In addition to being shorter than a single bond, the double bonds in ethylene don t twist the way single bonds do. In other words, the other atoms

Transcription

1 In addition to being shorter than a single bond, the double bonds in ethylene don t twist the way single bonds do. In other words, the other atoms attached to the carbons (hydrogens in this case) can no longer change their relative orientations by rotation because double bonds just don t undergo bond rotations. The reason they don t undergo bond rotations is that the pi bond has specific orientation requirements. In order for the p orbitals to overlap, they must be parallel. Rotation around sigma bonds can occur readily because it doesn t affect the overlap of the sigma orbitals, but rotation changes the overlap between p orbitals. When the p orbitals are perpendicular, as shown above, there is no bonding at all. Because pi bonds form when it is energetically favorable to do so, you can infer that it is energetically unfavorable to break a pi bond by rotation. So: it doesn t happen. One consequence of the need to align p orbitals to achieve and maintain overlap in a pi bond is that the atoms on the carbon atoms are all in the same plane. Thus, ethylene is flat. Okay. Now that we know about ethylene we are ready to talk about amide bonds. Amide bonds are bonds between amines and carbonyl groups (CO). We normally draw amide bonds with a single bond between the amine and the carbonyl group, which would seem to imply that the atoms are free to rotate past one another. However, amide bonds behave a lot like ethylene in that the atoms attached to the nitrogen and carbon groups are in the same plane and rotation is restricted. Furthermore, spectroscopic and crystallographic studies show that an amide bond is shorter than a typical N-C single bond. In order to understand this behavior better, we need to think about the bonding between nitrogen and the carbonyl carbon. 19

2 If we look at an amide bond, there is a nitrogen atom that is attached to a carbon atom, which is attached to an oxygen atom through a double bond. Earlier we explained that the latter kind of chemical structure is called a carbonyl (C=O). Due to the differences in electonegativity between carbon and oxygen, most of the electrons involved in the carbon-oxygen double bond spend more time around the oxygen atom, making the carbon atom slightly electro-positive. There is a tendency for the nitrogen to want to share its lone pair of electrons with the electropositive carbon atom of the carbonyl. The ability to be able to distribute the lone pair over two atoms creates a lower energy state. We call this situation resonance stabilization. Explaining in a rigorous way why delocalizing electrons lowers the energy of a molecule is complicated and requires a knowledge of quantum mechanics. For now it is sufficient to say that it is energetically favorable for electrons to be distributed over two or more atoms rather than concentrated on one atom. Resonance electron sharing between the nitrogen and the carbonyl carbon -- gives the amide bond 40% double bond character and 60% single bond character. It is important to realize that the resonance forms shown on this slide do not exist as discrete entities. Rather, the amide bond is a combination of both of these resonance forms. You should also note that the positions of the atoms in different resonance forms are identical. Only the positions of the electrons differ. Resonance forms are thus crude representations of probable distributions of electrons. By examining the resonance form on the right, we can see that a peptide bond is somewhat like ethylene -- planar. Thus, the resonance stabilization of the amide bond restricts the shape of the polypeptide chain the amide bonds are planar whereas the bonds on either side of the carbonyl and nitrogen are attached to tetrahedral carbons. Earlier we learned that double bonds are less free to rotate because doing so requires breaking the pi bond. Amide bonds can rotate, but it costs a lot of energy to break the partial pi bond, and so the rate of rotation is slow (high activation 20

3 energy barrier). The adjacent bonds have purely single bond character and are able to rotate readily. 20

4 There are actually two possible geometric isomers around the amide bond. Whether or not there are R groups attached to both alpha carbons flanking an amide bond, the peptide bond adopts what we call a trans conformation, where the alpha carbons are trans across the amide bond. This arrangement avoids the non-bonded repulsive interaction that exist in the corresponding cis isomer. Thus, amide bonds are flat and the preferred relative orientation of the larger substituents is trans. In a polypeptide chain, there are three different types of backbone bonds: the amide bond, which we have already talked about; plus the bond between the alpha carbon and the nitrogen, and the bond between the alpha carbon and the carbonyl. We will now look at the other two peptide backbone bonds. 21

5 We have talked about how an amide bond can be represented as a combination of two different resonance forms with the electrons localized on different atoms. You can see from the resonance structure on the right that electron donation from the nitrogen lone pair to the carbonyl carbon atom leads to a structure in which the nitrogen has a partial positive charge and the oxygen has a partial negative charge. This charge separation means that the amide bond has a dipole, which we represent by an arrow pointing in the direction of the partial negative charge. 22

