In this column, we continue the series, begun in issue 10(1), of surveys highlighting a small selection of recently completed doctoral theses with particular relevance to the fields covered by Rejuvenation Research.
1
–11
While it has become common for thesis work to appear in the general academic literature, it remains valuable to scan the thesis databases for important advances that one might otherwise overlook.
Clearance of Capillary Occlusions Improves Cortical Blood Flow and Cognitive Function in Alzheimer's Mouse Models
Jean Cruz, PhD, Cornell University
Alzheimer disease (AD) is characterized by a loss of cognitive function caused by the dysfunction and death of neurons and other cells in the brain. This cell injury is largely due to the toxic effects of aggregates of amyloid-beta (Aβ), which accumulates into dense plaques in the brain. Research in humans and in animals suggests that brain blood flow is reduced in AD by ∼30%. Although it likely contributes to cognitive impairment and disease progression, no physiological explanation for this hypoperfusion has emerged.
In part, studies of cerebral blood flow pathology have been limited by the inability to perform in vivo imaging of vascular function at cellular resolution. The focus of this thesis is to present the underlying cellular mechanism responsible for this hypoperfusion phenomena and Aβ clearance in AD mouse models. Chronic cranial windows and in vivo two-photon excited fluorescence microscopy were used to study cerebrovascular blood flow.
While no blood flow disruption in cortical arterioles or venules were observed, blood flow was found to be stalled in an average of 1.8% of cortical capillaries in mouse models of AD, as compared to 0.25% in wild-type controls. These capillary stalls appeared early in disease progression, before any amyloid deposition. We found that the majority of the occlusions were caused by leukocytes, which adhered tightly to the endothelium. Indeed, blocking neutrophil adhesion in AD mouse models led to the fraction of capillary stalls decrease by 70%, causing brain blood flow to increase by ∼30% and cognitive function to improve.
These data suggest a working model to explain the origin of hypoperfusion in AD: Aβ accumulation leads to increased production of ROS that stresses endothelial cells and leads to increases in inflammatory receptors on the vessel lumen. This vascular inflammation causes leukocytes to adhere and plug capillaries, resulting in decreases in perfusion. This blood flow deficit could contribute to dementia independently of the direct effects of Aβ and could also accelerate Aβ aggregation by decreasing clearance of Aβ monomers.
This research provides for the first time an explanation for the long known phenomenon of reduced blood flow in Alzheimer's disease, one which has an early and significant impact on the development of pathological phenotypes. Brief reviews of multiphoton microscopy (MPM) and Brain blood flow in AD are also presented.
Comment: The existence of a relationship between disease progression and impairments to the flow of fluid within the central nervous system—both in the vasculature and in the glymphatics—is well recognized in Alzheimer's, in several other forms of neurodegeneration, and in the “normal aging” of cognitive function,
12
and it is now estimated to contribute to at least 50% of all dementias worldwide. Despite this prevalence, the mechanism underlying the connection has not yet been convincingly established. An important role has been confirmed for pericytes, multifunctional cells of somewhat uncertain developmental origin embedded in the basement membrane of most capillaries throughout the body: This population shares many characteristics with the vascular smooth muscle cells, which play a corresponding role in larger vessels, including stimulus-driven contractile function, but are further specialized for the regulation of capillary angiogenesis, maintenance of the blood–brain barrier, and facilitation of the brain's immune response. On the other side of the vessel wall, pericytes also have a major neurotrophic role through their secretion of pleiotrophin; loss of this single factor in model animals is associated with rapid neuron death secondary to enhanced vulnerability to ischemia and excitotoxicity, a phenotype that can be remarkably rescued by pleiotrophin infusions even in the absence of any residual pericytes. Their location within the basement membrane places them directly in the clearance pathway of amyloid-beta from the parenchymal space, and it is thus quite unfortunate that they appear to be particularly susceptible to Aβ-induced toxicity—not only degenerating entirely at higher concentrations (one of several potential explanations for the severe vasculopathy induced by early anti-amyloid immunotherapies, as well as the indirect neurotoxicity of Aβ per se) but also contracting tonically in response to the elevated ROS levels associated with much milder accumulations (well before the appearance of detectable aggregates) as a result of a feedback loop that inhibits the production of the endogenous vasodilator nitric oxide by adjacent endothelium (consistently, inhibiting arginine—another major inhibitor of endothelial NO production—largely rescues normal tension
