Special Report: Do Stem Cell Telomeres Drive Most Idiopathic Pulmonary Fibrosis Cases?
Dysfunctional telomeres—molecular caps that keep chromosomes from unraveling like shoelace nubs—may be critical to idiopathic pulmonary fibrosis (IPF), which kills in two to three years.
According to a new study in Cell Reports, dysfunctional stem cell telomeres in particular are critical.
For the study, the team of the Spanish National Cancer Research Centre’s Mary Blasco, M.D., developed two mouse models of telomere dysfunction. In one, a mouse was engineered to conditionally express a disabled telomere gene in its type II lung alveolar (or stem) cells. In another, a mouse lacking, in all cells, the enzyme that protects telomeres—telomerase—was exposed to a low-dose toxin.
In both mouse types, robust IPF developed, the team reported.
“I like this paper a lot. It provides additional support to the role of the telomerase pathway in lung fibrosis, and especially highlights the role of the DNA damage as a mechanistic pathway" if repeated, Yale University pulmonologist Naftali Kaminski, M.D., told Bioscience Technology. Kaminski was uninvolved with the study.
“It also confirms the notion that targeted injury to alveolar type II epithelial [AEC2 stem] cells, and especially to their potential to replenish other epithelial cell populations, will lead to inflammation, aberrant healing, and lung fibrosis,” Kaminski said. “I find it very impressive that, more than ten years after Moises Selman, in a breakthrough review, proposed that enhanced epithelial cell predisposition to injury could potentially explain pulmonary fibrosis, cumulative experimental data—including this elegant paper—proves his speculation. This paper leads to significant mechanistic advances in our understanding of lung fibrosis," again, if it is repeated by other groups.
Blasco told Bioscience Technology: “We are very excited about the results.”
Johns Hopkins School of Medicine lung telomere expert Mary Armanios, M.D., warned that other groups have failed to get IPF using mouse models like Blasco’s second one. But a March Armanios study supports the idea that stem cell telomeres are key to IPF, she told Bioscience Technology. “It has been hypothesized, and there is recently published evidence supporting that, stem cell failure is sufficient to drive many features of age-related lung disease.”
Blasco’s first model “supports this premise.”
Shift away from inflammation
IPF strikes (mostly) men over age 50. It causes their lungs to stiffen with fibrotic scar tissue. Death often occurs two to three years post-diagnosis.
For decades it was thought IPF was caused by inflammation, immune cells run amok. But in recent years, papers by many teams, including Kaminski’s, have found inflammation to be secondary. Those papers indicate inflammatory cells are just rushing to the scene of a mysterious epithelial cell injury, one that first prompts alveolar epithelial cells to undergo hyperplasia—then slowly die. The prolonged death process signals fibroblasts, immune cells, and other scarring cells to move in, and deliver the last in a series of lethal blows to the lung.
IPF is likely caused by aberrant wound healing, not inflammation, many papers found.
Last summer the idea was bolstered by trials unveiling the first two drugs to slow IPF—pirfenidone and nintedanib--which they did by attacking, as predicted, aberrant wound healing. The FDA fast-tracked approval. But the side-effect ridden drugs only slow IPF, and only in some patients.
Armanios’ stem cell telomeres
Many groups, before and since, have tackled different pieces of the wound-healing puzzle. Armanios’ group has focused on a major piece: shortened telomeres. Telomeres shorten every time cells replicate. When telomeres get too short, cells die. In 2007, Armanios and Hopkins biologist Carol Greider published the first paper establishing an association between shortened telomeres and IPF. (Greider would win the 2009 Nobel Prize for her 1990’s work showing how telomerase protects telomeres.) In the 2007 paper, Armanios and Greider singled out mutations, in two key telomerase genes, oft seen in short-telomere, familial IPF biopsies: hTERT and hRT.
