. calretinin vs calcitonin). Given how important calcium regulation is, these cells might have different firing properties, and be functionally different. Even farther from the synapse, you can identify cells by secondary messengers like PKC-delta. To be honest, though, without a clear link between secondary messengers and actual function, I am skeptical whether they can truly be cell specific markers.

Besides identifying cell types based on their transcripts, you can subdivide cells based on the projection patterns as well. For example, the AgRP set pictured above is actually composed of multiple groups of cells the project independently to different brain areas.

At the moment, our ability to identify and manipulate cell types is lacking. We use brain atlases and microarrays to identify populations of cells, but don't really know the boundaries of each cell type. So how can we define them better?

Single neuron transcriptomics

My favourite answer is single neuron transcriptomics. The basic idea is that you can isolate the mRNA from a single neuron, amplify it via PCR, and then sequence it; this would give you a quantitative assessment of all the RNA that a single cell expresses. You then repeat this process on dozens or hundreds of cells chosen at random, and use clustering algorithms to see which cells have similar expression profiles, which would constitute a defined cell type.

A paper came out in Nature Biotechnology this year which illustrates the process in developing human cortex. They took tissue, dissociated it, and isolated individual cells using microfluidics. Once single cells were isolated in their little compartments, they were able to perform all the steps outlined above: lysation, amplification, and sequencing.

Clustering of cells based on single-cell transcripts. On the x-axis are individual genes. On the y-axis are individual neurons. The color of each rectangle is the gene expression in the cell. Neurons could be clustered into 4 groups. Some groups, like group 1, could be further subdivided.

From Fig. 3 in Pollen, et. al., 2014

Once they quantified expression for all the mRNA, they performed PCA to identify the most informative genes; then hierarchical clustering of cells using the 500 most PCA-loaded genes. Following clustering, they identified four major groups of neurons that corresponded to cells at different stages of cell division (see above). They were furthermore able to subdivide these clusters, like group 1, into smaller subsets at different stages of differentiation. Using this technique they found that two transcription factors, Egr1, and Fos, were more important in development and differentiation than previously realized.

While cell sorting is a great technique for developing cells, it is less useful for mature neurons which send neurites everywhere; if you were to try to shove a neuron in a cell sorting tube, you would rip off the axons and dendrites, and leak mRNA everywhere. To do single cell transcriptomics in adult neurons, James Eberwine's lab at UPenn developed a method they call TIVA (see figure below).

How TIVA works. "The TIVA tag is composed of several functional groups: biotin, Cy3, poly(A) tail binding 2′-F RNA poly(U) oligo (orange), photocleavable linker (PL), 2′-OMe RNA poly(A) oligo (yellow), Cy5, disulfide bond (S-S) and CPP."

1.) The CPP tag on TIVA allows it to penetrate cells; once TIVA is in the cytosol, CPP is cleaved. You can see which cells have taken up TIVA by monitoring Cy5 fluorescence while exciting Cy3 (Cy3/Cy5 is a FRET pair).

2.) To release the RNA binding sequence, shine light on a single cell, cleaving the PL linker. You can monitor this via the Cy3/Cy5 fluorescence ratio.

3) Once released, TIVA will bind to the polyA tails of mRNA.

4) You lyse the cell, then extract TIVA and associated mRNA via the biotin tag.

6) ???

7) Amplify and sequence

8) Profit!

Fig. 1 from Lovatt et al (2014)

They first verified that TIVA worked in cultures before turning to hippocampal slices. In slices they bath applied TIVA, and monitored its uptake in cells. (panel b, below). They then photocleaved the construct in a single cell (cell ii), and verified the cleavage using fluorescence (panel c, right). Finally, they sucked up the tissue in a pipette, extracted the cell, and sequenced. Using sequencing they were able to verify that they were sequencing neurons, as opposed to the surrounding tissue (panel d).

Single neuron sequencing in hippocampal slices.