6 The C alpha- nitrogen and C alpha -carbonyl bonds are single bonds and can rotate freely, at least in a short polypeptide. However, even short polypeptides have definite conformational preferences. That is, some combinations of angles around these bonds are preferred over others. The conformations that are high in energy are those that place side chains in close proximity. Thus, polypeptide chains have two kinds of rigidity. One kind is determined by the amide bond s strong preference for planarity, which results from favorable orbital overlap and which leads to high barriers to rotation. The other kind is determined by the desire to avoid steric clashes between atoms in the main chain and the side chains. You can have a steric clash -- a non-bonded interaction -- when electrons involved in a bond between two atoms get too close to electrons involved in an adjacent bond. You will see in the next lecture that some of the common shapes that peptides adopt can be predicted by considering non-bonded interactions (i.e., steric clashes) between side chains of the various amino acids and with the polypeptide backbone. The amino acid side chains, which we are about to discuss in detail, play a big role in the conformations that polypeptide chains can adopt. This is evident because if the amide bonds were all that mattered, all polypeptides would have the same conformations. 23

7 The individual amino acid building blocks all have different personalities. To understand the different personalities, we need to first classify the amino acids according to different descriptors, and then consider why there are multiple different amino acids in each category. The amino acids can be classified by size, charge, polarity, polarizablity or by the unusual conformational features that they impart on the polypeptide backbone (as we will see with glycine and proline). There are twenty natural amino acids and you should know the structures and designations (both three letter and one letter code) for each one. The natural amino acids all exist as the L-enantiomer in higher organisms but in bacterial systems D-versions of the amino acids are also observed. The nonpolar amino acids are those with aliphatic hydrocarbon side chains (one, methionine, also contains a sulfur). These amino acids are hydrophobic and have no dipoles, and are likely to be found in the interior of proteins (although alanine has such a small side chain that it is found both on the inside and on the outside). Even though we group these amino acids into a single category, you should be aware that they are all somewhat different. Methionine, for example, is much more flexible than isoleucine or valine, both of which have a methyl group on the side chain close to the peptide backbone (on the beta carbon - second carbon from carboxyl carbon). This methyl group leads to a greater restriction of conformations available to the peptide backbone because many more conformations would create unfavorable steric clashes. (The notes continue on the next page). 24

8 The polar amino acids are those that have polar groups, meaning they have groups with dipoles in their side chains. Remember that dipoles are the result of charge separation, which results from differences in electronegativity of bonded atoms. All the polar side chains can function as both hydrogen bond acceptors and donors because they all have available lone pairs and heteroatoms with attached hydrogens. There are other amino acids with polar side chains such as the acidic and basic amino acids, but we put these into separate categories because they contain full charges at physiological ph. (And you need enough categories so that you can remember all the side chains!) The charged side chains include the acidic amino acids, aspartic acid and glutamic acid, which are negatively charged at physiological ph, as well as the basic amino acid side chains, lysine, arginine, and histidine, which are positively charged at physiological ph. (You should be aware that cysteine has a pka of 9.0 and so it is easily deprotonated near physiological ph. It is the only polar amino acid that is readily ionized. You should also be aware that histidine has a pk a of 6.5 and it is the only charged amino acid that also contains a significant percentage of the neutral form at physiological ph. We will talk about both of these amino acids in more detail later.) (The notes continue on the next page). 25

9 There is also a group of aromatic amino acids, phenylalanine, tyrosine, and tryptophan. These amino acids have both polar elements and hydrophobic surfaces, and they have important spectral properties. Tyrosine and tryptophan, for example, absorb UV light strongly between nm and can be used to quantitate protein concentrations (because the amount of light that is absorbed at a particular wavelength depends on the amount of each of these two amino acid side chains in a particular protein). Tryptophan is also fluorescent, and because fluorescence is sensitive to environment, tryptophan can be used as a probe of changes in environment that occur near this amino acid. That makes it useful for studies of protein folding and protein-protein interactions. Finally, there are two amino acids that are very different from all the others with respect to their conformational properties. These amino acids are glycine and proline, which we will talk about in more detail in a minute. 26

