Striatal pathways tested

Not long ago, there was a study probing the function of the striatal circuit. Along the same lines, another recently published study reported the function of medium-spiny neurones from the direct and indirect pathway.
Medium-spiny neurones are the main output from the striatum. They project to the internal portion of the globus palidus (indirect pathway) and to the substantia nigra pars reticulata (direct pathway). In addition to the anatomical differences, medium-spiny neurones of the direct pathway express D1 type dopamine receptors while that of the indirect pathway express D2 type.
Bateup and her colleagues (2010), wanting to know the contribution of the medium-spiny neurones to the motor behaviour of the animal, designed a procedure to selectively disrupt each striatal pathway. They inserted a lox-P site in the gene of the dopamine- and cAMP-regulated phosphoprotein of 32 kD (DARPP-32). Then, they used the cre-recombinase under control of the D1- or D2-receptor promoter gene to selectively disrupt DARPP-32 disrupting the modulation of medium spiny neurons.
The corticostriatal synapse is still functional, but long-term potentiation vanishes when the DARPP-32 is absent. They also examined locomotor activity. Disrupting the direct pathway decreases motor activity, while disruption of the indirect pathway increases it. The second case is analogue to Huntington's disease. When they challenge the mice with cocaine, only the one with disrupted indirect pathway respond to cocaine by increasing the locomotor activity.
Finally, they performed a unilateral striatal leasion with 6OHDA and treated the mice with L-DOPA. The animals with disrupted direct pathway showed no dyskinesia. Thus, either “DARPP-32 might play a primary role in the generation of signalling abnormalities implicated in dyskinesia,” or diskinesia cannot manifest itself in animals with overall decreased motor activity.
At least for animals, these two papers (Bateup et al., 2010; Kravitz et al., 2010) constitute a clear proof of the striatal pathways in vivo.

Basal ganglia circuitry tested

It is often said that the circuit of the basal ganglia is just a convenient way to summarise the motor deficits but there is no proof that is real. A way to demonstrate its features would be to be able to make changes that are correlated with motor behaviour. This has been done in mice (Kravitz et al., 2010). The authors combined two techniques: expression of genes specific for direct or indirect pathway medium spiny neurones and stimulation of neurones using  light. After the standard controls (the genetic modifications do not alter the physiology of medium-spiny neurones etc.), they go for the optostimulation. Selective activation of the direct pathway increases ambulation, activation of the indirect pathway induces freezing. All of these done in vivo.
This paper should be mandatory reading for neuroscience students.

LRRK2

Four year ago, Parkinson's disease patients from families of Basque origin were found to carry a mutation which lead to the identification of the LRRK2 gene (previously known as Park8). The leucine-rich receptor kinase, dardarin, is a 286 kD modular protein containing seven modules, each of them with different function. Mutation occurred in a kinase, Roc, and Cor modules. So, it seems that some function related to these modules is altered in Parkinson's disease. But which one?

Newly generated transgenic mice give the clue. The mutation R1441G in the Ras in complex (Roc) GTPase domain produces Parkinsonian symptoms in the mice (Li et al., 2009). The mice develops bradikinesia by 10-12 months. The striatum also decreases the ability to release dopamine. To convince us of the results, Li and co-workers overexpressed the normal (WT) LRRK2 in other transgenic to find that they are normal. So the overexpression of protein is not what causes the parkinsonian symptoms in the mouse. The R1441G mutation doe not seems to cause too much damage. The authors find that dopaminergic neurones have sightly smaller soma and fragmented axons. Perhaps in the rush of publishing, they did not have time to study the function in more detail. They did study hyperphosphorylation of the tau protein. But tau is not the main feature of Parkinson disease.

So we still ignore how LRRK2 leads to Parkinson's disease. However, there have been some reports on how mutations at that position change the enzymatic activity. 

R1441C has been reported to decreases the GTPase activity (Lewis et al., 2007; Li et al., 2007) and to increase the kinase activity (West et al., 2005, Guo et al., 2007). Similar things happen for the mutant R1441G, it causes a decrease in the GTPase activity (Li et al., 2007). However, there has been a report in which none of the two mutations altered the kinase activity (Jaleel et al., 2007). An interesting insight comes from another study (Ho et al., 2008) in which LRRK2 is reported to interact with FADD, the death adaptor Fas-associated protein, related to caspase-8. Gloeckner and co-workers (2009) reported that LRRK2 can phosphorylate other kinases that mediate neurotoxicity and apoptosis. So the answer seems to be apoptosis after all.