TIVA, as presented in the paper, has a couple major downsides. First, it cannot be targeted to specific cell populations. This means that you would have to profile all the cells in a given region, rather than specific subsets your are interested in. If there are many small populations in a given area, identifying them would require large sample sizes. Second, the throughput for the technique appears to be one neuron per hemisphere; in the paper they sequenced on the order of tens of neurons. If you wanted to phenotype one hundred cells in a brain region, you would have to sacrifice fifty mice, which is a hefty burden. On the positive side, once you characterize a brain, you won't have to profile the region again.

Whatever the current caveats, I think single cell transcriptomics is the future. It will allow us to accurately profile cells entire transcriptome, and cluster then to truly identify cells that have similar identities.

The future

So what does this mean for you (or me), dear neuroscientist?

First, I would bet that whatever cell type you (or I) are working on is not actually a single homogenous cell type, but probably composed of multiple cell types. This should not stop you from doing experiments, since the experiments of today are always superseded by tomorrow; but if you encounter ambiguous results, this is likely one source.

Two, intersectional methods are going to be essential going forward. For example, the POMC neurons shown in the first figure can release glutamate, GABA, or neither. If you truly want to understand the subsets of POMC neurons, you would need to use cell lines specific for POMC and glutamate, and viruses that could recognize the combination. Experiments like this will certainly require more breeding, but will give a more precise answer. There is probably a nice faculty position for whoever can come up with an easy intersectional transgenic model.

Third, the future is nigh.There are grants on RePORTER using single-cell transcriptomics to characterize prefrontal cortex, retina and zebrafish, and the dorsomedial hypothalamus. Just in the couple of weeks I was mulling this blog post, a paper in Nature Neuroscience came out characterizing dorsal root ganglion cells. If you want to Better Know a Cell Type, now is probably the time to act.

Tl;dr: If you're a systems neuroscientist who hasn't kept abreast of genomics, just remember: single-cell transcriptomics is getting cheaper by the month, and is going to redefine our understanding of cell types in the next few years.

(If you would like more informed opinion, James Eberwine, who's lab developed TIVA, wrote a commentary in Nature Neuroscience earlier this year.)

References

Haubensak, W., Kunwar, P., ..., Anderson, D. (2010). Genetic dissection of an amygdala microcircuit that gates conditioned fear Nature, 468 (7321), 270-276 DOI: 10.1038/nature09553

Lovatt, D., Ruble, B., ..., Eberwine, J. (2014). Transcriptome in vivo analysis (TIVA) of spatially defined single cells in live tissue Nature Methods, 11 (2), 190-196 DOI: 10.1038/nmeth.2804

Pollen, A., Nowakowski, T., ..., West, J. (2014). Low-coverage single-cell mRNA sequencing reveals cellular heterogeneity and activated signaling pathways in developing cerebral cortex Nature Biotechnology, 32 (10), 1053-1058 DOI: 10.1038/nbt.2967

Deciphering the syntax of Nature article titles

We all want to publish in Nature. Papers in Nature are (supposed to be) the complete package: reliable results that show something novel; cool techniques; a famous corresponding author. And if you want to get one, you need a title that shows you are a refined gentleperson who belongs in the Nature club.

So to help you, dear blog reader, I have scoured the archives of Nature* to decipher the ideal form of Nature titles:

[research-y verb-ing] a neural circuit for [behaviour]

For example in hunger there are: Genetic identification of a neural circuit that suppresses appetite; and Deciphering a neuronal circuit that mediates loss of appetite. Or in anxiety there is Genetic dissection of an amygdala microcircuit that gates conditioned fear. Disambiguate is an underused verb here.

If you're feeling poetic, you can rearrange the elements. For example, you can try putting the neural stuff first, like The neural representation of taste quality at the periphery. Or in olfaction, there's Neuronal filtering of multiplexed odour representations.

If you are particularly concise, you can drop the verb altogether, and just combine a couple nouns with a preposition. The cells and peripheral representation of sodium taste in mice. Perception of sniff phase in mouse olfaction. Distinct extended amygdala circuits for divergent motivational states.

