Wednesday, 4 November 2015

pH, folding and catalysis

The recent level one class (MBB165) on the influence of temperature and pH on the enzymatic reaction catalysed by aryl sulfatase made me think about explanations of classic "physical chemistry" effects on proteins in general. School Biology and Chemistry classes introduce students to the idea of pH and temperature optima for enzymes. In the case of temperature, explanations are pretty straight forward: as the temperature of a reaction increases, so too does the rate. You will recall that the reason for this in a typical uncatalysed reaction is not simply an increase in the number of molecular collisions, but a shift in the proportion of molecules that have crossed the "activation energy" threshold. Hence there is an approximate doubling of reaction rate for every 10 degree increase in temperature. However, when an enzyme is present, the increase is a consequence )or function) not only of the activation effect, but is also a function of the intrinsic thermostability of the enzyme. So, enzymes exhibit temperature optima as a consequence of reaction kinetics and protein stability factors. 

How might you design experiments to dissect out these two components?

Can you think of an important enzyme catalysed reaction (used by thousands of labs every day) which relies heavily on thermal control?

The pH dependence of a reaction turns out to be a little more complex and you need to first consider the pH dependence of the side chains of the amino acids found in proteins. I wont discuss it here, but the same would be true for the bases found in RNA, in the case of catalytic RNA molecules, or ribozymes. That's for another time.

Amino acids and their pKas. All of the common amino acids except glycine are chiral (have a handedness). The major form found in Nature are the S-amino acid, or L-amino acids . [The mirror image of each amino acid can be found in Nature, although they are less common.] In the Table below, the third, fourth and fifth columns give pKa values of groups: -NH3 refers to the protonated alpha-amino group, the CO2H refers to the carboxylic acid group on the alpha-carbon, and the side chain is only relevant for a few amino acids. And this is a very important point. The impact of pH on protein structure and function is therefore limited to the acidic class (Glu and Asp), the basic pair; Lys and Arg and importantly, the amino acid histidine which has a "special" place in proteins, since its pKa lies in the neutral or physiological range.
In all cases below, pKas refer to the free amino acid, and not the amino acid incorporated into a polypeptide or protein chain. Usually, we assume that not much happens to the pKa of the side chain on incorporation. While this is not strictly true, it is a good starting point, until the value can be measured independently. Polar side chains such as Ser, Thr, Tyr and Cys are important as Hydrogen bond acceptors, but in order to promote their reactivity, they generally need to be "influenced" by adjacent (in spatial terms) side chains, cofactors and in particular metal ions. Perhaps the best example of the key role of a Ser residue and the sphere of influence of the active site residues is given by the Serine Proteases.

Amino Acid
Side chain
Glycine, Gly -H 9.78 2.35
Alanine, Ala -CH3 9.87 2.35
Valine, Val -CH(CH3)2 9.74 2.29
Leucine, Leu CH2CH(CH3)2 9.74 2.33
Isoleucine, Ile CH(CH3)CH2CH3 9.76 2.32
Phenylalanine, Phe 9.31 2.20
Tryptophan, Trp 9.41 2.46
Tyrosine, Tyr 9.21 2.20 10.46 5.65
Histidine, His 9.33 1.80 6.04* 7.58
Serine, Ser CH2OH 9.21 2.19
Threonine, Thr CH(CH3)-OH 9.10 2.09
Methionine, Met CH2CH2SCH3 9.28 2.13
Cysteine, Cys CH2SH 10.70 1.92 8.37 5.14
Aspartic Acid, Asp CH2CO2H 9.90 1.99 3.90 2.87
Glutamic Acid, Glu CH2CH2CO2H 9.47 2.10 4.07 3.22
Asparagine, Asn CH2CONH2 8.72 2.14
Glutamine, Gln CH2CH2CONH2 9.13 2.17
Lysine, Lys (CH2)4NH2 9.06 2.16 10.54* 9.74
Arginine, Arg 8.99 1.82 12.48* 10.76
Proline, Pro 10.64 1.95
*Refers to the conjugate acid.

In order to explain the pH dependence of enzymes we need to two major factors: (i) folding and stability and (ii) orientation and reactivity of active site side chains. These topics will be covered in great detail during your degree, but here is a taster of things to come.

