For years, transcriptional regulation took center stage in studies of gene expression, and the nuances of which transcription factor binds to which DNA sequence under which circumstances dominated the dialogue of molecular biologists.
In recent years, a supporting actor has threatened to steal the show: enter, post-transcriptional regulation. The idea that messenger RNA utilization could play an equally important role as transcriptional control in the regulation of gene expression was unthinkable to some critics. However, we now know that the translation of many, if not all, mRNAs is a dynamic process subject to multiple levels of fine-tuned controls. Sequences that influence mRNA utilization have been identified in virtually every part of the RNA message: in introns and exons of pre-mRNAs, and in coding and untranslated regions of mature mRNAs. RNA-binding proteins (RBPs) or noncoding RNAs (e.g., microRNAs) recognize and bind to specific RNA sequences, and these interactions can regulate the stability, splicing, nuclear export, subcellular localization, and translatability of mRNAs.
Because of post-transcriptional events, mRNA levels do not always correlate with the respective protein levels. Sequestration of mRNAs in ribonucleoprotein (RNP) complexes may enhance, or alternatively repress, translation without affecting steady-state RNA levels. Therefore, DNA microarray analysis of mRNA expression might not accurately reflect the levels of the encoded proteins. Researchers need a high-throughput technique to bridge the gap between transcriptomics and proteomics.
What is RIP-Chip?
Instead of being an epitaph for the mischievous cartoon chipmunk, RIP-Chip refers to the immunoprecipitation of RNPs from cell extracts and the subsequent microarray analysis of associated RNA molecules. RIP-Chip is the RNA analog of the more well-known ChIP-Chip (chromatin immunoprecipitation−microarray analysis), which identifies DNA targets of DNA-binding proteins. As a powerful method to probe mechanisms of post-transcriptional gene regulation, RIP-Chip has revolutionized the burgeoning field of ribonomics, or the genome-wide identification of RBP targets.
Most previous methods to identify RNA targets of RBPs were either limited in the ability to identify novel RNA targets or were in vitro techniques that could not be applied to cell extracts. The RIP-Chip method, first published in 2001 by Dr. Jack Keene’s lab at Duke University Medical Center,1 offers simultaneous, global detection of RNA targets of RBPs in cell extracts. Endogenous ribonucleoprotein (RNP) complexes are immunoprecipitated from cell extracts with an antibody against the RBP of interest. Then, the bound RNA is extracted from the RNP, purified, and identified by microarray analysis.
RIP Success Stories
Several researchers have employed RIP-Chip to identify RNA targets of their favorite RBPs. For example, Michael Whitfield, Ph.D., and colleagues at Dartmouth Medical School and the University of North Carolina at Chapel Hill used the method to identify mRNAs bound to the histone stem-loop binding protein (SLBP).2 SLBP binds to a conserved stem-loop at the 3’ end of histone mRNAs, thereby regulating mRNA stability and translation. Whitfield and colleagues used RIP-Chip to determine whether SLBP binds and regulates mRNAs other than the histone messages. The analysis revealed that the five classes of histone mRNAs were the protein’s exclusive targets in HeLa cells.
In the same study, Whitfield developed an in vitro variation of RIP-Chip known as “recombinant RIP-Chip” (rRIP-Chip). In this method, purified recombinant SLBP was incubated with purified total RNA, and SLBP and associated RNAs were immunoprecipitated. The RNA was purified and hybridized to a whole-genome microarray, allowing the researchers to identify messages that were enriched in the RNA fraction bound to SLBP. Somewhat surprisingly, the histone mRNAs were identified as the sole targets of SLBP both in vitro and in vivo. “We were concerned that we would obtain a lot of false positives with rRIP-Chip because you’re starting with a pool of free RNA, and the protein has the opportunity to bind a much larger set of RNA targets than it would in vivo,” Whitfield says. “However, at least with SLBP, the protein bound almost exclusively to its known targets.”??Whitfield thinks that rRIP-Chip represents a useful complement to RIP-Chip of cell extracts. “Endogenous RIP-Chip will always be the gold standard,” he says. “But rRIP-Chip could be used to search for all possible RNA targets of a protein.” Furthermore, rRIP-Chip identifies only direct interactions between the RBP and RNA, whereas endogenous RIP-Chip does not distinguish between direct interactions and interactions that occur through a protein bridge between the RBP and RNA.