10 This table shows you pk a values for all of the amino acids with ionizing side chains. As you can see, some of the amino acids exist in an equilibrium between two forms. (Recall: pk a = -log(k a ), and K a = [H+][A-]/[HA]). You have learned just two lectures back that when the ph of a solution is equal to the pk a of an acid, the concentration of the acid ([HA]) equals the concentration of the conjugate base ([A-]). For example, when the ph of a solution containing a protein is 4, 50% of all the aspartic acid side chains are protonated, and 50% are deprotonated at any one time. See if you can convince yourself that at physiological ph (ph 7), the conjugate base form of aspartic acid will dominate by a thousand fold over the free acid form. In general, if the pk a of an amino acid side chain is more than two log units from physiological ph, we assume that it exists almost entirely as either the charged or uncharged form (i.e., depending on what form it has at physiological ph). You will see later in the next lecture that histidine is a very special amino acid because the pk a of its side chain is very close to physiological ph. You should be aware, however, that the pk a s of amino acids in the active sites of enzymes can be different from what they are in water. You will hear from Prof. Robert Lue in detail about an example where aspartic acids in the active site of a protein are found in both their charged 27

11 and their uncharged forms. 27

12 Some amino acids are used more often than others in proteins. Leucine is the most common amino acid found in proteins and it is typically found in the interior of proteins. It is more abundant than isoleucine, which has similar a hydrophobic surface area. It is thought to be favored over isoleucine in many instances because it is not beta-branched (I.e., there is no methyl near the peptide backbone) and so it can accommodate a greater range of backbone conformations. Some amino acids are relatively rare in proteins and so their presence often indicates some functional importance. Histidine and cysteine, which we mentioned as being unusual for their categories on the previous slide (basic and polar respectively), are two of the least abundant amino acids, and we will talk more about them in a moment. 28

13 We are going to focus on four amino acids that are unusual for one reason or another. You should be aware, however, that all amino acids have their own special properties, and if you continue to learn about proteins you will become more and more familiar with the distinctions between amino acids and their uses in proteins. Glycine is the only amino acid without an R group (or R is hydrogen) and proline is the only amino acid with an R group on the nitrogen. These features give glycine and proline unusual conformational properties. Cysteine is the only amino acid that can form non-peptidic bonds in proteins, and histidine is the only amino acid that has two forms which are both present at physiological ph (ph 7). 29

14 Glycine is the simplest amino acid because its R group is a hydrogen. We learned earlier that peptides are not free to adopt all possible backbone conformations, but have distinct preferences that are due to the electronic properties of the amide bond and the desire to avoid steric clashes between side chains. Because glycine does not have a side chain, it allows conformations that don t form with other amino acids. When we talk about protein structure, you will come to appreciate the importance of glycine more. 30

15 Proline contains a side chain in which the R group on the alpha carbon is connected to the nitrogen. This has a couple of consequences. First, there is no hydrogen attached to the nitrogen of a proline in a polypeptide chain. Therefore, proline does not engage in the same kinds of hydrogen bonding patterns that other amino acids do. Second, there is not a strong preference for the trans conformation around the amide bond attached to proline, as there is with all other amino acids. All other amino acids contain a nitrogen with one R group and one hydrogen, and for steric reasons the R group has a strong preference to be trans to the alpha carbon attached to the amide carbonyl. Because nitrogen has two R groups, both conformations around nitrogen create steric clashes. There is simply no way to move the R groups away from the side chain of the adjacent amino acid. This brings the trans and cis forms closer in energy, and neither one is strongly favored or disfavored. Therefore, proline can adopt backbone conformations that are disfavored in other amino acids. It is important to note that the trans and cis forms do not interconvert between each other freely. The closeness in energy of these forms simply make them more easily interconvertable. There are a class of enzymes called proline isomerases that accelerate the rate of isomerization between trans and cis conformations of proline as a regulatory mechanism for switching on and off activity. 31

16 Cysteine contains a sulfhydryl group on the side chain. Sulfhydryls can be oxidized to form disulfide bonds. What you need to know is that two cysteine side chains can come together to form a sulfur-sulfur covalent bond. The formation of such disulfide bonds in proteins rigidifies the protein and can stabilize conformations that are otherwise not highly favored. 32

17 Your hair has lots of disulfide bonds that are important for mechanical strength. Beauty shops take advantage of the disulfide bonds in your hair. If you want curly hair but were born with straight hair, you can go the beauty shop and get a perm. Your beautician adds a reducing agent to your hair, which breaks the disulfide bonds. The reduced hair is curled around rollers, which places different sulfhydryl groups in proximity, and new disulfide bonds are formed when an oxidizing agent (usually hydrogen peroxide) is added. Now instead of straight hair, you have kinked hair. Two more thoughts: all those chemicals damage your hair and the people who love you usually love you the way you are... 33

