Review of Chemical Structure

As we saw last time, only a few atoms from the periodic table are used in making all biological relevant molecules. Also, by now you probably saw that only a limited number of repetitive structures are at the core of all biomolecules. Why? Hopefully we'll know why at the end of these series of lectures. Reviewing a little chemistry will help us a lot.

1. First, we been talking of arrangements of atoms and structures. What is a 'structure'?

The structure of a molecule is the result of the atoms that make up the molecule being arranged in a particular way. What is the structure of benzene? In geometrical terms, benzene has an hexagonal shape. Why are atoms arranged in a particlar geometry in the molecule?

First, think about the components that make up atoms and molecules. Positive nuclei and negative electrons. Electrons are very light compared to the nuclei. We can think of them as swarming like gnats in the space between the nuclei. What about the electron-electron repulsion? The fast moving electrons exhibit correlated movement. That means they can stay out of each other's way. The end result is that the nuclei 'see' negative charge between them, and the electrons 'see' positive charge around them. Consider a simple diatomic molecule like H₂. There will be a certain equilibrium separation of the positive nuclei. At this separation, there is optimal stabilization of the molecule - interactions between like charges are minimized, interactions between opposite charges are maximized. What happens to the energy of the molecule if we pull the nuclei apart or push them together? We can take a look at a Moorse curve to see that.

Electrons exist in quantized energy levels. Electrons that are in a particular energy level also take on a certain spatial arrangement. These are the atomic orbitals, s, p, d, as well as the hybridized orbitals, sp³, sp², sp, that should be familiar to you from general/priciples/organic chemistry. As we form more complex molecules involving more nuclei, quantum mechanics predicts the existence of a series of molecular orbitals, each of which is a 'home' for a pair of electrons. Some of these orbitals look like simple atomic orbitals, and keep electron density close to 'home base', near the nucleus. Others will place electron density in the space between nuclei, and it is these orbitals which we think of as forming bonds between the nuclei. It is the valence electrons, the ones with relatively higher energy and farther from the nuclei, which are involved in bonding. Nuclei that make four connections will generally exhibit bonding orbitals that are sp³ hybridized, with the bonds pointing toward the corners of a regular tetrahedron. Nuclei that have three connections will usually have bonding orbitals that are sp² hybridized, and directed towards the corners of a triangle. These arrangements keep the bonding electrons as far apart as possible.

3. How do we present molecular structure?

1. With molecular orbital (MO) diagrams. Physically the most realistic, not the most convenient. Little physical intuition. HOMO, LUMO, and all those other evil terms...

2. With bonding diagrams (including Lewis diagrams). This is the most common way of presenting chemical structures. ChemDraw lets you easily make very nice diagrams. You can even include a certain amount of three-dimensional information in these, using Fischer projections. We can write out all atoms, but since we deal mainly with carbon atoms, we'll use the shorthand notation (just lines, atom symbols go only for heteroatoms).

3. Using physical models, such as Darling models. These are perhaps the best, presenting a great deal of information, and letting you explore the conformations available to the molecule. BUT, you are limited to small molecules.

4. Using computer graphics. This is probably the best compromise available. It allows us to examine molecules in three-dimensions, to manipulate them interactively using a mouse or other tracking device, to rotate and scale as desired.

5. Unsolved problem - how do we deal with the VOLUME that a molecule takes up? Bonding diagrams ('stick figures') transparently convey bonding information, but do not reveal the space that a molecule occupies.

With these somewhat basic concepts, and looking at the periodic table, we can analyze the bond types that different atoms can have, as well as their strenghts.

FIrst, elements on the first groups of the table (1 and 2), like Li, Be, Na, Mg, K, Cs, etc., have only one or two valence (unpaired) electrons. We know that atoms get to be more stable if they have all their electrons paired up. Therefore, the easiest thing for these elements is to lose electrons to achieve a closed shell. They are very good at this, and most form cations relatively easy (low electron afinity). We say that they have low electronegativities. Stuff on the other end of the table (groups 16 and 17), like O, S, F, Br, Cl, etc., have almost full valence (6 or 7 electrons in outter shells). Thus, they want to chomp electrons from other atoms to get a closed valence shell. They form anions relatively easy. These ones have high electronegativities. As an aside, if we combine atoms with high electronegativity with atoms with low electronegativity we get salts (NaCl, NaBr, KCl, KBr, MgCl₂, CaCl₂), which have ionic bonds. If the electronegativities of both are not that different, but one is higher, we get polar covalent bonds, and if they are equal, we get non-polar covalent bonds.

This was moving sideways. What about moving down? The further down we go, the larger the atom gets. Inner shells (1s, 2px, 2py, 2pz) start to get filled up and new shells further away from the nucleus become the valence shells. In very simple terms, the positive nucleus has now a weaker grip over those electrons. This brings about two things. One, since electrons further away from the nucleus are less drawn to it, we may loose them more easily. That's why metals can form cations easily, and also why larger atoms (i. e., the 3d transition metals) can have more than one oxidation state: +2, +4, etc., etc.. Second, the bonds that these compounds make are weaker.

