Doug Barrick

The biases revealed in protein sequence alignments have been shown to provide information related to protein structure, stability and function. Recent studies using Potts models show that single-site biases and pair correlations can improve predictions of protein fitness, activity and stability compared to simpler models. Here we use a Potts model to design groups of protein sequences with different amounts of single-site biases and pair correlations and determine the stabilities of sequences from each group. Surprisingly, sequences excluding pair correlations maximize stability compared to sequences that maximize pair correlations, suggesting that pair correlations contribute to other aspects of protein fitness. Consistent with this interpretation, we find that for three enzyme families, activity is greatly increased by maximizing pair correlations. The finding that elimination of covariant residue pairs increases protein stability suggests a route to enhance stability of designed proteins, although this stability may be offset by reduced enzyme activity.

The Notch signaling pathway regulates cellular differentiation by activating transcription through an unusual heterotrimeric complex comprising the Notch receptor’s intracellular domain (NICD), the DNA-binding protein CSL, and the coactivator MAML. NICD has two binding sites for CSL, a short motif in the RAM region and an ankyrin domain (ANK), connected by an intrinsically disordered linker and which form a bivalent ternary complex with CSL and MAML. Although bivalency is required for maximal transcription activation, the energetic contributions of bivalency and heterotrimer formation within this essential complex are unknown. To elucidate the energetics of bivalency, we first determine the free energy of the CSL-ANK-MAML heterotrimer, using isothermal titration calorimetry and developing an obligate heterotrimer model to analyze the data. By comparing this heterotrimerization reaction with binding reactions involving different regions of RAMANK, we determine the energetic contribution of bivalency to heterotrimer assembly. We show that bivalency through the disordered linker increases the effective concentration of ANK, and that the bivalent interaction enhances occupancy of RAM and ANK at their binding sites on CSL by about three orders of magnitude. By redefining the standard state to a lower, more physiological protein concentration, we reveal the importance of the RAMANK intrinsically disordered linker for assembly of the Notch transcription activation complex. This work provides a framework whereby the energetic contributions of intrinsically disordered linkers to higher-order multivalent assembly may be analyzed.

Cooperativity is a central feature to protein folding and is important for protein design. Repeat proteins are good systems for quantifying the thermodynamic basis of cooperativity. Analysis of repeat proteins composed of identical consensus repeats show that repeats strongly drive the folding of their neighbors through extensive tertiary contacts. Here, we use the consensus approach to quantify the cooperativity of folding of spectrin repeat arrays. Spectrin repeats are unique among tandem repeat proteins in that they share an elongated α-helix that spans neighboring repeats. We generate a consensus spectrin repeat sequence and show that this sequence is structured by CD and NMR spectroscopy, and is considerably more stable than extant spectrin repeats. By generating pairs of consensus spectrin repeats, we find tandem repeats to be further stabilized, demonstrating cooperative stabilization by neighboring repeats. Using an Ising model to analyze single- and tandem spectrin repeat unfolding, we find that the consensus stability increase results from intrinsic but not interfacial stabilization. By introducing mutations and insertions at the boundary between consensus repeats, we find that cooperativity is driven primarily by helical propagation; to a lesser extent, helix propagation also stabilizes partly folded states where one of two repeats is unfolded.

The biases revealed in protein sequence alignments have been shown to provide information related to protein structure, stability, and function. For example, sequence biases at individual positions can be used to design consensus proteins that are often more stable than naturally occurring counterparts. Likewise, correlations between pairs of residue can be used to predict protein structures. Recent work using Potts models show that together, single-site biases and pair correlations lead to improved predictions of protein fitness, activity, and stability. Here we use a Potts model to design groups of protein sequences with different amounts of single-site biases and pair correlations, and determine the thermodynamic stabilities of a representative set of sequences from each group. Surprisingly, sequences excluding pair correlations maximize stability, whereas sequences that maximize pair correlations are less stable, suggesting that pair correlations contribute to another aspect of protein fitness. Consistent with this interpretation, we find that for adenylate kinase, enzyme activity is greatly increased by maximizing pair correlations. The finding that elimination of covariant residue pairs increases protein stability suggests a route to enhance stability of designed proteins; indeed, this strategy produces hyperstable homeodomain and adenylate kinase proteins that retain significant activity. Significance statement Recent methods for protein structure analysis and design have used sequence covariance to help predict protein structure, stability, and function. Here, by designing homeodomain and adenylate kinase sequences with different amounts of single-site bias and pairwise covariance, we find that stability is solely determined by single-site bias but not pairwise covariance. However, pairwise covariance makes an important contribution to catalysis in adenylate kinase. Our findings suggest a new way to generate highly stable proteins: by separating single-site biases from pairwise covariance, the single-site coefficients can be used to design proteins with stabilities even higher than those obtained by consensus design.

