We describe a protocol for generating proliferating and quiescent primary human dermal fibroblasts, monitoring transcript decay rates, and identifying differentially decaying genes.
Quiescence is a temporary, reversible state in which cells have ceased cell division, but retain the capacity to proliferate. Multiple studies, including ours, have demonstrated that quiescence is associated with widespread changes in gene expression. Some of these changes occur through changes in the level or activity of proliferation-associated transcription factors, such as E2F and MYC. We have demonstrated that mRNA decay can also contribute to changes in gene expression between proliferating and quiescent cells. In this protocol, we describe the procedure for establishing proliferating and quiescent cultures of human dermal foreskin fibroblasts. We then describe the procedures for inhibiting new transcription in proliferating and quiescent cells with Actinomycin D (ActD). ActD treatment represents a straightforward and reproducible approach to dissociating new transcription from transcript decay. A disadvantage of ActD treatment is that the time course must be limited to a short time frame because ActD affects cell viability. Transcript levels are monitored over time to determine transcript decay rates. This procedure allows for the identification of genes and isoforms that exhibit differential decay in proliferating versus quiescent fibroblasts.
Steady state levels of transcripts reflect the contribution of both transcript synthesis and transcript decay. Regulated and coordinated transcript decay is an important mechanism for controlling biological processes1,2,3,4. For example, transcript decay rates have been shown to contribute to the temporal series of events following activation by the inflammatory cytokine tumor necrosis factor5.
We have previously shown that the transition between proliferation and quiescence in primary human fibroblasts is associated with changes in the transcript levels of many genes6. Some of these changes reflect differences in the activity of transcription factors between these two states.
Changes in a transcript's decay rate can also contribute to changes in the expression level of a transcript in two different states7,8. Based on our earlier findings that the transition between proliferation and quiescence is associated with changes in the levels of multiple microRNAs9, we asked whether there is also a contribution from differential transcript decay in the changes in gene expression in proliferating versus quiescent fibroblasts.
In order to monitor transcript stability, we determined the rate of decay of transcripts genome-wide in proliferating versus quiescent cells. To achieve this, we monitored decay rates in proliferating versus quiescent fibroblasts by introducing an inhibitor of new transcription and monitoring the rate at which individual transcripts disappeared over time. The advantage of this approach is that, as compared to methods that simply monitor overall gene expression, by inhibiting new transcript synthesis, we will be able to determine the decay rate for these transcripts separately from the rate at which they are transcribed.
ActD treatment to inhibit new transcription and determine transcript decay rates has been successfully applied in multiple previous studies. ActD has been used to dissect the importance of RNA stability in the changes in transcript abundance that result from treatment with pro-inflammatory cytokines5. A similar approach has also been used to dissect the differences in transcript decay rate in proliferating versus differentiated C2C12 cells as they adopt a differentiated muscle phenotype10. As another example, global mRNA half-lives have also been detected in pluripotent and differentiated mouse embryonic stem cells11. In these examples, mRNA decay has been shown to be important for regulating transcript abundance and for the transition of cells to different cell states.
Applying the methods described below, we discovered changes in transcript decay rates in approximately 500 genes when comparing fibroblasts in proliferating and quiescent states12. In particular, we discovered that the targets of the microRNA miR-29, which is downregulated in quiescent cells, are stabilized when cells transition to quiescence. We describe here the methodology we used to determine decay rates in proliferating and quiescent cells. This methodology is useful for comparing global mRNA decay rates in two distinct but similar conditions when information about rapidly decaying genes is sought. It could also be used to address other questions such as the effect of cell culture conditions on transcript decay, for instance, in two dimensional versus three dimensional cultures. Decay rates can be determined genome-wide with methods such as microarrays or RNA-Seq. Alternatively, real-time qPCR or Northern blotting can be used to determine decay rates on a gene-by-gene or isoform-by-isoform basis. These rates can then be used to calculate the half-life of each monitored gene and to identify genes with decay rates that are different in two conditions.
All experiments described were approved by Institutional Review Boards at Princeton University and the University of California, Los Angeles.
1. Prepare Proliferating and Contact-inhibited Fibroblasts for ActD Time Course
NOTE: This protocol uses a timecourse with four timepoints. Three biologically independent samples can be collected per timepoint by collecting different tissue culture plates in one experiment, or the experiment can be repeated multiple times with different cultures of cells. In our experience, one 10 cm (diameter) tissue culture dish (one plate) provides sufficient RNA for analysis. If needed, multiple tissue culture dishes can be pooled for each sample to increase the number of cells at each timepoint. In addition, the same experiment can be performed with fibroblasts isolated from different individuals as truly independent biological replicates.
