Structural Biochemistry/Chemical Bonding/Hydrophobic interaction/Hydrophobicity scales

Hydrophobicity scales is a system used by biochemists who study amino acids to relatively define the hydrophobicity of amino acid residues. A hydrophobicity scale is typically within a negative to positive range. Values that are in the negative range are defined as not very hydrophobic, whereas the values in the positive range are defined as somewhat hydrophobic. Hydrophobicity scales also offer great insight as to the thermodynamics of the interactions that take place between lipids and proteins inside the cell membrane.

Hydrogen bonding between water molecules.


Hydrophobicity, also known as the hydrophobic effect, is the tendency and readiness of non-polar molecules, such as lipids, to associate themselves together in an aqueous solution while simultaneously excluding water molecules. The hydrophobic effect is created when lipids and other non-polar molecules disrupt the hydrogen bonding network of water and forces it reform itself around the non-polar molecule. The resulting effect is water forming a cage around that particular hydrophobic molecule. The hydrophobic effect plays a crucial role in the regulation of protein folding, as well as formation of lipid bilayers and the insertion of membrane proteins into non-polar lipid environment.

Different Hydrophobicity ScalesEdit

There are many different types of hydrophobicity scale. Below are five different ways to access the hydrophobicity level of a single amino acid residue or a single amino acid mutation of a polypeptide, as discussed by MacCallum in his hydrophobicity scales article. In the Radzicka–Wolfenden experiment, the sidechain of the amino acid arginine is placed into water and cyclohexane solvent. The concentration of the amino acid in each layer is used to calculate the free energy. The problem with this hydrophobicity scale is that cyclohexane, being more nonpolar, does not accurately represent the lipid bilayer of the cell membrane. The MacCallum et al.’s evaluation of the hydrophobicity scale is very similar to that of Radzicka’s. The only difference is that instead of using cyclohexane, MacCallum uses DOPC, which is a more accurate representation of the actual lipid bilayer. The problem is the same in that there is no backbone in this model. A “water defect” is seen because as some of the amino acids pass through the bilayer, its charges attracts and pulls the water with it. Wimley–White uses 1-octanol with water droplets and a pentapeptide Ace-WLxLL instead. They replace a single amino acid each time. This is a more realistic representation of the cell membrane because it takes into consideration the effects of the polypeptide backbone. They also measure the proportion of pentapeptide in each solvent. Moon–Fleming uses the protein OmpLA, which can be in an unfolded state in water or in a folded state when inserted inside the membrance.The equilibrium is shifted by mutating ompLA. Measurement is taken from using fluorescent spectrometry. Hessa et al. uses the protein leader peptidase. They attach an H-segment,a 19-residue-long chain, and 2 glycosylation sites on the H-segment to the leader peptidase. By assaying the glycosylation sites, they can predict if the H-segment have been inserted inside the membrane or not. These five experiments all measured the free energy transfers as the amino acid passes through the membrane. These hydrophobicity scales correlated very well with each other when normalized. Note: It should be kept in mind that these experiments do not fully represent the hydrophobicity inside the biological systems because they do not take into consideration the membrane proteins that are already presented inside the cell membrane, which could have interact with our amino acid.[1] The following image shows the system and the environment for each of the five experiments mentioned.

Add caption here

Importance of the Hydrophobicity ScaleEdit

Even though 20-30% of all proteins are membrane proteins, less than 1% of all structures known in the Protein Data Bank are membrane proteins. Knowing the hydrophobicity scale allows for prediction of the transmembrane protein sequence, as well as allows for a better understanding of the water-protein-lipid interactions.[1]

An example of the components of a typical transmembrane protein.

Protein-lipid interactionsEdit

The thermodynamic and microscopic details for lipid-protein interactions are very important in a number of vital biological factors.

One of such biological factor involves KvAP, which is the first crystal structure of a voltage-gated potassium channel. The structure of KvAP has led to numerous discussions in the biochemistry community about the interactions that take place between arginines, an amino acid, and lipids. This is due to the gating mechanism that the structure suggests in which positively charged arginines were exposed to the interior of the hydrophobic lipid bilayer of the cell membrane.

Another key biological factor in protein-lipid interactions is the action of anti-microbial peptides and cell-penetrating peptides. This is because anti-microbial peptides have specific amino acid sequences that are enriched in cationic and aromatic residues, whereas cell-penetrating peptides are rich in cationic residues.

The last major development is the successful crystallization and determination of the structure of the Sec-translocon system, which has the fundamental task of inserting membrane proteins into the membrane as soon as they are done synthesizing by the ribosome. The complexity of the Sec translocon system has raised questions about the thermodynamics of membrane insertion.