The Notch signaling pathway regulates cellular differentiation by activating transcription through an unusual heterotrimeric complex comprising the Notch receptor’s intracellular domain (NICD), the DNA-binding protein CSL, and the coactivator MAML. NICD has two binding sites for CSL, a short motif in the RAM region and an ankyrin domain (ANK), connected by an intrinsically disordered linker and which form a bivalent ternary complex with CSL and MAML. Although bivalency is required for maximal transcription activation, the energetic contributions of bivalency and heterotrimer formation within this essential complex are unknown. To elucidate the energetics of bivalency we first determine the free energy of the CSL-ANK-MAML heterotrimer, using isothermal titration calorimetry and developing an obligate heterotrimer model to analyze the data. By comparing this heterotrimerization reaction with binding reactions involving different regions of RAMANK, we determine the energetic contribution of bivalency to heterotrimer assembly. We show that bivalency through the disordered linker increases the effective concentration of ANK, and that the bivalent interaction enhances occupancy of RAM and ANK at their binding sites on CSL by about three orders of magnitude. By redefining the standard state to a lower, more physiological protein concentration, we reveal the importance of the RAMANK intrinsically disordered linker for assembly of the Notch transcription activation complex. This work provides a framework whereby the energetic contributions of intrinsically disordered linkers to higher-order multivalent assembly may be analyzed.

A protein sequence encodes its energy landscape—all the accessible conformations, energetics, and dynamics. The evolutionary relationship between sequence and landscape can be probed phylogenetically by compiling a multiple sequence alignment of homologous sequences and generating common ancestors via Ancestral Sequence Reconstruction or a consensus protein containing the most common amino acid at each position. Both ancestral and consensus proteins are often more stable than their extant homologs—questioning the differences between them and suggesting that both approaches serve as general methods to engineer thermostability. We used the Ribonuclease H family to compare these approaches and evaluate how the evolutionary relationship of the input sequences affects the properties of the resulting consensus protein. While the consensus protein derived from our full Ribonuclease H sequence alignment is structured and active, it neither shows properties of a well‐folded protein nor has enhanced stability. In contrast, the consensus protein derived from a phylogenetically‐restricted set of sequences is significantly more stable and cooperatively folded, suggesting that cooperativity may be encoded by different mechanisms in separate clades and lost when too many diverse clades are combined to generate a consensus protein. To explore this, we compared pairwise covariance scores using a Potts formalism as well as higher‐order sequence correlations using singular value decomposition (SVD). We find the SVD coordinates of a stable consensus sequence are close to coordinates of the analogous ancestor sequence and its descendants, whereas the unstable consensus sequences are outliers in SVD space.

Notch signaling in humans is mediated by four paralogous receptors that share conserved architectures and possess overlapping, yet non‐redundant functions. The receptors share a canonical activation pathway wherein upon extracellular ligand binding, the Notch intracellular domain (NICD) is cleaved from the membrane and translocates to the nucleus where its N‐terminal RBP‐j‐associated molecule (RAM) region and ankyrin repeat (ANK) domain bind transcription factor CSL and recruit co‐activator Mastermind‐like‐1 (MAML1) to activate transcription. However, different paralogs can lead to distinct outcomes. To better understand paralog‐specific differences in Notch signaling, we performed a thermodynamic analysis of the Notch transcriptional activation complexes for all four Notch paralogs using isothermal titration calorimetry. Using chimeric constructs, we find that the RAM region is the primary determinant of stability of binary RAMANK:CSL complexes, and that the ANK regions are largely the determinants of MAML1 binding to pre‐formed RAMANK:CSL complexes. Free energies of these binding reactions (ΔG_RA and ΔG_MAML) vary among the four Notch paralogs, although variations for Notch2, 3, and 4 offset in the free energy of the ternary complex (ΔG_TC, where ΔG_TC = ΔG_RA + ΔG_MAML). To probe how these affinity differences affect Notch signaling, we performed transcriptional activation assays with the paralogous and chimeric NICDs, and analyzed the results with an independent multiplicative model that quantifies contributions of the paralogous RAM, ANK, and C‐terminal regions (CTR) to activation. This analysis shows that transcription activation correlates with ΔG_TC, but that activation is further modified by CTR identity in a paralog‐specific way.

