Q: Dr. Lichtman, congratulations on your superb paper on regulatory T cell (Treg) changes in experimental hypercholesterolemia published in the July volume of Circulation. There are a number of highlights to this study but what were the key findings?
A: Thank you. Dr. Maganto-Garcia and co-authors worked very hard to obtain these interesting results.
There are three key findings from our study, which used the LDL receptor knockout mouse model of atherosclerosis. First, we found that Treg initially accumulate in aortas, along with effector T cells, as early lesions develop under diet-induced hypercholesterolemic conditions. Over a prolonged period of hypercholesterolemia, however, the Treg numbers in the aorta diminish, while effector T cells continue to increase. Second, prolonged hypercholesterolemia selectively impacted the phenotype of Treg and not effector T cells, including a reduced expression of functional selectin ligands, and impaired adhesion to aortic endothelium. Third, reversal of hypercholesterolemia by change in diet before the loss of Treg from lesions resulted in a sustained Treg presence and a reduction of effector T cells.
Q: You used an interesting animal model. Would you mind describing it? How difficult was it to generate?
A: We crossed LDLR KO mice with a reporter mouse line in which Treg are identifiable by their expression of a genetically encoded fluorescent marker. There are a several different Treg reporter lines described; the one we used was the FoxP3-GFP knock-in line developed in Dr. Vijay Kuchroo’s lab at the Brigham and Women’s Hospital. Stable expression of the transcription factor FoxP3 is a hall mark of Treg and Treg in FoxP3-GFP mice are identifiable by flow cytometry or fluorescence microscopy as green CD4+ T cells. These mice are invaluable for isolating viable Treg and studying their functions and phenotype.
Q: In this model then, it seems like you first directed your attention to the spleen and the circulating blood. What did you find with regards to the CD4+ and Treg cell population depending on the duration of the hypercholesterolemic diet? Where the trends the same in spleen and blood?
A: In cholesterol-diet fed LDLR KO x FoxP3-GFP mice we found that the number of Treg increased significantly from 4 through 20 weeks, while a more modest increase was seen in LDLR KO x FoxP3-GFP fed a control diet. Splenic effector CD4+ T cells did not increase in these same mice. No increase in Treg was seen in the spleens of wild type C57Bl/6 mice over the same time period. One interpretation of these results is that hypercholesterolemia has induced a systemic Treg response in peripheral lymphoid tissues. In the blood, the opposite trend was seen, suggesting a diminished trafficking of Treg out of lymphoid tissues, or decreased survival of Treg in the circulation.
Q: How did hypercholesterolemia affect regulatory T cell distribution at the level of the aortic arch then? You pursued a dual approach to this question, first by confocal immunofluoresence microscopy. Which antibodies did you use and why? Was the quantification easy? How did you quantify CD4 count in the lesions?
A: We performed en face confocal microscopy of the atherosclerosis-prone region of the lesser curvature of the aortic arch of cholesterol diet-fed LDR KO x -FoxP3-GFP mice to detect endogenously expressing GFP+ cells. The preparations were exogenously stained with anti-CD4 or anti-class II MHC reagents, followed by Alexa-555–labeled secondary antibodies and nuclear DAPI stain. This allowed us to confidently identify CD4+ GFP+ T cells (Treg) and CD4+ GFP- T cells ( effector T cells). The numbers of cells with each phenotype were manually counted in multiple fields and optical slices of each aorta. The quantification was tedious but otherwise relatively easy.
Q: Applying this technique, what happened to the CD4+ and Treg cell population in these atheroma which displayed a very nice progressive increase in lesion size over time? Did you assess any other aortic or vascular areas with this method?
A: Using the confocal technique, we found that lesional Treg decreased progressively at 8 and 20 weeks from a peak after 4 weeks of hypercholesterolemia, while total CD4+ T cells remained constant between 4 and 8 weeks, and then increased by 20 weeks. We restricted our confocal analysis to the lesser curvature of the arch, because this area is consistently involved in almost all LDLR-KO mice fed an atherogenic diet.
Q: You confirmed the microscopy findings using FACS on digested whole aortas. This seems to be a very elegant approach. Is this very easy to implement and does this generate fairly reproducible data? What are the caveats?
A: Collagenase digestion of aortas followed by FACS analysis of the released cells is a technique developed by Drs. Elena Galkina and Klaus Ley. The approach is not easy in that any one mouse aorta does not yield many cells, especially early in the development of atherosclerotic disease, and therefore many mice are needed for each experimental group, and digests need to be pooled to arrive at a single data point. Slow careful microdisection is also called for. Furthermore, some cell surface epitopes are destroyed by the collagenase digestion, limiting options for antibodies that can be used. In one of two separate experiments we performed with aortic digests, we stained for intracellular CD4 in order to avoid that problem. One general caveat is the fact that the precise anatomical source of the leukocytes that are obtained from the digest is not defined, and may include intimal lesional cells and adventitial cells. In our studies, we worked carefully to remove as much adventitial tissue as possible. Nonetheless, there is substantial data from the work of Galkina and Ley that in mouse aortic atherosclerosis, adventitial inflammation increases concomitantly with (and likely influences) intimal lesional inflammation.