Protein folding and stability is determined by well established principles of Physical Chemistry combined with the unique chemistry of biological polypeptides. As Anfinsen has discussed: the primary structure of a protein carries sufficient information to direct its folding into a uniquely active conformation. [There are some emerging caveats to this principle, but they can come later]. There are some difficulties encountered in the experimental investigation of protein folding and so despite Anfinsen's insight and choice of the "well-behaved" nucleases, detailed investigations of protein folding really became possible with the advent of high resolution NMR spectrometry. The second barrier to understanding the mechanism of protein folding has become known as Levinthal's paradox. In short, for a given polypeptide chain comprising say 200 amino acid monomer units, each monomer unit has the freedom to sample many conformations (rotation about bonds etc), and the time taken for such a polypeptide to sample all possible structures is incompatible with "biological time". Clearly, the cell has a solution and in level 3 you will come across lowest free energy states and the "folding funnel" model for protein folding. Let's simplify things here and consider the following stages of folding of a polypeptide chain emerging from the ribosome or a round bottom flask in the lab. First, the hydrophobic effect drives the polypeptide chain into several more compact forms, the acquisition of secondary structure elements often then leads to the formation of a metastable state and finally, the lowest free energy form of the functional protein is attained, usually in less than 1ms. I think it is pretty clear that whilst many secondary structure interactions involve main chain hydrogen bond donors and acceptors, the later stage tuning and stabilisation of the final structure involves side chain interactions. As a consequence, the potential for a pH influence becomes inevitable.The pH of the solvent can influence the properties of the active site of an enzymes by perturbing the ionisation of key residues. As you will have noticed, the pH profile of an enzyme is often (but not always) bell shaped. The peak of the curve represents the optimum pH for activity, whilst the descending "shoulders" tend to reflect the combined influence of pH on protein stability and activity:at extremes of pH, most proteins tend to denature. An interesting couple of examples to consider from the literature are the gastric protease pepsin and the Krebs Cycle enzyme fumarase. Pepsin is a proteolytic enzyme that is usually found at acidic pHs: interestingly as the pH of the solvent is raised from 5 to 7, pepsin begins to denature. Not surprisingly, it is an acid protease and aspartic acid is found at its active site.

Can you think of an aspartic protease that made the headlines as a potential drug target?

Fumarase was characterised in a lovely early pH study by Frieden and Alberty: two well known enzymologists. In a paper published over 50 years ago, they demonstrated the insight that could be obtained from a careful analysis of the pH dependence of an enzymatic reaction. Such studies underpin the broad view that ionisation of two or more key residues at the active site of an enzyme, is responsible for the pH profile. If we look at the active site of aryl sulfatases, not the one from the snail, Helix pomatia (the source used in the lab classes),but from a related enzyme found in bacteria (for which there is a very high resolution structure), it is immediately clear that it is stuffed full of ionisable side chains: see below.

The other interesting feature is the calcium ion at the centre. Can you think of a plausible mechanism for how such a clustering of residues could promote the removal of the sulphate group from the aromatic ring?

In summary, pH has the potential to influence protein function through structural perturbations and by modifying the degree of ionisation of key active site side chains. As you can probably tell, methods of analysis are needed that identify the contribution of all amino acids in a protein during catalysis. Such methods will come, but for now we will continue to make educated guesses to supplement experimental data.

Sunday, 1 March 2015

Molecular separation and detection: fundamental methods in Chemistry and Biochemistry

Last week I introduced the idea of using two dyes to illustrate the power of ion exchange chromatography (IEC). The dyes were used instead of amino acids, but served to illustrate how the net charge on a molecule can be used as a means of purifying (preparative ion exchange chromatography) or simply analysing mixtures of molecules, such as amino acids (LHS). Consider the 4 of the amino acids found in proteins: Glutamate (Glu), Aspartate (Asp), Tyrosine (Tyr) and Lysine (Lys). In addition to their shared amino and carboxylate groups (which in large part define their properties as free amino acids), Glu and Asp have a second carboxylate group as the side chain. Thus at neutral pH, they carry a net negative charge. By comparison, Lys is positive, owing to an additional amino group, whilst Tyr is polar, but neutral, owing to the hydroxyl group attached to its aromatic side chain. Look the structures up at the wikipedia site and satisfy yourselves that you appreciate the chemical properties of each amino acid. Forthe purposes of the lab sessions I will focus on IEC, since gel filtration chromatography (sometimes called gel permeation), elegant though it is, has limited application and affinity chromatography is specific (by its very nature) to single classes of proteins. You will also have last year's experimental report on the use of Ni-NTA Affinity chromatography for the purification of tagged GFP. See my Blog hereThe use of solvents and silica based resins in thin layer chromatography and high performance liquid chromatography, are discussed below. 