Armanios demonstrated the association between short telomeres and IPF in different ways over the decade. With Hopkins’ Elizabeth Blackburn, who won the Nobel with Greider, she placed IPF, in 2012, into the lead of what she dubbed “Telomere Syndrome,” an entire class of short-telomere disorders. (A short-telomere hunt began after Blasco and Greider created the first telomerase-deficient mouse in 1997.) As telomeres shorten every time cells divide, people born with short telomeres in all cells can be hit with many disorders mid-life. They can experience stem cell failure or apoptosis (quick cell death) in tissues with fast turnover, like bone marrow, liver, and skin, leading to aplastic anemia, cirrhosis, or premature hair greying. They see more cancers in those organs, like acute myeloid leukemia, non-Hodgkin’s lymphoma, squamous cell carcinoma, and hepatocarcinoma. In the slower pancreas, Type 2 diabetes can develop. In the very slow lung, a prolonged cell death called “senescence” occurs. Slow-dying stem cells emit help cries eliciting a veritable fibrotic-cell storm that clogs airways, stiffens walls, and disables lungs.
Read More: Newly-Found T Memory Stem Cells May Be Key to Gene Therapy
Armanios and Blackburn noted the Telomere Syndrome can hit earlier by generation, as inherited germ cell telomeres get ever shorter. This is “genetic anticipation.” If the syndrome hits one generation in their 60’s, the next may get hit in their 50’s. Diseases can be concurrent. While lung transplant is IPF’s only “cure,” IPF transplant patients can suffer complications like bone marrow failure due to anti-rejection chemos their short blood-telomeres can’t handle.
Another Armanios paper showed, in 2011, cigarette smoke accelerates senescence in short-telomere lung cells.
By now, she and others have shown telomere dysfunction occurs so often in human IPF—20 percent of familial IPF and three percent of sporadic IPF—that telomere-related mutations are “the most common identifiable genetic causes” of IPF, she wrote in a July 2014 Lancet. Gene mutations in proteins leading to shortened telomeres in human IPF now number six: hTERT, hTR, DKC1, TINF2, and (as of May) PARN and RTEL1.
And counting. For as Armanios told Bioscience Technology: “It has been estimated that at least half of IPF is related to telomeres; that half of all patients with IPF have very short telomeres. Telomeres are a huge clue to IPF.”
With stem cell pioneer Brigid Hogan’s evidence in 2013 that AEC2s are stem cells, the above has led some to speculate short telomeres in AEC2 stem cells in particular may prompt much IPF. For AEC2 stem cells sit between the toxic air and the inner lung. Buffeted by toxins, those stem cells turn over constantly to form and replace all damaged alveolar cells. For short-telomere stem cells, with limited lifespans, turnover is a risky proposition.
Still, despite leaps in understanding, the creation of short-telomere mouse models of IPF has stalled. Mouse models are key. Disease course—and drugs—aren’t well-tested without them. But mouse telomeres are longer. In 2007 and 2012, Tianjiu Liu et al and Amber Degryse et al reported repeatedly assaulting telomerase-knockout mice with the lung-damaging chemo, bleomycin.
They could not get IPF.
Armanios creates first relevant mouse model
This March, Armanios’ team—with Hogan—reported they’d made a different mouse model. As mice have longer telomeres, the Liu 2007 and Degryse 2012 papers featured later-generation mice (since each IPF generation has shorter telomeres). As that failed, Armanios’ teams used a different mouse, one with a conditionally mutated gene in the telomeres of AEC2 cells called trf2—which completely uncapped the telomeres.
Then they issued the mice a critical second blow: bleomycin.
This did, finally, generate senescence, and an IPF-like condition. They thus demonstrated, in an experimental setting, a mechanistic link between short telomeres in AEC2 stem cells and a fibrosis-like condition driven by AEC2 stem cell senescence. “Telomere dysfunction causes alveolar stem cell failure,” they called their paper.
In the paper, they also compared in vitro short-telomere AEC2 stem cells with non-stem short-telomere stromal cells. AEC2 stem cells with short telomeres stopped regenerating alveolar tissue. But short-telomere stromal cells were fine.
So they showed lung stem cells are more “exquisitely sensitive” to dysfunctional telomeres than other lung cells. They showed throwing toxins at those stem cells generates IPF-like fibrosis.
Still, Armanios’ team did not see full-blown IPF.
Blasco’s apparent “full-blown” IPF
Blasco was once a post-doc in Greider’s lab. When she got her own lab in Spain in 1997 (at the CNIO since 2003), she set out on a decades-long study of telomeres in many tissues and stem cells. In 2005, she published a paper in Science on telomerase and skin stem cells. In 2011, she was made CNIO director.