13
). However, the chronic low-level vasoconstriction implied by this response has not to date been shown to be sufficient to mediate the substantial loss of overall cerebral blood flow observed in clinical disease. The studies reported in this dissertation may resolve this inconsistency by attributing the pathological connection not to the biophysical effects of Aβ on pericytes, but rather primarily its immunological impact. An elevated level of inflammatory signaling from Aβ-exposed endothelial cells induces pericytes to upregulate adhesion molecules such as ICAM-1—facilitating leukocyte adhesion—and downregulate proteins responsible for maintaining stable intercellular junctions (which is likely to further increase localized oxidative stress, thanks to the superoxide-generating properties of hemoglobin on its escape from the circulation). Although a very reasonable response to pathogen-driven acute inflammation, ongoing activation of this mechanism is likely to be a major contributor to the capillary stalls described in this work, and thus in the longer term to the gradual hypoxic deterioration typical of neurodegenerative pathology; notably, although Aβ is clearly a catalyst of the process when present, the more general phenomenon of “inflamm-aging” presumably also drives pathology in the same manner. Current technology does not permit the assessment of cerebral capillary stalls in human patients—particularly not via noninvasive techniques that are appropriate for use in the preclinical stages of disease, where it would be most useful to identify—but the profound impact of inhibiting neutrophil adhesion on net blood flow reported here may be sufficient to permit early diagnosis by the latter, more accessible metric. Of course, any sustained inhibition of immune cell recruitment poses significant risks of opportunistic infection, but it may, nonetheless, be therapeutically viable in some patient groups given adequately intensive monitoring.
Dermato-Informatic Approaches to Understanding and Improving Lesional Diagnostic Expertise in Cutaneous Oncology
Roger Aldridge, PhD, The University of Edinburgh
Cutaneous malignancies represent a quarter of all new cancer diagnoses in the United Kingdom. The key to reducing the tumors' associated mortality and morbidity is early diagnosis and treatment. Prompt diagnosis remains predominately a clinical skill, but relatively little investigation of the cognitive psychology underpinning expertise in this domain has been undertaken. This thesis aims to improve understanding of these processes and investigate how lesional diagnostic expertise might be enhanced. A large database of diagnostically tagged images was captured specifically for this project. A series of separate studies were undertaken to give insight into how lesional diagnosis occurs and how it can be improved. The studies highlighted that nonanalytical pattern recognition (NAPR) is likely to predominate in distinguishing malignant and nonmalignant skin lesions and that the widely promoted rules advocating analytical pattern recognition (APR) are not effective for discriminating melanoma from benign pigmented lesions. The keystone to promoting the development of NAPR and, thus, diagnostic expertise would seem to be increasing a novice's personal library of examples with relevant feedback. Studies demonstrated that current undergraduate exposure was variable but universally sparse, so simulation by way of diagnostically tagged images was developed, which showed accuracy could be improved by increased exposure. This improvement occurred in both a content specific and dose responsive manner. These studies also highlighted that the learning curves for skin lesions are not uniform. Further studies demonstrated that the choice of images had implications on the development of diagnostic expertise; suggesting it was important that these images represent clinical practice rather than “classic” examples traditionally advocated for teaching purposes. In addition, studies highlighted the potential benefit of the 3D models developed during this project. Building on the idea that a personal catalogue of relevant referent images was crucial to enhanced diagnostic accuracy, prototype software was developed to exteriorise the experts' library of examples; in the tests described novices utilizing the software delivered superior accuracy than medical students on the completion of their undergraduate teaching. In summation, the work described shows that by utilizing dermato-informatic approaches lesional diagnostic competence can be improved significantly.