Separation of the four molecular species discussed above, presents a real challenge, but one that has been solved by applying a range of chromatography methods. A second important part of any separation method is detection. Glu and Asp differ from each other only by an a single methylene (CH2) group, but from the other two by possessing 2 carboxylate groups. Similarly, Lys differs from the others since it has two amino groups and finally Tyr has a polar aromatic side chain. Let us say that we had 4 separate samples of the 4 amino acids. How could we distinguish between them, rapidly and economically? There are several methods that can distinguish each of the amino acids, but these are not widely available in a typical Life Sciences Laboratory. Mass spectrometry will provide a clear answer, since it is possibly to distinguish between molecules differing by 1Da. Infrared spectroscopy measures differences in the vibrational characteristics of intramolecular bonding and NMR (above right) will readily distinguish between the four amino acids on the basis of differences in absorption characteristics of the protons in an applied magnetic field. One final point is that Tyr will absorb in the ultra violet region, since it has an aromatic ring as a side chain. 

TLC black ink.jpgNinhydrinAll of the above physical methods are not always available and sometimes samples may be available in too low a yield to obtain definitive results. One of the early methods employed for the analysis of amino acids was to combine paper chromatography, or later thin layer chromatography (TLC) with a range of solvents and to use the molecule ninhydrin, to provide a sensitive indication of the presence of an amino acid. Ninhydrin (top LHS) reacts with primary (and secondary) amines to give a purple colour (I think secondary amines are yellow). If all 4 amino acids are applied to a TLC plate and the plate dipped into a solvent reservoir and allowed to "develop", the amino acids can be separated owing to their differential "retention" by the stationary phase used for the separation (usually silica) on the TLC plate. The result of a typical separation by TLC is shown top right, for a mixture of dye molecules. The term TLC is used for a range of different resins and you can find a detailed discussion here. For the details of the separation of amino acids, follow this link. Briefly, separation can be achieved owing to the difference in the affinity of each amino acid for the silica and its relative solubility in the solvent. As you might imagine, it is easy to separate Tyr from Lys or Glu, but not Glu from Asp.  

On a separate note, because Lys side chains are commonly found on the surface of proteins, ninhydrin can be used in a spray to detect fingerprints. If a person presses their finger onto a surface, small amounts of protein are transferred from the dead skin. By reacting the surface with ninhydrin, otherwise invisible fingerprints are revealed in forensic analysis, as shown on the left. It should also be pointed out that ninhydrin (which is usually dissolved in acetone) is hazardous and must be used in a fume hood, with appropriate safety precautions (eg gloves, mask etc).

The most common method used for the separation of all 20 amino acids (well, 19 since technically Pro is an imino acid) found in proteins, is High Performance Liquid Chromatography (HPLC). We do have such an instrument in the laboratory, but the separation and detection of amino acids requires a specialised chromatography column and definitive detection requires the use of NMR (or one of the other physical methods). Nevertheless, HPLC is a powerful method which employs the principles of TLC in a cylindrical column, with sophisticated pumping systems which can deliver solvent mixtures precisely in order to separate small molecules on the basis of subtle differences in their affinity for the column resin and the solvents applied. See here for a detailed description.

Returning to the lab experiment, you should look up the structures of brilliant blue and safranin dyes, and decide which groups of amino acids share their net charge. Given that the column resin has a net positive charge, which dye should behave like Glu and Asp and which one like Lys? Secondly you should comment on whether the addition of 1M sodium chloride led to the displacement of the dye or dyes and explain your results. When I looked around the lab, everyone seemed to have obtained the expected results, but you must not only report and explain your findings, but you should comment on the choice of methods for separating amino acids and the limitations of the methods available with respect to the small differences in the chemistries of the amino acids. The importance of the definitive nature of the methods of detection should also be covered in your discussion. Finally, chromatography represents an excellent example of the predictability of scientific methodology. The sensitivity of the best methods of chromatography, to subtle differences in the chemical and physical properties of molecules, make it on the one hand a powerful methodology, but on the other places considerable demand on the operator in respect of attention to detail during an experiment!

