Epigenetic “biological aging clocks” capture aspects of aging largely distinct from those captured by telomere shortness and attrition rates. The two taken together — epigenetic “clock” measures and telomere integrity measures — may be more powerful as tools for assessing “biological aging” than either alone. The technologies exist that would now allow such studies to be undertaken.

Detailed analyses — first in yeasts and now more recently in human cells — have shown that the region of chromatin adjacent to the telomeric DNA-repeat tract influences the regulation of its adjacent telomere DNA repeat tract length. Small differences in the mean lengths have been found from one chromosome to another. Thus, each chromosome end will have a (usually small) difference in length and, as shown in some studies, a different rate of shortening.

However, the length of a given chromosome’s telomere varies so much from cell to cell that all the measured lengths of one chromosome’s telomere mostly overlap with those of another chromosome’s telomere. What remains unknown is whether these differences in mean lengths are mostly the same for all people, and whether these small differences have any biological significance in terms of chromosome stability or disease risk.

There have been studies of paternal telomere lengths and their heritability by offspring. These studies have shown a lot of variability between different cohorts, and so the significance of the heritability relationships seen between paternal and offspring telomeres is still not well understood.

G-quadruplexes have been identified in a great many locations in the human genome, including in many gene promoter regions. Their roles at such regions have been studied at the molecular level, but currently, how they might contribute to the aging mechanisms related to telomere maintenance is an open question.

Telomere maintenance in humans is a delicate balance: too much telomerase or too-active telomere maintenance can increase risks of certain cancers, while too little increases risks of many common diseases of aging, including cardiovascular disease. So it will take care and attention to unwanted side effects of any potential pharmaceutical intervention — particularly because such effects could take years to manifest. That said, telomere biology as a target should still be tried — thoughtfully — for diseases such as genetically-caused telomere deficiencies.

Telomere maintenance is an objective measure that can be used to add to statistical predictions about future disease risks in a population. Such measures can be useful — for example, in the face of political pressures — to provide objective quantitative evidence of the health benefits of social policies.

When I was a postdoctoral fellow in the research group of Joe Gall, I was determining the first DNA sequence of a telomere. The memorable moment occurred upon developing an X-ray film. This showed an image of radiolabeled DNA that visually told me that a telomere consisted of a simple repetitive DNA sequence — an unexpected result that led to many others, including our subsequent discovery of telomerase.

I was curious about how I could use the very early methods of sequencing DNA, which were just then emerging in the early 1970s, to figure out ways to sequence the very ends of chromosomes. It was known only that there was something special about what protected eukaryotic chromosome ends — known from decades of cytogenetic and genetic studies by pioneering researchers, notably Barbara McClintock. I was excited to find out whether there was anything special about the DNA itself at chromosome ends, now that methods existed that would make asking such a question possible. Once I started getting results, which from the outset were surprising, more and more mysteries emerged — and I have never lost interest in the biology of telomeres and their maintenance.

From my experience, the advice I would offer for choosing what to study is: think of an unanswered biological question which, because of a new technology, can now be addressed experimentally in a way that allows rigorous testing of hypotheses and ideas.

All 46 chromosomes are in territories, resembling separate small bags; each territory is size-proportional to chromosome size — some small bags, some large. No one knows how the territories are moved around, or the order; this is a major project for future research. We are currently working on this, but it is a hard problem.

The Cai paper from the Lai group is a major finding. I refer to her earlier finding on the separation of parental mitotic chromosomes with a detailed paragraph in my PNAS paper published in 2022 (Paper 3). The problem is that it is likely to be seen in some cells and in some stages of the cell cycle — not all — and only in primary cells; importantly, not in cancer cells.

The most challenging problem for Cryo-EM tomography (necessary to preserve cell structure) was the missing data arising from high-angle tilts. All tomography suffers from this missing data, which makes interpretation of Z-axis information essentially useless.

My suggestion to apply deconvolution — a method we and our UCSF colleagues had published — fixed this problem. It was implemented in collaboration with Michael Elbaum at the Weizmann Institute in Israel (see my first two PNAS papers, largely physics description papers covering the method). The chromosome data in Paper 3, studied by stereo for 3D, document the improvements and lead to likely structures.

Microscopy is basic science — understanding how the underlying cell biology (the chromosomes) is built, how it progresses through the cell cycle, and how genetic regulation takes place. This fundamental information is taken up by other scientists, usually the combination of bench scientists and clinician-MDs, and incorporated into their research, which may even extend to human studies. It is simply an approach to supplying knowledge for others to use, incorporate, and advance.

I have been working for about 50 years on chromosome architecture and structures, and there were quite a few high points — though it is hard to rank them. I have to give credit to the lab members, graduate students, and postdocs (since we work as a team) for many of the high points.

A recent very high point happened around 2011 when I suggested that a mathematical procedure — deconvolution — could fill in the missing tilt data for Cryo-EM, an intractable problem. Surprisingly, the method worked in just a few weeks. This insight is the basis of the recent five PNAS papers and the chromosome architecture that, for the first time, allows chromosome structures to be studied at different cell cycle points and transitions.

It is an interesting story. When I was a graduate student (1965–1970) at Caltech, Max Delbrück — a world-famous professor — took a few of us students aside and gave us a charge: our future destiny was to work on hard problems. How right he was.

After my postdoc with Fred Sanger at the MRC in Cambridge, England, working on early DNA sequencing, I realized that problem was essentially solved. I then decided that chromosome architecture — a very hard problem — needed work. It is indeed a hard problem, but we make credible progress. Much remains to be done.