Under NO CIRCUMSTANCES are you to mention the brain region, molecular marker, or techniques you used to [verb] your [behaviour]. If you are studying how photostimulating AgRP neurons induces feeding, don't mention photostimulation or AgRP (Deconstruction of a neural circuit for hunger). Otherwise, you might end up in Nature Neuroscience (AGRP neurons are sufficient to orchestrate feeding behavior rapidly and without training).

And, most importantly, keep it short. If you're studying taste receptors, go for something like An amino-acid taste receptor. Stuff like Gustatory expression pattern of the human TAS2R bitter receptor gene family reveals a heterogenous population of bitter responsive taste receptor cells goes in the Journal of Neuroscience.

If you have any other examples of the beautiful Nature titles, write in the comments!

*skimmed the tables of contents

Bonus titles:
The receptors and cells for mammalian taste.
Excitatory cortical neurons form fine-scale functional networksA family of candidate taste receptors in human and mouse
Short-term memory in olfactory network dynamics.The detection of carbonation by the Drosophila gustatory system
The cells and logic for mammalian sour taste detection.
The subcellular organization of neocortical excitatory connections.
The receptors and coding logic for bitter taste
The cells and peripheral representation of sodium taste in mice
The molecular basis for water taste in Drosophila

Channelrhodopsinning: Your light doesn't always do what you want.

Two years ago, I wrote a post about the common mistakes I notice in channelrhodopsin papers. Since then, two labs have developed improved photoactivatable chloride channels for inhibition, binary logic has been introduced to neurons, and you can use one vector to photostimulate and record from a neuron type. Beyond those headline advances, though, were some smaller papers that highlight some of the pitfalls of channelrhopsin use.

As always, given that I've published exactly one paper using ChR2, take these opinions with a Churymov-Gerasimenko comet of salt.

A crumbling pillar

In that post two years ago, I laid out my Two Pillars of Channelrhodopsin: always perform negative controls (I'm still surprised that this actually needed to be said); and always pulse your light.  In particular I was critical of a paper by Kravitz and Kreitzer wherein they used the PINP/phototag technique to record from medium spiny neurons (MSNs) in the direct and indirect pathways of the striatum. To identify, for example, direct pathway MSNs, they expressed ChR2 in those cells, recorded from them, and stimulated with 1 second long light pulses. Units that responded to light with short latency spikes (less than 15ms ) were considered identified. At the time I called this a "crap criterion," because I thought the time window was too long to precisely identify neurons.

A few weeks after my blog post, Kravitz and Kreitzer published a paper in Brain Research expanding on their technique. They wrote,

Medium spiny neurons have two properties that make them unsuited to identification protocols that require high spike fidelity [ed. note: pulsed light stimulation]. First, medium spiny neurons have very low excitability [for an example of MSNs' late firing properties, see the figure below, from this review by Kreitzer], and fire at low spontaneous rates in vivo. It is therefore difficult to drive them to spike reliably and at short latencies without using extremely high-powered illumination... Second, medium spiny neurons have variable membrane potentials which continuously fluctuate between approximately −50mV and −80mV in vivo. As we cannot monitor the membrane potential in our extracellular recordings, we do not know whether the cell is close to spike threshold when we deliver each laser pulse... While it would be simpler if identical techniques could be applied for optogenetic identification across multiple brain structures, it appears that optogenetic identification procedures will need to be optimized for each different brain structure, a situation that occurs with nearly all technical approaches in neuroscience.

MSNs are really hard to stimulate. Recordings from an MSN in response to current injection. MSNs only fire an action potential after prolonged stimulation.

Since then, I have recorded from MSNs in GPR-88 mice using an optrode, and I can confirm that neurons respond robustly to long square pulses (>10ms), but not to trains of short pulses. I was wrong, and I apologize to the Kreitzer lab.

Another interesting aspect of their review was the low yield. They recorded ten mice, and 143 well-isolated single units, and from that recorded only 19 PINP units. Some of the low yield was due to the difficulty of unit identification: of the multiunits recorded, around half were light responsive. (In my test mouse, my yield was slightly higher, ~7 cells, but perhaps my criterion are less strict.)