Comment:
There has been a welcome surge in research applying machine learning techniques to medical diagnosis, a strategy that has the potential to greatly reduce the time that human experts are required to spend on the routine work of classifying and interpreting samples—and when sufficiently refined, even to enhance the reliability of diagnosis available to patients (especially those whose local health care systems are under-resourced). That surge has been prompted to a large extent by broader developments in network-based methods that employ nonanalytical pattern recognition, and that markedly outperform older rule-based classifiers in a wide range of contexts. This interesting thesis raises two important points related to what seems to be an almost inevitable upcoming shift in common practice. First, automated diagnostic software is commonly evaluated by comparison to a panel of human experts, whose verdicts are treated essentially as ground truth. This work highlights one danger of making that assumption, at least without careful selection of the judges; the significant effect of simple exposure to real-world samples on human diagnostic accuracy suggests that the true gold standard will be found among those who perform the most active clinical work—who it seems must be inherently less likely to be willing and able to devote the substantial time required to tag libraries of training images! Consequently, for clinical institutions to responsibly adopt any particular software solution, claims of “human-equivalence” will need to be qualified by an objective assessment of the performance of the humans in question as compared with the experts on the institution's own staff (Given the obvious practical benefits of adopting such tools once their accuracy is established, the largest barrier to widespread deployment may actually be the reluctance of institutions and individuals to concede that they have been outmatched.). The second concern raised by this dissertation is the predicted negative effect of the widespread adoption of diagnostic software on human skill levels. Once acceptably accurate tools are available, there will be very little incentive for medics to classify samples by hand, and that reduction in exposure would over time (ceteris paribus) lead to a decline in typical human expertise. Although continuing education is a standard element of medical practice, a specific focus on intensive diagnostic “drills” may be required to maintain capabilities, as such work is eliminated from the normal clinical routine.
Furthering the Scope, Understanding, and Application of Proteolysis Targeting Chimera
Daniel Bondeson, PhD, Yale University
Modern biomedical research has unveiled many of the complicated processes that underlie life at the most basic of level of cells. This enterprise has shown reversible protein phosphorylation, mediated by kinases, is an integral process whose mis-regulation causes many diseases, including cancer. Current therapeutic strategies targeting kinases have been focused on inhibiting enzymatic activity, and this has led to many approved therapies that extend the life and livelihood of many people.
This thesis explores the limitations of these strategies and a novel chemical biology tool to overcome them. In the first chapter, a brief history of kinases highlights how modern thinking focuses on kinase activity but ignores additional functions of protein kinases. The onco-kinase BCR/Abl is one such example: non-kinase roles of this protein are implicated in maintaining the disease and preventing cure even when the kinase activity is efficiently inhibited. A second example is the pseudokinase ROR2 and pseudokinases in general. These proteins share common structural features of kinases yet are enzymatically inactive and participate in important signaling programs within the cell. These two examples illustrate how inhibition is a limited paradigm of drug discovery.
Chapter two exemplifies recent advances in a strategy to overcome the limitations of inhibition. This strategy is called proteolysis targeting chimera, or PROTAC, and is based on heterobifunctional small molecules which bind to the protein target and recruit it to E3 ubiquitin ligases. The target protein is then ubiquitinated and degraded. While previous iterations of PROTACs have been rather unimpressive, this chapter highlights the degradation of a protein kinase as well as a nuclear hormone receptor. These PROTAC molecules are unprecedented in their potency, selectivity, and drug-likeness. The chapter concludes by discussing the many recent examples of PROTACs and their application in research and, soon, therapeutic interventions.
Chapter three then asks a basic and important question about PROTAC design. In designing potent degraders, minor structural changes in the molecule can lead to drastic effects on protein degradation. This chapter explores that phenomena, first by using a model system in which PROTAC geometry is finely tuned for degradation. It is shown that the discriminating factor between poor and potent PROTACs is the ability to form a stable ternary complex between the target, the PROTAC, and the E3 ligase. The best PROTACs induce protein–protein interactions between the target and E3 ligase, stabilizing the complex and leading to more potent degradation. In the second part of chapter three, PROTACs are explored which bind to many different kinase targets, but only degrade a subset of possible targets. Again, the discriminating factor between degraded and non-degraded proteins appear to be protein:protein interactions unique to the degraded proteins. This chapter offers biophysical explanations for commonly observed phenomena, and aids in developing design principles for PROTAC molecules.
