A single molecule view of gene expression

Abstract

Analyzing the expression of single genes in single cells appears minimalistic in comparison to gene expression studies based on more global approaches. However, stimulated by advances in imaging technologies, single-cell studies have become an essential tool in understanding the rules that govern gene expression. This quantitative view of single-cell gene expression is based on counting mRNAs in single cells, monitoring transcription in real time, and visualizing single proteins. Parallel advances in mathematical models based on stochastic, discrete descriptions of biochemical processes have provided crucial insights into the underlying cellular mechanisms that control expression. The view that has emerged is rooted in a probabilistic understanding of cellular processes that quantitatively explains both the mean and the variation observed in gene-expression patterns among single cells. Thus, the close coupling between imaging and mathematical theory has established single-cell analysis as an essential branch of systems biology.

Figure 1. mRNA detection in single cells

Two common methods for single cell gene expression analysis using imaging: A) Single molecular resolution fluorescence in situ hybridization (FISH) uses synthetic oligonucleotides labeled at multiple positions with fluorescent dyes to detect single mRNAs. Multiple fluorescent probes are hybridized to paraformaldehyde fixed cells. FISH allows the detection of single mRNAs in the cytoplasm as well as nascent mRNAs at the site of transcription. On the right, yeast cells expressing MDN1 mRNA and mammalian CHO cells (hamster cell line) expressing a doxycyclin induced reporter are shown,. B) The MS2 system uses the specific interaction between the MS2 RNA hairpin and a fusion of a fluorescent protein and the MS2 phage coat protein to create a fluorescently labeled mRNA. Inserting multiple binding sites into an mRNA allows detection of single mRNAs in living cells. The MS2 system has been used in different organisms, shown here, for example, in E. coli to count single mRNAs (with permission from Ref 27) and in Dictyostelium to determine transcription kinetics in real time (with permission from Ref 30). Sites of transcription are marked by arrows.

Figure 2. A stochastic model of gene expression

A) Left, the central dogma of molecular biology – DNA to RNA to protein – is shown with rate constants of production and degradation: the rate of transcription ν₀, the rate of RNA degradation d₀, the rate of translation ν₁, and the rate of protein decay d₁. Once degradation, the RNA and protein are Right, a model of gene induction known as the telegraph model. The gene transitions between an inactive off state and an active on state (red line). From the active state, transcripts initiation events (vertical green lines) are separated by an average time interval ν₀⁻¹. B) The probability distributions (P_x) for each step of MDN1 expression, from left to right: nascent RNA at a transcription site (P_m, m = number of nascent chains), total cellular mRNA (P_r, r = number of mRNA), proteins/mRNA (P_n, n = number of proteins/mRNA), and total protein/cell (P_N, N = number of proteins/cell). The gray symbols are published data; the red lines are theoretical fits using the equations shown below each panel. The data for nascent transcripts and total mRNA is from . There is no experimental data for proteins/mRNA for MDN1. The data for total protein/cell was reported as a mean and variance (σ²) ,, shown here as a Gaussian distribution. The parameters in the probability distributions are shown in panel A. τ is the time to synthesize an MDN1 transcript. Γ denotes the gamma function. Upon degradation, neither the RNA nor the protein is considered in the probability distribution, so the state of degraded RNA and degraded protein is indicated by the symbol ϕ. Two additional symbols are used for simplicity: a is the ratio of transcription rate/protein degradation rate (ν₀/d₁); b is the translation rate/RNA decay (ν₁/d₀), also known as the protein burst size. For MDN1, a=19, b=29.