2. ActD Time Course
3. Isolation of Total RNA from Phenol-guanidine Isothiocyanate Solution Lysate
4. Analysis of Transcript Abundance
5. Decay Rate Constant Determination
We have previously reported the results of microarray analyses of transcript decay rates in proliferating and contact-inhibited primary human fibroblasts over an 8-hour time course12. A list of genes with a significant change in transcript stability comparing proliferating and contact-inhibited fibroblasts is provided in Supplementary Table 1. The fluorescence intensities at time zero and over a time course after ActD treatment are provided. Genes were included on the list by determining the genes that have significantly different slopes in proliferating and contact-inhibited conditions using an ANOVA F-test. Examples of genes with different decay rates in proliferating and quiescent fibroblasts are shown in Figure 5.
Figure 1: Schematic of protocol for establishing proliferating and contact-inhibited plates for ActD time course analysis. A day-by-day description of the necessary steps to establish cultures of cells for ActD treatment is provided, assuming four timepoints will be used for the time course. For simplicity, one plate is shown per timepoint, but additional plates can be added to generate biological replicates. The protocol can also be adjusted to add more plates of cells if more than four timepoints are desired. Please click here to view a larger version of this figure.
Figure 2: Schematic of ActD time course and decay rate calculations in quiescent compared to proliferating human fibroblasts. The expected fraction of mRNAs remaining over time for a hypothetical transcript after treatment with a transcriptional inhibitor is shown. Below, sample plots of the fraction of mRNA remaining over time are provided for genes that are more stable in quiescent than proliferating cells, and genes that are more stable in proliferating than quiescent cells. For each datapoint, only a single value is shown for clarity. In the actual protocol, there will be three datapoints for each timepoint. This figure has been modified from Elizabeth Pender Johnson's Ph.D. thesis with her consent. Please click here to view a larger version of this figure.
Figure 3: Photographs of mixing phenol-guanidine isothiocyanate reagent and chloroform before and after centrifugation. The mixture is shown when the phenol-guanidine isothiocyanate reagent and chloroform are initially combined, after mixing and after centrifugation. After the mixture is centrifuged, the bottom phase, which is red, is the organic phase. The interphase in the middle is white and cloudy. The top phase is the aqueous phase and it is clear. The RNA is in the aqueous phase, which should be collected with a pipette for precipitation. Please click here to view a larger version of this figure.
Figure 4: Sample RNA gel to monitor RNA quality. An illustration of an RNA gel showing samples with intact RNA. Intact RNA has prominent bands for the 28S and 18S rRNAs. Please click here to view a larger version of this figure.
Figure 5: Examples of genes with different decay rates in proliferating versus quiescent fibroblasts. Fluorescence intensities (gene expression) over time are shown for four genes that decay differently in proliferating versus quiescent fibroblasts. Please click here to view a larger version of this figure.
Supplementary Table 1: Gene-specific fluorescence intensities in proliferating and quiescent fibroblasts. Data are provided for time zero and over a time course after ActD treatment in proliferating and contact-inhibited fibroblasts. P values, q values, and false discovery rates are provided. Table is reprinted from its original publication in BMC Genomics12. Please click here to download this file.
Quiescence can be induced by external signals including withdrawal of mitogens or serum, lack of cell adhesion, and contact inhibition. Contact inhibition, one of multiple possible methods for inducing quiescence, is a highly evolutionarily conserved process in which cells exit the proliferative cell cycle in response to cell-to-cell contact. We focus here on contact inhibition as an example of a method to induce quiescence. Previous studies have reported that cell-cell contact can affect microRNA biogenesis25, thus, the results provide information on one quiescence method and additional studies are required to determine which changes are associated with quiescence per se.
We used primary human diploid fibroblasts that we isolated from neonatal skin. We initiated these experiments with fibroblasts that were passaged fewer than 10 times. We discarded later-passage fibroblasts because they senesce. While this protocol focuses on proliferating and quiescent dermal fibroblasts, the procedures described can be used to monitor decay rates in different types of cells and under different conditions.
We monitored mRNA decay rates genome-wide using a transcription shut-off time course. Shut-off time courses are the most widely used method for decoupling mRNA decay from mRNA transcription in order to calculate mRNA half-lives1,26. In the method described here, ActD, a drug that complexes with the B-form of DNA to block elongation of RNA Polymerase II, is used to arrest transcription of mRNAs27. The rate that transcripts decrease in abundance over an 8-hour time period can be determined as the transcript's decay rate.