Having shown PROTAC molecules to be a strategy for potent protein degradation in chapter two and enhancing the understanding of that platform in chapter three, chapter four returns to the two examples listed above. First, potent degradation of BCR/Abl is achieved through a PROTAC designed to target the allosteric site. Next, these compounds are used in initial assays to explore functions of BCR/Abl that are affected by either inhibition or degradation of the protein. Finally, initial studies in patient-derived stem cells are presented. While the viability of these cells is reduced by PROTACs, more nuanced work must be done to highlight differences between degradation and inhibition of BCR/Abl.
Second, initial efforts are made to develop ligands for the pseudokinase ROR2. While these compounds may not have activity on their own, they could be converted into PROTAC molecules which would deactivate all functions of ROR2. A thermal shift assay is used to identify potential ligands of ROR2 which bind with modest affinity. Future work will explore these compounds as well as developing high-throughput screens for pseudokinases in general. While previous iterations were limited in potency, this study demonstrates that PROTAC molecules can be versatile chemical tools.
While outside the scope of this thesis, PROTACs also show promise as therapeutic interventions. By degrading the entire protein rather than just inhibiting one functionality, PROTACs may expand what is currently considered druggable. Many literature examples point to this possibility. With the first PROTAC molecules soon to enter clinical trials, this study highlights the reasons for the considerable excitement surrounding this technology.
Comment: The human genome encodes hundreds of thousands of proteins, of which only a subset have enzymatic activity; the functionality of the remainder is thought to hinge instead on their interaction with the rest of the proteome. Small-molecule drugs have demonstrated generally high efficacy in modulating the former class, thanks in no small part to the inherently compact and well-defined active site that must be bound to inhibit the operation of an enzyme. However, interfering in protein–protein interactions with small molecules has proven drastically more difficult, primarily because the interacting surfaces are much larger; this greater contact area produces a much stronger net affinity, such that even a tightly bound small molecule may struggle to prevent the association, as well as offering up fewer unique structural features to target in search of clinically acceptable specificity. This disparity has led to the concept of the “druggable proteome”—the subset of the full proteome that can be effectively targeted by small molecules, and that unfortunately excludes a great many species of known or suspected pathological relevance. Insofar as simply removing such a protein from the cell constitutes a viable therapeutic approach (an assumption that holds best for those with relatively few interacting partners, such that the side-effects of their deletion are minimal), the most popular technique currently available to achieve this in living animals is RNA interference (RNAi). Identified in the late 1990s, RNAi has marked limitations—in particular the fragility and hydrophilicity of the therapeutic agent, necessitating the use of sophisticated delivery systems—but thanks to promising early results has been the focus of intense study, and very recently has begun to move into the clinic. However, a major strength of RNAi—its action at the level of messenger RNA, such that a single molecule of the agent can prevent the production of numerous copies of its target—does also constitute a weakness in the context of those targets that differ only post-translationally from related nonpathological species (where the group are products of the proteolysis of a single precursor, varying folds of a single peptide chain, or bear a specific post-translational modification
14
for example). Such groups, which are particularly prominent in the context of age-related disease, can only be inhibited en masse by RNAi; that lack of selectivity imposes a major and seemingly unavoidable limitation on the scope of the technique. PROTACs, first described only a few years after RNAi, directly induce the elimination of their protein target by labeling it for degradation via the proteasome. Consequently, they have the unique potential to selectively delete distinct products of a single mRNA within living cells. This dissertation summarizes and extends the body of work performed to date on this exciting class of therapeutics, and it crucially also clarifies the fundamental rules of their operation—groundwork that we very much hope will accelerate their increasingly overdue translation.
