Session R49: Physics of Genome Organization: From DNA to Chromatin II

Award for Outstanding Doctoral Thesis Research in Biological Physics Talk: Single-stranded nucleic acid elasticity arises from internal electrostatic tension

The strong negative charge of single-stranded nucleic acids (ssNAs) means that an understanding of their biological roles (e.g., RNA folding) must account for electrostatic effects. ssNA flexibility means that, unlike for double-stranded nucleic acids, the characteristic structural and electrostatic length scales are in competition. We used single-molecule force spectroscopy to measure ssNA elasticity in the force range corresponding to these competing length scales. From these data, we developed an elastic model for flexible polyelectrolytes based on the internal electrostatic tension arising from the repulsion between backbone charges. This model better describes the elasticity of ssNAs than does the traditional Odijk-Skolnick-Fixman theory of electrostatics-dependent persistence length.   

Boundary-driven scaling of Turing patterns and bacterial chromosomes

The interplay between the internal architecture of a cell and its confining boundary is of fundamental importance for development and survival. We study the principles of such interplay by developing new nanofabrication and microfluidic tools to precisely manipulate bacterial cell shape and size, which we then combine with genetics, microscopy, and computational tools to quantitatively dissect organizational patterns of proteins and chromosomes. Two intriguing scaling phenomena have emerged in our experiments. First, Min protein oscillations, a division regulator, adapt the length scale of their dynamic gradients to cell length, a property unexpected from their Turing mechanism that historically were expected to have a fixed wavelength. Second, chromosome size scales with cell size nonlinearly, a property that has not been described theoretically. We will propose the underlying principles that drive these phenomena and also invoke discussions on the general advantage of organizational mechanisms in cells that exploit boundary effect.   

Shannon Information Entropy in the Genetic Code

The canonical genetic code is the nearly universal language for translating the information stored in DNA into proteins, and has evolved a considerable measure of robustness to single-letter mutations. Shannon entropy measures the expected information value of messages. As with thermodynamic entropy, the Shannon entropy is only defined within a system that identifies at the outset the collections of possible messages, analogous to microstates, that will be considered indistinguishable macrostates. This fundamental insight is applied here to amino acid alphabets, which group the twenty common amino acids into families based on chemical and physical similarities. By calculating the normalized mutual information, which measures the reduction in Shannon entropy conveyed by single nucleotide messages, groupings that best leverage the fault tolerance of the code are identified. The relative importance of properties related to protein folding - like hydrophobicity and size - and function, including side-chain acidity, can also be estimated. This approach allows for the quantification of the average information value of nucleotide positions, which can shed light on the severity of hereditary and de novo genetic mutations.   

Using Self-Assembling DNA Complexes to Examine the Relationship between DNA Organization and Transcriptional Efficiency

DNA organization within a cell is multifaceted and dynamic. Not only does chromatin structure vary with position, as there is significant heterogeneity in the degree of local compaction within the global DNA complex, but it also changes with time, e.g. showing dependence on the growth stage of a cell. Much of this variability is presumed to arise through diverse, non-equilibrium interactions with proteins, whose binding dynamics and motor activities can, e.g., propagate forces throughout the weakly-crosslinked DNA network. As DNA organization is innately coupled to gene accessibility, the structural changes induced by active processes are expected to have important effects on gene expression. To begin probing this incredibly complex relationship, we use a simple in vitro system – comprised of self-assembling DNA nanostars, DNA gene templates, and RNA polymerase – to examine how transcriptional activity influences DNA organization and, conversely, how DNA structure can modulate genetic output.   

Functional Consequences of Methylating the Unmethylated Yeast Genome

Cytosine methylation in DNA is an important epigenetic mark in many higher eukaryotes, established in vivo by a variety of DNA methyltransferase enzymes (DNMTs). The genome-wide methylation patterns these enzymes establish have complex relations with gene transcription, chromatin architecture and histone modifications, only some of which may be attributable to direct cellular response to information contained in methylation patterns. To disentangle fundamental causes and consequences of DNA methylation from secondary correlations, we profiled the transcriptomes and methylomes produced by combinatorial knock-in of DNMT genes in Pichia pastoris, a yeast species lacking DNA methylation machinery. We find that all active DNMTs preferentially methylate the linker DNA separating nucleosomes along the chromatin fiber and produce distinct methylation patterns at differentially expressed genes. We study variations in methylation patterns established by different DNMTs and use a time-course experiment to capture adaptation of the yeast cells to de novo methylation distinguishing changes in gene expression directly caused by DNA methylation from those caused by stress response.   

RNA Base Pairing Determines the Conformations of RNA Inside Spherical Viruses

Many simple RNA viruses enclose their genetic material by a protein shell called the capsid. While the capsid structures are well characterized for most viruses, the structure of RNA inside the shells and the factors contributing to it remain poorly understood. We study the impact of base pairing on the conformations of RNA and find that it undergoes a swollen coil to globule continuous transition as a function of the strength of the pairing interaction. We also observe a first order transition and kink profile as a function of RNA length. All these transitions could explain the different RNA profiles observed inside viral shells.   

