A new mechanistic insight into fate decisions during yeast cell aging process

Abstract

Despite massive technological advances in mammalian models in recent years, studies in yeast still have the power to inform on the basic mechanisms of aging. Illustrating this, in Nan Hao’s recent article published in the journal Science, he and his lab use microfluidics and fluorescent imaging technology to analyze the dynamics and interactions of aging mechanisms within yeast cells. They focused in on the Sir2 gene and the heme activator protein and, through the manipulation of these two molecular aging pathways, were able to determine that yeast cells can undergo one of three modes of aging, with one of them having a significantly longer lifespan than the others. These findings provide unexpected insights into mechanisms of aging, apparently as regulated fate-decision process, and open up avenues for future research.

The field of geroscience is one of the most challenging yet important fields of biology. Research on aging holds the promise of prolonging human longevity as well as improving the health and quality of life for elders (Kennedy et al., 2014). The latter is especially important as age-related chronic diseases impose a huge and increasing burden on global healthcare.

Previous studies in the aging research field have led to identification of many genes and pathways that influence the aging process (Fontana et al., 2010; Kenyon, 2010; Smith et al., 2008; Wasko and Kaeberlein, 2014). However, much less has been learned about the dynamic cross-talks of these pathways in control of the aging process. These pathway interactions might underlie the variability observed in the lifespans of genetically identical cells (Lopez-Otin et al., 2013). Recent advances in technologies, such as microfluidics, time-lapse imaging, and computational analysis, have enabled researchers to monitor and probe the aging process from a multitude of angles, leading to new fascinating results (https://doi.Org/10.1016/j.tma.2019.09.0022.

Microfluidics is the science of designing contraptions or devices to manipulate liquids in a highly confined area. Recently, this technique has been successfully implemented to investigate the replicative aging in the budding yeast S. cerevisiae (Chen et al., 2017; Polymenis and Kennedy, 2012). Yeast replicative aging is measured as the number of cell divisions that a mother cell undergoes before it eventually dies, and has been a genetically tractable model for the aging of mitotic cell types (He et al., 2018). For over 60 years since its discovery, the traditional approach for determining yeast replicative life span (RLS) requires manual dissection of daughter cells from mother cells after each cell division and hence is labor-intensive and with low throughput (Steffen et al., 2009). In addition, it does not allow real-time tracking of molecular and cellular processes in live cells during aging. To overcome these limitations, microfluidic devices have been developed to automate mother-daughter separation and, more importantly, to allow RLS measurements to be interfaced with time-lapse fluorescence microscopy, which enables tracking of the dynamics and interactions between aging-related genes and pathways throughout the lifespans of individual cells.

Using this technology, researchers have started to uncover the hidden mechanisms of aging processes obscured by traditional population-based methods (Crane et al., 2019; Jin et al., 2019; Li et al., 2017; Liu et al., 2017; Morlot et al., 2019; Yang et al., 2015). A recent article from Nan Hao’s lab is a good example as it demonstrates the power of the single-cell approach for transforming aging research (Li et al., 2020). In the study, they focused on the interactions between chromatin instability and mitochondrial dysfunction, two major factors that contribute to cellular aging. Using microfluidics coupled with time-lapse microscopy, they observed two different types of phenotypic change during aging of isogenic yeast cells (Figure 1A). The first one was defined as “mode 1” aging. In this mode, mother cells produced daughters with an elongated shape later on in life and were characterized by enlarged and fragmented nucleoli. In contrast, the second form of aging, defined as “mode 2”, had mother cells that produced small, round daughter cells for their entire lifespan and were characterized by deterioration in mitochondrial health before death. To characterize the molecular pathways underlying these two age-related deteriorations in organelle health, they further probed the SIR2 gene as well as the iron-containing compound heme in individual aging cells.

Figure 1, Aging modes in yeast cells.

(A) Wild-type cells exhibited two modes of aging at roughly an equal rate, mode 1 is characterized by nucleolus deterioration and elongated daughter cells whereas mode 2 is characterized by mitochondrial deterioration and smaller, round daughter cells. (B) Sir2 gene overexpression causes a third path called mode 3 aging in which both HAP4 and rDNA silencing are maintained, leading to a long-lived mode of aging with elongated daughter cells. (C) Overexpression of both Sir2 and HAP4 leads to an even higher rate of cells undergoing mode 3 of aging.

