Experimental design: getting from cloning a new gene to finding out its role

You have cloned a new gene X. Patch clamp recording from HEK293 cells overexpressing X revealed that X-transfected cells express large ionic currents when these cells are mechanically stimulated (i.e. by touching the cell with glass rod); no such currents were seen in untransfected cells. Immunostaining of rat tissues with antibodies against X revealed that X is highly expressed in a subset of large dorsal root ganglia and trigeminal ganglia neurons and also in the auditory hear cells. Design the experimental strategy to test if gene X is an ion channel, what other tissues it is expressed in and its physiological role. 

Cells that express ionic currents when one recombinant gene is implanted but no current in native cells means that this gene is likely an ion channel. The current direction gives an indication of the role of the gene as a large downward inflection would suggest a role in depolarisation. However, without further experimental evidence, a whole cell patch clamp recording cannot confirm that gene X is an ion channel. The following will discuss the techniques for identifying whether it is an ion channel, tissue expression and the function of the X protein.

Is it an ion channel and what does it conduct?

Before adding the recombinant channel, a green fluorescence protein tag sequence can be added to the gene. This would allow viewing of the plasma membrane through Total Internal Reflection Fluorescence Microscopy (TIRF) which emits light at an angle where the wavelengths are all reflected away from the specimen but some photons diffuse through and excite fluorophores only on the PM. So there will only be a strong fluorescent signal from the protein if it is present on the PM. If this was not the case it is unlikely to be an ion channel. Using native HEK cells as a control will make for easy comparison.

Ion channels have a distinct single-channel recording profile so a cell-attached patch clamp to isolate one channel could be done and record from the single channel whilst mechanically stimulating the cell. An ion channel would provide a current trace like in figure 1. 

Also this technique discounts the chance that the gene is an ion-pump because a single pump recording is too low to be detected. siRNA should be used to reduce the expression of the mRNA of the gene and then redo whole-cell voltage clamp recordings and compare the first recombinant cell recordings as a control. If the current is significantly reduced it is likely that the gene codes an ion channel. A control using scrambled siRNA can be used which as it has no effect on X-mRNA so if the current still decreases then it is not due to X-mRNA knockdown as this non-specific action is a limitation of siRNA.

Identifying the ions it conducts can begin with adding NMDG, if the current is ablated then the channel is a cation channel. If not, it is likely a chloride channel. A cation channel could be non-selective so the channel can then be treated with tetrodotoxin (Na+), tetraethylammonium (K+) or ruthenium red (Ca2+) separately, if the current is reduced by any of these then the channel conducts that ion. If all do then it is non-selective. However, not all subtypes of each cation channel are inhibited by these blockers, such as NaV1.7. Therefore, the results can be confirmed by removing each cation from solution and measuring the current.

Expression in other tissues:

Immunohistochemistry and electrophoresis can be used in conjunction to support the results of each test. Being as this is a newly cloned gene it is unlikely there is an antibody for it. One could be produced in normal animal-anti-animal technique or an epitope could be genetically introduced into the gene of transgenic species. Tissue slices are taken and stained with the specific primary antibody to bind and then a secondary, fluorescent antibody. The tissue will then fluoresce under microscope if the tissue expresses the protein. Occluding primary antibody acts as a control to show the non-specific binding of secondary antibodies.

SDS-PAGE electrophoresis denatures and adds uniform charge to the proteins of a sample from each tissue. The proteins are separated by size then radio or fluorescent antibodies are used on the film which will appear as a blot under x-ray or microscope if the protein is expressed in the tissue. The control for this will be to normalise the expression level compared to actin as it is expressed at similar levels in all tissues.

Testing the physiological role of gene-X:

It is wise to use siRNA knockdown first because the siRNA can be delivered to the dorsal root ganglion via viral mediated injection and reduce the expression levels of the gene so behavioural and physiological changes can be tested. It is especially useful to do first because the same individual can be used as the control and experimental group which reduces the variability in response and reduces the number of animals being used at this stage. If this stage produces positive results then knockout mice can be used to compare.

Flanking the gene with LoxP and having KO mice only express Cre under the NF200 promoter so the gene would only be lost in large diameter DRG neurons. After doing so it is possible to view any behavioural changes and because it is likely a mechanically activated ion channel, use tests such as Von Frey filament to measure changes to innocuous touch. Post-testing, DRG neurons can be dissected and stained to identify any changes to the neuron physiology.

Evaluation of this experimental design:

The methods being used are ones common to this type of research and yield reliable, reproducible results that are easily interpreted without the need for extensive normalisation calculations. In addition, the experimental design that is laid out uses at least two methods for collecting data so each will support the results from the other or present discrepancies that can be dealt with rather than drawing false conclusions. Also, with fairly few experiments, the tissue expression, which ion is conducted, the mechanism by which the channel is activated (mechanical) and the function can be identified. However, the specific way mechanical stimuli open the channel is not identified.

Transfection of genes is only 10% so large cultures are required for success, also GM and transgenics are expensive and have stringent guidelines for use. TIRF is also an expensive technique. Another issue is that low levels of expression in tissues are not detected by these methods. If required RNA-sequencing could be used instead to detect low levels. Cysteine linked optogenetics would be useful for identifying the channel function but can only be used if the channel is ligand activated, as this channel is mechanical, it cannot be used.

In summary, gene-X can be identified as an ion channel through distinct traces when measuring single channel currents and using siRNA to knockdown the gene to view the change in current. The ion it conducts can be identified by removing specific ions from the solution and measuring the change in current. Electrophoresis will show the tissues expressing gene-X and knocking out the gene will identify the function of the gene through the changes that occur when it is not expressed. 

Weekend Post!!

I will be releasing another post close to/over the weekend which is something a little bit different but is often forgotten about a little bit when we are reading papers and getting information and generally trying to remember all the information thrown at us. 

That is how the researchers set up an experiment to eventually reach the conclusions at the end. 

Therefore I am going to post a short piece that shows a simple experimental design that gets us from having cloned a novel gene to identifying what that gene does, what it is and where it is. 

I will include the question so people can attempt it and think of some other methods that could be used as well or instead. 

(it is not an essay about methodology which would be boring as anything for you to read, it is a short simple idea of experimental design)

Drug Treatments to Enhance the Cognition in Patients with Attention Deficit Hyperactivity Disorder

Attention deficit hyperactivity disorder (ADHD) is a cognitive deficiency characterised by problems with executive functions, specifically attentional control and inhibition. This lack of executive control leads to wide ranging effects such as a short attention span, particularly when learning and listening (Barkley, 1997). This leads to forgetfulness and a high incidence of careless mistakes. The loss of inhibition leads to hyperactivity and impulsiveness, this manifests itself as fidgeting, constant interrupting and excessive talking and movement. There are significant pathological features that are associated with ADHD, identified through functional magnetic resonance imaging and positron emission tomography. These imaging techniques (shown in figure 1) show a smaller brain overall and less brain tissue, Castellanos showed the brain to be up to 5% smaller (Castellanos et al., 1994). There are also fewer connecting fibres in the corpus callosum (Catherine, 1994), this would contribute to the difficulty in acting appropriately as there is less information passing between contralateral sensory and association areas. In addition there is evidence that the caudate nuclei are asymmetrical in a way not in concordance with those not suffering from ADHD because the right caudate has a lower volume than normal (Castellanos et al., 1994). Genetics have been implicated in ADHD and the treatments discussed later provide significant evidence that mutations in these genes play a role in ADHD development. Dopamine receptor gene for DRD4 and dopamine reuptake transporter gene for DAT1 have been shown to have polymorphisms in ADHD, treating with methylphenidate relieves symptoms by increasing dopamine present (Thapar et al., 2013). The same applies for the serotonin reuptake transporter 5HTT and receptor HTR1B (Thapar et al., 2013) suggesting gene mutations are the main contributing factors in ADHD. Premature birth and brain damage, both pre and postnatally also lead to ADHD.  