For the new paper, she jumped into the fray, and tried to create the elusive AEC2/short telomere/IPF mouse model. She created mice lacking a related gene for telomere capping in type II alveolar cells: trf1. But, noting Armanios’ mice got such a high bleomycin dose that 60 percent died in two weeks, she tried only uncapping the telomeres: no bleomycin. The result: most developed “massive pulmonary fibrosis,” she reported.
She moved to another model, since—as with Armanios—hers was not believed to mimic human IPF closely enough since it lacked short telomeres. Blasco’s group took 2nd and 4th generations of telomerase knockout mice like those used by Liu et al and Degryse et al. Noticing those two used much bleomycin, which she felt may have killed the mice before they could get IPF, she gave a smaller amount to her short-telomere, telomerase knockout mice.
As with her other model, these mice contracted “full-blown pulmonary fibrosis,” her team reported.
“In one model we deleted a component of shelterin (trf1), specifically in AEC2s,” Blasco told Bioscience Technology. “We disrupted telomere capping in this cell type. This was sufficient to trigger progressive and lethal pulmonary fibrosis. To our knowledge, this is the first time it has been demonstrated that telomere dysfunction is sufficient to trigger pulmonary fibrosis in the absence of other damages. This happened in the absence of telomere shortening.” It was also the first time damage to AEC2s alone was sufficient, she said.
Again, this wasn’t seen as physiologic. So in their second model Blasco’s team set out to “trigger pulmonary fibrosis in a setting of short telomeres owing to telomerase deficiency. This is the scenario in humans with telomerase mutations.”
“We knew from Armanios work that short telomeres per se did not trigger the disease in telomerase-deficient mice, so we combined short telomeres with a low-damage toxin dose that would not trigger pulmonary fibrosis in wildtype telomerase-proficient mice. We did this by combining short telomeres with low doses of bleomycin. This is the first demonstration that low doses of damage can induce pulmonary fibrosis in a context of telomerase deficiency and short telomeres.”
Blascos’s next step is “to find therapeutic strategies to ‘cure’ pulmonary fibrosis in this second mouse model.”
Reservations
Kaminski agreed the work may end up known, if repeated, as the first demonstrating telomere dysfunction is sufficient to trigger pulmonary fibrosis absent other damage—although “many have shown an association with short telomeres, or telomere mutations, in humans.”
He also said that while he “fully agrees” that “injury to AEC2s is sufficient to cause fibrosis in mice, I don’t think we know for sure that they are the origin of the disease in humans, as multiple cell types are involved and behave abnormally,” like bronchial basal and other progenitor cells.
“Targeted telomerase abnormalities to AEC2s is not the situation in our patients,” he said. “If they have a mutation, it is in germ cells, so all cells have it.” Given this: “Why are fibroblasts accumulating? Are they less sensitive to DNA damage? We still have to solve the cell death paradox in IPF: How come one cell type (epithelial) dies more and the other (fibroblast) lives longer? Is there a way we can answer this using genetically modified mice?”
Kaminski also contested a definition. The Blasco team “refer to ‘IPF’ in figure legend, the introduction, and the discussions. But IPF is characterized histologically by the pathological pattern of usual interstitial pneumonia, which includes peripheral distribution, temporal heterogeneity, accumulation of abnormal extracellular matrix, myofibroblasts foci, epithelial cells hyperplasia, and formation of honeycomb cysts on CT. The lesions they show – as much as can be judged from the paper — are not IPF lesions. They show extensive lung fibrosis and inflammation, found also in other mouse models of fibrosis.”
Kaminski said fine-tuning the model may get them “closer to the human disease. But they do not provide any such evidence. This does not reduce the value of their results. But from what could be gleaned from the paper, this is pulmonary fibrosis in its broadest definition. Not IPF.”
Still, Kaminski said, the paper may lead to “a better mechanistic link between telomerase dysfunction and fibrosis.”
Response
Blasco responded that “it is true patients with ‘telomere syndromes’ who develop IPF have germ-line mutations in telomerase in all tissues, and not only in AEC2s. This is also the case in the telomerase knockout mouse used in our paper, in which we induce pulmonary fibrosis with low dose bleomycin. We discuss that this mouse model would be instrumental to find new therapeutic strategies to treat pulmonary fibrosis associated with short telomeres.”