Protein membrane structureEdit

The membrane bilayer is a highly heterogeneous layer with large gradients in density and polarity on a nanometer length scale. The membrane lipid bilayer can be divided into four major regions. The hydrophobicity decreases and the hydrophilicity increases moving along each region. In the first region, which is the center of the bilayer, is very hydrophobic and is significantly disordered with properties similar to decane. In the second region, the lipid tails are more ordered and have a higher density, with features that are very similar to a polymer. In the third region, there is a diverse mixture of functional groups, most of the head group density, as well as water. In the fourth and last region, it is very hydrophilic because it is defined as mostly water that is perturbed by the lipid layer. This layer can be very deep, depending on the cellular conditions.[2]

Comparison of Hydrophobicity ScalesEdit

Different hydrophobicity scales have been developed in the biochemistry community to study the lipid-protein interactions that take place at the membrane protein. Each hydrophobicity scale has been developed independently of one another using different techniques to study these lipid-protein interactions.[1]

Radzicka-Wolfenden small molecule partitioning scaleEdit

One of the earliest hydrophobicity scales was developed by Radzicka and Wolfenden to study the folding of globular proteins. This scale was based on the partitioning of small molecule analogs of amino acid sidechains between a polar layer (water), and a lipophilic layer (cyclohexane). This particular scale is relevant for membrane partitioning due to the fact that the center of the membrane has physiochemical properties similar to those of bulk hydrocarbon.[3]

Sidechain analogs of amino acids were added to a biphasic system of water and cyclohexane. After the system reached equilibration, the concentrations of both water and cyclohexane were both measured. This ratio in the concentrations of water and cyclohexane are directly proportional to the free energy of transfer.

Although the Radzicka-Wolfenden hydrophobicity scale is simple, this scale does not accurately reflect real cellular conditions because lipid bilayers do not resemble isotropic solvents and sidechains by themselves ignore important aspects of protein structure.[4]

MacCallum et al. molecular dynamics potential of mean force scaleEdit

MacCallum et al.'s molecular dynnamics potential of mean force scale focuses on molecular dynamics simulations in order to calculate the distribution of the Radzicka-Wolfenden sidechain analogs (as described above). Instead of using cyclohexane as the lipophilic layer, a more realistic bilayer called 1,2-dioleoylsn-glycero-3-phosphocholine bilayer (DOPC) was utilized instead. Because these are computer simulations, the local environment is able to be known in the molecular level.

One of the most important things about this particular model is the formation of water defects in the bilayer, which are local deformations in the membrane which allow for water to penetrate into the bilayer core and keep polar and charged groups hydrated in the partitioning of polar and charged molecules into the lipid bilayer membrane.[5]

Wimley-White pentapeptide-based hydrophobicity scalesEdit

Wimley and White developed a peptide-based system to derive a thermodynamic scale between amino acid side-chain residues and lipids. [6]. The pentapeptide which Wimley and White used was Ace-WLxLL, where x can be any of the 20 naturally occurring amino acids. Water was used as the polar layer and 1-octanol was used as the lipophilic layer, and the partitioning between the two layers was measured. The partitioning between water and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) was also measured via equilibrium dialysis and reverse-phase high-performance liquid chromatography (HPLC).

One of the caveats of this particular scale is that its focus is explicitly interfacial instead of the bilayer core like the other scales. In order to achieve a better understanding of this scale, one will need to dwelve into the microscopic level. However, compared to the other small molecule scales, this scale is more realistic but emphasizes on the interactions of amino acid side-chain residues with either the water-lipid interface or a heterogeneous octanol environment with a hydrophobic environment for neutral amino acids like lycine and a more hydrophilic environment for charged amino acids like arginine. [7][8]

Moon-Fleming OmpLA folding/refolding scaleEdit

Moon and Fleming developed a hydrophobicity scale that is based on the reversible in vitro equilibrium between the water soluble unfolded state and membrane-inserted folded state of outer membrane phospholipase A (OmpLA) [9]. By making mutations in OmpLA, equilibrium between the folded, membrane-inserted state and the unfolded state in solution can be shifted and subsequently measured by fluorescent spectroscopy.

The three-dimensional structure the outer membrane phospholipase A (OmpLA).

This experiment compares a well-defined folded membrane state with an unfolded state in solution and measures the thermodynamic equilibrium between the two. One of the complications is that the bilayer used, 1,2-dilauroyl-sn-glycero-3-phosphocholine (DLPC), is relatively thin and unstable.

Hessa et al. Sec translocon hydrophobicity scaleEdit

Hessa et al. developed a hydrophobicity scale using a previously developed membrane protein insertion assay using the small membrane protein leader peptidase. They engineered two glycosylation sites and a 19 amino acids long residue, also known as the H segment, which can be inserted into the membrane as a transmembrane helix depending on its hydrophobicity. By assaying the two glycosylation sites, the insertion state of the H segment can be determined [10]. An apparent equilibrium between the inserted and non-inserted state of H segments can be achieved by modulating the sequence.

The insertion of the H segment involves the Sec translocon, which is a cellular machinery that either inserts of secretes a given H segment into the membrane as it gets synthesized by the ribosome. Hessa et al. used this to develop a transmembrane prediction method that relies on a linear combination of single residue results to predict whether or not a helix of an amino acid residue will get inserted into the cellular membrane [11].