Stimulants as an ADHD treatment

Methylphenidate is a potent stimulant, known by the brand name Ritalin©, it is effective in 80% of ADHD patients and shown to noticeably improve symptoms. It is a noradrenalin-dopamine reuptake inhibitor that acts most potently on dopamine reuptake membrane transporters to block them. Methylphenidate has a pharmacophore (figure 2A) that binds to DATs at one site and binds in three ways (Volz, 2008). The nitrogen from the amine forms a hydrogen bond, the ester bond forms one or two hydrogen bonds with arginine on DAT and the phenyl ring forms a hydrophobic interaction by slotting into a hydrophobic binding pocket on DAT (Volz, 2008). Figure 2B shows dopamine entering between transmembrane domains 4 and 5 of the DAT1 transporter, this causes a conformational change which releases dopamine into the cell, binding of methylphenidate changes the confirmation of the transporter such that dopamine can no longer pass through the transporter. This prevents the reuptake of dopamine back into the presynaptic neuron, this means the concentration stays high enough to act on post synaptic receptors for a prolonged period of time.

Dopamine is a modulatory neurotransmitter involved primarily in inhibition so methylphenidate restoring adequate dopamine to the mesocortex returns the ability to attend to stimuli. It also causes increased inhibition in the prefrontal cortex, this improves the persons ability to filter out any distractions and inhibit inappropriate behaviours and therefore enhances cognition to within normal range (HuntMD, 2006). Motor problems such as fidgeting are also inhibited through increased dopamine in the nigrostriatal pathway.

In a healthy person, methylphenidate will overstimulate dopamine receptors and cause schizophrenia-like symptoms, such as catatonia from too much motor inhibition and delusions and hallucinations (Chaudhury, 2010) therefore diagnosing ADHD must be accurate. In ADHD patients, these schizophrenia-like symptoms occur mainly in overdose but can appear at therapeutic doses (Mosholder et al., 2009), but more common adverse effects are vomiting, headaches and tachycardia (Wishart et al., 2006).

Other stimulants include Dexamphetamine and Lisdexamfetamine, these are amphetamine stimulants that are used in ADHD when methylphenidate has been contraindicated. Lisdexamfetamine is the inactive prodrug administration of dexamphetamine. They act by reversing all monoamine reuptake transporters compared to methylphenidate which is selective for noradrenalin and dopamine (Wallace, 2012). These drugs also increase the release of monoamines by increasing the activity of vesicular monoamine transporter 2 (VMAT2) which is an antiporter, moving monoamines into vesicles in exchange for protons, as well as inhibiting monoamine oxidase to prolong the action of monoamines (Wallace, 2012). These amphetamines improve function in the right caudate nucleus and prefrontal cortex, benefiting both motor and cognitive deficits (Spencer et al., 2013). They show clinical efficacy in a similar number of cases to methylphenidate (Parker et al., 2013), However, are used less frequently due to worse adverse effects, including mood swings, aggression and hyper-excitability (Punja et al., 2012).

 Atypical antidepressants as an ADHD treatment

Mirtazapine is an atypical antidepressant with a tetracyclic structure and the overall function of increasing the action of serotonin/5-hydroxytyptamine (5-HT) and noradrenalin. It affects ADHD by acting against the mutations in 5-HT transporter and the HTR1B receptor mentioned above. Mirtazapine works by blocking 5-HT2 and 3 receptor subtypes, these are post synaptic G protein coupled receptors (GPCRs) and ion channels respectively. Whilst simultaneously increasing 5-HT1 receptor activity. These are pre and post synaptic GPCRs that decrease cellular responses through inhibiting the cyclic adenosine monophosphate second messenger pathways via phosphodiesterase (Anttila and Leinonen, 2001). It has been suggested that increasing the serotonin to bind to 5HT1 receptors decreases cellular response further, thus reversing the emotional dysregulation in ADHD patients, such as aggression and being antisocial (Davis and Wilde, 1996). Mirtazapine has been shown to be <40% effective in the most effected age groups and 50% effective in adults (figure 3A). 5-HT1B and 5-HT2B receptor types have very similar structures, so mirtazapine blocking one and activating the other is intricate. Figure 3B shows that the binding pocket of 5-HT2B is smaller than in 5-HT1B so mirtazapine blocks the pocket, whereas in 5-HT1B mirtazapine binds in the pocket to modulate the receptor, still leaving space for the binding of serotonin (Davis and Wilde, 1996).

In a healthy person there is no beneficial emotional regulation and it does not increase mood, so it only works in someone who has depression or ADHD (Schüle et al., 2002). However, whether administering to a healthy person or ADHD patient, it elicits the same adverse effects. One of the most significant adverse effects is serotonin syndrome, this is an excessive increase in serotonin leading to tremors, severe hypotension and hyperthermia which can be fatal (Boyer and Shannon, 2005). Other adverse effects include tachycardia, headaches and extreme drowsiness.

Mirtazapine is not the only atypical antidepressant to be used to treat ADHD. Atomoxetine is a noradrenalin reuptake inhibitor used in the treatment of depression. In addition, it is also licenced as a second line treatment for refractory ADHD, when stimulants have been contraindicated or ineffective (Ghuman and Hutchison, 2014). For example, it is used in those at risk of amphetamine addiction because it has a low potential for abuse (McDonagh et al., 2011). Atomoxetine has similar effectiveness as mirtazapine at 40% (Ghuman and Hutchison, 2014) however, there are less sedative effects and no abrupt withdrawal effects (McDonagh et al., 2011) so is a safer drug for children. Additionally, it has fewer interactions with alcohol and other drugs than mirtazapine. The precise mechanism of action of atomoxetine is unclear but it elicits most of its effects in the frontal cortex and nucleus accumbens on presynaptic noradrenalin transporters.

Summary

ADHD has very significant cognitive and emotional impairment, which leads to the loss of attention and excessive motor action such as being easily distracted and fidgeting respectively. These symptoms are associated with the pathological features of a smaller brain size, significantly fewer fibres in the corpus callosum and changes in the right caudate nucleus as shown by Castellanos’s team and the Catherine studies. The treatments for ADHD focus on reversing the symptoms and not on the pathological features because little can be done about these. Methylphenidate is a dopamine, noradrenaline reuptake inhibitor that acts primarily on the DAT1 dopamine transporter to prevent dopamine re-entering the presynaptic membrane to prolong the activity of dopamine. This leads to cognitive enhancements to normal functioning. Mirtazapine increases the action of serotonin on 5-HT1, blocking 5-HT2A and 5-HT2B at the same time. The adverse effects are significant for a child, however, the benefits the drugs have on the cognitive and emotional symptoms exceed the adverse effects.

The next post over the weekend!!

My next post will be on ADHD and will be written in a report style with full Harvard citations, it is more written like this to get those of you who are reading and want to read scientific papers to get an idea of how to read and extract information needed to formulate your own informative paper.

I will not do many of these I promise, but they will all be as interesting as my other posts!

