Autopsy of a Transcriptome: Zombie Genes and the Non-Believers

It’s a pretty universal question:

What happens to us after we die?

To question our own mortality is human. Through the lens of biology, we can further question our existence by asking:

What happens to our bodies after we die? What does it mean for our cells to be alive after we are no longer living?

We know that our cells don't die when we die; this is what makes organ donation possible. But what does death of an organism mean for an individual cell? How does each cell decide how long to stay alive and when to shut down for good? Some researchers believe they may have found a way to answer this question using genomics. Enter the thanatotranscriptome and the discovery of zombie genes, genes that “wake up” in our cells after we die. But like any theory supposing an afterlife, human or cellular, many are skeptical.

Cells produce readouts of genes, called RNA transcripts, to make proteins. The pool of RNA transcripts within a cell or tissue is called the transcriptome. The thanatotranscriptome (thanato-: death) represents all of the RNA transcripts that are present when or produced after an organism dies. By comparing the thanatotranscriptome at different points in time, researchers gain insight into what kind of cellular changes occur after death. One article published this past June, entitled “Thanatotranscriptome: genes actively expressed after organismal death” (which we covered here at Signal to Noise), examined thanatotranscriptomes over many days and concluded that cells “wake up” a program of specific genes even days after death [1]. They found genes whose RNA transcript abundance increased in the pool at 48-96h postmortem, which led the authors to believe that there are transcriptional programs that turn on after death, hence: zombie genes.

What’s notable about this paper is not the paradigm-shifting finding that zombie genes may exist, but that unbeknownst to the public, the scientific community actively retaliated against the supposition.

The media presented the conclusions of the paper as a new scientific discovery:

 

However, scientists in the field were less than convinced:

In the public sphere, the findings were portrayed as wins for the forensic and organ transplant disciplines. If cells produce certain RNA transcripts at known time intervals following death, we might be able to determine when someone died, or how viable a donated organ is. However, these technologies are dependent on our ability to reliably interpret postmortem gene expression patterns. The authors found an increase of certain transcripts at later time points, which can be interpreted many ways. Unfortunately, the experimental design of this study left much room for interpretation. There are always caveats to transcriptome analyses that are not often publicly discussed. Some of the caveats are presented below, but ultimately, the authors were unable to prove that the cell “wakes up” these genes by making new RNA transcripts. 

 

What the paper suggests: Active transcription, aka zombie genes “wake up” after organismal death

FIGURE 1: ACTIVE TRANSCRIPTION. AFTER THE ORGANISM DIES, THE CELL MAKES MORE COPIES OF RNA TRANSCRIPT A.

FIGURE 1: ACTIVE TRANSCRIPTION. AFTER THE ORGANISM DIES, THE CELL MAKES MORE COPIES OF RNA TRANSCRIPT A.

Active transcription means that the cell continued making copies of RNA transcript A after death, which is why it is more abundant at 48 hrs (Figure 1).
 

Alternative explanation #1: RNA degradation/decay

FIGURE 2. DEGRADATION OF OTHER TRANSCRIPTS IN THE POOL MAKE RNA TRANSCRIPT A APPEAR MORE ABUNDANT OVER TIME, THOUGH NO NEW TRANSCRIPTS ARE PRODUCED.

FIGURE 2. DEGRADATION OF OTHER TRANSCRIPTS IN THE POOL MAKE RNA TRANSCRIPT A APPEAR MORE ABUNDANT OVER TIME, THOUGH NO NEW TRANSCRIPTS ARE PRODUCED.

RNA transcripts are made by the cell, but also degraded. When we measure RNA in a sample, we look at the relative abundance of a transcript - how much RNA transcript A is in the whole pool? If RNA transcripts degrade at different rates over time, the pools between time points will inherently vary [2]. Figure 2 illustrates how RNA transcript A can be interpreted as more abundant at 48 hr when accounting for degradation.

The relative abundance of RNA transcript A is the number of transcripts divided by all RNA transcripts in the pool. RNA transcript A makes up 4 of the 14 total transcripts in the live cell. Its relative abundance is 4/14, or 28.6%. At 48 hours, its relative abundance is 4/8 or 50%. Without knowing the absolute total number of molecules in each pool (we empirically cannot), this increase from 28.6% to 50% might suggest that transcript A is higher at 48 hours because the cell produced more RNA. However, we cannot rule out that RNA is not being produced, and the difference arises from other transcripts in the pool degrading. 

 

 

 

 

Alternative explanation #2: Cell type heterogeneity

FIGURE 3. THE EFFECT OF CELL TYPE WHEN COMPARING DIFFERENT TISSUE SAMPLES. DIFFERENCES IN TRANSCRIPT B AT 48hr DUE TO VARIABLE COMPOSITION. TRANSCRIPTOMES ARE BEST STUDIED AT THE INDIVIDUAL CELL-TYPE LEVEL. 

FIGURE 3. THE EFFECT OF CELL TYPE WHEN COMPARING DIFFERENT TISSUE SAMPLES. DIFFERENCES IN TRANSCRIPT B AT 48hr DUE TO VARIABLE COMPOSITION. TRANSCRIPTOMES ARE BEST STUDIED AT THE INDIVIDUAL CELL-TYPE LEVEL. 

Another major caveat to the thanatotranscriptome study, and with all transcriptome studies analyzing RNA from tissue, is the mixed/heterogenous property of tissue. Tissue is made up of many different types of cells, each with its own unique pattern of gene expression. In conjunction with the previous caveat, different cell types also degrade RNA at different rates [3]. Figure 3 depicts tissue samples made up of orange and blue different cell types, where RNA transcript B is only produced by blue cells.

The two different tissue samples were taken at different time points. The composition of the two tissues vary, where the blue cell type makes up 40% of the tissue in sample #1 and 80% of the tissue in sample #2. These differences may be meaningful to biology (different cell types have different lifespans, so some cell types persist longer than others especially after death), but they might also vary from the sampling process and human error. The bottom line is that uneven mixtures make things difficult to compare. When analyzed, there would be a greater abundance of RNA transcript B in the tissue from 48 hours. However this increase isn't due to increased production of RNA transcript B, but an effect of the sampled tissue at 48 hours having more blue cells.

 

Despite the skepticism, the concept of the thanatotranscriptome is not dead yet (pun intended). The field of transcriptomics is still growing, and the concept of the thanatotranscriptome is still new. Deconstructing a cell's shut down process would have clear utility: what are the decisions a cell must make to stay alive, starting from the instant the organism stops living? Uncovering these patterns still might be useful, but the scientific community must first reach consensus on what's real and what isn't. And that consensus has yet to be reached. This paper examined thanatotranscriptomes to find mechanisms that all cells share before dying, and discovered an abundance of “zombie gene” transcripts in late postmortem samples. But the truth is, just because we observe transcripts differences, it still doesn’t mean cells are “waking up” these genes over time.


Chantle Edillor (@chantleedillor)
Social Media Coordinator, Signal to Noise Magazine
PhD Student, Human Genetics

 

References
[1] Pozhitkov A.E. et al. Thanatotranscriptomegenes actively expressed after organismal death. bioRxiv 058305; doi: http://dx.doi.org/10.1101/058305 (2016).
[2] Chen H. et al. Genome-wide study of mRNA degradation and transcript elongation in Escherichia coli. Mol Syst Biol. 11(1), 781(2016).
[3] Babapulle C.J. et al. Cellular changes and time since death. Med Sci Law. 33(3), 213-222 (1993).