While a yield of ~2 units per mouse seems low, if you run the numbers it makes sense. In many brain regions, half of the units you record will be from non-target neurons, and of the target population only a fraction of the neurons will express ChR2 at a high enough concentration to be usable. So in the end, only ~10-30% of the units you record, if you are doing everything right, will express ChR2. If you record 20-30 units / mouse, which seems reasonable to me, you would end up with a high of 2-9 neurons per mouse. If your brain region is smaller, these yields would drop.

Channelrhodopsin can inhibit too

While the Kreitzer lab showed how long light pulses can work, former Duke post-doc, now Baylor PI, Ben Arenkiel put out a paper that highlighted why, in general, I prefer pulsed light. They expressed channelrhodopsin in a wide variety of neuron types throughout the brain (mitral cells, cortical pyramidal cells, interneurons, and more), and patched onto them. While they were recording, they stimulated with trains of light, varying the frequency and the duration of each pulse (from 1 ms to 49 ms (or near constantly)).  They found that, as you increase the pulse duration, stimulation fidelity decreases, and you in fact can begin to inhibit neurons.

Recording from CRH neuron in the olfactory bulb. They stimulate the neuron at 20Hz with light pulses (top) or current injection (bottom); and vary the width of the pulses from 10ms (left panels) to 49ms (right panels, ~100% on time). At the short pulse width, the neurons are able to follow the 20Hz stim. With the long stimulus, the neurons do not fire action potentials, and are actually inhibited. From Herman, et. al. (2014)

The idea that prolonged stimulation could cause inhibition is not new - we've known about shunting inhibition for a while - but now we have evidence that it can happen with channelrhodopsin. More intriguing, the response depended on neuron type. Most neurons, presumably with low basal firing rates, were inhibited by prolonged light, but a subset of fast-spiking neurons were actually even more excited by constant photostimulation.

Plot of firing rate versus pulse duration for two types of neurons photostimulated at 20Hz. Most neurons decrease their firing rate at long pulse duration, like the cortical neurons shown in D. Some, however, increase their firing rate (E). From Herman, et. al. (2014)

My conclusion from this paper is the same as that from Kreitzer's review, "optogenetic identification procedures will need to be optimized for each different brain structure." You need to record from your neurons of interest in slice, or if you can't, ask those weirdo slice physiologists down the hall for a solid. You need to know how fast you can stimulate reliably, and how long your light pulses should be. Otherwise, you might be inhibiting the neurons you are trying to stimulate, or making them fire faster than you thought you were.

Spikes != synaptic release

So you've recorded from the cell type you want to stimulate, and ran the cells through their paces to choose a stimulation paradigm. You're good to go, right?

A couple weeks ago in joint lab meeting, a post-doc presented data where she recorded from a pair of neurons: one neuron expressing ChR2, and the other downstream. She stimulated at 10-20Hz for a short while, and saw something like this:

In the middle trace, the cell expressing channelrhodopsin easily follows the 10Hz train. However, in the bottom trace, the downstream cell receives a strong EPSC for the first light induced action potential, but the EPSCs get smaller as the vesicle pool is depleted, until they all fail. Then, after the stimulus is ended, there is a refractory period where synaptic activity is absent. 

This is, of course, an obvious result if you think about it. Yet I hardly ever see it mentioned. Just because your neuron can follow your channelrhodopsin stimulation doesn't mean that it's actually releasing vesicles.

So be careful out there channelrhodopsinners. Choose the right stimulus paradigm, make sure it's not inhibitory, and hope that the axon terminals can keep up with the action potentials. Your light doesn't always do what you want.

References:

Herman AM, Huang L, Murphey DK, Garcia I, & Arenkiel BR (2014). Cell type-specific and time-dependent light exposure contribute to silencing in neurons expressing Channelrhodopsin-2. eLife, 3 PMID: 24473077

Kravitz, A., Tye, L., & Kreitzer, A. (2012). Distinct roles for direct and indirect pathway striatal neurons in reinforcement Nature Neuroscience, 15 (6), 816-818 DOI: 10.1038/nn.3100

Kravitz AV, Owen SF, & Kreitzer AC (2013). Optogenetic identification of striatal projection neuron subtypes during in vivo recordings. Brain research, 1511, 21-32 PMID: 23178332

Questions of Taste

Some time ago, Neuroecology asked people in the twitterverse what the biggest questions in their field are. While I'm no longer working on taste, these are the three big questions to my mind. Take these with a rock of salt, as I'm not 100% up to date on the literature, and these are just my opinion.