Huntington's Disease

Huntington's disease is caused by a polyglutamine repeat which is a type 1 trinucleotide repeat disorder. A polyglutamine repeat is the adding of the CAG triplet codon which codes for the amino acid glutamate when the cell divides. The Huntingtin protein is present on the gene locus 4p16.3 and under normal conditions does still contain CAG repeats, the normal non-pathological range of repeats is 6-34 repeats of CAG. It becomes pathological at 36-121 repeats. It is inherited but also progressive so the number of repeats the previous generation has could actually increase with the next generation. It is therefore possible for the previous generation to be in normal range with 34 repeats but the next generation could have 36 repeats and develop Huntington's disease.

Lets break these ranges down a little more because there are a lot of gaps between ranges. Up to 26 repeats never causes HD, people with over 40 repeats will develop HD later in life. Those with 36-39 repeats may or may not develop HD, it is impossible to tell if or when they will get symptoms. These leaves a sizeable gap at 27-35 repeats which is the intermediate range. This range is where the person within this range does not get the disease but their offspring could have an expanded CAG repeat to a range that could or will lead to HD.

There are little to no symptoms until midlife often, onset is usually between 30 and 50 and if the repeat is high, into the hundreds, juvenile onset occurs and it is very severe. Death is usually 10 to 20 years but around 5 years if it is juvenile. Death is often due to respiratory failure but high instances of suicide are seen as well as infection at least in part due to the difficulty in being hygienic.

The huntingtin protein is a cytoplasmic protein and has its function in the brain, mostly in the cortex and striatum so these areas become most affected by the polyglutamine repeat in the huntingtin protein. This leads to a specific set of symptoms, in 90% of cases a chorea occurs, chorea is characterised by jerky involuntary movements especially in the face. Bradykinesia is a lack of movement and dysphagia is trouble swallowing. However, the main symptom is other movement symptoms that progress to an ataxia. Ataxia is a complete loss of control of most bodily movements. Loss of short term memory and a level of psychosis. There are also significant changes in personality, people can quickly change to become agitated and irritable and aggressive. Depression, mania and complete social withdrawal can also occur.

Not only is this disease inherited but it is also autosomal dominant which means only one parent needs to have or carry the disease for the offspring to also develop it. The the striatum inclusions of plagues in the cytoplasm and nucleus of neurons is rare but in HD the huntingin protein mutates at the N terminus which then aggregate in the nucleus and cytoplasm of neurons as well as in the dendritic spines. Noticeable changes to the dendritic spines appear in HD, at mid stage the GABA/enkephalin medium spiny neurons are first affected but many spiny neurons have abnormal bends in the distal dendrites and increases in branching of the dendrites. The spines coming from these dendrites increase in number and increase in size. The spines are the parts on the dendritic branches that form synapses with the presynaptic neurons. This increase could explain the involuntary, uncoordinated movements, because an action potential from presynaptic neurons activate more spines causing more frequent action potentials. In late stage HD the dendrites become truncated, leaving some pre-synaptic terminals behind and the spines decrease in density so there are less synapses along the truncated dendrite. Despite this there is still an increase in number, which suggests that degeneration of neurons in the striatum leads to a compensatory response of the neurons that are left to increase spine density and dendritic branch density.

If you remember back to the pathways post, the changes in GABA transmission make sense. The striatum releases GABA to inhibit the external globus pallidus but this does not occur in HD so the external globus pallidus uses GABA to inhibit the sub-thalamic nucleus so it cannot use glutamate to activate neurons in the internal globus pallidus which would usually be inhibiting the thalamus vi GABA to prevent the thalamus signalling to the cortex to cause movement, this is the indirect pathway so inhibiting it leads to unwanted, involuntary movements. In addition to this, there is an area of the striatum that signals to the substantia nigra pars compacta using GABA and enkephalin to reduce the amount of dopamine released to the striatum. If this GABA release declines excess dopamine will be released to the striatum, in the direct pathway this means that GABA is released to the internal globus pallidus inhibiting its neurons from releasing GABA to the thalamus thus allowing the thalamus to signal to the cortex to allow movement, but because it is excess dopamine it would lead to excess and most likely uncoordinated movements because the frequency of signal would be much higher leading to chaos where timing of a movement appear.

Having said all this, the neurons that contain enkephalin and GABA and project to the external globus pallidus (indirect pathway) are more vulnerable than the medium spiny neurons that contain substance P and GABA and project to the internal globus pallidus. This suggests that the indirect pathway dysfunction increasing involuntary movement is more involved than the direct pathway. Another interesting point is that the huntingtin protein is expressed ubiquitously throughout the body and brain yet the damage lies very selectively in the striatum. It is more selective than that because interneurons seem to be largely unaffected and why do GABA/enkephalin medium spiny neurons become altered first?

Huntingtin protein:

I have mentioned this and we are basically saying that this protein is causing this debilitating disease by mutating, then I have talked about the affects this mutation has on the neurons leading to neuropathology, but what I have not done is talked about the pathology of the protein. That is because the function of huntingtin has not been assigned (we do not know) and the pathogenesis is poorly understood. The protein has phosphorylation sites near the N terminus that regulate clearance of the protein and a sequence for ubiquitin binding to degrade huntingtin and a site for sumoylation which regulates the proteins stability and activity. There is a sequence next to this which allows the protein to associate with mitochondria, golgi and ER. There is also a site for regulating the trafficking of vesicles and phosphorylation sites for cleavage of the protein, aggregation and vesicular transport. As it is also a nuclear and cytoplasmic protein it also contains a nuclear export signal to allow it to enter the cytoplasm. So do these things translate to clues about function? The regulator of vesicle trafficking and the protein having an association with HYP-C suggest it has involvement in intracellular trafficking and retrograde axonal transport (transport of things along the cytoskeleton back to the nucleus). It has association with proteins involved in vesicle endocytosis and membrane recycling plus it is present in synaptic vesicles, so probably has involvement in these two processes. Finally, it has been shown to inhibit acetyltransferase activity which decreases the level of histone acetylation at H3 and H4 when it its aggregated form. Acetylating histones loosens their coil on DNA allowing for faster transcription, so huntingtin could be involved in reducing the rate of transcription (histone deacetylase inhibitors reverse their deacetylation of histones).

Evidence suggests that mutated huntingtin proteins aggregate and then become involved in disrupting transcription. It was first shown that mutated huntingtin directly uncoupled the Sp1 promoter from its TAF4 coactivator by binding to TAF4 thus stopping transcription that is initiated by Sp1 promoters. After this it was shown that TAF4 is present in TFIID which interacts with CREB. Interfering with CREB interactions would reduce the transcription of a wide range of protein mRNA in neurons, therefore interfering with CREB will usually cause neurodegeneration. It has been shown that PGC-1alpha is heavily downregulated by mutant huntingtin in early HD and it also has a CRE site for CREB. PGC-1alpha is a transcriptional coactivator that regulates oxidative phosphorylation in mitochondria. Therefore, interfering with the transcription of this coactivator leads to mitochondrial dysfunction. Striatal neurons seem to be very vulnerable to mitochondria dysfunction because decreased striatal metabolism is seen years before HD symptoms appear. PGC-1alpha inhibition limits striatal neurons ability to respond to the metabolic demands needed to keep responding to movement signals, they may be more vulnerable because constantly inhibiting unwanted movement requires lots of ATP through oxidative phosphorylation in the mitochondria. The lack of ATP available causes neurons to lose their function in the brain and cellular functions are lost that require ATP eventually leading to neuron death (neurodegeneration). In addition to this there are direct toxicities from mutant huntingtin, it targets processes such as axonal transport which require high ATP levels and mutant huntingtin with a higher number of repeats also changes the depolarisation of mitochondrial membranes. Further evidence is shown in postmortem studies of HD patients as the expression of PGC-1alpha in stiatum is 30% lower than normal but no significant change was found in the level in the hippocampus or cerebellum.

