Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2011 Jul 5;7:506.
doi: 10.1038/msb.2011.32.

Cell-to-cell Variability of Alternative RNA Splicing

Affiliations
Free PMC article

Cell-to-cell Variability of Alternative RNA Splicing

Zeev Waks et al. Mol Syst Biol. .
Free PMC article

Abstract

Heterogeneity in the expression levels of mammalian genes is large even in clonal populations and has phenotypic consequences. Alternative splicing is a fundamental aspect of gene expression, yet its contribution to heterogeneity is unknown. Here, we use single-molecule imaging to characterize the cell-to-cell variability in mRNA isoform ratios for two endogenous genes, CAPRIN1 and MKNK2. We show that isoform variability in non-transformed, diploid cells is remarkably close to the minimum possible given the stochastic nature of individual splicing events, while variability in HeLa cells is considerably higher. Analysis of the potential sources of isoform ratio heterogeneity indicates that a difference in the control over splicing factor activity is one origin of this increase. Our imaging approach also visualizes non-alternatively spliced mRNA and active transcription sites, and yields spatial information regarding the relationship between splicing and transcription. Together, our work demonstrates that mammalian cells minimize fluctuations in mRNA isoform ratios by tightly regulating the splicing machinery.

Conflict of interest statement

The authors declare that they have no conflict of interest.