In 2001, Jack Keene, the father of RIP-Chip, founded the company Ribonomics, which is exploring the use of RIP-Chip methodologies to diagnose and treat diseases such as diabetes and breast cancer. For example, Ribonomics scientists are examining changes in RNP complexes with glucose-regulated events in pancreatic beta cells. In response to glucose stimulation, beta cells produce insulin, which requires rapid changes in gene expression. Some of these changes in protein expression appear to occur at the post-transcriptional level. “We’re looking at acute changes in mRNAs moving into and out of RNP complexes in response to glucose,” says Bentley Cheatham, Ph.D., Director of Biology at Ribonomics. “What’s really exciting is that you can see these immediate changes by RIP-Chip, whereas they go undetected by a total transcriptomic analysis because the steady-state levels of the messages are not changing within this very short time frame.”Interestingly, Ribonomics scientists have discovered that the RNA-binding characteristics of a beta-cell RBP appear to be regulated by glucose-stimulated phosphorylation events.
The researchers hope to use RIP-Chip to link various signaling pathways to RBP function. “RIP-Chip is widely applicable; it can be utilized in any cell system or pathway. We are developing reagents and user-friendly methods that will enable research laboratories to readily implement RIP-Chip in their cell-based system of interest,” says Cheatham. “As more researchers become aware of RIP-Chip and it gains in popularity, RIP-Chip is going to be an important tool to look at some of these post-transcriptional changes.” Barry Henderson, Ph.D., Director of Platform Technologies at Ribonomics, says, “I think we’ve only begun to envision the potential applications of RIP-Chip.”
RIP: Current Limitations
Although our intention is not to rip on RIP-Chip, we must point out that, like any relatively new technology, RIP-Chip suffers from some limitations. A dearth of antibodies against RBPs represents a substantial bottleneck. Cheatham says, “We have a growing list of antibodies that are going through QC. All of our current antibodies are useful for western blotting to detect and quantify the RBP of interest, but getting antibodies that immunoprecipitate and isolate the RNP complexes is a stumbling block.” Whitfield hopes to circumvent the antibody generation issue by exploring the use of epitope-tagged proteins in RIP-Chip. “We don’t want to be in a position where we have to make antibodies against every RBP we want to study,” he says. “However, we don’t know whether the tagged protein will behave exactly like the endogenous one in RIP-Chip, so we need to do some studies comparing the two.”A second major bottleneck is data analysis. According to Cheatham, “With RIP-Chip, we’re dealing with some non-traditional problems in microarray data analysis.” Unlike standard differential expression analyses that compare total RNA levels between two different experimental conditions, RIP-Chip compares RNAs enriched in a particular RNP with total RNA. As a result, standard microarray analyses and normalizations don’t work for RIP-Chip. Ribonomics is developing a package of data analysis methodologies that will be made public as the company moves toward commercialization.
Whitfield agrees that data analysis is currently the rate-limiting step. He says, “My lab spends probably twice as much time on the RIP-Chip analysis as we do on the experiments.” Whitfield thinks that the data analysis quandary will be resolved by the development of novel algorithms, in addition to the emergence of the “interdisciplinary scientist.” He says, “We really need scientists who are as comfortable doing experiments as they are sitting at the computer doing analysis.” An issue related to RIP-Chip data analysis is the weeding out of false-positive RBP-RNA interactions. As mentioned earlier, endogenous RIP-Chip does not distinguish between direct interactions and RBP-RNA interactions that occur via protein bridges. False positives can also result from procedural errors. According to Whitfield, “Sorting out the false positives is not a trivial procedure. We ultimately tracked the false positives in our study to defects in homebuilt array manufacture that resulted in cross-contamination of different spots with histone mRNA.” When the RNA samples were hybridized to a commercial array platform (Agilent) that was more strictly quality-controlled, the false positives were eliminated. Another potential limitation of RIP-Chip is that different RNPs may require different experimental conditions for immunoprecipitation.