Cadherins

The striatum is a complex structure. The origin of this complexity is its homogeneity. Structurally, the striatum can be compared to a noodle soup, mostly เส้นหมี่ full of medium-spiny neurones with a round soma resembling ลูกชิ้น and occasionally here and there, a เกี้ยวทอด, a giant cholinergic neurone.

The striatum lacks of obvious inner structure. There are no cytoarchitectonics differences that might help to design some sort of Broadmann areas. There is no evident anatomical unit like the Purkinje cell-climbing fibre of the cerebellum, of the CA1-CA3 pyramidal cells of the hippocampus. Nevertheless, the striatum ought to have some organisation. After all, the frontal cortex projects to the rostral striatum, sensorimotor cortex to the dorsolateral portion, and the parietal cortex to the caudal tail. This anatomical division is an expression of the non-motor functions of the striatum.

The patchy nature of the striatum is the closest approximation to a global structure. Here, the striatum is considered as a Swiss cheese, in which the matrix will be the cheesy part, and the striosoma will be represented by the holes. The matrix of the striatum can be revealed by somatostatin, acetycholinesterase, enkephalin, of calbinding stainings. The islands that form the striosoma can be visualised with opiate receptors, substance P, or tyrosine hydroxylase stainings. The current thinking (Gerfen, 1992) holds that the medium-spiny neurones of the matrix projects to the globus palidus lateralis forming the indirect pathway. The medium-spiny projection neurones of the striosoma, on the other hand, project to the globus palidus medialis, forming the direct pathway.

Another approach to define molecular territories in the striatum is to examine the molecular texture of the neurons. Extracellular-matrix molecules (here matrix means the outside of the cells no relation to the striatal matrix) that act as a glue sticking cells together. Examples of such molecules are cadherins, or calcium-dependent cell adhesion molecules. Hertel et al. (2008) studied the striatal expression of 12 cadherins; four true cadherins (Cdh4, Cdh7, Cdh8, and Cdh11), and eight protocadherins (Pcdh1, Pcdh7, Pcdh8, Pcdh9, Pcdh10, Pcdh11, Pcdh17, and Pcdh19). Except for Cd11 and Pcdh1 (that are exclusive to the striosome), all other cadherins are expressed in the matrix. Some cadherins had weak expression (Pcdh8), other were expressed everywhere (Cdh11), while some have high expression rostrally decreasing caudally (Pcdh7, Pcdh17, Pcdh11), while other have exactly the opposite pattern (Pcdh19). Thus, cadherins may define territories in the basal ganglia in the same way that ephrins do in the lateral geniculate nucleus guiding the axons coming from the appropriate regions of the retina. In the striatum there are also ephrins that play a role in delineating the matrix-striosome compartments.

Dendrites caught releasing dopamine

Synaptic transmission is the process by which a nerve cell releases a neurotransmitter substance to cause the excitation of a nearby neuron. Sir Bernard Katz described the details of this process in the neuromuscular synapse long time ago. When it seems that there is nothing new about a process recounted in thousand of textbooks, surprising results come from examining dopaminergic cells in the substantia nigra pars compacta (SNc).

The SNc is not part the striatum; it is located in the midbrain, more caudally. It provides the dopaminergic inputs that modulate the corticostriatal synapse by mechanism that are not completely clear. Results from the group of Laurent Venance in France (Vandercasteele et al., 2008) show surprising finding in the way that dopaminergic cells contact each other. They examined synaptic transmission between dopaminergic cells. This may sound strange, but it happen that dopaminergic cells not only modulate the corticostriatal synapse, but in doing so, they talk to each other. Recording from two cells shows that action potentials in one cell produce a hyperpolarising response in the other cell. The obvious explanation would be that they release GABA. However, when bicuculline was added, the response did not change. The authors added many other things to the pairs. Once they added the D2-receptor blocker raclopride, then the depolarisation vanished. So the transmission is clearly dopaminergic.