Figures

Figure 1
Figure 1
Schematic for identification of candidate mRNAs. A series of steps were taken to identify suitable genes for smFISH imaging of alternatively spliced isoforms including searching for genes with two isoforms which have the same 5′ start site and differ substantially in sequence (>800 bp), verification of isoform presence in multiple databases and/or in the literature, and qRT–PCR verification of isoform presence.
Figure 2
Figure 2
The alternative splicing of CAPRIN1 and MKNK2 imaged at single-molecule resolution. (A) The mutually exclusive mRNA sequences of CAPRIN1 isoform 1 and isoform 2 were hybridized to tens of short oligo probes coupled to Alexa555 (green) or Cy5 (red) dyes, respectively. (B) The entire mature mRNA sequence of MKNK2 was hybridized with Cy5 (red) probes except for the unique sequence of isoform 1 (Alexa555 probes, green). Isoform 1 was detected as a colocalized spot in both channels (yellow). (C, D) Example images of CAPRIN1 (C) and MKNK2 (D) in HeLa and Rpe1 cells, respectively. Images are maximum intensity projections covering the entire z axis of the cells. Nuclei are stained with DAPI (blue). Scale bars, 10 μM. Background subtraction (ImageJ) applied to merges. (E, F) Maximum intensity projections of cells hybridized with no probes (E) or with 48 randomly scrambled, 20 bp, Cy5-coupled probes (F).
Figure 3
Figure 3
Detection of active sites of transcription using single-molecule imaging. Transcription sites actively transcribing CAPRIN1 or MKNK2 mRNA were identified as colocalized nuclear spots whose signal was either beyond the linear range of detection (4096 AU on a 12bit camera) or at least 1.5 times brighter than the mean spot intensity in both channels. (A) Maximum intensity projections of an Rpe1 cell which containing an active CAPRIN1 transcription site. Arrows point to the spot that fulfills the criteria for detection as an active transcription site. This spot is also the highest intensity spot for each channel. Scale bars, 10 μM. (B) Intensity plots of the lines in panel (A) show that mRNA spot signals are substantially above background. (C) Intensity histograms of all identified spots (mRNAs) in each channel. Arrows point to the spot depicted by the yellow arrows in panel (A). The dashed gray line points to the mean background intensity for each channel (Alexa555, 452 AU; Cy5, 407 AU). (D) Active transcription sites were not detected following treatment with the transcription inhibitor actinomycin D.
Figure 4
Figure 4
Non-alternatively spliced CAPRIN1 mRNA is present away from transcription sites. (A, B) Maximum intensity projections of CAPRIN1 (A) and MKNK2 (B) mRNA in HeLa cells containing an identified transcription site (red arrows). Enlargements show enrichment of colocalized nuclear spots (yellow) around the transcription sites of both genes. One spot is depicted by a white arrow in each enlargement. For CAPRIN1, these yellow spots represent mRNA that has not undergone alternative splicing. For MKNK2, in addition to non-alternatively spliced mRNA, these spots also represent isoform 2. Scale bars, 5 μM. Background subtraction (ImageJ) applied to MKNK2 enlargement. (C) Plot showing the abundance of detected transcription sites in HeLa and Rpe1 cells. Bars represent mean±s.d. (n=3). (D) Histogram of mRNA distance from transcription sites for CAPRIN1 shows enrichment of unspliced mRNA up to several microns away from transcription sites. n depicts number of cells per plot. Data represent mean±s.d. (n=3). Source data is available for this figure at www.nature.com/msb.
Figure 5
Figure 5
FIA variability is close to the theoretical minimum in Rpe1 but not in HeLa cells. (A, B) Histograms of isoform variability for CAPRIN1 (A) and MKNK2 (B). Bars represent mean±s.e.m. (n=3). (C, D) FIA is not correlated with total cytoplasmic mRNA. An example scatter plot of FIA1 versus total mRNA is shown (D). Bars represent mean±s.d. (n=3). (E) Plot quantifying the increase in measured FIA variability above the theoretical minimum established by a binomial partitioning of isoforms. Bars represent mean±s.d. (n=3). Source data is available for this figure at www.nature.com/msb.
Figure 6
Figure 6
Sources of cell-to-cell variability in fractional isoform abundance. (A) Schematic of sources of cell-to-cell variability in FIA. (BD) Fluctuations in synthesis rate are a source of FIA variability, particularly for MKNK2, but cannot explain the increased variability in HeLa cells compared with Rpe1 cells. (B) Schematic demonstrating how synthesis rate fluctuations, in this case transcriptional bursting, together with differential isoform lifetimes can lead to FIA fluctuations and thus increase FIA variability. (C) Measurements of isoform lifetimes show differential isoform stability primarily for MKNK2. Data represent values from a fitted curve±s.e.m. (D) Measured total cytoplasmic mRNA abundance variability is greater than that predicted in the absence of synthesis rate fluctuations (see Materials and methods and Supplementary Theory). (E) Total cytoplasmic mRNA, but not FIA, is correlated with DAPI intensity. (FH) SRSF1 knockdown causes a large increase in FIA variability beyond the theoretical minimum (binomial isoform partitioning). Efficient knockdown of SRSF1 was confirmed by western blotting (F), and led to decreased isoform 1 abundance and increased isoform 2 abundance (G). (H) FIA variability compared with binomial isoform partitioning. Bars in (D, E, G, and H) represent mean±s.d. (n=3). Source data is available for this figure at www.nature.com/msb.

Comment in

Similar articles

See all similar articles

Cited by 44 articles

See all "Cited by" articles

References

    1. Allemand E, Batsche E, Muchardt C (2008) Splicing, transcription, and chromatin: a menage a trois. Curr Opin Genet Dev 18: 145–151 - PubMed
    1. Barash Y, Calarco JA, Gao W, Pan Q, Wang X, Shai O, Blencowe BJ, Frey BJ (2010) Deciphering the splicing code. Nature 465: 53–59 - PubMed
    1. Bengtsson M, Stahlberg A, Rorsman P, Kubista M (2005) Gene expression profiling in single cells from the pancreatic islets of Langerhans reveals lognormal distribution of mRNA levels. Genome Res 15: 1388–1392 - PMC - PubMed
    1. Buxade M, Parra-Palau JL, Proud CG (2008) The Mnks: MAP kinase-interacting kinases (MAP kinase signal-integrating kinases). Front Biosci 13: 5359–5373 - PubMed
    1. Chang HH, Hemberg M, Barahona M, Ingber DE, Huang S (2008) Transcriptome-wide noise controls lineage choice in mammalian progenitor cells. Nature 453: 544–547 - PMC - PubMed

Publication types

MeSH terms

Feedback