Most of the initial demonstrations of RIP-Chip efficacy have analyzed relatively tight RNA-RBP binding interactions. Weaker interactions might need to be stabilized in some way. In fact, some RNPs can dissociate and reassemble with other (perhaps nonphysiological) targets after cell lysis.3 Reversible formaldehyde crosslinking can be used to cement RNA-RBP interactions prior to cell lysis. However, crosslinking can decrease cell lysis efficiency, increase background, and introduce sequence bias. Furthermore, the starting material (which can be whole-cell lysates, nuclear extracts, or polyribosomes) can influence the RIP-Chip results. Whitfield says, “SLBP from whole-cell lysates doesn’t precipitate histone mRNA, but when we IP SLBP from polyribosomes, we get very nice binding. We don’t understand why that difference exists.” Therefore, each RBP to be analyzed might require specific optimization of the RIP-Chip protocol.
Future for RIP-Chip
With interest in post-transcriptional regulation accelerating at a breakneck speed, RIP-Chip is poised as a powerful discovery tool that is rapidly gaining in popularity and accessibility. Although most RIP-Chip applications have addressed mRNA-RBP interactions, the explosion in research on non-coding RNA such as microRNA is likely to drive the development of microarrays tailored specifically for RIP-Chip. Whitfield says, “Up until now, we’ve been using microarrays that are, for all intents and purposes, mRNA expression microarrays.” Ideally, microarrays for RIP-Chip would allow the detection of both coding and non-coding RNAs, as well as distinct splice variants. The combination of broader-coverage arrays with improvements to the RIP-Chip protocol might allow the pinpointing of precise RBP sites on RNA molecules. “Right now, we’re pulling down the entire message and hybridizing it to cDNA arrays,” Whitfield says. “I think the next steps are to add an RNA-shearing or -degrading step after the immunoprecipitation to eliminate the portion of the message that flanks the RBP binding site, and to have probes designed to more specifically localize these binding sites.”
Whitfield’s lab is exploring methods to more precisely identify cis-acting regulatory sequences in RNAs targeted by RBPs, including next-generation, high-throughput sequencing. Cheatham says, “RIP-Chip is clearly adaptable to next-gen sequencing. I don’t think there are going to be any big technical hurdles in linking these two technologies.” However, Ribonomics researchers are currently focusing their efforts on improving microarray-based downstream analyses. Henderson says, “We are pursuing the broadest possible commercialization of the RIP-Chip method, and that includes reagents and services. We see that as the key to wider adoption of the method.” Cheatham agrees, saying, “We really want to provide a cohesive package of methodologies to enable an investigator to jump right into RIP-Chip and get some novel data.”
- Tenenbaum, S. A., Carson, C. C., Lager, P. J., and Keene, J. D. (2000) Identifying?mRNA subsets in messenger ribonucleoprotein complexes by using cDNA?arrays. Proc. Natl. Acad. Sci. U.S.A. 97: 14085-14090.
- Townley-Tilson, W. H. D., Pendergrass, S. A., Marzluff, W. F., and Whitfield, M. L. (2006) Genome-wide analysis of mRNAs bound to the histone stem-loop binding protein. RNA 12: 1853-1867.?
- Mili, S., and Steitz, J. A. (2004) Evidence for reassociation of RNA-binding proteins after cell lysis: implications for the interpretation of immunoprecipitation analyses.