An important limitation of ActD-based approaches for monitoring transcript decay is that ActD prevents global transcription, which will affect the cells' viability. This should be kept in mind when interpreting ActD-based data, as decay rates are being determined in cells that are more stressed than their native state. Another important concern is that the effects of ActD may be different for cells in different states, including proliferating versus contact-inhibited fibroblasts. One approach for minimizing the side effects of ActD treatment is to keep time courses short in duration. For this reason, we selected an 8 h time course.
There are alternative approaches to ActD for monitoring transcript decay rates28,29. Yeast strains with temperature-sensitive RNA polymerase II alleles have been used to monitor transcript decay rates30. A variation on this approach called "anchor-away" has been used in yeast to achieve conditional depletion of RNA polymerase II in response to rapamycin treatment31,32. It is also possible to monitor decay in cells with active transcription. For instance, thio-substituted uracil nucleotides can be introduced into cells, incorporated into mRNA, and subsequently detected33,34,35. With this approach, the rate at which different transcripts are synthesized can be determined in cells that are continuing to synthesize mRNA. Such approaches have a significant advantage over ActD treatment because they are not toxic. In our experience, however, there can be variability in the recovery of transcripts with these methods. ActD treatment is a straightforward and reproducible approach for monitoring transcript decay.
A few steps in the protocol provided are critical for its success. One important issue is preventing RNA degradation during RNA isolation. At step 4.1, the RNA should be monitored for degradation. If the RNA does not contain two clear peaks for 18S and 28S rRNAs (Figure 4), this is a sign that transcript degradation has occurred. Some instruments provide RNA Integrity Number (RIN) scores ranging from 1 to 10. Samples with RIN scores in the 7-10 range are acceptable for further analysis. If the RNA is degraded, then it is important to take more precautions to prevent RNA degradation, such as ensuring that solutions and sterileware are RNase-free, and that gloves are worn when handling pipettes for RNA. Water can be pretreated with 0.1% diethyl pyrocarbonate (DEPC), incubated for at least 2 h at 37 °C, and autoclaved for at least 15 min. Note that DEPC cannot be used with Tris-based buffers. RNase decontaminating solutions can also be used to remove RNase from work surfaces, centrifuges, and pipettes, and reagents that inhibit RNases can be added to samples.
Another critical step is the separation of the phases of the phenol-guanidine isothiocyanate solution. It is important to ensure that only the top phase is collected for RNA processing. If the resulting RNA includes residual phenol, the ratio of absorbance at 230, 260, and 280 nm will not be in the ratio 1:2:1 as expected. Higher than expected absorbance at 230 nm can indicate phenol contamination. If this occurs, the RNA can be ethanol precipitated again and washed several additional times to remove any residual phenol.
Finally, when the RNA is pelleted and the supernatant is removed, the pellet can be sticky, thin, and easily disturbed. At this point, it is important to take extra care in taking off the liquid. It may be important to visualize the tube with good lighting to locate the pellet and make sure it is not disturbed. Added glycogen can help to ensure the pellet is visible at this point.
The selection of time points for the time course of decay is also an important issue to consider. While the four timepoints we selected allowed us to monitor decay for over 10,000 genes, some genes did not significantly decay during the 8 h time course. If short-lived transcripts are a high priority, the investigators may want to add additional sample collections prior to 120 min, for instance at 30 min. For better assessment of long-lived transcripts, the investigators may want to add additional sample collections at later timepoints.
Another potential issue could result if the decay rates do not follow the expected distribution. For instance, there may be too many or too few genes showing a log-linear decay rate over 8 h. If this happens, it is helpful to monitor the decay rates of genes that are known to be short or long-lived in order to determine whether they follow the expected pattern. If genes have a positive rather than negative slope over time, they can be excluded from further analysis as they do not demonstrate a log-linear decay rate. A final issue could arise if the cells in the two conditions being compared have very different responses to the ActD such that many genes decay faster in one condition. In this case, it may be helpful to refer not only to the difference in decay rate between the two states, but also to the spike-in controls that we suggest including prior to RNA isolation to provide information on absolute decay rates in the two states.
The ability to monitor transcript decay rates between proliferating and quiescent cells or any two similar states has the potential to provide further insight into the importance of transcript decay in regulating physiological changes. Applications include cells with and without oncogenic activation, before and after senescence or viral infection, with and without knockdown of a specific gene, or in two different states of differentiation. The findings can be coupled with RNA-Seq to provide information about isoform-specific changes in transcript decay rates, which may provide insight into the changes in isoform expression observed as cells undergo physiological transitions.