Some good news from this is that lentiviral injections of PGC-1alpha into the striatum of transgenic mice and found that neuronal volume was 27.8% higher in the striatum of those injected over those not injected. Suggesting some neuroprotective effect of PGC-1alpha and use as a potential treatment.

Whether mutated huntingtin is a true histone deacetylase remains to be answered but treatment of HD with HDAC inhibitors in mice is showing a countering affect to mutant huntingtin and it certainly has some sort of involvement in inhibiting transcription. The drug Selisistat has found to be safe and tolerable in phase II trials but as far as I can tell is still in phase III, I will update this if I find any information.

Current treatment: 

There is no cure as of yet and neurodegenerative disorders seem nearly impossible to cure so far but it is always just a matter of time. The main treatments are for the psychiatric symptoms and tetrabenazine is used for the chorea to control the involuntary movements. it is unknown how tetrabenazine reduces chorea but it is believed to be because it is a reversible inhibitor of vesicular monoamine transporter 2 (VMAT2) which means monoamines, importantly dopamine, but also noradrenalin, histamine and serotonin cannot be loaded into synaptic vesicles so there is a lower release of monoamine to the post synaptic membrane meaning fewer action potentials. This is a pretty good hypothesis to me, it just has not been proven.

Antidepressants are used for depression and haloperidol has been used for the psychosis and hallucinations, however chlorpromazine or sulpiride are more used now. These may have some effect on the movement disorders because they antagonise the D2 receptor so have a duel effect, this is why they are more used than the better drug clozapine because it is a D4 antagonist so only treats the psychosis. Lithium ca be used to prevent the mania/mood swings and benzodiazepines will help with any anxiety and excess movement because they are sedatives.

Future therapy:

I am going to mention an important future therapy target for HD but the diagram shows a few more that are also viable.

Neurotrophic factors- The first is brain derived neurotrophic factor (BDNF), in HD context BDNF usually get released to the striatum by cortical neurons to aid in neuronal survival in the striatum. BDNF is also released in the nigrostriatal pathway and in HD the levels of BDNF are reduced compared to normal, probably due to reduced transcription (CREB is needed at promoter site 4). It is shown that the protein HAP1 interacts with BDNF but the level of interactions is reduced in the presence of mutated huntingtin. Huntingtin has been shown to facilitate BDNF transport and mutant huntingtin stops this facilitation. Additionally, BDNF elicits its effects through tyrosine related kinase B receptor (TrkBR) to have effects of transcription. In HD the number of TrkB receptors are downregulated, it is unclear how mutant huntingtin downregulates TrkB but the TrkB gene is under CREB regulation. It seems to be a future therapy with studies using adenoassociated viral injection to locally increase BDNF levels in the striatum and it has shown to be neuroprotective and and aid motor function, but this has been known over 10 years. Annoyingly, in 2014 new evidence came to light that showed no decrease in the level of BDNF produced in the cortex and the level received by the striatum is normal and the abundance of TrkB is normal. However, the same study also shows that the TrkB signalling was impaired rather than it being downregulated. Whay was signalling impaired?

Well BDNF also acts on a receptor known as p75NTR which directly acts on GDP dissociation inhibitor which inhibits the action of GTPase (RhoA). RhoA also activates associated kinases to phosphorylate PTEN which is a phosphetase that dephosphorylates the ring of phosphoinositide (PIP3) to stop its signalling cascade. PIP3 is part of the TrkB signalling so its inhibition could explain the TrkB signalling impairment in HD. Inhibition of the p75NTR receptor, The RhoA associated kinase and PTEN separately all restored TrkB function and synaptic plasticity in HD mice. There was no upregulation of p75NTR so the effects must be due to increased BDNF binding to the receptor or upregulation somewhere else. As it turns out, PTEN expression is much higher in HD mice, it is unclear why but it may be due to mutant huntingtin altering the proteins trafficking causing higher concentrations. The therapy is moving towards inhibiting PTEN and/or p75NTR.

New Post Alert

I will be releasing a post on Huntington's disease on Wednesday this week, so look out for that, it is interesting and hopefully will help anyone who knows a sufferer to understand a bit more about it and the things that can be done.

Parkinson's: Therapies

I am going to start this with a table just showing what I am going to cover in the green column and I will probably mention what is in the blue column. The red column is just there for anyone who wants to help someone with PD as some hings that can be done to help keep the person moving and healthy.

As you might expect, there is a timeline for what drugs to use at what stage in the progression of PD. Levodopa used to just be the treatment for most of the disease, but now it usually begins with MAO-B inhibitors alone, once these begin to fail, Amantadine, then COMT inhibitors or MAO-B inhibitors are used in combination with Levodopa. In late stage PD, dopamine agonists are used, often still with L.DOPA even though this is much less effective at this point. Anticholinergics are used to reduce hand tremor for much of PD but they are used late stage also to aid the dopamine agonists.

Monoamine oxidase-B inhibitors (MAOIs):

There are two types of MAO, A and B, A are more effective at metabolising serotonin and B are more effective at metabolising dopamine and noradrenalin. Drugs such as selegiline are selective for MAO-B and inhibit its action in the neuron and astrocytes to prevent the metabolism of dopamine into homovanillinic acid by covalently binding to the MAO-B enzymes to stop them catalysing reactions in both pathways needed for homovanillinic acid production. Selegiline enters the active site of MAO-B where it is converted into its active form but this form covalently binds to the active site of the enzyme during the catalytic phase preventing dopamine entering the active site. This action is irreversible so new MAO-B must be synthesised to break down dopamine so more dopamine is available to act on receptors. Some evidence shows MAOIs as antioxidants thus making them neuroprotective too which means it should slow the loss of dopaminergic neurons. This drug is usually used as a monotherapy to begin with and then combined with levodopa.

COMT inhibitors (COMTIs): 

I am going to talk about these before L.DOPA because they act on the same two pathways of dopamine metabolism as MAOIs but they act on a different enzyme, catechol-O methyl transferase, which also converts dopamine to homovanillinic acid. These are newer drugs, the most common of which is entacapone. COMT catalyses the transfer of S-adenosyl-L-methionine residues methyl group to a phenyl group in catechol structured molecules, such as dopamine. Entacapone selectively and reversibly inhibits COMT to stop the metabolism of dopamine. This means more dopamine is available and active at the synapse and it is active for longer. The idea of COMTIs and MAOIs is to counter the loss of dopaminergic neurons by increasing the dopamine available in the neurons that are left. I just mentioned that COMT acts on any catechol structure so not only does it act on other catecholamines like adrenaline but L.DOPA (the drug and the chemical that already exists in the body) also has a catechol structure. This precursor to dopamine is either converted to dopamine by decarboxylase or converted to 3-O-methyl DOPA by COMT, stopping this pathway is an extra effect of COMTIs because it means that L.DOPA (and the drug) is converted to dopamine not methyl DOPA. This is an added advantage of this drug and explains why it is more effectively used in combination with levodopa.

This image shows what I am talking about when I say COMT and MAO-B is used in both pathways to convert dopamine to homovanillinic acid, they are just used at different stages.

