Local fitness landscape of the green fluorescent protein

By K. Sarkisyan, D. Bolotin, Margarita V. Meer, D. R. Usmanova, A. Mishin, G. Sharonov, Dmitry N. Ivankov, N. Bozhanova, M. Baranov, Onuralp Soylemez and 11 more. 2016

622 citations recorded by the source corpus.

Fitness landscapes depict how genotypes manifest at the phenotypic level and form the basis of our understanding of many areas of biology, yet their properties remain elusive. Previous studies have analysed specific genes, often using their function as a proxy for fitness, experimentally assessing the effect on function of single mutations and their combinations in a specific sequence or in different sequences. However, systematic high-throughput studies of the local fitness landscape of an entire protein have not yet been reported. Here we visualize an extensive region of the local fitness landscape of the green fluorescent protein from Aequorea victoria (avGFP) by measuring the native function (fluorescence) of tens of thousands of derivative genotypes of avGFP. We show that the fitness landscape of avGFP is narrow, with 3/4 of the derivatives with a single mutation showing reduced fluorescence and half of the derivatives with four mutations being completely non-fluorescent. The narrowness is enhanced by epistasis, which was detected in up to 30% of genotypes with multiple mutations and mostly occurred through the cumulative effect of slightly deleterious mutations causing a threshold-like decrease in protein stability and a concomitant loss of fluorescence. A model of orthologous sequence divergence spanning hundreds of millions of years predicted the extent of epistasis in our data, indicating congruence between the fitness landscape properties at the local and global scales. The characterization of the local fitness landscape of avGFP has important implications for several fields including molecular evolution, population genetics and protein design.

Subfields: Genetics, Molecular Biology, Computational Biology & Bioinformatics

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Genetics

In Molecular Biology

In Computational Biology & Bioinformatics

Citations in the corpus

Evolutionary context-integrated deep sequence modeling for protein engineering (2020, methodology): “For example, to predict the fitness of quadruple mutants of the avGFP protein [54] , we trained separate models using the fitness data of single, double, triple or all three orders of mutants.”
Evaluating Protein Transfer Learning with TAPE (2019, neutral): “[43].”
Low-N protein engineering with data-efficient deep learning (2020, methodology): “…collection of random sequences, and subsequently projected representations of the experimentally characterized random mutant sequences of avGFP from Sarkisyan et al. (2016) 42 and the single mutants of TEM-1 β-lactamase from Firnberg et al. (2014) 48 onto the first and second PCs of both eUniRep…”
Epistasis and Adaptation on Fitness Landscapes (2022, background): “…studies have measured phenotypes other than competitive fitness such as monoculture growth rate (e.g., Pinheiro et al. 2021), fluorescence (e.g., Sarkisyan et al. 2016), antibody binding (e.g., Adams et al. 2019), minimal inhibitory concentration in the presence of a drug (e.g., Weinreich et…”
Exploring evolution-based & -free protein language models as protein function predictors (2022, background): “Fluorescence: This is also a protein-level regression task, predicting the log-fluorescence intensity of the protein sequence [34].”
AlphaFold2 can predict structural and phenotypic effects of single mutations (2022, background): “…three data sets from high-throughput experiments, covering three distinct phenotypes: (i) green ﬂuorescence is measured for 2,312 GFP sequences; 17 (ii) blue and red ﬂuorescence is measured for 8,192 sequences linking mTagBFP2 (mostly blue) and mKate2 (mostly red); 18 (iii) foldability and…”
Experimental assay of a fitness landscape on a macroevolutionary scale (2017, background): “This synergistic epistasis may reflect the free energy of the protein 10,47,48 , as evidenced by a correlation between the fitness potential and the impact of substitutions on the free energy of His3p ( Supplementary Fig.”
Toward machine-guided design of proteins (2018, methodology): “” Cells with differentially functional GFPs can be separated using Fluorescence-Activated Cell Sorting (FACS) combined with next-generation sequencing to characterize libraries of ≈ 10(5) sequence-function pairs [6, 7].”
ProteinBERT: a universal deep-learning model of protein sequence and function (2021, neutral): “Biophysical properties Fluorescence Continuous Global 21,446 (Sarkisyan et al., 2016; Rao et al., 2019)”
Protein Representation Learning by Geometric Structure Pretraining (2022, background): “Fluoresence landscape prediction (Sarkisyan et al., 2016) This task aims to predict the logfluorescence intensity of mutants of the wild type green fluorescent protein (GFP).”