0. To what extent, and where, is taste a labeled line system versus a combinatoric one?

This is the grand debate in the field. On one hand you have Charles Zuker pushing the idea that labeled lines - that each of the basic taste qualities is represented separately with its own "line" - extend from taste cells in the periphery to deep into cortex in the form of segregated hotspots. On the other hand, everyone else in the field believes that taste is a combintarial sense, which means that the lines for individual tastes begin to overlap early in the system, perhaps even in taste cells tongue, and certainly in the brainstem and cortex. (There is an interesting inside/outside dichotomy in people's opinions. Most people outside the taste field assume Zuker is correct, while everyone inside assumes he's wrong.)

I personally think the evidence right now is that there are labeled lines in the periphery which combine in the central nervous system, probably near the NTS. But the three questions below are aimed at answering this overarching question in the periphery and cortex.

1. What in God's name is going on in the taste bud?

Due to the work of Zuker, Margolskee, and others, we have identified the basic taste receptors and the cells that express those receptors. To wit, there are GPCR expressing cells ("type II") that are responsible for sweet, umami, and bitter; sour seems to be detected by PKD channels in presynaptic "type III" cells; and salty seems to be detected by ENaCs on glial-like "type I" cells. Simple enough, yes?

None of these cells form conventional synapses.

The three main types of taste receptor cells in the taste bud. Note that none of the cells forms a canonical synapse with afferent fibres.
From Chaudhari and Roper, 2011

I believe the figure above is accurate and reflects the current understanding of connections in the taste bud. Glial-like cells do not appear to have direct connections with other cells in the taste bud, and may signal via glia-like mechanisms. Receptor cells are wrapped by afferent nerves, but do not form synapses with the nerves. Rather, they appear to release ATP which activates the fibres. And pre-synaptic cells do indeed form synapses, but it's not clear whether they actually synapse directly onto the afferent fibres. There is also some cross-talk between the receptor and pre-synaptic cells in the form of serotonin and P2Y.

How the cells in the taste bud interact with each other and send information to the central nervous system will have fundamental implications on question zero. If the cells are talking to each other in the taste bud, it means the labeled lines are not independent at the first step. And, well, knowing how your sensation is transduced is fundamentally important.

2. Is bitter a single labeled line?

In 2000, the Zuker lab published two papers reporting that the T2R family of GPCRs is the bitter taste family. In Adler et. al. (2000), they asked the question of how many of these T2Rs are expressed in a given taste cell. To measure this they performed in-situ hybridization with 2, 5, or 10 probes for bitter receptors. They hypothesized that if bitter taste cells only expressed a subset of T2Rs, the 10 probe stain would reveal more cells than the 2 or 5 probe stain. They wrote,

"the demonstration that different mixtures of 2 or 5 probes detected as many positive cells as the mix of 10 suggests that each positive cell expresses nearly the full complement of T2Rs. If we assume that each receptor signals via the same pathway, and that the patterns of receptor expression delineate the logic of taste coding, these results indicate that there be limited functional discrimination between T2R-positive cells."

If this were true, one would hypothesize that our discrimination between foods comes from olfaction.

The next year the Roper lab (Caicedo and Roper, 2001) performed calcium imaging on taste buds from intact tongues, blind to cell type. They applied five different bitter compounds, at varying concentrations, and found that individual cells responded to only a subset of the different bitters. In other words, single cells were able to discriminate between different bitter compounds, which would imply each cell has a different complement of bitter receptors. This sort of result has been found in recordings from the NTS and PBN.

Receptive field of different taste cells (y-axis) to five different bitters (x-axis). Many cells respond primarily to one bitter compound (black blocks), while some cells respond to two or more bitters (blue, green, and red blocks).
From Caicedo and Roper, 2001.