Levodopa (L.DOPA):

This is the most common and well known drug used to treat PD. The reason the precursor drug is used and dopamine not just given to patients is for one simple reason. Dopamine is a large molecule and the blood in the brain is almost completely sealed off from the brain itself. It is a phenomenon known as the blood brain barrier and it stops all but essential molecules entering the brain. Its evolutionary function is to stop pathogens and toxins etc... entering the brain and killing us. As I said, dopamine is to large to enter by diffusion and it is not an essential molecule because the brain makes its own dopamine so it has no transporter. However, L.DOPA is an essential molecule because the brain needs it to make dopamine, therefore it does have a transporter in the blood brain barrier so can enter easily.

There are some problems and I am going to point some out before going into the mechanism of action. Levodopa enters through the intestine no problem so is orally bioavailable, however because it is a dopamine precursor it is a monoamine and MAOs are abundant in the intestine walls, so 90% of levodopa is metabolised by MAOs. So 10% of the dose has entered the blood and of that 10% only 10% enters the brain because the other 90% is either metabolised in the blood or enters peripheral cells because dopamine is used throughout the body. This poses a further problem, the peripheral cells using dopamine are fine, so there is now excess dopamine in the periphery acting on receptors which causes vomiting and anorexia as well as many other side effects. The good news is that most dopamine receptors in the periphery are D2 receptors and a drug called carbidopa inhibits D2 receptors but is incapable of crossing the blood brain barrier, therefore it limits the peripheral effects of levodopa without compromising the central effects.

The mechanism of action is basically to put the precursor in the dopaminergic neurons and let the cell use it in its normal functioning. Levodopa enters the dopaminergic neurons in the substantia nigra that remain and is converted to dopamine via dopa decarboxylase. The overall effect being the same as the other drugs mentioned, compensate for the loss of neurons by increasing the output of those neurons that remain. If used too early it can lead to excess dopamine which is the believed cause of schizophrenia, so levodopa can cause psychosis but clozapine is an antipsychotic that inhibits D4 receptors, reversing the psychosis without effecting the treatment because motor movement is controlled by D1 and D2 receptors.

In the first image, amantadine is mentioned, many of you will have herd of amantadine and not know why and some of you will know that amantadine is an antiviral drug used to treat flu. It is used in PD because it increases the release of dopamine from vesicles in the synaptic cleft, because well no-one really knows and it acts on M2 proteins in flu viruses so there is no link there either but it works because science. It is useful in early stages but can be combined with L.DOPA which amazingly has showed to increase the effectiveness of levodopa to 79% of people responding to treatment up to late moderate stages. Statistically, it is significant at 0.05.

Dopamine agonists: 

During late stage PD all of the above become in effective because there is only so much you can increase the output of one neuron and eventually the neuron loss is so extensive that those that are left cannot increase their levels to counter that loss. So a different approach to just increasing the output of neurons is needed.

That approach is to activate the dopamine receptors in the striatum largely independently of whether the post synaptic membrane is being innervated by a dopaminergic neuron from the substantia nigra or whether that neuron has been lost. To do this, dopamine receptor agonists are needed, preferably ones specific to either D1, D2 or both (to reduce the chance of psychosis and to limit the areas the drug is activating. Earlier I said that dopamine cannot be used because it is too large to cross the blood brain barrier, agonists are just molecules that activate the receptors, they do not have to be the same or even very similar to dopamine. The agonists used are Bromocriptine which is selective for D1 and D2 and crosses the BBB and pergolide which is D2 selective and crosses the BBB. As you might expect they have similar side effects to levodopa due to the periphery but carbidopa will also help here. Their mechanism of action is just to mimic the action of dopamine in its absence.

Anticholinergics:

These can be used throughout PD for those with a hand tremor, benzhexol is very effective at reducing the hand tremor and they work by inhibiting the muscarinic receptors at the neuromuscular junctions. This is because acetylcholine release to the muscle at the NMJ is what initiates muscle movement. The lack of dopamine acting on D2 receptors in the in the striatum reduces the inhibition of unwanted movement which translates to increased acetylcholine which is enough to cause a tremor in the hand. Antimuscarinics block the muscarinic cholinergic receptors on the muscle to reduce the unwanted movement.

In late stage PD anticholinergics are useful because they inhibit cholinergic neuron activation in the striatum which opposes some dopaminergic action. This means a lower dopamine concentration is needed to elicit the same effect as the usual concentration of dopamine due to the lack of opposing influence of cholinergic neurons.

This diagram shows simply the mechanisms of the drugs above to just give a visual idea of what I am talking about. for the most part it is accurate but it is worth noting that COMTIs also do what MAOIs do and that MAO-B is mostly a neuronal enzyme so should be inside the neuron not in the extracellular matrix.

Deep brain stimulation:

I am not going to go into too much detail on this but in essence it causes action potentials in groups of neurons in the area of the electrode to cause a response, or it can hyperpolarise so in PD it potentially hyperpolarises the striatum to reduce the tremor and other movements. But honestly, so far there is no one clear or believed theory of how DBS works so I am not going to reel off  bunch of equally refutable hypotheses that may or may not be correct. Although, the results show that fully understood or not, it certainly works.

This will the most amazing thing you will have seen all day, probably the most amazing thing for quite a while. This is an amazing scientific and medical achievement not only because of how well it works but also the simplicity of use and that it is not just someone lying there with wires coming out of their head, which is sometimes peoples view of DBS but it is allowing people to have a relatively normal life. The difference really is staggering. Link just in case: https://www.youtube.com/watch?v=mO3C6iTpSGo

That is it for Parkinson's disease.

Parkinson's disease: Theories on the cause

In a general sense PD occurs due to a gradual and progressive loss of dopaminergic neurons specifially in the the substantia nigra which in turn causes a depigmentation of the area because the neurons here cause the area to appear black. The difference can be seen clearly with the naked eye. In addition to this, extensive gliosis occurs which is the infiltration of glial cells, mostly microglia into the area. This happens in areas of damage and neuron loss in a repair attempt and acts as a neural inflammatory response. These are the main hallmarks of PD as well as the always present Lewy bodies. These plaque balls exist inside the nucleus of the neuron (so are not the same as amyloid plaques which are extracelular). There are occlusions consisting of mainly alpha-synuclein and ubiquitin as well as other intracellular proteins that aggregate as the neuron begins to die.

Alpha-synuclein is not a very large protein, only 140 residues. It is generally unfolded, but normally randomly folds to a coil due to its hydrophobic domain. This occurs due to a lysine, threonine, lysine, glutamic acid, glycine, valine (KTKEGV) repeat in the N-terminus domain which leads to the formation of 2 alpha helices. These features suggest it is a membrane bound protein. The hydrophobic domain causes oligomerisation of multiple copies and converts the structure to anti-parallel beta sheet once aggregated. C-terminus has a serine that if phosphorylated, changes the hydrophobic nature and distribution of charge which contributes to its oligomerisation (this makes up a lot of each Lewy body. The significance of Lewy bodies suggests that it probably is not just general damage due to them leading to PD. So what is it?

It could be because alpha-synuclein regulates apoptosis via 3 other proteins: 14-3-3 which regulates cell viability, BAD which is pro-apoptosis and Bcl-2 which is anti-apoptosis. BAD will bind to Bcl-2 to inhibit it and allow apoptosis to begin. (apoptosis is controlled cell death, something your body does at least millions of times a day, but much less in the brain, in general it is safe and needed but unregulated apoptosis is not good).