Q: How well did the FACS findings correlate with the findings in the ascending aorta? Did you observe similar lesions in the descending aorta? Did you do cross-over studies so that at the end you would have a mix from both aortic regions in both groups?
A: We had very good correlation between the confocal findings in the aortic arch and two separate studies in which we FACS analyzed digest of the entire aorta distal to the root. In one of the FACS studies we identified Treg by endogenous GFP, and in the other, we stained the cells for intracellular CD4 and FoxP3. In all three studies there was a significant decrease in Treg numbers between 4 and 8 weeks. We included a 20 week time point in the confocal study, and in one FACS study, and in both, Treg numbers continued to decrease at this late time point. Between 4 and 8 weeks, effector CD4+ T cell numbers remained constant in the confocal study, but they increased in the FACS study. Effector T cells increased significantly between 8 and 20 weeks in both confocal and FACS studies.
We did not do a cross over study, because we used the entire aorta for the collagenase digests, in order to ensure we obtained enough T cells for the analyses.
Q: It appears then that the ratio of Treg:Teff decreased over time and that this decrease started relatively early on. Is this correct? By the way, were any baseline studies on the above and other parameters pursued? Could it be that these dynamics started even within the first 4 weeks?
A: Yes. There is good evidence from studies of various disease models that the ratio of Treg to effector T cells is the critical parameter in assessing how effective regulation of pathogenic T cell responses will be. In our work, a decreasing Treg:Teff ratio was a consistent finding at both 8 and 20 weeks, analyzed by confocal and FACS methods.
In 4-6 week old LDR-KO mice fed a chow diet (no cholesterol/low saturated fat), we found no atherosclerotic lesions in the aorta, and we could not detect any Treg or effector T cells in the aortic wall. It is certainly possible that a decrease in Treg migration, or Treg viability within lesions, might have begun before 4 weeks. Because the lesions are quite small even at 4 weeks (perhaps because of the favorable Treg:Teff ratio) we shied away from trying to quantify T cells at earlier time points.
Q: You then looked more closely into the mechanisms of how prolonged hypercholesterolemia reduced Treg accumulation. You found a reduction in cell viability as well as a reduction in adhesive and transmigratory capacity. You called this a change in the phenotype of Treg. Please tell us more about this.
A: Our data suggest that Treg are influenced by hypercholesterolemia or by the systemic inflammatory response that hypercholesterolemia induces, reducing migratory functions. The unusual lesional microenvironment appears to reduce Treg viability more than effector T cell viability. Our FACS studies and ex vivo analyses of viable Treg and effector CD4+ T cells obtained from the FoxP3-GFP mice consistently show that Treg migratory/adhesive phenotype is affected by the prolonged hypercholesterolemia, but the effector T cells are not affected in the same way.
The data in the paper do not address another possibility, which we are actively investigating now. It is possible that the inflammatory environment of the plaque may cause a re-differentiation of Treg into T effector T cells, including IL-17 producing Th17 cells. Treg plasticity is now well documented, and the abundance of innate cytokines in the plaque, including IL-1 secreted in response to cholesterol, may favor a change in Treg phenotype to Th17 cells.
Q: Is there any involvement of the bone marrow in this process?
A: If you are asking if the changes in Treg are a result of an effect of hypercholesterolemia on hematopoietic development, I would think not. A bone marrow problem should influence both Treg and effector T cells, because the common precursor to all T cells is in the bone marrow. T cells do not develop in the marrow, but rather progenitor cells of all T cells in the bone marrow migrate to the thymus. Both natural Treg and naïve T cells develop from common precursors in the thymus. Induced Treg differentiate in peripheral lymphoid tissues from the same naïve T cells that effector T cells are derived from.
Q: What are the underlying molecular mechanisms and pathways responsible for this change in functional capacity and viability? What is the trigger mechanism – is it hypercholesterolemia alone?
A: This is an important question which we have not yet answered. We hypothesize, based on some new data, that the innate inflammatory response to hypercholesterolemia is the culprit. Dendritic cells and macrophages are likely to be primary responders, which then influence T cells. The relevant pathways will be the cellular responses to damage associated molecular patterns (such as cholesterol crystals) and to cytokines.
Q: How did dietary reversal affect the dynamics?
A: Our data demonstrate that changing a cholesterol-rich diet to a cholesterol-free diet we were able to maintain high Treg numbers in the aorta and prevent the decrease in Treg binding to selectins and aortic endothelium. Mice were fed a cholesterol diet for 4 week and then the mice were fed a cholesterol-free diet for an additional 4 weeks in order to reverse the hypercholesterolemia. We confirmed that the diet switch resulted in significantly sustained peak Treg numbers in the aorta that correlated with a decreasing ratio Treg:Teff in aortas over a period of sustained hypercholesterolemia and a decrease of the lesion size. These results indicate that the changes in Treg dynamics that may have begun by 4 weeks are reversible by reducing cholesterol.
