Levitt Beta Sheet Propensity Values for Amino Acids Explained

Amino acids have different preferences for forming beta sheets, a common type of protein secondary structure. The Levitt beta sheet propensity values for amino acids are numerical scores that quantify this preference, helping researchers predict how a protein will fold. In short, these values rank amino acids from those that strongly favor beta sheets (like Valine and Isoleucine) to those that strongly avoid them (like Proline and Glycine).

How Propensity Values Are Determined

Propensity values come from statistical analysis of known protein structures. Researchers like Michael Levitt counted how often each amino acid appears in beta sheets versus other structures across a large database of solved protein structures.

The calculation is straightforward. A value above 1.0 means an amino acid is found in beta sheets more often than average. A value below 1.0 means it appears less often. The higher the number, the stronger the preference for a beta sheet conformation.

These values describe a statistical tendency, not an absolute rule. A single value does not guarantee an amino acid will or will not form a beta sheet in a given protein. The local sequence context and the overall protein environment also play major roles.

The Classic Levitt Scale

The original Levitt scale, published in 1978, remains a standard reference for protein structure prediction. The values are normalized so that the average for all amino acids is 1.00.

Amino Acids That Favor Beta Sheets

These amino acids have bulky, hydrophobic side chains that pack well in the pleated sheet structure.

• Valine (Val): 1.70 — The strongest beta-sheet former. Its branched side chain fits neatly.
• Isoleucine (Ile): 1.60 — Another strong former with a similar branched structure.
• Phenylalanine (Phe): 1.35 — The aromatic ring provides good hydrophobic packing.
• Tyrosine (Tyr): 1.29 — Similar to phenylalanine but with a hydroxyl group.
• Tryptophan (Trp): 1.25 — The large indole ring favors sheet formation.
• Leucine (Leu): 1.22 — A common hydrophobic residue with a moderate preference.
• Cysteine (Cys): 1.19 — Can form disulfide bonds, which sometimes stabilize sheets.
• Threonine (Thr): 1.19 — A polar amino acid that still shows a slight preference.
• Methionine (Met): 1.05 — Essentially neutral, close to the average.

Amino Acids That Avoid Beta Sheets

These amino acids are small, flexible, or have side chains that disrupt sheet packing.

• Proline (Pro): 0.38 — The strongest beta-sheet breaker. Its rigid ring structure cannot adopt the required backbone angles.
• Glycine (Gly): 0.53 — Too flexible. It introduces too much backbone mobility for a stable sheet.
• Asparagine (Asn): 0.67 — The polar side chain can form hydrogen bonds that compete with sheet formation.
• Serine (Ser): 0.75 — Small and polar, similar to asparagine.
• Aspartic Acid (Asp): 0.80 — Charged side chains often prefer helix or loop regions.
• Glutamic Acid (Glu): 0.87 — Similar to aspartic acid.
• Lysine (Lys): 0.87 — A long, flexible charged side chain.
• Arginine (Arg): 0.90 — Also charged and flexible.
• Alanine (Ala): 0.97 — Very close to neutral. Its small methyl group does not strongly favor either structure.
• Histidine (His): 0.87 — The imidazole ring is polar and can be charged.

A Quick Worked Example

Consider a short protein segment: Val-Ile-Phe-Gly. According to the Levitt scale, the first three residues all have values above 1.0. This segment is very likely to form a beta strand. In contrast, the segment Pro-Gly-Ser-Asn has values of 0.38, 0.53, 0.75, and 0.67. This segment is very unlikely to be part of a beta sheet.

Practical Applications in Research

Propensity values are a basic tool in computational biology. When predicting how a protein folds from its amino acid sequence, beta-sheet propensity scores help assign secondary structure. Software algorithms scan a sequence and look for stretches of high-propensity residues to predict beta strands. These predictions are combined with other factors like hydrophobicity and helix propensity for a full structure model PubMed.

Researchers designing new proteins or modifying existing ones use these values to stabilize or destabilize beta sheets. If a protein needs a stronger beta sheet, engineers might mutate a neutral residue like Alanine (0.97) to Valine (1.70). If they want to break a sheet, they might introduce a Proline (0.38) or Glycine (0.53). This approach is common in designing stable enzymes or therapeutic antibodies.

Some genetic diseases are caused by single amino acid changes that disrupt protein folding. If a mutation replaces a beta-sheet-favoring residue with a breaker, it can destabilize the protein’s structure. Researchers use propensity scales to assess the structural impact of such mutations UniProt.

Limitations and Context

The Levitt scale is a useful approximation, but it has limits. It was derived from a relatively small set of protein structures compared to modern databases. Newer scales exist, and modern machine learning methods often outperform simple propensity tables.

The values do not account for long-range interactions. A beta sheet forms between two or more strands that may be far apart in the sequence. The propensity of a single residue matters less than the cooperative stability of the entire sheet. A short stretch of low-propensity residues can still form a sheet if it pairs with a strong complementary strand.

Propensity values also ignore the effect of neighboring residues. The context of the sequence can shift preferences. For example, a polar residue like Threonine (1.19) might be more likely to form a sheet if it is flanked by strong formers like Valine and Isoleucine.

Key Takeaways

• Levitt beta sheet propensity values rank amino acids by how often they appear in beta sheets in known protein structures.
• Valine (1.70) and Isoleucine (1.60) are the strongest beta-sheet formers; Proline (0.38) and Glycine (0.53) are the strongest breakers.
• The values are used in protein structure prediction, protein engineering, and analyzing disease mutations.
• The scale is a statistical guide, not an absolute rule. Local context and long-range interactions also determine actual structure.
• Modern prediction methods have largely replaced simple propensity tables, but the Levitt scale remains a clear and useful teaching tool.