If BAD is phosphorylated at a serine residue then 14-3-3 can bind to inhibit BAD, thus disinhibiting Bcl-2 to prevent apoptosis. Alpha-synuclein will interact with 14-3-3 to prevent it binding to BAD thus promoting apoptosis. However, this theory only explains PD if alpha-synuclein is being overexpressed. There is evidence that mutated alpha-synuclein is more likely to lead to PD due to incorrect folding. In addition, the aggregation of alpha-synuclein may interfere with ubiquitin which would lead to oxidative damage and therfore cell death, this may explain the presence of ubiquitin in Lewy bodies. Ubiquitin proteins have many roles so if this is the case then it could be many things. There is also the possibility of oxidative stress directly due to alpha-synuclein or oxidation that is independent of these mechanisms or the cause of these mechanisms. (e.g. oxidation could change the expression of alpha-synuclein).


Other potential cell death mechanisms:

Having already mentioned it lets talk about the theory of misfolding alpha-synuclein  before going into the details it is important to note that this theory gives an explanation as to how cell death spreads to other neurons. In the above theory the overexpression would have to occur in all neurons but not start until around 55 (this highly rules out just a genetic answer) but the misfolding theory suggests that the misfolded alpha-synuclein leaves the neuron in vesicles and enters other neurons via endocytosis. It acts like a virus in a way, infecting more cells which is why the disease progresses rather than alpha-synuclein just being overexpressed in all cells which would cause the patient to go from fine to late stage PD. So this theory so far is more attractive. The theory states that misfolding causes the protein to be non-functional and hyper-aggregate which is what causes the formation of Lewy bodies, which then transfer to nearby neurons.

Protein misfolding is not as uncommon as it sounds with an estimated 1 in 3 translated, folded proteins are incorrect, whether it be in folding or damaging events after folding there are many opportunities for misfolding. For the most part molecular chaperones called heat-shock proteins can refold proteins or mark them for proteolysis. Usually ubiquitin tags targeted proteins forming a polymer that can be degraded by 26S proteasome. The misfolding of synuclein can form pore structures for other Lewy bodies to exit and 'infect' other cells. In PD synuclein aggregates heavily which can make the ubiquitin-proteasome pathway redundant because it is too aggregated to enter the proteasome pore to be degraded but cells have another system in place to prevent these aggregations causing damage. It is called macroautophagy and is a process of an autophagosome engulfing the aggregated protein and entering the lysosome via endocytosis where the whole thing is degraded. Autophagosomes are large enough to internalise whole organelles so can deal with large aggregated proteins but in PD the synuclein aggregates to a point that neither process works which causes either necrosis or apoptosis, leaving Lewy bodies behind. Once in the ECM they attract microglia and inflammation which causes more damage and increases symptoms.

ubiquitin-proteasome pathway: I have spoken a little about this just but I want to go into ubiquitin. Ubiquitin has 3 enzymes, E1 which is an activating enzyme, E2, a conjugating enzyme and a ligating enzyme- E3 also called parkin protein. This protein catalyses the oligomerisation of ubiquitin and proteins that need degrading. There are a few missence mutations that can appear in parkin that lead to misfolding of the enzyme preventing it from interacting correctly with ubiquitin and change its solubility leading to progressive aggregation and toxicity. These mutations are often described as recessive disorders but increasing evidence suggests only one allele needs to be mutated to massively increase the risk of the PD. This certainly shows some evidence for it being at least a cause of PD, in addition, evidence suggests that an increase in age naturally decreases parkin solubility causing its aggregation, it would also mean that the synuclein aggregates cannot be degraded in the proteasome pathway because the E3 cannot ligate the ubiquitin. So this may be beginning to provide some insight into the formation of sporadic PD, something we still know very little about.

Oxidative stress and mitochondrial dysfunction: These are two separate mechanisms but very closely inter-related. Mitochondria are the primary cause for the generation of reactive oxygen species in cells. ROS are formed mainly through electron passage through complex I and III and the main ROS are superoxide radicals, which are one electron being added to oxygen during oxidative phosphorylation. For the most part superoxide dismutase converts it to hydrogen peroxide which is then detoxified by catalase. Unfortunately, is iron ions are present, hydrogen peroxide can be converted into hydroxyl radical due to a Fenton reaction and this molecule is highly reactive and causes massive damage in cells. Low ATP production corresponds to increased ROS level and it seems that postmortem the mitochondrial complex I has very low activity in the substantia nigra neurons suggesting a cause of PD to be a deficiency in mitochondrial complex I leading to ROS which cause cell death. It has also been found that the catalytic subunits of complex I contained oxidised proteins, correlating to decreased electron transfer, thus low ATP, increasing oxidation through complex I. this is an interesting loop and is possible, but it poses a chicken or the egg debate, how did the first oxidation occur to start this? Another question to ask is how PD mainly affects dopaminergic neurons when nearly all cells contain mitochondria.

An important thing to note is that mitochondria are more bacterial than they are eukaryotic, there is a substantial body of evidence that suggests that they are in fact prokaryotic organisms that entered eukaryotic cells millions of years ago and act in a symbiotic relationship because they produce ATP at a rapid rate and in huge numbers and they gain protection and nutrients from eukaryotic cells. This is supported by the fact their DNA is circular as in bacteria and mitochondria DNA only comes from your mother, not your father because they reproduce independently of the rest of the cell and carry their own DNA so the mitochondria from the egg reproduce rather than having genes from both parents mix to create it, which is the case with all other organelles.

Mitochondrial DNA encodes 13 proteins which are all subunits for ATP production mechanisms, it also encodes tRNA and rRNA for the protein synthesis of these genes. Mitochondria are not very big and much of the space is taken up for its function so this means that any ROS generated are always in close proximity to DNA. The close proximity accompanied by the lack of DNA protection (nucleotide excision repair, replaces damaged DNA in each cell at least 10,000 times a day but is not present in mitochondria) and no protective histones means mutations are possible and the presence of ROS makes mutations even more likely. The mutations in substantia nigra are often deletions and this type of mutation is not seen in the mitochondria in other cells like pyramidal cells, even in aged brains, suggesting that these deletions are specific to dopaminergic neurons which may then increase their susceptibility to oxidative stress, thus potentially explaining why dopaminergic neurons in the nigra are lost and not other neurons in other areas.

The final Parkinson's post will be on some of the treatments.

Apologies

Sorry for the hiatus, with bank holiday and working last week I have been very busy. I am working on the next Parkinson's post and because I will be back at uni very soon, I am going to endeavour to do one topic a week from now on, so only one or two posts but hopefully I will get through quite a few things.

Parkinson's Disease: Pathways

Before discussing Parkinson's as a disease I must first explain the role of dopamine specifically in movement more so than memory but memory will come into it a little later and will make sense if you remember back to the synaptic plasticity posts.

Dopamine is an important biogenic amine present in the hippocampus but most the dopaminergic are most abundant in the pars compacta of the substantia nigra. This area of the brain stem is a collection of nuclei rich with melanin which is why it appears black (this is also why the retina is black). Anyway these dopaminergic neurons project to all over the brain especially the hippocampus and the rest of the limbic system in the mesolimbic pathway, this pathway controls an animals reward pathway so high dopamine in this pathway is a good thing. The mesocortical pathway projects to the dorsolateral prefrontal cortex which is involved in higher brain function including emotional control, motivation and higher cognitive function, so as you might expect, more is better but regulation still needs to be tight, interestingly this pathway begins at the ventral tegmental area as does the mesolimbic pathway and this area is dopaminergic and very near the substantia nigra . Another other main pathway for dopamine is the nigrostriatal pathway. This is the pathway from the substantia nigra pars compacta to the striatum.

