They say you can’t teach an old dog new tricks, but it turns out you can use new epigenomic tricks to tell you how old a dog is. The heuristic that to get a dogs age in human years, you multiply by seven is a little simplistic, and due for an update using epigenomic insights. The lab of Trey Ideker at the University of California, San Diego wanted to define epigenetic aging in dogs and compare it to known DNA methylation changes in humans with age.
Existing tools for the dog genome were not suitable for the task however. RRBS and WGBS cover a large number of CpGs, but they don’t always get the exact same CpG sites at low read depth. The authors wanted to be able to make evolutionary comparisons of high-quality dog methylomes with other mammals. They wanted to identify CpGs in dogs that are syntenic with CpGs on human Illumina arrays, creating a high quality panel of dog methylomes with substantial coverage of CpGs noted in prior human studies. They developed a method called Synteny Bisulfite Sequencing (SyBS), designed to capture ~90,000 of the ~232,000 CpG sites conserved from dog to human. Capture oligonucleotides specific to dog genome regions syntenic to 450k probes were used to enrich these regions for targeted bisulfite sequencing. They collected blood from a cohort of 104 Labrador retrievers ranging from 0.1-16 years and compared it to 450k data from the blood of 320 humans ranging from 1-103 years of age. Here’s what they dug up:
- SyBS provides an average coverage of 163X in the dogs, about 13-fold greater than RRBS
- 54,469 CpGs were well profiled in both species and used for further analysis
- They compared the methylomes of dogs and humans using several correlation-based methods, for instance plotting the age of each dog methylome against the average age of the five nearest human methylomes and vice versa
- This method identified a clear logarithmic relationship: human_age = 16 ln(dog_age) + 31
- This relationship agrees with the rapid early development of dogs, when compared to humans, which then slows in adulthood and puts major physiological milestones at the same epigenetic time, such as the emergence of teeth, and average life expectancy
- Next, they identified 394 genes at which methylation changed over lifetime in dogs, humans, and mice
- These CpGs are enriched near transcriptional start sites, and have functions related to development, specifically synapse assembly, anatomical patterning, and polycomb repressor targets
- Finally, they constructed a conserved epigenetic clock using these 394 CpGs, which is accurate at predicting chronological age in each species
These finding go well beyond being able to tell how old your furry companion really is and should offer more than a new consumer epigenomics market for man’s best friend. The results build on the ability of the epigenetic clock to act as a diagnostic readout in a single species, allowing a cross-species translation of the physiologic state of aging. This could be used in many aging studies to better translate comparisons between model organisms and humans.
Go see who let the dogs out over at Cell Systems, July 2020.