Optimization and Application of Synthetic High-Density Lipoprotein (sHDL) System in Atherosclerosis and Glioma Therapy
Dan Li, PhD, University of Michigan
Overwhelming evidence indicates that higher levels of high-density lipoprotein cholesterol (HDL-C) correlate with reduced risk of coronary heart disease (CHD). HDL can efflux excess cholesterol by reverse cholesterol transport (RCT). Hence, reducing acute plaque accumulation by direct infusion of cholesterol-free synthetic HDL (sHDL) has generated considerable interest. sHDL consists of a phospholipid bilayer held together by apolipoprotein A-I (ApoA-I). Due to the high manufacturing cost of recombinant ApoA-I, ApoA-I mimetic peptides complexed with a variety of lipids have been studied as treatments for various pathologies. However, the best methods of administration and formulation remain controversial. For sHDL products consisting of ApoA-I mimetic peptides like ETC-642, rapid elimination can limit their clinical application and subsequent development. Thus, prolonging the circulation time of sHDL can potentially improve its antiatherosclerosis effect. In addition, designing novel nanoparticles mimicking sHDL can eliminate the need for the ApoA-I protein/peptide component in sHDL while preserving the core pharmacological activity of sHDL.
We first studied the influence of administration route and lipidation of ApoA-I mimetic peptide 22A on plasma peptide levels, cholesterol mobilization, and lipoprotein remodeling in vivo. The mean circulation half-life for 22A-sHDL (T1/2 = 6.27 h) was longer than for free 22A (T1/2 = 3.81 h). The amount of 22A absorbed by the vascular compartment after intraperitoneal (IP) dosing was ∼50% for both 22A and 22A-sHDL. The strongest pharmacologic response was observed after intravenous (IV) injection of 22A-sHDL compared to IP injection and administration of free peptide. Both the route of administration and the formulation of 22A significantly affected the peptide's pharmacokinetic and pharmacodynamic properties. Following this study, sHDL surface modification with polyethylene glycol (PEG) was investigated for its potential to extend sHDL circulation in vivo. The circulation half-life of sHDL was extended both by adding more PEG or using PEG of longer chain lengths. Addition of PEG also increased the AUC for the phospholipid component of sHDL, leading to higher mobilization of free cholesterol in plasma due to prolonged circulation and increased stability.
To extend this research into biomimetic nanomaterial development, we formulated nanomicelles (NanoMCLs), structural nano-mimetics of sHDL with small particle size (12–14 nm) and a hydrophobic core and hydrophilic exterior. NanoMCLs were shown to be functionally similar to sHDL and exhibited up to 14-fold more efficient inhibition of inflammatory cytokine release in vitro compared to sHDL. When administered as a 6-week treatment to ApoE-/- mice fed a high-fat diet, sHDL and NanoMCL reduced atheroma by 21% and 40%, respectively. In addition, NanoMCL treatment significantly depleted atheroma macrophages.
Lastly, the application of sHDL as an anticancer drug delivery system was explored for treatment of glioma. Chemotherapeutic agent docetaxel (DTX) and immune-stimulatory toll-like receptor-9 (TLR-9) agonist cholesterol-CpG1826 (cholCpG) were co-incorporated in sHDL nanoparticles. The sHDL composition was optimized to maximize DTX retention in plasma. In a murine glioma model, intracranial injection of DTX-sHDL-cholCpG system exhibited significant antitumor efficacy with 20% of animals surviving past 90 days.
In summary, this thesis systemically studied the effect of sHDL lipid composition on cholesterol mobilization and established sHDL-PEG and NanoMCL systems to improve the antiatherosclerosis effect of sHDL. In addition, DTX-sHDL-CpG nanoparticles were developed and applied in glioma therapy. We have shown that sHDL is a versatile nanoparticle with utility in atherosclerosis treatment and drug delivery.