5. Why then C, N, O, P, S, and H in biomolecules?

The first thing people realized from looking at biological molecules is that Nature has used few of the elements in the Periodic Table. There is obviously a reason for this. We can explain this using the concepts we outlined above. As we've been repeating over and over, we need to have the ability to form pretty large molecules with certain stability. Also, certain 'zones' of the molecules should be able to disassemble and reassemble under certain conditions rather easily. Lets see what we can do with the elements we have in the periodic table.

With elements on either extreme of the periodic table we saw that we can form mainly ionic compounds (salts). These type of compounds don't get to be the size we need in biomolecules. Furthermore, they (usually) disolve in water... If we go to heavy atoms, we more or less have the same problem. Even if we are somewhat in the middle of the table (say, Si or Ge), their size makes their bonds weak, and polymers of these compounds are not common or unstable.

What about Nitrogen? it has 5 electrons in the outer shells, two paired up in one sp³ orbital and three for bonding in the other three. We call the paired up electrons the 'lone pair'. If we try to bring two Ns together, the negative charges of the lone pair will repel, making very weak bonds (~40 Kcal/mol). Also, N tends to make N₃ (with a triple-bond) because is a lot more favourable (~250 Kcal/mol) in a rather explosive way. If we look at Boron, it has empty valence shells (electron deficient - there are only three electrons for four valence orbitlas). Again, weak bonds unsuitable for biomolecules. If we go down (Al, Si, P, S), we already saw we get bigger atoms that make weaker bonds.

Lets take a look at carbon. It is the smallest atom that can make 4 bonds. In doing so, it fills up all of its valence orbitals. Also, since it is small the bonds it can make with other carbon atoms are pretty OK energetically (in the boundary of 90 Kcal/mol). Therefore, carbon is the only compound that can make large polymers. Since we can have up to four other atoms (either carbon or others) bound to each carbon atom, we have the posibility to branch out. These turns out to be good in terms of possible compounds - From all the compouds known to man, approximatelly 90% are carbon (organic) compounds...

6. What types of bonding do we observe in biological molecules? What is the mechanical behavior of different bond types?

We have single or s bonds, which direct electron charge along a single line that connects the nuclei. An example is the sp³ hybridization of the bonding orbitals in ethane. The most important features of single bonds are:

Molecules that include rotatable bonds are able to adopt multiple conformations. We'll see about this next time.

Then we have p bonds, which place electron density on either side of a plane of symmetry of a molecule - simple examples are the double bond (see ethylene) and aromatic rings (see benzene). In the case of carbon, they are formed with the p orbitals. We always have a s bond first (formed either by an sp² or sp hybridized carbon), and then the p bond. The most important features of these bonds:

In each case, we can represent the bond by a mechanical model. This is a representation using spring-like elements. We can also include rotational barriers in our models.

Aromatic rings are always systems that can be represented by alternating single and double bonds. Do you remember the aromaticity rules?

This concepts will be very important when analyzing both the chemical and structural characteristics of biological molecules

7. Functional groups and reactivity

Under physiological conditions, carbon only compounds are pretty stable, way to stable to be useful because they would not react with anything (both the C-H and the C-C bonds are strong, 99 and 83 Kcal/mol). Furthermore, they are really non-polar and greasy, and don't disolve well in water were everything is taking place. We need to add 'handles' to the carbon chains that will allow us (actually, enzymes in the cell) to do chemistry with them and make them have at least some water solubility. What if we used heteroatoms here and there? This will give us certain 'regions' of the molecule that will have marked chemical differences to the rest. This gives usually more reactivity to that part of the molecule. Also, it will give polar character to certain regions of the molecule. Additionally, we can even put a couple of different heteroatoms together is some places (-O-P-O- in phosphates, for example). These handles, known to you from organic chemistry as functional groups, are were 99% of the chemistry occurs in biological systems.

We will see that the same functional groups are used over and over again when we examine the structure of biological compounds. These form a chemical language that we must learn if we want to easily communicate about these compounds.

Functional groups that appear in biological systems include:
 

Hydroxyl	R-CH2-O-H
Carbonyl	aldehyde, R-CH2-CO-H
Carbonyl	ketone, R-CH2-CO-R1
Carboxyl	R-CH2-CO-O-H
Ester	R-CH2-CO-O-CH2-R1
Ether	R-CH2-O-CH2-R1
Amino	R-CH2-NH2
Amido	R-CH2-CO-NH2
Gunadino	R-CH2-HN-CNH-NH2
Imidazole	R-CH2-[-C=CH-N=CH-NH-]*
Sulfhydryl	R-CH2-S-H
Disulfide	R-CH2-S-S-CH2-R1
Phosphoryl	R-CH2-PO3H2

We can co even further, and group carbon-only chunks as functional groups:
 

Phenyl	R-CH2-[-CH=CH-CH=CH-CH=CH-]*
Alkyl	R-CH2-[-CnH2n-]-CH3

* In these one the bracket indicates that the functional group is cyclic.

Next time we'll see how some of these functional groups assemble together, what compounds they form part of, and the chemical reactions that involve them.

Prepared by Randy J. Zauhar and Guillermo Moyna, 1999.