The non-coding RNA composition of the mitotic chromosome by 5'-tag sequencing

Abstract

Mitotic chromosomes are one of the most commonly recognized sub-cellular structures in eukaryotic cells. Yet basic information necessary to understand their structure and assembly, such as their composition, is still lacking. Recent proteomic studies have begun to fill this void, identifying hundreds of RNA-binding proteins bound to mitotic chromosomes. However, by contrast, there are only two RNA species (U3 snRNA and rRNA) that are known to be associated with the mitotic chromosome, suggesting that there are many mitotic chromosome-associated RNAs (mCARs) not yet identified. Here, using a targeted protocol based on 5'-tag sequencing to profile the mammalian mCAR population, we report the identification of 1279 mCARs, the majority of which are ncRNAs, including lncRNAs that exhibit greater conservation across 60 vertebrate species than the entire population of lncRNAs. There is also a significant enrichment of snoRNAs and specific SINE RNAs. Finally, ∼40% of the mCARs are presently unannotated, many of which are as abundant as the annotated mCARs, suggesting that there are also many novel ncRNAs in the mCARs. Overall, the mCARs identified here, together with the previous proteomic and genomic data, constitute the first comprehensive catalogue of the molecular composition of the eukaryotic mitotic chromosomes.

Figure 1.

Purification of metaphase chromosomal-associated RNA. (A) Overview of the protocol. (B) Protein composition determined by SDS-PAGE after incubation with buffers containing the indicated increasing concentrations of salt. (C) Ultrastructure of metaphase chromosomes after incubation with the 0.2 M NaCl buffer. (D) Detection of β-actin mRNA using RT-PCR in different fractions of the sucrose gradient. (E) RT-PCR of U3 snRNA and β-actin in the mCARs (m) and cytoplasmic RNA (c) samples.

Figure 2.

Strategy to prepare the 5′ tag library for Illumina sequencing.

Figure 3.

Basic characterization of the sequenced LS and HS samples. (A) The LS (left) and HS (right) tag cluster length distributions. (B) Distribution of the chromosomal locations from which the clusters are derived. The blue (red) bars indicate cluster locations from the LS (HS) samples.

Figure 4.

Genomic characteristics of mCARs. (A) Pie chart showing the percentage of mCAR reads in the coding (blue), noncoding (red), or unannotated (green) region of the genome. (B) Conservation analysis of lncRNA mCAR (left) and all annotated lncRNAs (right). (C) Conservation analysis of the snoRNA mCARs (left) and all annotated snoRNAs (right).

Figure 5.

Conservation analysis of the ID4 mCARs. (A) Phylogenic analysis of the ID4 mCARs. The three sequences with the greatest number of reads are indicated with a red diamond. The scale bar reflects evolutionary distance. (B) Sequence comparison of 20 randomly chosen ID4 mCARs (top) and 20 randomly chosen ID4 sequences from the whole population of ID4 sequences in the genome (bottom). Similar results are obtained with the entire population of ID4 mCARs and 100 randomly chosen ID4 sequences from the whole genome population (Supplementary Figure S3).

Figure 6.

(A) qPCR validation for the HS- and LS- samples relative to metaphase cell total RNA. (B) qPCR validation between the HS-, LS- samples and cytoplasmic RNA. (C) RNA FISH validation of the mCAR localization of MALAT1 and snoRNA FR137451 to the mitotic chromosome. RNA FISH using sense probes was used as a negative control. Scale bar: 5 μm.