Cell Cycle and Sleep Wake Cycle Questions

Question

[tfhahn]

Some general questions about cyclical gene expression as it could hold the key for truly understanding and subsequently completely abolishing aging.

1.) What is actually the advantage of constructing co-expression and regulatory networks from time-series data compared to non-time series data?

2.) Can time-series data add new dimensions to our findings that other data cannot?

3.) Could it be that the relatively transient very brief, but periodically reoccurring and highly oscillating mRNA levels, which may fluctuate synchronously for some genes while asynchronously fluctuate for others.

4.) Will relative timings between the oscillation periods for different genes change as the yeast ages?

5.) Will the relative expression change between different genes over time? Are the time periods of oscillations equal for all genes?

6.) Could it be that the highly regulated, very rapid, but equally transient mRNA oscillations serve any purpose , other than progressing through the cell cycle stages and checkpoints with the only final objective to maximize replication output or may it have some still undiscovered benefits?

7.) How come that even most genes, which belong to pathways that are generally believed to remain unaffected by the cyclical fluctuations of the cell cycle seem to follow its two major temporal anti-correlated expression motifs very strongly?

8.) Which biological properties other than replication are lost or altered when the cell cycle stops?

9.) Are the amplitude or period of the cyclical expression change with advancing age?

10.) How is the cyclical behavior of the cell cycle similar to, or, different from the sleep-wake-cycle?.

11.) Does yeast have a sleep-wake-cycle too?

12.) Are there any experiments or datasets exploring the cyclical expression, which can be assumed for the sleep-wake-cycle?

13.) In humans, many of our cells, especially the very rapidly dividing polypoetic stem cells, red blood precursor cell, bone marrow cells, stomach lining and immune cells, keep dividing very rapidly. Hence, even in humans, cell cycle induced fluctuations should be more noticeable. Has the human or mouse transcriptome ever been analyzed in short enough time intervals for detecting the relatively short but surprisingly intense fluctuations driven by the cell cycle?

14.) If we keep spreading our time intervals to far apart for detecting these transient periodically reoccurring cell cycle driven concentration changes aren't we at risk of missing something very essential? .

Could you please email me links to cell cycle and sleep-wake cycle datasets?

Where can I find proteomic cell cycle datasets for yeast?

Please email me directly at [e-mail] because I cannot see well enough to effectively communicate on this platform since the links, on which I need to click to reply are of too low contrast for me to find. You can also Skype me at tfh002 or call me at my cell phone at + 1 [phone]

Thanks in advance for considering helping me out

Best regards,

Thomas Hahn

Answer 1

[Jean-Karim Heriche]

Answer part 1:
First a few general considerations on time series in biology: Biological processes are dynamic so it makes sense to take this into account and get an idea of how the system under consideration evolves over time. Fixed-time point analyses miss a lot and can lead to the wrong conclusions. For example, imagine a treatment applied to cells that causes a group of genes to be transiently downregulated then upregulated then returned to the same level as control cells. Depending on when your single time point data collection happens, you get different results. The same applies to analyses of time series with inadequate time sampling. Another advantage of time series is that they sometimes allow you to distinguish between primary effects and consequences of these. You also need to consider that there is a stochastic aspect to most biological processes that accounts for some differences between individual cells. Generating time series data from populations effectively averages the values which could mask some effects. If possible it is preferable to get time series for single cells and if needed, average these after appropriate alignment. As an example, consider a time series measurement made on two cells, one cell gives a peak at two minutes and the other cell a peak at three minutes. If you average them, you would get two smaller peaks one at two minutes and one at three minutes or maybe no peak at all if the individual peaks are overlapping.

[Jean-Karim Heriche]

Splitting my answer to go past the 5000 character limit.

[WouterDeCoster]

You may want to add numbers to the posts, as the order will not remain the same when you get upvotes.

[Jean-Karim Heriche]

I didn't think about that. Good idea.

Answer 2

[Jean-Karim Heriche]

Answer part 2:

Your questions are a bit vague but I'll try to answer as best I can.

1.) What is actually the advantage of constructing co-expression and regulatory networks from time-series data compared to non-time series data?

See above

2.) Can time-series data add new dimensions to our findings that other data cannot?

Yes, see above.