It is also true her other mouse model deleted trf1 only in AEC2s. But “by using this other mouse model, we wanted to demonstrate telomere dysfunction stemming specifically from this cell type is sufficient to trigger pulmonary fibrosis in mice. This suggests telomere dysfunction of AEC2s could be at the origin of this disease.”
Regarding fibroblasts, she said, fibroblasts can show longer telomeres than corresponding epithelial cells. Epithelial cells from breast have shorter telomeres than stromal fibroblasts, she said
“Together, our results suggest telomere shortening and environmental insults in the alveolar epithelium can provoke pulmonary fibrosis, while fibroblasts have no need to proliferate until injury appears,” Blasco told Bioscience Technology. “Thus, in patients with telomerase mutations and short telomeres, epithelial cells are removed from the tissue when they became apoptotic or senescent, then AEC2s proliferate to regenerate the alveoli until they cannot proliferate anymore due to telomere exhaustion. Thus, regeneration fails because telomere shortening is never solved, and the problem persists.”
It is unknown, she said, “whether activated myofibroblasts, responsible for the fibrosis, are recruited from circulation or are resident fibroblasts which activate with epithelial damage. These questions about fibroblast provenance remain unresolved. However, in both the telomerase-deficient, and trf1-deficient, mouse models, we demonstrated the presence of activated myofibroblast foci in the regions with fibrosis.”
Finally, Blasco believes she created an IPF model, “not another model of lung fibrosis in the mouse. Our telomerase knockout mouse model for lung fibrosis fits the criteria established by Raghu et al for IPF diagnosis. We found obvious interstitial pneumonia with evidence of multifocal fibrosis. These fibrotic foci deform lung architecture and provoke honeycombing structure next to non-affected areas. Active fibroblast areas were also found with extracellular matrix deposition with increased amount of myofibroblasts.”
Kaminski disagreed with the last point. “The term IPF denotes a very, very specific medical condition, a diagnosis that requires significant expertise and a multidisciplinary team of experts,” he said. “Per definition, IPF is idiopathic: the injury is unknown. So a model that includes bleomycin cannot be considered IPF. IPF is defined--in the manuscript they quoted--as a specific form of chronic, progressive, fibrosing interstitial pneumonia of unknown cause, occurring primarily in older adults, limited to the lungs, and associated with the histopathologic and/or radiologic correlate of Usual Interstitial Pneumonia (UIP). The definition of IPF requires the exclusion of other forms of interstitial pneumonia, including exposure (like hypersensitivity pneumonia), autoimmunity (like rheumatoid arthritis) or any other environmental exposure, medication, or systemic disease. The authors need to quit talking about IPF and start speaking about UIP. And they need to convince us that what we see histologically in the lungs of mice is indeed UIP. They would need to provide the histological slides to pathologists who are experts in IPF, who will do the proper blinded analyses. The images provided in the manuscript cannot be interpreted as UIP, but they clearly contain fibrosis.”
Study needs repeating
The results need repeating.
“The results of this study must be interpreted with caution, as several groups over the past decade have looked at bleomycin sensitivity in mice with short telomeres but found no susceptibility,” Armanios told Bioscience Technology (referring to the 2007 Liu and 2012 Degryse knockout mouse studies). In the Degryse study, “there was no difference after low-dose bleomycin was given for prolonged periods.” In the Liu study, telomerase was required for bleomycin-induced lung fibrosis in mice.
Blasco responded: “In Degryse et al., they used a higher bleomycin concentration than the one used in our experimental settings. Degryse used either 40 or 80 μg per mouse, and in some experiments even treated the mice with 40 μg every two weeks for eight doses. We used a concentration of 0.5 mg/kg, equivalent to 12.5 μg per mouse considering an average body weight of 25 mg. Indeed, if we use higher bleomycin doses (either 1 mg/kg (25 μg per mouse [but still far lower than Degryse and Liu]) or with 2.5 mg /kg body weight (62.5 μg per mouse), we also see pulmonary fibrosis in wild-type mice.