The nigrostriatal pathway consists of two sub-pathways, the direct and indirect, The direct pathway is responsible for facilitating movements that have been instigated, so movements we want, whether it be kicking a ball or unconsciously retracting the arm after touching a hot kettle. The direct pathway begins in the substantia nigra and the dopaminergic neuron enters the putamen and then the caudate nucleus where it synapses with GABAergic neurons and substance P neurons which innervate the globus pallidus internal segment and the substantia nigra pars reticularis. The GABA and substance P neurons contain D1 dopaminergic neurons in the post synaptic membrane which are excitatory through the activation of adenylate cyclase and therefore cAMP and also increase phospholipase C level which leads to calcium release and therefore an action potential.

Activating the GABA receptor leads to an inhibition of GABA neurons in the globus pallidus internal segment and pars reticularis. These GABA neurons usually innervate the ventral anterior and ventrolateral nuclei in the thalamus. These nuclei signal to the motor cortex to say it is okay to make the movement which the cortex turns into movement. So at rest the GABA neurons from the globus pallidus and pars reticularis inhibit the thalamus to stop movement. But in initiating movement the GABA neurons are inhibited by the GABA neurons from the striatum that dopamine activated. Thus allowing the thalamic neurons to signal to the cortex via glutamate. It is in effect inhibition of an inhibition, this is known as disinhibition.

This video explains the direct pathway in basic terms quite well

So far this is the best diagram I have found to explain it but it all takes some getting used to.

The indirect pathway still has glutamate entering the striatum which allows GABAergic neurons to fire to inhibit the globus pallidus external segment preventing it from releasing GABA to the sub-thalamic nucleus meaning it is active to release glutamate to the globus pallidus internal segment. This activates it to release GABA to inhibit the ventral tier nuclei so the motor cortex cannot initiate movement. Dopamine inhibit this pathway by binding to D2 receptors in the striatum to stop GABA release to the external segment meaning this is active to inhibit sub-thalamic nucleus via GABA. This means no glutamate release to the internal segment meaning it cannot release GABA to the thalamus so it is not inhibited and therefore can signal to the motor cortex allowing movement to be initiated.

In Parkinson's disease these pathways are defective due to a loss of dopaminergic neurons in the substantia nigra pars compacta, meaning that the GABA neurons in the direct pathway have the D1 receptors not being activated by dopamine so the GABA cannot be released to inhibit the internal global pallidus so the thalamus is not disinhibited and so no movement occurs. The lack of dopamine disinhibits the indirect pathway so movements are inhibited. This is akinesia, a serious lack of movement and difficulty initiating movement.

One of the main symptoms of early Parkinson's is a resting hand tremor, treating Parkinson's by increasing dopamine rarely fixes this symptom suggesting it is not directly related to the loss of dopamine. It has been shown that serotonin deficiency in the red nucleus could be the cause or contributing factor to the resting tremor, but the truth is, as of yet we are not sure what the cause is. Patients also often have an isometric tremor and this can be explained because the reduction in dopamine makes it difficult to recruit motor units to perform a movement, this is hypometria and the body attempts to correct this by multiple bursts of activity in the basil ganglia to increase muscle activity, but the bursts will cause a tremor as the movement is being executed because the movement is not flowing but appearing in bursts of activity. In a way it is like watching a movie by rapidly pressing play and pause.

That is about it for pathways and their relations, it can be a bit to get your head round or you'll be a lucky one that gets it straight away, but if not, read over it a few times whilst following the diagrams and it will become clear.

Problem with post

Apologies for the excessive delay, but the Parkinson's post has hit a snag, there is a large discrepancy in the literature and I really do not want to make a post that will confuse anyone or potentially be wrong. I am working on it and hope it will be up today, in case that is not the case I am going to write and publish something interesting today regardless.

update because it looks suspiciously like I am ignoring the blog

I am in the process of putting together a Parkinson's disease post and I will be covering some more neurodegenerative disorders and a post on ADHD and an amazing technique called Clarity. In other words, to me synaptic plasticity is incredible but to some people is boring, but it will appear as a staple in neurodegenerative disorders and pain so it was important to cover that in detail before doing the more interesting things. (I probably will not cover pain for a little while).

syaptic plasticity: post synaptic

In the last post I talked about implicit memory and the presynaptic modifications that allow these memories to exist.

In this post I am going to go into some details about explicit memory, In essence, the things you think of when someone says memory are the explicit ones, like what we call our actual memories and learning facts, being able to recall and learn. So if you read anything I write and remember just one thing, your brain will have made the following changes in a few synapses in the hippocampus. 

This is what the hippocampus looks like, inputs enter from the cortex (usually association cortices after being consolidated with all different senses) to the dentate gyrus which signal to the CA3 neurons via mossy fibres, CA3 neurons also receive direct input from the cortex. The CA3 neurons then signal to CA1 neurons via the schaffer collateral pathway, the synapses between this pathway and CA1 neurons are also innervated by dopaminergic neurons acting as interneurons.

Hippocampal neurons release glutamate so the post synaptic membrane consists of AMPA, Kainate and NMDA receptors. AMPA and Kainate receptors are sodium selective ion channels that open in response to glutamate. The opening of these channels and sodium entering will depolarise the membrane, if this depolarisation reaches +40mV an action potential ensues. NMDA receptors are different in that, when glutamate binds (along with glycine) the NMDA receptor opens however its conductance is 0 because a magnesium ion is blocking the entry, so although open it is not contributing anything to depolarisation. Magnesium ions are positively charged so the depolarisation of the membrane repels the ion pushing it out of the NMDA channel. by the time this occurs the action potential will be over. But if the next action potential is in quick succession then the magnesium cannot re-enter in time and the NMDA receptor can allow calcium into the neuron. There is more detail on this in the first synaptic plasticity post. 

The influx of calcium controls the first 2 mechanisms:

The first is the activation of Src tyrosine kinase, calmodulin dependant protein kinase II and protein kinase C which will phosphorylate readily releasable pools of vesicles filled with AMPA and Kainate receptors causing their insertion into the membrane as well as phosphorylating existing receptors which will keep them in the activated state longer. The increase in channel number means that when glutamate is released, more channels can be opened and they will be open longer, therefore the signal has been potentiated. This is immediate long term potentiation but it does not last long, only minutes. 

The second is changes to translation, because the AMPA changes cannot last long so another process needs to adapt to strengthen the synapse. The main 'synaptic plasticity' post debunks a few myths and oversimplifications mRNA and ribosomes so I will not talk about it here. Protein synthesis requires initiation factors to bind to the 5' (prime) untranslated region and other proteins such as transport factors bind the 3' untranslated region of the mRNA. Ribosomes will interact with the initiation factors to regulate the rate of synthesis and will read the mRNA code to make a polypeptide. There are three important regions in the ribosome, the aminoacyl site where the tRNA binds to the mRNA, the peptidyl site is just after which is where the amino acid on the end of the next tRNA forms  peptide bond with the previous amino acid in the sequence. Finally is the exit site which is where the tRNA detaches from its amino acid and the mRNA. Initiation factors control these areas b changing the shape to increase the rate at which the mRNA moves through the ribosome. The initiation factors are also interacting with their binding proteins and when interacting with binding proteins they cannot be interacting with the ribosome, so this acts as inhibition.