Comment:
Lipids and other hydrophobic molecules are transported through the bloodstream principally inside lipoproteins, short-lived particles bounded by a single layer of cholesterol-rich phospholipid membrane that is, in turn, stabilized by members of the apolipoprotein family—the precise complement of which determines the identity and behavior of the particle. There are five generally recognized classes of lipoproteins, but only the most compact, HDL, is responsible for removing such materials from cells with a surplus (via the ABCA1 transporter), whereas the larger particles play various roles in delivering them to areas of metabolic demand. Although excess lipids pose a potential threat to any cell, the most important instance from a clinical perspective is that which occurs in atherosclerosis, the progressive deterioration of circulatory function considered the leading cause of death in the developed world (and that is increasingly prevalent in developing countries also). In atherosclerotic vasculature, an overload of lipid and cholesterol within the vessel wall is associated with the transformation of macrophages into doomed “foam cells,” accompanied by chronic localized inflammation (aggravated by other sources of chronic inflammation),
15,16
which eventually leads to the formation of a scar-like plaque that obstructs blood flow. Facilitating the removal of the excess lipid and cholesterol present in such plaques is an obvious potential intervention, and infusions of HDL (or its hallmark apolipoprotein ApoA-1, which when administered in isolation binds to and remodels endogenous lipoproteins to improve their functionality) seem an obvious way to achieve that with minimal side-effects. However, although very promising results were demonstrated in early clinical trials around the turn of the century using the genetic variant ApoA-1 Milano—associated with extraordinary resistance to cardiovascular disease in human carriers, and with its robust reversal in model animals—the translation of this approach has been stymied by the cost and complexity involved in delivering full-length recombinant proteins. This thesis optimizes a semisynthetic alternative employing 22A, a peptide over an order of magnitude smaller than full-length ApoA-1 yet with similar HDL-organizing properties (one of a family of peptides originally developed to elucidate structural features of the full-length ApoA-1, but which unexpectedly demonstrated the ability to recapitulate its overall function), before moving on to prepare an entirely synthetic substitute constructed from phosphatidylcholine with varying levels of PEG modification. Both technologies are shown to have clear promise for therapeutic application. It should be noted that the causal relationship between fat accumulation and inflammation in atherosclerotic lesions is still not well understood, although both processes are enhanced by the presence of oxidized lipids and cholesterol (which simultaneously interfere with the efflux of their non-oxidized relatives, and also directly stimulate an inflammatory response). The selective degradation or removal of those toxic species is expected to be necessary for the lasting resolution of atheroma, as in many other forms of age-related disease.
17
However, it remains to be seen whether established plaques will, in fact, regress fully after the removal of the toxins that drove their formation, or whether the combination of the sheer over-concentration of “normal” lipid also present—combined with late-stage features such as fibrosis and ectopic calcification—will remain beyond the reach of endogenous repair mechanisms. Should the latter be the case, agents such as those introduced here will be essential to a comprehensive therapy; but either way, they should at the very least help to delay the onset of more recalcitrant pathology.
Regulatory T Cell Enriching Microparticles for Promoting Tolerance in Vascularized Composite Allotransplantation
James Fisher, PhD, University of Pittsburgh
Vascularized composite allotransplantation (VCA) is an emerging field encompassing transplantation of limbs and face. In clinical VCA, combination therapy with two or more immunosuppressive drugs is standard. Nonetheless, side-effects associated with the administration of lifelong, high-dose, multidrug immunosuppression continue to hamper wider implementation of VCA. Greater feasibility, wider acceptability, and routine applicability of VCA will only be realized if the risks of transplant rejection and chronic immunosuppression are minimized. An immediate goal is the exploration of novel strategies that achieve donor-specific immune hyporesponsiveness via local immunomodulation, minimizing or eliminating the need for systemic immunosuppression.
Interestingly, the cells of our bodies have evolved to utilize a host of strategies to maintain immunological homeostasis, representing a level of sophistication that dwarfs current attempts at immunosuppression. As a hallmark example, our bodies contain a subset of lymphocytes called regulatory T cells (Treg) that play a critical role in establishing and maintaining immunological homeostasis. However, because Tregs are found in low numbers throughout the body, strategies to harness them for therapeutic use have thus far focused on ex vivo expansion followed by in vivo re-administration. Though promising, the clinical implementation of these approaches is faced with numerous logistical and regulatory hurdles.
To this end, our group has developed synthetic approaches utilizing cell-sized, biocompatible, biodegradable microspheres composed of poly(lactide-co-glycolide) acid (PLGA) capable of releasing factors that can locally enrich for Treg. Specifically, PLGA microparticles that release the Treg-recruiting chemokine CCL22 (referred to as Recruitment-MP) were fabricated and tested for their ability to prevent allograft rejection in a rodent model of hindlimb transplantation. Indeed, Recruitment-MP was able to prolong hind limb survival indefinitely and promote donor antigen-specific tolerance.