"However, we wanted to use the maximal bleomycin dose that does not induce pulmonary fibrosis in wild-type mice, but could be sufficient to induce pulmonary fibrosis in telomerase knockout mice. To establish this, we performed a titration analysis (0-0.1-0.5-1-2.5 mg/kg body weight) in wild-type mice. By doing this, we wanted to address whether short telomeres represent a susceptibility factor to develop pulmonary fibrosis triggered by low doses of a DNA damaging agent such as bleomycin. Treatment with [that higher bleomycin dose, which is still lower than Liu’s or Degryse’s] (> 1 mg/kg) induces DNA damage above the tolerated threshold in wild-type mice, leading to pulmonary fibrosis in both wild-type and telomerase-knockout mice. However, the low bleomycin dose used in our work is well tolerated by wild-type mice, while causing pulmonary fibrosis in telomerase-knockout mice that present a basal DNA damage due to short telomeres.”
She added: “Liu et al. used a bleomycin concentration of 40 μg per mouse and observed a telomerase activity dependence for pulmonary fibrosis development. These results stand in contrast to both our work and Degryse’s et al. Degryse et al. state that `The reason for this discrepancy between our results and those from Liu et al. remains unclear even though we used similar mice and similar bleomycin dose.’ In our work we carry parallel experiments with wild-type and telomerase knockout mice of different generations, and never observed any protection from fibrosis development by the telomerase deficiency. This is mentioned in our paper.”
All told, she said, the much higher doses of Degryse and Liu may be “too high, in that they may lead to lung damage in multiple cell types, but not generation of fibrosis.”
Another concern of some researchers is that Blasco’s 2nd and 4th generation knockout mice exhibited the same amount of fibrosis. Blasco responded: “We show that both G2 and G4 knockout mice show short telomeres. We think that is why we see pulmonary fibrosis in both G2 and G4. From previous work we know there is increasing embryonic lethality of the telomerase-deficient mice with increasing generations, which means that in later generations (G4) we are probably selecting for mice with the longest telomeres, which may explain why there is not such a big difference with G2.” That is, very short telomere mice start dying off at G4.
Another concern: in another paper, she showed trf1 inhibition had an anti-lung-cancer effect. Yet in the IPF paper, inhibition of trf1 had a pro-fibrotic effect.
Blasco noted the models are different. In the lung cancer model, her team deleted trf1 “by intra-tracheal inoculation of adeno-Cre, in principle targeting a certain percentage of different lung cell types. At the same time, we activate K-Ras. In that model, we did not see signs of pulmonary fibrosis, and did not study which percentage of AEC2s delete trf1. In the current model, we delete trf1 only in AEC2s, not other cell types. We do this very efficiently, eliminating most of these cells.”
Blasco told Bioscience Technology that, in both of her IPF models—the telomerase and trf1 knockouts—her team is likely “impairing stem cell function, as both telomerase and trf1 have been shown to be important for adult stem cells in literature. We think it is the combination of damage triggered by short or dysfunctional telomeres, together with an impairment of lung regeneration that contributes to trigger the fibrotic phenotype. Regarding the importance of apoptosis versus senescence for triggering the phenotype, we are currently addressing that.”
Concluded Kaminski: “I think both models in the new Blasco paper support the role of telomeres in pulmonary fibrosis, and the second model establishes a link between short telomeres in AEC2s and pulmonary fibrosis. I have no doubt that this mechanism is important in patients with familial pulmonary fibrosis, as well as in some of the patients with IPF--though I would still be cautious about the extent. The most common genetic association with IPF is a mutation in the promoter of a mucine gene (MUC5B). This discovery by (the University of Colorado's) David Schwartz has been now replicated many times on thousands of patients, and its effect size is way beyond anything else. Studying how this variant affects telomere length may be of interest to really understand sporadic IPF.”
Potential clinical approaches
In May, a report in PLOS showed that overexpressing telomerase protects human lung epithelial cells from bleomycin-induced apoptosis, adding this may lead to IPF therapies.
Still, the team behind the “Liu et al” paper has continued to argue that telomerase—not telomerase deficiency—is needed for fibrosis, if they note this may apply to mesenchymal, not epithelial cells. Stay tuned for a Bioscience Technology story on potential future telomere syndrome therapies.