Q: This is quite fascinating. How transferrable are these data to human conditions?
A: There is ample evidence that Treg are present in human atherosclerotic lesions, and that Treg numbers in the blood and in lesions change with changes in human plaque inflammation, such as in acute coronary syndrome. Furthermore, there is evidence that reduced Treg correlate with enhanced disease activity in various human immune/inflammatory diseases. Whether or not the benefit of pharmacologic or dietary cholesterol reduction is in part due to maintenance or enhancement of Tregs is still a matter of speculation. However, the concept that Treg:Teff ratios are important in suppressing pro-atherogenic T cell responses should guide the development of anti-inflammatory therapies that specifically improve these ratios.
Q: Do you think that other risk factors such as diabetes and hypertension affect Treg cells as well?
A: Let me turn the question around, i.e. do Treg affect diabetes and hypertension? Various animal studies and some clinical evidence suggest they do. The contribution of inflammation to the metabolic syndrome and Type 2 diabetes is quite striking, and it appears that Treg are important players in these processes, at least in mouse models. The contribution of T lymphocytes to hypertension is also an area of active investigation. Thus, in addition to directly reducing pro-atherogenic T cell responses, sustaining Treg responses may beneficially ameliorate diseases that are important risk factors for atherosclerosis.
Q: What would you recommend as a therapeutic approach, simply cholesterol lowering or other specific therapies?
A: Our findings certainly support the current standard of aggressively reducing cholesterol. Statins have interesting pleiotropic effects on T cell function and may contribute directly to changes in Treg:Teff ratios, independent of blood cholesterol lowering. Perhaps some statins may work better than others in this regard. Of course our findings in a mouse model cannot yet serve as basis for any recommended change in therapy, but ongoing studies of the mechanisms by which plaque inflammation may impair sustained Treg responses could lead to more directed anti-inflammatory approaches.
Q: What do you see as the implications of this study and what are the next steps?
A: As I mentioned before, this study highlights an important therapeutic goal: to maintain high Treg:Teff ratios in the arterial wall. However there are still some questions that we need to answer about the mechanisms responsible for the changes on Treg function and numbers. One hypothesis is that cholesterol loaded antigen presenting cells influence Treg phenotype. Our next step is to focus in how the unique inflammatory microenvironment of the atherosclerotic plaque alters Treg differentiation and function.
Q: Thank you so much for this interview, Dr. Lichtman! What a great study – congratulations!
A: It has been my pleasure. Thank you for the opportunity to discuss our work.
Brief commentary to the paper by Maganto-Garcia et al.
"Dynamic changes in regulatory T cells are linked to levels of diet-induced hypercholesterolemia"
Insights into the interplay between metabolism and immunity
Three major findings:
• After 20 weeks, HFD, Tregs in aorta decrease upon diet challenges and there is increased atherosclerosis and heightened serum cholesterol levels. Foxp3+ cells in spleen increase and lymphocytes stained for Treg/GFP+ decrease.
• CD4+ Tregs are enriched in the aorta at 4 weeks but become dramatically less at 8 weeks. The reduction of Tregs in the aorta at 8 weeks is due to increased apoptosis.
• Migration and adhesion of Tregs to the aorta is impaired due to lower expression levels of GITR, E-selectin, PSGL1, P selectin, LFA-1, and CD62L after 20 weeks on diet.
• Diet switching causes restoration of Treg function/migration to the aorta and reduction of atherosclerosis and serum lipids!
This study eloquently defines new paradigms for the interplay between immunity and metabolism and in particular, its role in atherosclerosis. These conceptual shifts are that dietary factors impact Treg functionality and Tregs need to function properly to curtail disease onset. Hypercholesterolemia impairs Treg cells´ ability to migrate into the aorta, not only underscoring the importance of Tregs in atherosclerosis progression but implicating that lipids hinder Treg action. The reversal of the severity of atherosclerosis by eliminating the high cholesterol diet and subsequent restoration of Tregs in the aorta, solidifies the authors’ scientific story. This study will be a stepping stone for future investigations into the mechanisms by which dietary lipids affect the balance between inflammation and healing, and hence the balance between proatherogenic and atheroprotective immunity.
Robert Badeau and Goran K. Hansson
Center for Molecular Medicine
Karolinska Institutet, Stockholm, Sweden
European Society of Cardiology
European Heart HouseLes Templiers2035 Route des CollesCS 80179 Biot
06903Sophia Antipolis, FR
Our mission: To reduce the burden of cardiovascular disease
© 2017 European Society of Cardiology. All rights reserved