Calcium entry through NMDA receptors will activate the kinase, mTORC1 or protein kinase C which phosphorylate eIF4E binding protein which prevents it binding to the eIF4E initiation factor. Stopping this interaction means that the initiation factor binds to the ribosome for longer thus increasing the translation of proteins such as AMPA receptor subunits which will keep synaptic strength high until transcription increases, which is the long term change.

So this last mechanism is changes to translation, and it is largely the same as in pre-synaptic, but it is activated differently.
I mentioned earlier that the schaffer collateral pathway innervates the CA1 neurons along with dopaminergic neurons. Well when signals are coming in rapid succession (repeating something increases the chance of memory) the dopaminergic neurons release dopamine to the CA1 neurons where they bind to either D1 or D5 receptors which are Gs protein coupled receptors so they cause the increase in adenylate cyclase, leading to an increase in cAMP levels which will activate rotein kinase A by binding to the regulatory unit to release the catalytic component. The PKA then enters the nucleus and phosphorylates CREB so it can bind to CRE on the promoter region and have CBP bind, this process will cause the acetylation of histones, freeing many genes to be more rapidly transcribed. Some of these genes will add to the existing synapse, so will increase the number of mRNA for AMPA receptors etc... to keep the synapse strong, but some will form a new synapse in that area with the presynaptic membrane, this synaptic growth coupled with strengthening of the existing synapse is what makes these explicit memories long lasting, sometimes for a lifetime. 

Synaptic plasticity: presynaptic neuron

The long post 'synaptic plasticity' covers it a little more generally and talks as if it is all one synapse, this was just to show what it would be like if all synapses used all mechanisms. As it is, one set of memory uses presynaptic mechanisms and one set uses post synaptic mechanisms.

Implicit memory uses presynaptic mechanisms and is an older system evolutionarily speaking. Explicit memory is a newer form of memory, using post synaptic mechanisms and is exclusively mammalian. It is exclusively mammalian because it is memory related to facts and events (semantic and episodic memory respectively) and so is a newer memory form, it also includes spatial memory. Implicit memory consists of motor memory and perceptual memory, it is habituation, sensitisation, classical conditioning and operant conditioning. If anyone ever has trouble working out if what you're looking at is implicit or explicit memory ask yourself 2 questions:

1. Is it conscious (explicit) or unconscious (implicit)- learned fear or kicking a ball is implicit. remembering where you left your keys or remembering which way to turn because you recognise the pub on the crossroads is explicit.

2. Which area of the brain is involved? if it is the hippocampus or medial temporal lobe it is explicit (hippocampus only found in newer species like mammals), if it is the cerebellum, reflex pathway (grabbing to stop falling) or the amygdala (fear and recognition (often mainly recognition of friend or foe in older species)) then it is implicit memory.

I am just going to talk about presynaptic plasticity with some images, I will then do another post for post synaptic. You are best off reading the previous post for all the information as these will just dive into the details of certain specifics. The other post will give you a more rounded understanding before reading these.

This image shows the basic action of the synapse, action potential depolarising the neuron opens calcium channels which act to release neurotransmitter from vesicles which diffuse to the post synaptic membrane, interact with receptors on sodium channels causing them to open and depolarise that neuron.

In implicit memory it is the presynaptic membrane that is altered. In the previous post I stated one alteration and explained how that was only short term memory, this would not be enough for implicit memory because it is only short term and I have said that presynaptic modifications are the only thing that cause changes to implicit memory.

This short term mechanism is the increase in vesicle fusion an therefore neurotransmitter release. What occurs is repeated stimulation of the presynaptic neuron depolarises a branch of that neuron enough that it depolarises a serotonergic interneuron. This interneuron has an output connection to the same presynaptic neuron that activated it but the connection terminates at the synapse with the post synaptic neuron (most likely a motor neuron). This interneuron is modulatory because under normal conditions it is not active, but upon repeated stimulation it becomes active and releases serotonin to the presynaptic terminal. The serotonin will bind to 5-HT4, 5-HT6 or 5-HT7 because all of these receptors for serotonin are Gs protein coupled receptors which activate Gs protein. A Gs protein once activated will increase the level of activated adenylate cyclase which catalyses the conversion of ATP to cAMP. cAMP binds to the regulator unit of protein kinase A allowing the catalytic unit to break free thus becoming active. The catalytic unit of PKA phosphorylates calcium channels which causes them to remain open longer thus allowing more calcium in. Calcium binds to synaptotagmin which allows synaptobrevin on the vesicle to bind to SNAP25 on the membrane causing vesicle fusion and so glutamate release. Increasing the level of calcium will cause it to bind to calmodulin, causing calmodulin to bind to Munc13 proteins to increase their interaction with syntaxin, increasing this interaction increases the number of primed vesicles meaning more vesicles ready t have calcium bind to syntaxin. The increased neurotransmitter release strengthens the synapse. However, phosphorylation is transient and all the activation of molecules is reversed over a period of minutes to hours.

Therefore what needs to happen is more high frequency stimulation of the presynaptic neuron to increase the serotonin release from the interneuron. When this happens it increases the level of activated protein kinase A so some of it can diffuse into the nucleus rather than going to phosphorylate calcium channels. Once in the nucleus PKA phosphorylates CREB, allowing it to bind to the promoter region at the cAMP response element site and allows CREB binding protein to bind to CREB. Once this has happened it acetylates the histones that make up the chromatin which loosens their interacting with the DNA making up the genes which makes it easier for RNA polymerase and transcription facotrs other than CREB to bind to the promoter region. This will happen in many genes which will ultimately cause the growth of new presynaptic terminals which strengthens the communication between the two neurons.

The image below shows how extensive this growth of new synapses is between a sensory (A and B) and motor (1 and 2) neuron after a memory has been formed.

This particular memory is the implicit memory type sensitisation. What has happened is in (A) and (1) an aplysia (type of sea snail) is sensing an innocuous touch (not harmful) to its siphon, it is then retracting its gill because of it (gill withdrawal reflex)

In (B) and (2) the siphon has been shocked with electricity (tactile/noxious stimulus) causing the gill to retract much further, then a few minutes later the siphon has had an innocuous touch applied (the gill retacts much further than before the shock) in (B) this process has been done repeatedly for a few days (for example) and this has caused the interneuron to release serotonin and CREB to be activated a lot which has caused the formation of many new synapses between the sensory neuron receiving the innocuous touch and the motor neuron causing the gill to withdraw. so when the siphon is just touched, many of these synapses are activated which repeatedly depolarises the motor neuron causing the gill to withdraw further than before. Therefore, fear has led to an increased response of the gill which has been remembered by having more synapses formed.

As you can see in (2) the motor neuron has increased synapse growth too, to match up to the presynaptic change. So when I said that in implicit memory on the presynaptic changes occur, technically I was lying, well sort of anyway. Because the post synaptic membrane will have small nubs coming off them, not by the active zone but in the areas close to the presynaptic neurons, and the membrane here will have proteins in them called neuroligin which extend out in to the extracellular matrix. When the growth occurs presynaptically the membrane will contain a protein called neurexin which reaches out into the ECM too. These two proteins interact which causes the post synaptic neuron to extend to that area creating a new synapse.

As an extra bit of information. Using an inhibitor of CREB stops all this growth and long term memory completely but the short term process involving PKA is unaffected.