Over the last decade, the Meyerhof lab has been investigating bitter receptors, and has performed two different tests that indicate bitter taste cells only express a subset of T2Rs. In Behrens et. al., 2007, they performed in-situ experiments similar to Adler, and found that going from one bitter receptor probe to two probes roughly doubled the number of labeled bitter cells. In Voigt et. al., 2012, they generated T2R131-GFP, and found that GFP only labeled ~50% of the gustducin-labelled bitter cells (gustducin is a common downstream messenger of the bitter GPCRs).

At this point, I would normally write off the bitter-is-a-single-line theory as Zuker being Zuker. Except bitter discrimination behaviour supports the idea. To test bitter discrimination, experimenters first normalize the bitterness by choosing concentrations that elicit similar aversion for each bitter. Then they ask animals to discriminate between these two bitter compounds. In both flies (Masek and Scott, 2010) and rats (Spector and Kopka, 2002), the subjects fail to discriminate between equally aversive bitters, supporting the idea that there is a single bitter labeled line.

Rats discriminating between two tastes. Rats are able to discriminate bitters (Qui: quinine; Den: denatonium) from salts (NaCl and KCl). However, when asked to discriminate between two bitters (final testing sessions), the rats are not able to discriminate. From Spector and Kopka, 2002.

These experiments are not completely clean to my mind. One might imagine that there are two ranges of bitter tastes: at high concentrations, bitters taste bad and are aversive; and at lower concentrations, bitter tastes may be less aversive, and more informative. By testing at aversive concentrations, they may be saturating receptors, and not be in the discriminatory range. These experiments, of course, are hard to design.

Personally, I lean towards the idea that bitter is a multi-lined label. The physiology is straightforward. Humans who are anosmic are still able to discriminate between tastes. Yet no one has demonstrated convincing behaviour that bitters can be discriminated without olfaction.

3. How do cortical neurons respond to mixtures of tastes? And mixtures of taste and smell for that matter?

When you eat a hamburger, you don't just taste sweet ketchup, umami meat, and sour pickles; the different taste lines merge together, and combine with the smell of the toasted bun, and non-taste oral cues like the temperature and texture of the ground meat, to form a whole. Which is a long way of saying that while "taste" is a single sense, flavour involves multisensory integration.

Unlike the previous two questions, there really isn't that much to talk about in terms of past research. A couple papers have looked at how binary mixtures are represented in the brainstem and cortex, but they've all been (rightfully) published in "specialty" journals. It is a difficult question to study: done well, one would want to record many neurons' responses to large numbers of mixtures. Hopefully someone can manage it sometime soon.

References

Adler, E., Hoon, M. a, Mueller, K. L., Chandrashekar, J., Ryba, N. J. P., & Zuker, C. S. (2000). A novel family of mammalian taste receptors. Cell, 100(6), 693–702.

Caicedo, A., & Roper, S. D. (2001). Taste Receptor Cells That Discriminate Between Bitter Stimuli. Science, 291(5508), 1557–1560. doi:10.1126/science.1056670

Chaudhari, N., & Roper, S. D. (2010). The cell biology of taste. The Journal of Cell Biology, 190(3), 285–96. doi:10.1083/jcb.201003144

Masek, P., & Scott, K. (2010). Limited taste discrimination in Drosophila. Proceedings of the National Academy of Sciences of the United States of America, 107(33), 14833–8. doi:10.1073/pnas.1009318107

Spector, A. C., & Kopka, S. L. (2002). Rats fail to discriminate quinine from denatonium: implications for the neural coding of bitter-tasting compounds. The Journal of Neuroscience, 22(5), 1937–41.

Voigt, A., Hübner, S., Lossow, K., Hermans-Borgmeyer, I., Boehm, U., & Meyerhof, W. (2012). Genetic labeling of Tas1r1 and Tas2r131 taste receptor cells in mice. Chemical Senses, 37(9), 897–911. doi:10.1093/chemse/bjs082