18 An ionization reaction is an equilibrium. When the ionized and neutral forms of a species are equal in concentration, the pk a is equal to the ph (because the ratio of the concentrations of the protonated to deprotonated forms is simply 1). From titration experiments (which you will learn in lab), we can determine the pk a of histidine ( ). In a titration experiment, we add base to a solution containing protonated histidine until all the protonated histidine is completely neutralized. When we add half the amount of base required, only half of the protonated histidine would be deprotonated, while the other half remains protonated. At this point, the concentration of protonated histidine is equal to that of deprotonated histidine. Since ph = pk a when these two concentrations are the same, we can determine the pk a of histidine by simply measuring the ph of the solution when the concentration of protonated histidine is equal to that of deprotonated histidine. 34

19 35

20 Histidine has a pk a of about At the pk a, half the population of histidine is protonated, and thus bears a positive charge, and half is not protonated and is neutral. That means that at physiological ph (around ph 7.0), a substantial proportion of histidine is protonated (around 10%), which is why we consider it a basic amino acid. A substantial proportion of histidine is not protonated at physiological ph but has an available lone pair that can become protonated. The neutral form of histidine can thus act as a base at physiological ph while the protonated form of histidine can act as an acid at physiological ph. The ability to act as both an acid and a base at physiological ph makes histidine a very important amino acid side chain for catalyzing many enzymatic reactions. You will learn more about the role of ionizable amino acid side chains in enzymes a little later in the course. 36

21 Enzymes provide a striking example of how amino acid side chains cooperate to allow for chemical reactions to occur on a biologically reasonable timescale. On this slide are pictures of three enzymes that break amide bonds in proteins. For example, chymotrypsin is involved in the digestion of proteins in your gut; acrosin is involved in facilitating the fertilization of an egg by a sperm; and Factor X is involved in blood clotting. As you will learn, enzymes are catalysts, molecules that accelerate the rate of a reaction without affecting the equilibrium. As you have learned, amide bond hydrolysis is a favorable reaction, meaning the equilibrium lies towards products, i.e. the free acid and amine rather than the amide. However, amide bond hydrolysis is also a very slow reaction. It has been estimated that the uncatalyzed rate of amide bond hydrolysis is somewhere between 7 and 400 years in sterile water at neutral ph. Obviously, you couldn t live if it took hundreds of years to digest your ham sandwich (or to clot your blood when you cut yourself). Your body needs enzymes that can accelerate the rate of amide bond hydrolysis so that it occurs in microseconds rather than years. These enzymes use precisely positioned amino acid side chains to accelerate the rate of amide bond hydrolysis. 37

22 Many enzymes that hydrolyze amide bonds, such as the three shown on the previous page, have strikingly similar catalytic mechanisms even though they are unrelated (i.e., they do not have a common ancestor). There are only a few possible ways to hydrolyze an amide bond using the twenty available amino acids, and so many enzymes have converged on the same solution. One solution that is used by many enzymes involves a catalytic triad, a precise arrangement of three amino acids in the active site. The catalytic triad involves serine, histidine, and aspartic acid, and these amino acids play key roles in protonation and deprotonation reactions that are essential to the mechanism. In order for these amino acids to be arranged so that they can interact in just the right way, and in order for the enzyme to bind the substrate containing the amide bond to be cleaved, the enzyme must be folded in a way that orients all the necessary amino acid side chains (and backbone) in three dimensions. Before we talk in detail about how enzymes function, we are going to talk about protein structure and folding. 38

23 39

24 What does Humpty Dumpty have to do with protein folding? 40

25 In the last lecture, we talked about the components of proteins, the amino acids, and how they are connected. We talked about conformational properties of the three different kinds of bonds in the peptide backbone and pointed out that polypeptide chains can only adopt certain conformations due to a combination of electronic and steric reasons. Today we are going to talk about the four levels of protein structure and how proteins fold. 41