DNA sequencing based biophysical characterization of cell-free DNA in the blood circulation sheds light on its origin

A large number of small fragments of cell-free DNA circulate in human blood. These molecules are the product of cell death across the body, and offer a unique window into health and disease. Here, we have used next-generation DNA sequencers to probe the biophysical properties of cell-free DNA (cfDNA) in blood plasma, urine, and peritoneal fluid. We analyzed more than 250 samples and show that the biophysical properties of cell-free DNA, including fragment length, genome alignment, and dinucleotide frequency, depend strongly on the biochemical properties of the extracellular environment. We further show that cell-free DNA in urine, plasma and peritoneal fluid comprises “footprints” of proteins bound to the genome. The displacement of chromatin organization protein complexes, such as nucleosomes, near transcription start sites and strong protection signals of transcription binding factors, can be used to identify the tissues that contribute cfDNA toward the mixture in different bodily fluids. Last, we discuss the properties of non-human cfDNA, i.e. bacterial and viral cfDNA, in these different fluids and demonstrate that unbiased sequencing enables comprehensive monitoring of infection.   

Fundamental Sources of Codon Bias

Differences in genome-wide frequencies of synonymous codons have been attributed to various biophysical mechanisms such as tRNA concentration and ribosomal affinity. Main features in the bias appear to be due to a balance between protein production efficiency and accuracy. In an effort to develop a comprehensive theoretical model of this phenomenon, we have developed a quasi-species evolutionary model which captures several of these main influences and predicts the frequencies of genomic codons with high accuracy. The model consists of a deterministic description of the evolution of independent codons under the influence of biophysically-motivated forms of selection, and as a result, the model parameters hold physical interpretability.   

Programmable and Orthogonal Gene Circuit Elements

While CRISPR may be better-known as a gene editing tool, it can also be used as a tool for selective repression of genes using a catalytically-dead CRISPR nuclease. Here, we exploit both the freedom in the CRISPR identification region as well as the freedom in promoter design to create a set of 64 synthetic ‘barcoded’ promoters, each of which is keyed to a specific CRISPR guide RNA. These synthetic promoters can then be used to drive chosen genes in modular fashion. By driving GFP, we measure the on/off ratio of these so-called CRISPRgate circuit elements. We then take a multiplexed approach using ‘randomized’ promoter barcodes in order to demonstrate orthogonality of the CRISPRgates on a large scale. This is done using a sacB/tetA cassette, a counterselection scheme which slows cell growth in the presence of sucrose and fusaric acid. Cells with matched CRISPR RNA and promoter barcodes preferentially survive, and by using multiplexed sequencing, we demonstrate that these CRISPRgate circuit elements are indeed orthogonal on a massive scale. We then explore how these orthogonal circuit elements can be exploited to create complex genetic circuits.   

Dynamics of Active Rouse Chains

We consider how active forces modeled as non-thermal random noise affect the average dynamical properties of a Rouse polymer. As the power spectrum of the noise is not known we keep the analytical treatment as generic as possible and then present results for a few examples of active noise. We discuss the connection between our results and recent experimental studies of dynamics of labeled DNA telomeres in living cells, and propose new chromatin tracking experiments that will allow one to determine the statistical properties of the active forces associated with chromatin remodeling processes.   

Programmable and bistable CRISPR-based toggle switches in Escherichia coli

Recent developments and advances in CRISPR-Cas (Clustered regularly interspaced short palindromic repeats and CIRSPR-associated proteins) systems have ushered a new generation of powerful genetic engineering tools in synthetic biology. In particular, a catalytically ‘dead’ version of CRISPR-Cas proteins that lack nuclease activity can essentially function as a logic NOR gate by selectively binding to a promoter sequence and preventing initiation of transcription by RNA polymerase. In this work, we create programmable, bistable genetic toggle switches using pairs of mutually repressible orthogonal CRISPR-based NOR gates. Another pair of convergent positioned aTc-inducible and IPTG-inducible promoters that can program/reset the state of each toggle switch are wired into a single plasmid and transformed into E. coli cells. We show that CRISPR-based toggle switches can remain into a fixed state for many generations, and a statistical mechanical model is developed to predict a bistability criterion for CRISPR-based toggle switches. We experimentally test these predictions and we outline an experimental framework to tailor the sequence in the promoter to produce numerous pairs of toggle switches that can work in parallel and with limited cross-talk.   

Polymer Modeling of Dynamics and Disrupted Organization of Fragmented Chromatin

Various biological factors help maintain genome organization, but the genome’s own inherent polymeric structure and connectivity provide the foundation for its complex and dynamic architecture. We probe the extent to which the connectivity of the underlying polymer structure maintains chromatin organization through analysis of Hi-C genome contact maps and polymer molecular dynamics simulations. We cut the chromatin structure into ~10 kb fragments by treating nuclei with restriction enzymes and generate contact maps of the resulting structures via Hi-C after different time delays. We observe loss of compartmentalization and alterations to the contact probability scaling across multiple length scales after time delays of minutes to hours. To explain these observations, we construct a block copolymer simulation model for chromatin, which forms spatially distinct compartments via phase separation under normal conditions. Upon progressive fragmentation of the polymer, compartmentalization is progressively lost and other physical properties of the model genome are perturbed. The model illustrates how the underlying connectivity of chromatin along with other biophysical factors act together to maintain spatial organization of the genome.   

Evolution of Nested Folding States in Compression of a Strongly Confined Semiflexible Chain

We use Brownian dynamics (BD) simulations to probe the physics of non-equilibrium polymer
compression in extreme nanoconfinement. In our system, modelled on the ‘nanodozer assay’, a gasket
translating at a fixed sliding speed impinges on a nanochannel extended chain. In square channels
with diameter much smaller than the chain persistence length, we find that chain compression
proceeds through a unique folding kinetics driven by repeated double fold nucleation events and
growth of nested folds. We show that the folding kinetics can be understood by coupling a theory
for deterministic contour spooling across the folds with a dynamically varying energy landscape for
fold nucleation. These findings are critical for understanding compression of nanochannel confined
DNA in the sub-persistence length (Odijk) regime.