It's fun to transcribe with Fun30: A model for nucleosome dynamics during RNA polymerase II-mediated elongation

Abstract

The ability of elongating RNA polymerase II (RNAPII) to regulate the nucleosome barrier is poorly understood because we do not know enough about the involved factors and we lack a conceptual framework to model this process. Our group recently identified the conserved Fun30/SMARCAD1 family chromatin-remodeling factor, Fun30^Fft3, as being critical for relieving the nucleosome barrier during RNAPII-mediated elongation, and proposed a model illustrating how Fun30^Fft3 may contribute to nucleosome disassembly during RNAPII-mediated elongation. Here, we present a model that describes nucleosome dynamics during RNAPII-mediated elongation in mathematical terms and addresses the involvement of Fun30^Fft3 in this process.

Keywords: Fun30; RNAPII elongation; chromatin remodeler; histone chaperone FACT; nucleosome barrier; nucleosome disassembly.

Figure 1.

Mathematical approach for estimating nucleosome disassembly and the relation among nucleosome disassembly, reassembly, and loss during RNAPII-mediated elongation. (A, B) Scatterplots showing the correlation between RNAPII occupancies in wild-type cells and the levels of Fun30^Fft3- (Fig. 1A) or FACT-mediated nucleosome loss (Fig. 1B) at transcribing regions of genes with high RNAPII occupancies in wild-type cells (red; n = 519) or at transcribing regions of other genes (blue; n = 4,630). According to the RNAPII occupancies at transcribing regions in wild-type cells, the top ∼ 10% of protein-coding genes were selected as having high RNAPII occupancies. Data were obtained from 5,149 protein-coding genes of duplicate samples. Correlation coefficients (r) were calculated by the Pearson method. In Fig. 1A, P-values for the linear regression between RNAPII occupancies in wild-type cells and Fun30^Fft3-mediated nucleosome loss at genes with high RNAPII occupancies, other genes and total protein-coding genes are 1.35 × 10⁻³¹, 1.57 × 10⁻⁶ and 5.32 × 10⁻⁴⁸, respectively. In Fig. 1B, P-values between RNAPII occupancies in wild-type cells and FACT-mediated nucleosome loss at genes with high RNAPII occupancies, other genes and total protein-coding genes are 5.82 × 10⁻²⁰, 0 and 0, respectively. (C) Schematic model showing the relation among RNAPII-mediated nucleosome disassembly (D), nucleosome loss (L), and nucleosome reassembly (R). (D) Scatterplot showing the correlation between the rate of histone H3 exchange in wild-type cells and the Fun30^Fft3-mediated nucleosome loss at transcribing regions of genes. Data were obtained from 5,149 protein-coding genes of duplicate samples. The correlation coefficient of 0.41 was calculated by the Pearson method. P-value for the linear regression is 7.42 × 10⁻²¹¹. (E) Lineplot analysis in which the contribution of Fun30^Fft3 to nucleosome disassembly (D) was estimated by calculating the point at which RNAPII occupancy at transcribing regions was maximally correlated with D. When RNAPII occupancy and nucleosome disassembly show their maximal correlation (r = 0.71), Fun30^Fft3 accounts for 84% of total nucleosome disassembly during RNAPII elongation. (F) Scatterplots showing the correlation between RNAPII occupancies in wild-type cells and the estimated levels of Fun30^Fft3-mediated nucleosome disassembly at transcribing regions. Data were obtained from 5,149 protein-coding genes of duplicate samples. The gene clusters are the same as those described in Fig. 1A and B. Correlation coefficients (r) were calculated by the Pearson method. P-values for the linear regression between RNAPII occupancies in wild-type cells and Fun30^Fft3-mediated nucleosome disassembly at genes with high RNAPII occupancies, other genes and total protein-coding genes are 2.33 × 10⁻⁵⁹, 0 and 0, respectively.

Figure 2.

Schematic model showing the overlapping and distinct roles of Fun30^Fft3 and FACT in modulating nucleosome dynamics during RNAPII-mediated elongation. Fun30^Fft3 travels with elongating RNAPII, probably by associating with RNAPII or transcription-elongation factors. During RNAPII-mediated elongation, Fun30^Fft3 facilitates nucleosome disassembly to reduce the nucleosome barrier at transcribing regions. Meanwhile, FACT cooperates with Fun30^Fft3 to reduce the nucleosome barrier by inducing H2A-H2B dimer disassembly or ‘nucleosome breathing’ (which occurs without H2A-H2B dimer disassembly). After RNAPII passes, the destiny of an evicted nucleosome is determined in a FACT-dependent manner: it can be restored by FACT-mediated nucleosome reassembly or, in the absence of FACT activity, it can be lost from chromatin.