The mechanism of action seems to be inhibition of the hyperpolarisation-activated current Ih (also called HCN, the same current that makes the pacemaker in heart cells). The transmission between SNc neurones disappears after adding ZD7288, the classical Ih blocker. In addition, they show that membrane conductance decreases during transmission, and depolarisation of the post-synaptic cell (that would close the HCN channels) also decreases and eventually eliminates the transmission. These new findings support curious early observations (Chéramy et al., 1981, Chéramy et al., 1983) that SNc dendrites can release dopamine.

Working memory

The ability to remember things for a short time is called working memory. Some people can remember many things, while other just a few. Think about an actor playing Julius Cesar this evening and Macbeth tomorrow. He must be able to remember quite a bit, but just for the performance. Recent research indicates that working-memory capacity depends on the striatal function, in particular on dopamine. It is known that as we lose our D2 receptors, we also start to lose our mind.
Why some people has larger working memory capacity than others? The first answer comes from just looking at the striatum in healthy volunteers (Roshan et al., 2008). These authors asked eleven ladies to listen some sentences to test their memory abilities later. At the same time, they injected them with a modified tyrosine (6-[18F]fluoro-L-m-tyrosine) that can give a PET signal when converted to dopamine. As they checked the ladies in the PET scan, they found that those with larger working memory capacity have a striatum with a superb ability to make dopamine.
These observations mean that the faster you can delivery dopamine, the better your working memory would be. But dopamine is not a magic thing that produce an action by itself. It needs to bind to a receptor to produce a neuronal response. So, one would expect that a person with more dopamine receptors would produce better neuronal response at the striatum. In fact, there are two types (or classes) of dopamine receptors in the striatum. The D2 receptors produce a decrease in cAMP. i.e. an inhibitory response.
Now, it gets more interesting. There are two forms of D2 receptor; a long and a short form (due to alternative spicing). The short form is probably located presynaptically, and modulates glutamate release. It turns out that there is a polymorphism of D2 receptors. This means that my gene that encodes for D2 receptor may be different from your gene. Zhang and co-workers (Zhang et al., 2007), actually found that there are 23 places where the gene differs (single-nucleotide polymorphysm of SNP). Some of these forms favour the long D2 receptor, some the short. For example the SNP19 can have GG or GT at a particular position. The GT would favour the D2 receptor long form. A fMRI of those carrying the GT form revealed a more active caudate when the subjects have to remember things. Unfortunately these authors did not determine the memory capacity of their subjects.
The take home message seems to be the more active the dopaminergic transmission at the striatum is, the larger the working memory capacity. The bad news is that we are not all equal.

Cataloguing the striatum

How does the brain works is the result of not only the neural connectivity, but also of its molecular components. A brain region, then, may contain unique proteins that confers it a specific function.

The cDAN microarray is a technique to identify messenger RNAs (ribonucleic acid), indicating the expression of certain genes. If the gene is being transcribed, the mRNA will be there. Microarrays can detect gene expression on a large scale (i.e., detect the expression of many genes). Ghate and coworkers (Ghate et al., 2007) screen 24,000 genes to determine which of them is expressed in the striatum. To do that, they extracted mRNA from the striatum and whole brain, and converted it to cDNA. Then, they determine which genes are expressed in the striatum more than twice more than in the whole brain. They also look for genes whose expression is undetectable in the whole brain, but substantial in the striatum. Using this procedure, they found 14 genes the expression of which was restricted to the striatum. Some of these genes have been described before, such as Gpr88 and RAR-beta. But also they found new genes. For example H3076B11 is expressed exclusively in the striatum. Its function is currently unknown.

To discover new genes is very important. The next step will be to find out the cellular distribution. This can be done, for example, using the single-cell PCT technique combined with electrophysiological recordings that use the firing patters of a neurone to give a precise identification of the cell in the living brain slice. It will be interesting to know which genes are expressed in the projection neurons (medium spiny neurones). One would like to know which genes are related to the firing pattern. Do fast spiking interneurones express a particular set of genes? These questions may be answered soon. Stay tuned.