Investigation of the Chromosome Regions with Significant Affinity for the Nuclear Envelope in Fruit Fly – A Model Based Approach

Nicholas Allen Kinney,Alexey V Onufriev,Igor V Sharakhov,Freddie Salsbury

doi:10.1371/journal.pone.0091943

Abstract

Three dimensional nuclear architecture is important for genome function, but is still poorly understood. In particular, little is known about the role of the “boundary conditions” – points of attachment between chromosomes and the nuclear envelope. We describe a method for modeling the 3D organization of the interphase nucleus, and its application to analysis of chromosome-nuclear envelope (Chr-NE) attachments of polytene (giant) chromosomes in Drosophila melanogaster salivary glands. The model represents chromosomes as self-avoiding polymer chains confined within the nucleus; parameters of the model are taken directly from experiment, no fitting parameters are introduced. Methods are developed to objectively quantify chromosome territories and intertwining, which are discussed in the context of corresponding experimental observations. In particular, a mathematically rigorous definition of a territory based on convex hull is proposed. The self-avoiding polymer model is used to re-analyze previous experimental data; the analysis suggests 33 additional Chr-NE attachments in addition to the 15 already explored Chr-NE attachments. Most of these new Chr-NE attachments correspond to intercalary heterochromatin – gene poor, dark staining, late replicating regions of the genome; however, three correspond to euchromatin – gene rich, light staining, early replicating regions of the genome. The analysis also suggests 5 regions of anti-contact, characterized by aversion for the NE, only two of these correspond to euchromatin. This composition of chromatin suggests that heterochromatin may not be necessary or sufficient for the formation of a Chr-NE attachment. To the extent that the proposed model represents reality, the confinement of the polytene chromosomes in a spherical nucleus alone does not favor the positioning of specific chromosome regions at the NE as seen in experiment; consequently, the 15 experimentally known Chr-NE attachment positions do not appear to arise due to non-specific (entropic) forces. Robustness of the key conclusions to model assumptions is thoroughly checked.

Highlights

Unlike enzyme proteins, which usually adopt the same unique three-dimensional (3D) shapes in all cells, the conformational states of chromatin fibers are not nearly as compact and ordered; the basic principles governing these conformational states are only beginning to emerge through computation and experiment [1,2,3,4,5,6,7,8,9,10]
The model answers three questions: Are there additional statistically significant chromosome-nuclear envelope (Chr-nuclear envelope (NE)) attachment regions? If there are additional Chr-NE attachment regions, do they correspond to heterochromatin? Does confinement of the polytene chromosomes in a spherical nucleus alone favor the positioning of specific chromosome regions at the as seen in experiment? Our model demonstrates that the geometric effects of chromosome confinement inside a spherical nucleus alone do not bring about specific Chr-NE attachments
We interpret this as validation of our modeling approach; we further checked the robustness of our modeled chromosome territories to the non-equilibrium modifications of our self-avoiding walk (SAW)

Summary

Introduction

Unlike enzyme proteins, which usually adopt the same unique three-dimensional (3D) shapes in all cells, the conformational states of chromatin fibers are not nearly as compact and ordered; the basic principles governing these conformational states are only beginning to emerge through computation and experiment [1,2,3,4,5,6,7,8,9,10]. Recent computational studies have demonstrated that chromosome organization in the nucleus may strongly depend on the degree of (spherical) confinement [13]: increasing the degree of confinement mimicked the effect of increasing chromosome looping probability, reinforcing the idea that the boundary conditions of the nucleus matter. These [13], and the results of other studies [3,14,15,16,17,18,19,20], have raised the possibility that the boundary conditions of the nucleus, chromosome topology, and non-specific (entropic) forces may be sufficient to account for the organization of chromosomes in the nucleus of some Metazoans. Chromosome looping, potentially brought about by the degree of confinement [13], has been linked to gene expression levels; higher chromosome looping probability was associated with higher local chromosome density and lower transcriptional activity in a recent study [21]

Objectives

Methods

Results

Discussion

Conclusion