In the heart but not the blood: Transcriptional robustness in the aging hematopoietic system

Last year we learned (mostly from the Vijg lab) that aging in the heart is associated with a dramatic increase in transcriptional noise: Even in the case of genes that don’t change mean expression levels on average between old and young animals, there is a significant rise in the scatter (i.e., variability) in gene expression.

This may not be a general feature of all tissues, however, as we now learn from Warren and Rossi et al. (who came heavy…check out the last three names on that author list: Kim, Weissman, and Quake). In the hematopoeitic system, no apparent increase in stochasticity of gene expression can be observed over the course of aging. (Note that the first few sentences contain quite a shot across the bow of Vijg and co-workers):

It has been proposed that cumulative somatic mutations contribute to the aging process by disrupting the transcriptional networks that regulate cell structure and function. Experimental support for this model emerged from a recent study of cardiomyocytes that showed a dramatic increase in the transcriptional heterogeneity of these long-lived postmitotic cells with age. To determine if regulatory instability is a hallmark of aging in renewing tissues, we evaluated gene expression noise in four hematopoietic cell types: stem cells, granulocytes, naïve B cells and naïve T cells. We used flow cytometry to purify phenotypically equivalent cells from young and old mice, and applied multiplexed quantitative reverse transcription–polymerase chain reaction to measure the copy number of six different mRNA transcripts in 324 individual cells. There was a trend toward higher transcript levels in cells isolated from old animals, but no significant increase in transcriptional heterogeneity with age was found in the surveyed populations. Flow cytometric analysis of membrane protein expression also indicated that cell-to-cell variability was unaffected by age. We conclude that large-scale regulatory destabilization is not a universal concomitant of aging, and may be of significance as an aging mechanism primarily in nonrenewing tissues.

Quick summary: The Vijg lab looked at aged cardiomyocytes and found an increase in genomic rearrangement and transcriptional noise. This multi-lab collaboration checked out aged hematopoietic cells and found none.

So: is stochasticity of gene expression a feature of aging only in non-renewing tissues? The Vijg study also looked at peroxide-treated fibroblasts (and found noise), but it’s not clear whether this is a fair model for aging. Conversely, one could also ask: is the lack of age-related transcriptional noise a peculiarity of the high-turnover, highly diverse hematopoietic lineage? Studies of more non-renewing cells like neurons, fast-renewing cells (gut epithelia), and renewal-competent cells (hepatocytes; fibroblasts from actual old animals) will all be required to fill in the picture.

Taking all of these results at face value, my question is this: Do heart cells accumulate genomic rearrangements and noise faster than hematopoietic cells, or do hematopoietic cells simply have an efficient way of culling mutants and epigenetically dysregulated “noisy” cells from the herd?