Finally, an alternative strategy to locally enrich for Tregs via their induction from Naive T cells was described. Treg inducing microparticles (TRI-MP) that release T GF-β, Rapamycin, and IL-2 were also able to promote long-term graft survival and (importantly) induce donor-specific tolerance. Taken together, the Treg-enriching formulations described herein are synthetic systems that take inspiration from nature's mechanisms for resolving inflammation and have the potential to prevent aberrant inflammation in the context of allotransplantation.
Comment: Aging is the consequence of a lifelong accumulation of molecular damage throughout the body's tissues, progressively compromising normal function and setting the stage for age-related disease. Although the resolution of that damage at the level it occurs is in general expected to be the most effective and safest approach, the sheer variety of such biochemical lesions implies that it will still be many decades—at best—before a sufficiently broad portfolio of reparative therapies exists to render every species of damage harmless. In the interim, for all organs other than the brain, an alternative strategy is the bulk replacement of deteriorated tissue with a healthy substitute. However, such transplants face the fundamental obstacle of immune rejection, manageable only via sustained immunosuppression—which, of course, carries its own significant risks, most obviously an increased vulnerability to infections. A transplant engineered from the patient's own cells
18
(or cell therapy employing a nonimmunogenic population
19
) can, in theory, avoid this predicament, but such techniques are still in their infancy for most applications—and autologous tissue engineering fundamentally introduces an irreducible delay in the process, rendering that method unsuitable for urgent cases. The need for a reliable tactic to induce immunological tolerance to a graft has been recognized for decades, with a particular focus falling on regulatory T cells since their identification in the mid-1990s; this population has a marked capacity to delay or prevent graft rejection in vivo, and some Treg-based therapies have now demonstrated safety and are beginning to enter phase II clinical trials. However, cell therapies are fundamentally more difficult and expensive to develop and administer than less complex agents; although the delivery of exogenous Tregs may indeed be effective, it is probably not the optimal route for wide application (and insofar as it entails the expansion of patient-derived cells, will also be unsuitable for those in urgent need). Despite initial challenges in their isolation and characterization, our improving grasp of the signals that regulate the regulators now presents the alternative possibility of harnessing endogenous Tregs to promote transplant tolerance. This thesis constitutes a very exciting proof of principle for that option, demonstrating that an indefinite prolongation of vascularized graft survival can be achieved in mice by localized recruitment and/or induction of recipient Tregs through readily manufactured and administered drug-eluting PLGA microparticles. Should any similar agent be shown to have comparable efficacy in humans, its clinical deployment will profoundly transform transplant therapy—releasing the immune-matching bottleneck on donated organs and paving the way for “off-the-shelf,” immediate use of engineered replacement tissues as such become available. Locally Treg-enriching materials may also have potential in the prevention of autoimmune disease, particular where the reaction is specific to a particular organ—as in the case of classical type 1 diabetes and its slower-onset form LADA (“latent autoimmune diabetes of adults”), rheumatoid arthritis, primary biliary cholangitis, and so forth—further strengthening the argument for their swift development.
Tau Prion Strains Induce Distinct Pathological Phenotypes In Vivo
Sarah Kaufman, PhD, Washington University in St. Louis
Tauopathies are a diverse set of neurodegenerative diseases that feature the progressive accumulation of aggregated tau in the brain. Recent work suggests tau fibrils can propagate along neuronal networks and template the aggregation of natively folded tau protein (“seeding”). This prion-like spread is thought to underlie the progression of pathology observed in tauopathies. Prion diseases are caused by templated misfolding and assembly of prion protein into different amyloid conformations or strains. These distinct strains likely underlie the phenotypic diversity observed in human transmissible spongiform encephalopathies. During my dissertation, I tested whether tau can form unique prion strains that produce different pathological phenotypes in vivo. In collaborative work, we found that tau repeat domain can form different amyloid conformations that give rise to distinct morphological and biochemical phenotypes in culture. Inoculation of these strains into a tauopathy mouse model induced different patterns of histopathology that were stably propagated upon serial passage through three mouse generations. Aggregated tau from the final mouse generation retained the same conformations as the original strain inoculum, and induced identical cellular phenotypes after reintroduction into cell culture. I next showed tau strains can differentially target specific brain regions, and induce different rates of spread of tau pathology through the brain. To extend this work to human tauopathies, I developed an assay to quantify the level of seeding activity in fixed human tissue sections. I found that seeding activity progressively accumulates in Alzheimer's disease and progressive age-related tauopathy patients, beginning in the transentorhinal and entorhinal cortex before advancing to distant, anatomically connected brain regions. This work confirms tau can act as a true prion in vivo, and suggests different tau strains may underlie the progression of pathology and phenotypic diversity observed in human tauopathies.