Another useful application for this method with the Sec translocon is to test whether a particular sequence from a voltage-gated potassium channel would insert into the membrane, despite having multiple arginine and other polar residues [12].

Similarities between the hydrophobicity scalesEdit

Despite having used different environments and methods to test for hydrophobicity, each of the five aforementioned derived hydrophobicity scales all exhibit a well-defined correlation with one another [1]. For instance, the Radzicka-Wolfenden and MacCallum scales correlate very well with each other and yield almost identical absolute free energy differences. The Wimley-White scale measures the interactions in the heterogeneous environment of water and 1-octanol, whereas the Moon-Fleming and Hessa et al. scales measure properties that are directly related to membrane protein insertion and stability.

The absolute magnitudes of the Wimley-Hessa-Moon scales and the MacCallum-Radzicka scales differ by a significant amount. Despite this fact, all five of these scales all point to the same result [13].


Although the experiments that have been conducted have had similar results in an attempt to mach the reaction of lipids and their membrane proteins, the experiments were overall very simple and not necessarily replicators of a true physiological setting, and therefore do not give much insight to true physiological cellular membranes. The experiments are simply correlated and not necessarily exact, for biological membranes contain a diverse mixture of lipids as opposed to the single component bilayers used. Membrane proteins play up to a 25% role in cellular membranes and these are not included within the experiments. Additionally, charged or polar molecules typically distort the lipid-water interface. Moreover, the scales used in the above experiments have only considered single amino acid residues, while in a true biological setting, multiple amino acid residues are present. The free energy calculations (the free energy being the amount of energy measured from the transfer of each amino acid between a polar, water environment and a lipid bi-layer environment) used to determine the scales in each of the experiments were also very different and this is why a scale factor was used to compare the experiments. Overall, true biological systems have very different settings involved, which have yet to be matched more closely to true settings. Despite the discrepancies within the research of hydrophobicity scales, there are many profound reasons to continue investigating this rather complex and relatively unknown topic, as discussed below.


Researching hydrophobicity and establishing a scale is rather important in creating the ability to discuss the implications of the different scales, as well as in finding a single scale to utilize. The scales help to predicting membrane protein-lipid interactions as well as membrane protein structure, of which little is known. Because many drugs interact strictly with membrane proteins, understanding protein-lipid interactions through measuring hydrophobicity could give great insight to how to cure or treat illnesses through new methods that are not yet known and could be more effective. This would be most applicable in mechanisms of antimicrobial and cell penetrating peptides. Also, hydrophobicity reveals side chain-lipid interactions.


  1. a b c d MacCallum, J.L. and Tieleman, D.P. (2011) Hydrophobicity Scales: A Thermodynamic Looking Glass into Lipid–protein Interactions. Trends in Biochemical Sciences. 12, 653-661
  2. Engleman, D.M. et al. (1986) Identifying nonpolar transbilayer helices in amino acid sequences of membrane proteins. Annu. Rev. Biophys. Biophys. Chem. 15, 321-353
  3. Wolfenden, R. (2007) Experimental measures of amino acid hydrophobicity and the topology of transmembrane and globular proteins. J. Gen. Physiolo. 129, 357-362
  4. Radzicka, A. and Wolfenden, R. (1988) Comparing the polarities of the amino-acids - side-chain distribution coefficients between the vapor-phase, cyclohexane, 1-octanol, and neutral aqueous-solution. Biochem. 27, 1664-1670
  5. MacCallum, J.L. et al. (2008) Distribution of amino acids ina lipid bilayer from computer simulations. Biophys. J. 94, 3393-3404
  6. Wimley, W.C. and White, S.H. (1996) Experimentally determined hydrophobicity scale for proteins at membrane interfaces Nat. Struct. Biol. 3. 842-848
  7. Franks, N.P. et al. (1993) Molecular oganization of liquid n-octanol: An X-ray diffraction analysis. J. Pharm. Sci. 82, 466-470.
  8. Wimley, W.C. et al. (1996) Solvation energies of amino acid side chains and backbone in a family of host-guest pentapeptides. Biochem. 35, 5109-5124
  9. Moon, C.P. and Fleming, K.G. (2011) Side chain hydrophobicity scale derived from transmembrane protein folding into lipid bilayers. Proc. Natl. Acad. Sci. U.S.A. 108, 10174-10177
  10. Hessa, T. et al. (2005) Recognition of transmembrane helices by the endoplasmic reticulum translocon. Nature 433, 377-381
  11. Hessa, T. et al. (2007) Molecular code for transmembrane-helix recognition by the Sec61 translocon. Nature 450, 1026-1030
  12. Hessa, T. et al. (2005) Membrane insertion of a potassium-channel voltage sensor. Science 307, 1427
  13. Gumbart, J. et al. (2011) Free-energy cost for translocon-assisted insertion of membrane proteins. Proc. Natl. Acad. Sci. U.S.A. 108, 3596-3601