The authors have nothing to disclose.
HAC was the Milton E. Cassel scholar of the Rita Allen Foundation (http://www.ritaallenfoundation.org). This work was funded by grants to HAC from the National Institute of General Medical Sciences Center of Excellence grant P50 GM071508 (P.I. David Botstein), PhRMA Foundation grant 2007RSGl9572, National Science Foundation Grant OCI-1047879 to David August, National Institute of General Medical Sciences R01 GM081686, National Institute of General Medical Sciences R01 GM0866465, the Eli & Edythe Broad Center of Regenerative Medicine & Stem Cell Research, the Iris Cantor Women’s Health Center/UCLA CTSI NIH Grant UL1TR000124, and the Leukemia Lymphoma Society. Research reported in this publication was supported by the National Cancer Institute of the National Institutes of Health under Award Number P50CA092131. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. HAC is a member of the Eli & Edythe Broad Center of Regenerative Medicine & Stem Cell Research, the UCLA Molecular Biology Institute, and the UCLA Bioinformatics Interdepartmental Program.
Centrifuges for microcentrifuge tubes capable of reaching 12,000 x g and 4°C | |||
Equipment for running agarose gels | |||
Sterile tissue culture plates and conical tubes | |||
Dulbecco's Modified Eagle Medium | Life Technologies | 11965-118 | |
Fetal bovine serum | VWR | 35-010-CV | |
Sterile serological pipets, pipettors and pipet tips for tissue culture | |||
Individually wrapped, disposable Rnase-free pipettes, pipette tips and tubes for RNA isolation and analysis | |||
Disposable gloves to be worn when handling reagents and RNA samples | |||
RNaseZap | Invitrogen | AM9780 | For decontaminating work surfaces from RNase |
2.0 ml eppendorf tubes | |||
Trypsin-EDTA Solution 10X | Millipore Sigma | 9002-07-07 | |
Sterile PBS | Life Technologies | 14190-250 | |
Trizol | Thermo Fisher Scientific | 15596018 | |
Actinomycin D | Millipore Sigma | A1410-2 mg | inhibits transcription |
Sterile DMSO | Fisher Scientific | 31-761-00ML | solvent for actinomycin D |
Chloroform | Thermo Fisher Scientific | ICN19400225 | MP Biomedicals, Inc product |
Isopropanol | Fisher Scientific | BP2618500 | molecular biology grade |
Ethanol | Fisher Scientific | BP28184 | molecular biology grade |
RNase-free Glycogen (20 mg/ml aqueous solution) | Thermo Fisher Scientific | R0551 | carrier for RNA precipitation |
TURBO DNA-free Kit | Thermo Fisher Scientific | AM1907 | removes DNA with DNase, then, in a subsequent step, inactivates DNA and removes divalent cations |
Agarose | Thermo Fisher Scientific | ICN820721 | MP Biomedicals, Inc product |
Loading dye for RNA gel | Thermo Fisher Scientific | R0641 | suitable even for denaturing electrophoresis |
Millenium RNA markers | Thermo Fisher Scientific | AM7150 | RNA ladder |
One-color RNA Spike-in Kit | Agilent Technology | 5188-5282 | Example spike-in control for Agient microarray analysis |
Tris base | Fisher Scientific | BP152-1 | molecular biology grade, for TAE buffer |
Glacial acetic acid | Fisher Scientific | A38-500 | for TAE buffer |
EDTA | Fisher Scientific | BP120-500 | electrophoresis grade, for gels |
Ethidium bromide | Millipore Sigma | E7637 | for molecular biology |
External RNA Controls Consortium RNA Spike-in Mix | Thermo Fisher Scientific | 4456740 | Spike-in control for RNA Seq |
TruSeq Stranded mRNA Library Preparation Kit A (48 samples, 12 indexes) | Illumina | RS-122-2101 | for RNA Seq |
96-well 0.3 ml PCR plate | Thermo Fisher Scientific | AB-0600 | for real-time qPCR |
Microseal B adhesive seals | Bio-Rad | MSB1001 | for real-time qPCR |
Rnase/Dnase-free Reagent Reservoirs | VWR | 89094-662 | for real-time qPCR |
Rnase/Dnase-free Eight tube strips and caps | Thermo Fisher Scientific | AM12230 | for real-time qPCR |
SuperScript Reverse Transcriptase | Invitrogen | 18090010 | for real-time qPCR |
AMPure XP Beads | Beckman Coulter | A63880 | for real-time qPCR |