Wednesday, 28 January 2015

Enzyme rates: the good, the bad and the ugly!

This week we started measuring the activity of Glutamate Dehydrogenase in solution, using the reduction of NAD+ to NADH, which gives an increase in absorbance at 340nm. Having made up the stock solutions of glutamate and NAD+, the enzyme (diluted from a concentrated suspension in saturated ammonium sulphate) in phosphate buffer at pH7.6, was added at a volume determined by you to give good initial rates. So what is a "good" initial rate? 

The Ugly! I am going to use the "Good the Bad and the Ugly" in reverse order, extending the use of a literary concept, to identify the "Goldilocks" initial rate! This is an unacceptable initial rate. Here something has gone wrong. This might present as no observed change in absorbance: you forgot to add an ingredient, say. (Familiarise yourself with the height of a 1ml volume in your cuvette!). The absorbance starts shooting up in the absence of enzyme. This can occur if your solutions (which may be on ice) create condensation on the optical face of the cuvette. And then, there is the passing increase, as a hair or piece of chocolate floats past the light beam! These represent examples of the ugly side of initial rate measurement! They are all a result of a lack of care!

The Bad! A bad rate is a rate that is either too fast or too slow to capture. I am referring here to rates measured on general lab spectrophotometers, and in our case we are using good quality specs but in a simple manual mode. Therefore you will take absorbance readings against a blank sample every 15/30s. If the rate is too sow, ie the absorbance changes by 0.001 at each reading, this is bad for productivity (but it may be good for the determination of the tangent). However, on balance we try to obtain good slopes, in a reasonable time scale ( maybe 10 minutes). If the rate is too fast, this is worse. The slope is hard to determine and the reaction is over before you have had time to write down the first absorbance reading!

Good! When the data are plotted (absorbance versus time) and the first 50% of the plot lies on an approximate 45 degree straight line, you have obtained the Goldilocks conditions! This is a balance of the substrate concentrations and the amount of enzyme added to initiate the reaction. The rate shown below was obtained by someone and is in my view satisfactory. I want to replace this with a post of your rate measurements tomorrow, to highlight the best result in the class and to improve on the one below! If you could show me your initial rates, I'll choose the best!

Saturday, 24 January 2015

We have a lovely bunch of coconuts in the lab!

I was delighted to see the natural product experimental planning taking shape this week. Last year, lemons proved the most popular source of potential ant-bacterials, but this year coconuts are topping the bill! It has prompted me to use coconuts as the backdrop to help with your experimental planning. The first thing I noticed was an overwhelming rush of enthusiasm to make the plant extracts. However, how many of you weighed the item? Did you measure the volume of the coconut (or kiwi fruit etc)? If you did, how did you do it? How did you extract the "milk"? What was the volume? How did you store it?

These were the questions that I asked several groups and the reason for asking is simple. Whenever you embark on an experiment, you must plan to record all of your observations. Before you extract the coconut milk, weigh the coconut. You should then compare (for example) 3 coconuts. Are they the same weight? What is the average weight? The same applies to the volume (think ancient Greeks taking baths!). Ask yourself why these measurements are important.

Now we come to the fluid inside the coconut: the so called coconut milk. Milk is usually defined as the fluid from a mother's breast or from the udder of a cow. It is a mixture rich in proteins and fats. So what is in coconut milk? You should consider the coconut carefully. Define its structure, consider the biological origin of its distinctive structures and then think about the source and composition of the "milk". You should then plan how you make various extracts, with reference to the chemical composition of the various parts of the "nut".

The search for lauric acid (above) was a topic of discussion with two groups. This is one of the therapeutic molecules found in coconut milk. If I tell you that chemistry is full of alternative names for the same thing (trivial and systematic), you may be surprised to find that lauric acid, which is claimed to be of therapeutic value, is very similar to the detergent we use to denature proteins, sodium dodecyl sulphate. So how can two very similar molecules have such different properties! 

What I want you all to do is to measure everything you possibly can before you commit to any destruction of the plant or fruit. Take photos and annotate the structures. Research what is known about your choice of plant and then consider whether you are likely to find water soluble or lipid soluble antibacterials....and how you might plan for both! But don't lose that enthusiasm!!!