Summary: Notch signaling is mediated by four paralogous receptors with conserved architectures and overlapping yet non-redundant functions. Notch signaling generates a transcriptional activation complex (NTC) wherein the N-terminal RAM and ankyrin repeat (ANK) regions of the Notch intracellular domain (NICD) bind transcription factor CSL and recruit co-activator Mastermind-like (MAML). To better understand paralog-specific differences in Notch signaling, we analyzed the thermodynamics of binary and ternary NTCs for all four Notch paralogues and chimeric constructs. We find that RAMANK:CSL stability (ΔG _RA ) is primarily determined by the RAM region while MAML binding to preformed RAMANK:CSL complexes (ΔG _MAML ) is largely determined by the ANK region. We determined transcriptional activation data for the paralogous and chimeric NICDs and analyzed the data with an independent multiplicative model. This analysis shows ternary complex stability (ΔG _TC , where ΔG _TC = ΔG _RA +ΔG _MAML ) correlates well with transcriptional activations and provides insights into contributions of RAM, ANK and the C-terminal regions to Notch signaling.

A protein sequence encodes its energy landscape - all the accessible conformations, energetics, and dynamics. The evolutionary relationship between sequence and landscape can be probed phylogenetically by compiling a multiple sequence alignment of homologous sequences and generating common ancestors via Ancestral Sequence Reconstruction or a consensus protein containing the most common amino acid at each position. Both ancestral and consensus proteins are often more stable than their extant homologs - questioning the differences and suggesting that both approaches serve as general methods to engineer thermostability. We used the Ribonuclease H family to compare these approaches and evaluate how the evolutionary relationship of the input sequences affects the properties of the resulting consensus protein. While the overall consensus protein is structured and active, it neither shows properties of a well-folded protein nor has enhanced stability. In contrast, the consensus protein derived from a phylogenetically-restricted region is significantly more stable and cooperatively folded, suggesting that cooperativity may be encoded by different mechanisms in separate clades and lost when too many diverse clades are combined to generate a consensus protein. To explore this, we compared pairwise covariance scores using a Potts formalism as well as higher-order couplings using singular value decomposition (SVD). We find the SVD coordinates of a stable consensus sequence are close to coordinates of the analogous ancestor sequence and its descendants, whereas the unstable consensus sequences are outliers in SVD space.

Cooperativity is a central feature of protein folding, but the thermodynamic and structural origins of cooperativity remain poorly understood. To quantify cooperativity, we measured guanidine-induced unfolding transitions of single helix-hairpin-helix (HhH)₂ repeats and tandem pairs from a seven-repeat segment of Methanopyrus kandleri Topoisomerase V (Topo V) to determine intrinsic repeat stability and interfacial free energies between repeats. Most single-repeat constructs are folded and stable; moreover, several pairs have unfolding midpoints that exceed midpoints of the single repeats they comprise, demonstrating favorable coupling between repeats. Analyzing unfolding transitions with a modified Ising model, we find a broad range of intrinsic and interfacial free energies. Surprisingly, the G repeat, which lacks density in the crystal structure of Topo V without DNA, is the most stable repeat in the array. Using nuclear magnetic resonance spectroscopy, we demonstrate that the isolated G repeat adopts a canonical (HhH)₂ fold and forms an ordered interface with the F-repeat but not with the H repeat. Using parameters from our paired Ising fit, we built a partition function for the seven-repeat array. The multistate unfolding transition predicted from this partition function is in excellent agreement with the experimental unfolding transition, providing strong justification for the nearest-neighbor model. The seven-repeat partition function predicts a native state in which three independent segments ("stability islands") of interacting repeats are separated by two unstable interfaces. We confirm this segmented architecture by measuring the unfolding transition of an equimolar mixture of these three separate polypeptides. This segmented structural organization may facilitate wrapping around DNA.