3.) Could it be that the relatively transient very brief, but periodically reoccurring and highly oscillating mRNA levels, which may fluctuate synchronously for some genes while asynchronously fluctuate for others.

Yes in principle.

4.) Will relative timings between the oscillation periods for different genes change as the yeast ages?

It depends on so many things. It may be true for some genes under some specific conditions.

5.) Will the relative expression change between different genes over time? Are the time periods of oscillations equal for all genes?

Yes to the first question and no to the other.

6.) Could it be that the highly regulated, very rapid, but equally transient mRNA oscillations serve any purpose , other than progressing through the cell cycle stages and checkpoints with the only final objective to maximize replication output or may it have some still undiscovered benefits?

Expression patterns like oscillations do not necessarily reflect a biological function. It could just be the way things are set up. In particular, gene expression is often uncoupled from when the gene product is performing its function. For protein-coding genes, what matters most for function is post-transcriptional regulation.

7.) How come that even most genes, which belong to pathways that are generally believed to remain unaffected by the cyclical fluctuations of the cell cycle seem to follow its two major temporal anti-correlated expression motifs very strongly?

I am not sure what you're referring to here but I think the answer lies in what I wrote above. Many transcription factors required to control expression of cell cycle genes are also involved in controlling genes in other pathways. If it is not detrimental to the cell to have cycles in these other pathways, there's no pressure to evolve an additional regulatory mechanism.

8.) Which biological properties other than replication are lost or altered when the cell cycle stops?

It depends in which way it stops. Is it due to stress/damage or differentiation or senescence or quiescence ? It also probably depends on the cell type. For example some differentiating cells stop proliferating but not growing (i.e. accumulate mass) while others also stop growth. However, even when cells stop proliferating, replication doesn't always stop (i.e. in endocycling cells)

Answer 3

[Jean-Karim Heriche]

Answer part 3:

9.) Are the amplitude or period of the cyclical expression change with advancing age?

I don't know if there's a correlation with age but even if there were it could be due to some other associated but unrelated factor. As an example of what I mean, consider bacteria that live in the gut. If you expose them in the lab to an increase in temperature, you would see expression of temperature-related genes but you may also see expression of genes associated with low oxygen because what these bacteria normally encounter in the gut is higher temperature associated with low oxygen so the two sets of genes are coupled because there's no evolutionary pressure on producing separate regulatory mechanisms.

10.) How is the cyclical behavior of the cell cycle similar to, or, different from the sleep-wake-cycle?.

I don't know. Do cells sleep ? At the level of a whole organism, it's easy to find example where the two are unrelated. Plenty of cells don't cycle at all whether the organism sleeps or not. Also an apparent association between the two cycles could be due to other factors. For example, when sleeping, the organism usually doesn't perform some functions (e.g. eating) which may indirectly influence cell proliferation.

11.) Does yeast have a sleep-wake-cycle too?

How do you define sleep at the cellular level ?

12.) Are there any experiments or datasets exploring the cyclical expression, which can be assumed for the sleep-wake-cycle?

I think so. For example I think there's some literature on gene expression during sleep/wake cycles in Drosophila and mouse. You may be able to find relevant data sets in ArrayExpress or GEO.

13.) In humans, many of our cells, especially the very rapidly dividing polypoetic stem cells, red blood precursor cell, bone marrow cells, stomach lining and immune cells, keep dividing very rapidly. Hence, even in humans, cell cycle induced fluctuations should be more noticeable. Has the human or mouse transcriptome ever been analyzed in short enough time intervals for detecting the relatively short but surprisingly intense fluctuations driven by the cell cycle?

I am not sure I see what you're asking here. Your question affirms the existence of "relatively short but surprisingly intense fluctuations driven by the cell cycle" so if this is the case then this fact must have been documented otherwise, it's just speculation. Also you don't say what is supposed to be fluctuating.

14.) If we keep spreading our time intervals to far apart for detecting these transient periodically reoccurring cell cycle driven concentration changes aren't we at risk of missing something very essential? .

This goes back to some of the considerations I gave at the beginning. In the ideal case, you would want to monitor biological processes in real-time. Failing that, the observations are usually time-lapse with the interval being a compromise between various practical factors (instrument speed, costs ...) weighted by what we know is a relevant time scale.

It is not clear what biological question you're interested in but I would suggest that yeast is not a good model organism for studying sleep cycles (at least in the way I understand sleep).