Reading the abovementioned article¹, I was disappointed by the idea (which is shared by most researchers) that Mus musculus has very long telomeres. In reality, this a characteristic of “classical” laboratory strains (Figure 1). Nonetheless, since the authors successfully reached their aim, I was surprised by the following contradiction: laboratory mice show the conserved character of having methionine 492 (M492) in RTEL1, but have not the conserved character of “short” (15-20 kb) telomeres, while Mus spretus has not the conserved M492, but has the conserved character of short telomeres. So, I checked for RTEL1 in all other Mus species and subspecies (identifying new orthologs in species for which only genome assemblies are available) and found that M492 is conserved in all species (Figure S1 in Supplementary material). Finally, I used Mus spretus genome assemblies to predict de novo the protein sequence and I found a surprise: RTEL1 in Mus spretus has M492 (Figure 2). Then I found that also the sequence present in the Ensembl genome database (www.ensembl.org) is identical to the one predicted by me (Figure S1 in Supplementary material). Because of all these findings, it can be said that Mus spretus protein sequence of RTEL1 (deposited by Ding et al.² and used by Smoom et al. in their article¹) in the GenBank database shows a wrong amino acid in position 492. Thus, the amino acid 492 of RTEL1 is not responsible for the difference in telomere length between laboratory mice and Mus spretus (as well as between laboratory mice and all other mice).

Because of a false premise (i.e., K492 in Mus spretus), the Telomouse does not mimic the condition in the Algerian mouse. Nonetheless, it is a useful inbred strain (congenic with the popular C57BL/6) with human-like telomere length. Moreover, it could recapitulate the RTEL1-related Hoyeraal-Hreidarsson syndrome (HHS), since many patients with this disease show the M492I variation³. Someone could object that Telomouse is a bad model for recapitulating HHS, because mice have telomerase activity and humans have not. However, this is an over-simplification (although very popular): in reality, many human cells (especially, but not limited to, hematopoietic stem cells) show telomerase activity⁴.

Beside the potential use of Telomouse, the results obtained by Smoom et al. open up new intriguing questions. How much the telomeres of Telomouse fibroblasts would shorten in vitro if grown at physiological oxygen levels (i.e., 5% instead of the usually employed atmospheric 20%). Is the telomere shortening observed in vitro and between generations of Telomouse ascribable only to a decrease in telomerase activity at telomeres? Or the reduced activity of RTEL1 has consequences also on Alternative Lengthening of Telomeres (ALT)? The authors found no differences in telomeric recombination (indicative of ALT) in vitro, but this does not exclude differences, for example, in oocytes and cleavage stage embryos. In these, in fact, telomere elongation has been observed also in telomerase-null mice and ALT activity has been reported⁵.

Another important question, not strictly related to Telomuse, is the link between RTEL1 and the presence of very long telomeres in laboratory mice. Smoom et al. already indicated that M492 is present in rodent species (excluding Mus spretus, according to them) with short telomeres and thus cannot explain telomere length differences between these and laboratory mice¹. Here I demonstrated that this is valid also for Mus spretus. Moreover, I found that there is no amino acid that is conserved in short-telomeres Mus species and is modified in long-telomeres strains (Figure S2 in Supplementary material). Therefore, differences in telomere length cannot be ascribed to differences in RTEL1 protein sequence. Nonetheless, the Rtel1 gene was at first discovered as a chromosome locus responsible for telomere length differences between Mus spretus and BALB/c⁶ mice and a subsequent study showed that in crosses between Rtel1^+/- laboratory mice and Mus spretus, the ones with Rtel1⁺ allele had long telomeres and the ones with Rtel1^- had short telomeres². Different explanations can be proposed: 1) the study of Ding et al.² had some unidentified error and RTEL1 is not responsible at all for differences in telomere length; 2) differences between Mus species/strains are due to differences in the relative abundance of RTEL1 isoforms; 3) differences in telomere length are due to the Rtel1 gene, but the variations are in the promoter (causing different levels of transcription) and not in the coding sequence. I found that among the conserved (in wild species) nucleotides in the promoter, only two are modified in laboratory strains: one of the two is a C>T modification, changing a -CAAC- into a -CAAT-. However, although attractive, it does not really resemble a CCAAT box and seems to far from the transcription start site. In any case, these hypotheses can be easily tested by quantitative PCR and Western blot, in order to assess relative abundance of variant transcripts and protein isoforms.

The last issue (which in my personal opinion is the most important) regards telomere length in Mus musculus. The mystery of how and why long telomeres appeared in laboratory strains has not been solved: it seems like it is not due to inbreeding (Figure 1) and many hypotheses can be raised on domestication of mice and increase of mating period, ALT and telomerase activity in germ cells, epigenetics and so on. But beside these questions, one thing should be clarified: Mus musculus (and even Mus musculus domesticus) has NOT long telomeres. This is a characteristic of laboratory strains (for the history of their ancestry, see⁷) and not of the species (or subspecies). Not grasping this aspect can lead to errors and false conclusions in studies on inter-species comparison or when physiological and life history traits are regressed against telomere length. For example, because of the plausible link between telomeres and aging, many researchers have regressed telomere length^8-9 (or telomere shortening rate^10-11) against maximum lifespan. However, the telomere length used for Mus musculus is always the one of laboratory strains (40-100 kbp, heavily influencing every regression) and the maximum lifespan is the one of wild mice (4 years¹²). Similar considerations can be made on studies on the evolution of telomeres in mammals⁹. In conclusion, when dealing with telomere length, attention should be paid to the differences between Mus musculus as a species and laboratory strains.