SIR2 is a well-documented longevity gene (Gartenberg and Smith, 2016). It encodes a lysine deacetylase, Sir2, that mediates ribosomal DNA (rDNA) silencing. Loss of rDNA silencing is related to nucleolar enlargement observed in aged cells. On the other hand, heme is a compound containing iron, which directly functions in the electron transport chain in mitochondria and also turns on the heme activator protein (HAP) transcriptional complex which induces genes for maintaining mitochondrial biogenesis (Buschlen et al., 2003). It has been proposed that the depletion of heme may play a large role in mitochondrial decay and dysfunction in aged cells (Atamna et al., 2002). Using genetically encoded fluorescent reporters, they simultaneously monitored both rDNA silencing and intracellular heme level in the same cells and found that the aging processes of mode 1 and mode 2 cells diverged early in life towards two discrete ending states with anti-correlated rDNA silencing and heme levels. This observation is consistent with mutual inhibition between rDNA silencing and heme pathways, a model that was further supported by genetic perturbations of SIR2 and HAP.

Based on these results, they devised a mathematical model where both Sir2 and HAP undergo positive autoregulation while simultaneously inhibiting each other. This model may underlie the two different modes of aging where each mode is characterized by high levels of one of the two molecules and a low level of the other. Importantly, the model can also help explain a very interesting observation from genetic perturbation experiments, in which 2-fold overexpression of SIR2 led to the emergence of a new mode (mode 3) of aging (Figure 1B). Cells that underwent mode 3 aging were able to simultaneously maintain high levels of rDNA silencing and heme, and as a result, they retained a normal cell cycle rate throughout the entire life span and had a much longer RLS than mode 1 and mode 2 aging. Further, guided by the model, they overexpressed both SIR2 and HAP and enriched the fraction of mode 3 cells in the isogenic cell population, leading to a significant extension to the lifespan, much more dramatic than overexpression of either SIR2 or HAP alone (Figure 1C).

This study represents an important conceptual advance in our understanding of yeast aging for the following reasons:

1. Previously cellular aging has been viewed as a passive damage accumulation process, albeit one in which the rate of damage accumulation is quantitatively controlled. The authors provided evidence that cellular aging, at least the replicative aging in yeast, can be considered as a regulated fate-decision process in which single cells age by one of two mutually exclusive and qualitatively distinct cellular trajectories associated with different life spans and their fate choice can be modulated by various genetic or chemical interventions. Whether fate choice can be environmentally affected, for example by nutrient availability, remains to be determined.

2. The network structure and computational model from this study established a theoretical framework for understanding the yeast aging process from a systems perspective. This framework provides insights into some long-standing paradoxes in the field. For instance, 2-fold overexpression of SIR2 can extend the life span, whereas a further increase in SIR2 expression shortens the life span. This was an intriguing but puzzling observation from previous studies (Kaeberlein et al., 1999), which can be nicely explained using the model from this study: While 2-fold overexpression of SIR2 induces long-lived mode 3 aging, excessive SIR2 overexpression promotes short-lived mode 2 aging by strongly activating the high Sir2, low HAP pathway. Importantly, this framework can not only help clarify previous results not previously interpretable with a traditional reductionist view, but also can serve as a foundation for researchers to generate new hypotheses and design future experiments.

3. This study clearly demonstrated how a quantitative understanding of pathway interactions and network structures can help guide the rational design of intervention strategies for promoting longevity. Due to the complex interactions among aging pathways (e.g. mutual inhibition), perturbations of individual genes or factors often have modest effects. Hence, a cocktail approach that targets multiple genes or factors will be needed and computational modeling is especially useful for determining potential targets for this approach. As shown in the case of this study, overexpression of HAP does not extend the life span to a statistically significant degree, making HAP not a good candidate for intervention. However, when combined with SIR2 overexpression, HAP overexpression can extend the life span in a synergistic manner, much more dramatic than the sum of the effects from both perturbations. This result cannot be intuitively anticipated without modeling. In a simplistic linear view of aging pathways, the effects of perturbations should be largely additive. Given the complexity of the networks underlying aging, we envision that the use of computational modeling, as illustrated in this study, will become necessary for understanding the biology of aging and for developing effective interventions.

Finally, this study highlights the sophisticated nature of the processes involved in aging, but also raises new questions that deserve further attentions. For example, why are yeast cells programmed to diverge in their aging paths? Is it associated with physiological advantages? How do other aging-related factors or genes, such as TOR and PKA, interfere with the fate decision process? Are there additional modes of aging if other factors are taken into consideration? Are the framework and mechanisms from this study applicable to the aging of other organisms? Why does only the overexpression of Sir2 cause this long lived mode 3 of aging and why does HAP not cause a mode 4 of aging? Addressing these questions will be important for better mechanistic understanding of single-cell aging, a necessary step towards promoting healthy longevity in humans.

Highlights –

• Microfluidics and fluorescent imaging technology to analyze aging mechanisms of yeast

• Focused in on the Sir2 gene and the heme activator protein

• Yeast cells can undergo one of three modes of aging, with one of them having longer lifespan than the others.

• Insights into mechanisms of aging, apparently as regulated fate-decision process