26 42

27 When we talk about protein structure, it is useful to distinguish between the four different levels of structure. The first level of structure is the primary sequence, which is the linear sequence of amino acids; in this picture the letters represent the different amino acids and the colors represent the different properties of the side chains (hydrophobic, hydrophilic, charged). By convention, we always write the primary sequence of a protein from the N-terminus to the C-terminus. So on this slide, phenylalanine would be nearer the N-terminus, and tyrosine would be nearer the C-terminus. The second level of structure is the secondary structure which refers to the local structure adopted by stretches of contiguous amino acids. We will talk about two types of secondary structures that are commonly found in proteins: alpha-helices and beta-sheets. The third level of structure is the tertiary structure - the fold of a single polypeptide chain. Tertiary structure forms because the regions of secondary structure interact with each other in precisely defined ways to form a distinct shape. Finally, the fourth level is the quaternary structure, which is the interaction of individually folded polypeptide chains to form a higher order complex. Many proteins are made up of multiple polypeptide chains (i.e. hemoglobin). In this picture, each of the subunits in a tetrameric protein complex are colored differently. 43

28 The folded structure of a protein determines the function of the protein. Proteins have well defined three-dimensional structures and for many proteins we know what these structures are through high resolution NMR (nuclear magnetic resonance spectroscopy) or X-ray crystallography. Different classes of proteins are composed of common structural elements (beta strands, alpha helices, turns) in different proportions and assembled in different ways. The three proteins shown here are cytochrome b 562, which forms a helical bundle comprised of alpha helices; LDH (lactic acid dehydrogenase), which is comprised of a twisted sheet of parallel beta strands sandwiched between alpha helical segments; and a fragment of an antibody, which is comprised of antiparallel beta strands. Turns and loops connect the different elements of secondary structure in each of these proteins. In all three structures there are some bits of yellow which represent unstructured (random coil) regions of the peptide chain. You do not need to remember details about any of these proteins. We have previously talked about amino acids and how they are put together to form polypeptide chains. We learned that polypeptide chains are restricted to certain conformations due to the properties of the amide bond and the desire to avoid unfavorable non-bonded interactions between the alpha carbons and side chain groups. Now we need to understand the forces that stabilize the configuration of small regions of the peptide chain and how interactions among the stabilized regions define a unique three-dimensional structure. 44

29 When we compare the structures of many proteins, we see that proteins consist of small subunits - either helices or beta strands -- connected through turns or loops of various lengths and organized in particular ways. We represent beta strands as arrows, with the origin indicating the amino terminus of the strand and the arrowhead indicating the carboxyl terminus. Beta strands are typically organized into sheet-like structures for reasons you will learn about shortly. Helices are represented as helical (corkscrew) structures and are typically right handed due to the chirality of the amino acid side chains (L). Helices and beta sheets are very different, as you will see. For example, as the next slide shows, they have very different peptide backbone angles. 45

30 Linus Pauling was one of the greatest scientists of the twentieth century. He made fundamental contributions to understanding the structure and function of biomolecules, including enzymes. In the early 1950s, he was studying the structure of acetamide and related amides because he was interested in protein structure. X-ray experiments had shown that many proteins gave a periodic diffraction pattern, suggesting a regular structure, but the best crystallographers of the day, including Bragg, were unable to rationalize the pattern. Pauling thought that polyamide model systems might provide insight into how peptides organize. Shown here is a representation of a synthetic polyamide polymer that Pauling studied. Pauling noticed in crystal structures of this polymer that the amide bond was flat (planar) and that the dipoles of the carbonyls and amides in adjacent strands were aligned. Based on his observations about this synthetic polymer, he predicted that polypeptides comprised of natural amino acids could fold into helical structures in which the dipoles are aligned. The suggestion that biopolymers of amino acids could form helices apparently also influenced the thinking of Watson and Crick to make a similar proposal for polymers of nucleic acids. 46

31 If we look at an alpha-helical region of peptide structure, the polypeptide chain is organized in a corkscrew with all the carbonyl oxygens along the helix pointing towards the carboxyl terminus and all of the amide hydrogens pointing towards the amino terminus. This arrangement of carbonyls and amide nitrogens with respect to the helix axis is a consequence of the particular combination of peptide backbone angles, and it means that hydrogen bonds can form between residues in one turn of a helix and residues in the succeeding turn. In an alpha helix, the dipoles of the carbonyl of residue i are perfectly aligned with the NH group of residue i+4 and so these two groups form a strong hydrogen bond (we call this an i,i+4 hydrogen bond). Since the carbonyl is the last of the three atoms in amino acid i and the amide nitrogen is the first of three atoms in the i +4 residue, there are 3.6 residues per turn of the helix and there are 13 atoms participating in each twist of the helix (counting from the carbonyl oxygen through to the amide hydrogen atom). There is a 1.5 Å rise per residue in an alpha helix, and so you can easily estimate the length of a helix from the number of residues within the helix. For example, a twenty residue helix would be approximately 30 Å long. Within a helix, all of the hydrogen bonds between amide nitrogens and carbonyls are satisfied except at the two termini of the helix. The first four amide nitrogens and last four carbonyls are unpaired in an alpha helix. Because there are unsatisfied hydrogen bonding partners at the amino and carboxy termini, and all the dipoles in the helix are pointing in the same direction, there is a significant partial positive charge at the amino terminus and a correspondingly large negative charge at the carboxyl terminus. Therefore, alpha helices have large macrodipoles, and the electrostatic fields at each end of the helix can participate in attractive interactions with oppositely charged moieties. The side chains are arrayed around the helix axis and there are no unfavorable 47

