Commentary on Some Recent Theses Relevant to Combating Aging: October 2016

CO₂/HCO₃ ⁻/pH Sensing Soluble Adenylyl Cyclase Regulates Lysosomal pH

Nawreen Rahman, PhD, Weill Medical College of Cornell University

Lysosomes, the degradative organelle of the endocytic and autophagic pathways, function at an acidic pH, but the molecular processes that determine lysosomal pH are not understood. In particular, no lysosomal pH-sensitive signaling enzymes have yet been identified. Here we show genetically and pharmacologically that in the absence of bicarbonate (HCO₃ ⁻ )-regulated soluble adenylyl cyclase (sAC), lysosomes fail to properly acidify leading to accumulation of autophagic vacuoles (AVs). Due to the ubiquitous presence of carbonic anhydrases (CAs), which instantaneously equilibrate carbon dioxide (CO₂), HCO₃ ⁻ , and intracellular pH (pHi), sAC serves as a physiological sensor of HCO₃ ⁻ , CO₂, and pHi. We show that the sAC role in lysosomal acidification is dependent on its regulation by intracellular HCO₃ ⁻ in a CA-dependent manner, consistent with sAC serving as a pH sensor regulating lysosomal pH. Thus, sAC is the first, and thus far only known, pH-regulated signaling enzyme that affects lysosomal pH. Because intracellular vesicles such as lysosome are functionally dependent on acidification, this work has broad implications.

Comment: The function of lysosomes in the recycling of cellular waste material has been well established for many decades, with its significance underlined by the severe phenotypes observed in the congenital lysosomal storage diseases—as well as a growing association between lysosome impairment and several major age-related disorders. This function is strictly dependent on the maintenance of a highly acidic intraorganellar pH of around 4.75, which both directly accelerates the degradation of most biomolecules and is required for proper function of the numerous lysosome-specific hydrolases. Although the concept of the lysosome as a simple “bag of acid” has been effectively abolished by a long series of findings placing it alongside mTOR at the center of a complex regulatory web that dictates cellular responses to nutrient levels, the mechanisms responsible for maintaining its fundamental acidity remain incompletely understood. Full elucidation of the biochemistry has been greatly impaired by the lack of any method for simultaneously determining lysosomal pH and membrane potential. Undoubtedly, the V-type ATPase acts as the primary proton pump, expending ATP to import H⁺ into the organelle. This process generates a transmembrane potential and would rapidly stall in the absence of a dissipative counter-ion flux. There is strong evidence that chlorine is the counter-ion, with the ClC-7 antiporter considered the best candidate for the mediating protein. However, even this remains controversial; ClC-7 knockout mice do have a phenotype characteristic of lysosomal storage diseases, but their lysosomal pH is virtually identical to wild-type littermates. Meanwhile, no element of the pathway yet described allows for control of the final pH. Soluble adenylyl cyclase is expressed in all human tissues and is well characterized as a regulator of extracellular acidification by specialized cell types in the kidney, inner ear, bone, and epididymis. While the exact mechanisms vary, a common effector pathway involves increased trafficking of V-type ATPase to the relevant cell membrane in response to cyclic AMP produced by HCO₃ ⁻-stimulated sAC. (It is worth noting that while many other enzymes do directly respond to H⁺, the movement of free protons through the cell is retarded by their ready binding to other molecules; the equilibration of CO2/HCO₃ ⁻/H+ by carbonic anhydrases is so fast by comparison that it increases the effective diffusion rate by two- to fivefold.) The discovery in this thesis that the very same enzyme is necessary for lysosomal acidification represents a major step forward in this challenging field, although much remains to be explored—not least the question of how a protein thought to be localized to the cytoplasm can monitor pH levels within the lysosomal lumen.