Comment:
The hypothesis that infectious agents composed solely of protein could exist and be involved in transmissible spongiform encephalopathies was first proposed in the 1960s by Alper and Griffith, building on work from a decade earlier by Carleton Gajdusek on the ability of the disease kuru to pass from humans to chimpanzees (studies that earned Gajdusek a Nobel Prize in 1976). Griffith actually proposed three variants of the hypothesis, of which the second—that the protein involved could directly convert healthy copies of itself into the pathological form—was confirmed in the 1980s by the group of Stanley Prusiner, who coined the term “prion” (protein infection) and in 1997 received his own Nobel prize for his contributions. Until the middle of the last decade, all human prion diseases were believed to be caused by variants of the prion protein—the function of which remains unclear, although its extensive expression in nervous tissue may explain why all such diseases primarily display neurological symptoms. However, it has always been notable that prions are invariably associated with the formation of amyloid aggregates, raising the question of whether other amyloidoses might also be driven by prion species. Initial evidence for a prion form of α-synuclein in multiple system atrophy was uncovered in 2015, when nucleic acid-free brain extracts from MSA patients were shown to propagate the pathology in cell and mouse models (interestingly this was shown at the same time to be untrue for patients with Parkinson's disease, suggesting that a distinct strain of α-synuclein is involved in the two conditions). This thesis, based on work published by the author since 2014, extends the same findings to tau—the hallmark proteotoxic agent in a wide range of neurodegenerative conditions (a role that may be heavily related to the apparent positive feedback between tau-mediated cell death and cellular senescence
20
). Here, tau conformations (strains) derived from patients exhibiting a range of tauopathies are shown to drive distinct pathological features (such as cell-type specificity, rate of spread, degree of glial involvement, and more) and to be persistent across multiple serial passages through mice, classic features of a true prion. Of course the most prevalent neurodegeneration—AD—is typified more by the role of amyloid-beta than by tau, despite the definite involvement of the latter. An explanation for that connection does appear to finally be taking shape, after the recent discovery that distinct amyloid-beta strains can cross-seed distinct tau strains; AD brain-derived Aβ42 induces the seeding of compact tau fibrils, which self-seed efficiently, whereas Aβ42 from the brain of nondiseased individuals (identically purified and at the same concentration) induces the seeding of diffuse fibrils that are inefficient self-seeds (It should be noted that it is not at all certain that amyloid-beta itself is a true prion, at least at the monomer level; the evidence for the existence of strains in its case is instead at the level of oligomers, with those formed in the presence of excess fatty acids particularly well described and specifically associated with the development of cerebral amyloid angiopathy.). The recent and rather overdue acknowledgment of the disorder termed “primary age-related tauopathy,” defined by the presence of neurofibrillary tangles of tau identical to those seen in AD but in the absence of amyloid-beta plaques, further supports the theory that toxic Aβ oligomers catalyze a tau-centric process that can also occur sporadically in their absence. These findings do not change the damage-repair strategy for such pathologies—selective degradation or removal
21
of the toxic species—but will, nonetheless, certainly be crucial to the development of agents to achieve that goal in clinical practice, as well as potentially enabling differential diagnosis via strain identification in the very early stages of illness (at which point clinical symptoms are minimal or absent, and largely indistinguishable across the spectrum of tauopathies) when intervention is likely to be the most successful.