32 non-bonded steric interactions. 47

33 Beta strands organize very differently from alpha helices. Beta strands adopt a zig-zag conformation and the carbonyls and amides of successive residues within the strand point in opposite directions. Because of the extended conformation, there is no opportunity for hydrogen bonds to form within a beta strand. Instead, beta strands organize into larger structures called beta sheets in which hydrogen bonds form between adjacent strands. As the name suggests, beta-sheets create a relatively flat sheet-like surface, with side chains arrayed on both faces (see picture). There are two main classes of beta sheets parallel and anti-parallel. In parallel sheets, the beta strands are aligned in a parallel fashion (i.e., the N-C direction of each strand is the same), whereas in anti-parallel strands the N-C chains run in opposite directions. The hydrogen bonding network of a sheet is not always perfectly defined and multiple variations can exist where the sheets can be twisted. Some beta strands can assemble into never-ending sheets as new strands are added to each side of the sheet. This can produce insoluble aggregates. In Alzheimer s disease a protein in your brain is cleaved, and one of the resulting pieces forms an infinite beta sheet that aggregates to form plaques. The presence of these plaques is correlated with tissue death. 48

34 We have seen that proteins are comprised of alpha helices and beta strands, and we have learned various facts about their structures and properties, but one question we have not addressed is why some proteins are largely helical whereas others are largely comprised of beta sheet or mixed alpha/beta structures. Christian Anfinsen proposed several decades ago that the tertiary structure adopted by a particular protein is a direct consequence of its primary sequence. That is, proteins somehow find the folded conformation that is lowest in energy -- meaning the conformation in which favorable interactions are maximized and unfavorable interactions are minimized. To test this hypothesis, Anfinsen decided to denature a protein -- make it unfold -- and then see if it folds back into its native conformation in a test tube. He chose as the protein to study an enzyme called ribonuclease A, which cleaves the phosphodiester linkages of RNA. You will learn about these linkages in later lectures. This protein is relatively small, and because it is an enzyme, proper folding can be assessed by evaluating enzymatic activity. 49

35 Anfinsen took two samples of ribonuclease and added both reducing agent and 8 M urea to each of them. The reducing agent was added in order to break the four disulfide bonds that stabilize the structure of ribonuclease. Urea is a chaotropic agent, a small molecule that causes proteins to unfold. At high concentrations, it alters the extensive hydrogen bonding network between water molecules (chaotropic agents contain both hydrogen bond donors and acceptors) and thus changes the balance of forces that promote protein folding. For example, in the presence of high concentrations of urea, the hydrophobic effect is no longer a driving force for folding because the entropy of the solvent is high. (Don t worry: you will learn about the hydrophobic effect in detail in a few slides.) The reduced, urea-treated proteins samples unfolded, or became denatured. Anfinsen then took one of the samples and dialyzed (removed) away the urea. As the concentration of urea decreased, the protein folded into a compact structure and Anfinsen treated it with oxidizing agent to promote disulfide bond formation. The protein was tested for activity and was found to have 90% of the activity of the original sample, implying that most of the proteins had folded into their correct shape. As a control, Anfinsen took the other sample and oxidized it to promote disulfide bond formation first and then removed the urea. The point of carrying out this control was to establish that urea had, in fact, denatured the protein. This sample was only 1-2% as active as the original. By oxidizing it before removing the chaotropic agent, Anfinsen trapped conformations containing incorrect disulfide bonds. These bonds prevented the protein from adopting its native structure again. This experiment showed that the native structure of ribonuclease will form following denaturation provided that premature oxidation is prevented. Therefore, the protein somehow finds its low energy conformation. Anfinsen concluded that the information required to fold a protein into its native, low energy conformation is contained within its primary sequence. 50