Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

New Tool Profiles mRNA Ends of Specific Cells in Mouse Brain

20 Sep 2017

by Marina Chicurel

This article appears by special arrangement with Alzforum. See original article with additional links/commentary.

Researchers have worked out a technique to capture snapshots of the 3' ends of mRNA species in distinct cell populations derived from intact tissue. In the September 13 Neuron, Robert Darnell at the Rockefeller University in New York and colleagues use the method to analyze brain cells, and report on unique 3' end profiles of mRNAs in excitatory and inhibitory neurons, as well as in astrocytes and microglia. They show examples of how alternative polyadenylation contributes to protein diversity between cell types, and between the resting and activated states of microglia.

  • cTag-PAPERCLIP maps the 3' ends of mRNAs in specific cell types.
  • cTag-PAPERCLIP avoids cell fractionation.
  • Different cell types use different polyadenylation sites.
  • A change in polyadenylation of the ARAF kinase is linked to microglial activation.

Scientists urgently want to dissect the roles of different cell types in health and disease. In Alzheimer’s disease research, tracking microglia has become a priority in the wake of genetic findings pointing to these cells as key contributors. Microglia-specific markers are allowing researchers to isolate microglia and study their gene-expression profiles (Bennett et al., 2016), and single-cell sequencing techniques have enabled discovery of rare subtypes relevant to disease.

But these studies require the disruptive process of cell isolation, which includes mashing up the brain, exposing tissues to enzymes, and running cells through a FACS machine or a column. During these manipulations, which take hours, short-lived mRNAs decay, cellular processes break off, stress-related genes turn on, and microglia become activated (Cardona et al., 2006).


Building on a technique called “poly(A) binding protein-mediated mRNA 3' end retrieval by crosslinking immunoprecipitation,” or PAPERCLIP for short, the new method circumvents cell isolation. First author Hun-Way Hwang at the University of Pittsburgh invented PAPERCLIP to map the 3' sequences of polyadenylated mRNAs (Hwang et al., 2016). The sequences are of interest because they help control how long mRNAs survive, where they are in a cell, and their translational efficiency. 

PAPERCLIP uses an antibody against the polyA-binding protein (PABP) to pull out mRNAs by their 3' tails from intact tissues, after shining ultraviolet light to crosslink the protein to the RNA (see image above). Since PABP, a protector and regulator of mRNA function, binds to the polyA tails of nearly all mRNAs, the antibody targets essentially all mRNAs present at the moment of crosslinking. The researchers then trim the mRNAs, run the crosslinked complexes on a gel, and isolate those in the desired size range. After digesting away the PABP with proteinase K, and purifying the RNA bits, the samples are ready for high throughput sequencing. “It allows us to look at the net aggregate of all RNAs with different polyA tails, without disturbing the cells at all,” said Darnell.

In their new paper, Hwang and colleagues add cell type specificity to PAPERCLIP. The researchers used the cre-lox system to generate mice expressing PABP with a green fluorescent protein (GFP) tag under the control of promoters specific to brain cell types. Hwang designed a hybrid construct with sequences from the last exon of the PABP gene (Pabpc1) linked to the GFP gene, so that the Cre recombinase would swap the endogenous PABP exon for the tagged version. He then bred mice carrying the construct to mice expressing Cre recombinase in either excitatory neurons (Camk2a-Cre), inhibitory neurons (Gad2-Cre), astrocytes (mGfap-Cre), or microglia (Cx3cr1-Cre). Using tissues from these mice, the researchers cast a pair of GFP antibodies to pull out mRNA tails from specific cell types.

Comparing mRNAs retrieved from microglial cells with cTag-PAPERCLIP to those obtained from FACS-based microglial isolation, Hwang observed that both approaches yielded similar results for mRNAs encoding microglial markers, but the cTag-PAPERCLIP samples expressed lower levels of activation markers and immediate early genes.

The new study also illustrates how cTag-PAPERCLIP can be used to study mRNA regulation. “During pre-mRNA splicing, the choice is made of when to put on a polyA tail, giving rise to different transcripts,” said Darnell. Hwang found 27 genes that shifted their polyA sites in microglia from mice treated with lipopolysaccharide, a microglial activator, compared to untreated controls. One of these genes, encoding the ARAF kinase, switched from producing more of a short transcript than a long one to producing more of the long one after activation. The long transcript encodes the functional kinase, while the short one encodes a dominant-negative regulator. “It turns out that the differential use of polyA sites is quite important. It can lead to all kinds of functional differences,” said Darnell.

Could researchers studying neurodegenerative disease add cTag-PAPERCLIP to their toolbox? Darnell thinks so, noting the technique could be used to probe alternative polyadenylation in specific cells in mouse models of disease. Tristan Li, Stanford University in California, added that it might also provide insight into aging. However, Li cautioned it is difficult to predict if cTag-PAPERCLIP would inform studies about microglia’s role in disease. Whether polyadenylation plays an important role is still unknown, and the change in the ratio of ARAF transcripts appears to be modest, Li noted. In addition, although Li considers the Cx3cr1 promoter used by Hwang “the best available” to target CRE expression to microglia, circulating monocytes and macrophages in various tissues also use it.

Could cTag-PAPERCLIP be used to examine gene expression profiles, independently of 3' tail lengths? Li noted that classic RNA sequencing methods rely on sampling RNAs along their entire lengths to get overlapping, confirmatory data, hence he worried that this heavily 3' biased method may be less reliable. For his part, Darnell noted that by reducing the RNAse treatment, longer pieces of RNA can be retrieved to identify species with greater certainty.

Darnell thinks the biggest limitation is that cTag-PAPERCLIP can be applied only to mice. “The elephant in the room is the mouse,” he joked. He is optimistic, however, that bioinformatics analyses of large human genetic datasets will help translate the mouse findings into meaningful insights for humans.