Engineering Hypertrophic Chondrocyte-Based Grafts for Enhanced Bone Regeneration

Jonathan Bernhard, PhD, Columbia University

Bone formation occurs through two ossification processes, intramembranous and endochondral. Intramembranous ossification is characterized by the direct differentiation of stem cells into osteoblasts, which then create bone. Endochondral ossification involves an intermediate step, as stem cells first differentiate into chondrocytes and produce a cartilage anlage. The chondrocytes mature into hypertrophic chondrocytes, which transform the cartilage anlage into bone. Bone tissue engineering has predominantly mimicked intramembranous ossification, creating osteoblast-based grafts through the direct differentiation of stem cells. Although successful in specific applications, greater adoption of osteoblast-based grafts has failed due to incomplete integration, limited regeneration, and poor mechanical maintenance. To overcome these obstacles, inspiration was drawn from native bone fracture repair, creating tissue-engineered bone grafts replicating endochondral ossification.

Hypertrophic chondrocytes, the key cell in endochondral ossification, were differentiated from mesenchymal stem cell sources by first generating chondrocytes and then instigating maturation to hypertrophic chondrocytes. Conditions influencing this differentiation were investigated, indicating the necessity of prolonged chondrogenic cultivation and elevated oxygen concentrations to ensure widespread hypertrophic maturation. Comparing the bone production performance of differentiated hypertrophic chondrocytes to differentiated osteoblasts revealed that hypertrophic chondrocytes deposit significantly greater volume of bone mineral at a higher density than osteoblasts, albeit in a more juvenile form. When implanted subcutaneously, the hypertrophic chondrocytes stimulated turnover of this juvenile template into compact-like bone, whereas osteoblasts proceeded with processes similar to bone remodeling, generating spongy-like bone. Implanting these tissue-engineered constructs into an orthotopic, critical-sized femoral defect saw hypertrophic chondrocyte-based constructs integrate quickly with the femur and facilitate the creation of significantly more bone, resulting in a successful bridging of the defect. The success of hypertrophic chondrocyte-based grafts in overcoming the failures of tissue-engineered bone grafts demonstrates the potential of endochondral ossification-inspired bone strategies and prompts its further investigation towards clinical utilization.

Comment: Bone is unusual in having the capacity to regenerate fully under typical conditions, without forming the permanent, mechanically detrimental scars characteristic of other injured tissues. However, this process is successful in vivo only for lesions below the “critical size,” defined somewhat circularly in 1986 as “the smallest size intraosseous wound in a particular bone and species of animal that will not heal spontaneously during the lifetime of the animal” (although for reasons of practicality, a looser definition in terms of the length of a given study is now common). For long bones, the critical size is generally equivalent to a gap of two to three times the shaft diameter; larger defects instead fill with fibrous tissue, which inhibits osteogenesis and provides little or no functional support. Such injuries were historically treated using a native bone graft to bridge the wound site, but this approach faces inevitable challenges—autografts implicitly create a further defect elsewhere in the skeleton, whereas allografts pose the usual problem of immunological rejection. For much of the last three decades, efforts have been made to provide a credible alternative to native bone by recapitulating intramembranous ossification, the developmental mechanism that generates bone in utero. Unfortunately, grafts produced in this manner are already fully mineralized by the time of implantation. Consequently, there is little opportunity for the graft vasculature to connect with that of the host, and poor perfusion leading to necrosis of the graft core is a common complication. Meanwhile, the rigid contact sites that are surgically enforced between grafted and host bone bear minimal resemblance to a natural wound, which impairs proper biomechanical integration. Endochondral ossification—typical of adult bone healing—has been somewhat neglected in this context, aside from attempts to use soluble or scaffold-bound factors to encourage healing across super-critical defects. Yet, as this thesis rightly highlights, the intermediate cartilaginous stage that features in the endochondral pathway offers tissue engineers an opportunity to prepare flexible, immature grafts that are more capable of interacting with host tissues to achieve full vascularization and mechanical fusion. Such products will be particularly valuable in the context of age-related bone degeneration, where natural healing is significantly impaired—notwithstanding some efforts to reverse that trend via supplementation ^12,13 —and where the weakness of the surrounding bone makes robust integration especially critical.

Expanding the Accuracy, Resolution, and Breadth of Cell-Free DNA Investigation

Matthew Snyder, PhD, University of Washington

When cells die, they do not simply vanish without a trace. Instead, they leave behind fingerprints of their genetic and epigenetic identities in the form of cell-free DNA (cfDNA), or the scant amount of highly fragmented DNA circulating in human plasma. As the detritus of apoptotic and necrotic cell death in multiple tissues throughout the body, this class of molecule serves as a powerful biomarker for noninvasive detection and monitoring of disease processes and physiological conditions, including pregnancy, organ transplantation, and a growing number of cancers.

Despite this promise, current methods for interrogating cfDNA are challenged by limited resolution, imperfect accuracy, and constrained breadth. Taken as a whole, these factors restrict the set of conditions that might in principle be detected or monitored with this molecular evidence. In this dissertation, I directly address these limitations with the goal of expanding the scope and precision of the “liquid biopsy,” or the noninvasive monitoring of health status through cfDNA analysis.

I first address the limited resolution of cfDNA testing in the context of pregnancy by developing statistical methods for inference of the entire fetal genome at the single-nucleotide level, including both inherited and de novo variation. I show that the use of parental haplotypes and maternal cfDNA in a hidden Markov model can yield highly accurate prediction of inherited fetal genotypes. I next determine that the length of parental haplotype blocks is a key parameter driving the prediction accuracy, and demonstrate a method for increasing block length and downstream inference. I explain how these approaches, coupled with improved methods for detection of de novo variation, open the door to a single, noninvasive test with the possibility of prenatal detection of more than 3000 highly penetrant single-gene disorders.

I next demonstrate a method for improving the accuracy and positive predictive value (PPV) of noninvasive screening for fetal aneuploidy. In the most popular screening methodologies, PPVs of cfDNA-based tests are limited by a combination of the low incidence of trisomic pregnancies and the small number of false-positive tests, in which truly euploid pregnancies are incorrectly classified as aneuploid. I investigate the causes of false-positive test results in a small cohort and determine that maternal copy-number variants (CNVs) substantially contribute to the burden of spurious findings. I further develop a statistical framework for quantifying the likely impact of maternal CNVs by size and tested chromosome. I then propose a straightforward method for addressing this analytical limitation and improving test accuracy.

Finally, I develop a new approach to disentangle the various tissues or cell types contributing to cfDNA in a biological sample, potentially expanding the breadth of physiological conditions that can be monitored in this way. Here, I show that the locations of cfDNA fragment endpoints evidence the positions of proteins on the DNA in vivo in the contributing cells, and use these endpoints to infer the spacing of nucleosomes and transcription factors genome wide. I demonstrate that these positions correlate with gene expression profiles and use these data to model cell-type contributions in healthy individuals, where the expected myeloid and lymphoid cell lineages are recovered. I then apply this analytical framework to a cohort of individuals with advanced cancers and recover the tissue-of-origin of the primary tumor for a subset of the cancers.

Comment: The diseases of old age are distinguished by an initially very gradual accumulation of damage, largely compensated over most of the life span by homeostatic mechanisms. This slow pathogenesis poses a serious challenge for medicine; by the time a patient reports symptoms, a great deal of deterioration has already occurred and restoring fully healthy function is near-impossible. Detecting the earliest stages of disease would offer far greater therapeutic leverage, but is inherently challenging given the small magnitude of the changes and (typically) their confinement to a specific tissue or cell type. Removing these barriers requires a diagnostic technique that is cheap and noninvasive enough to apply on a routine, longitudinal basis—eliminating the substantial confounding effects of interindividual variation—and yet capable of detecting minor pathological changes in arbitrary body compartments, such as the localized loss of muscle in sarcopenia. ¹⁴ Over the last few years, dramatic advances in the analysis of cfDNA have begun to indicate that this may indeed be achievable, at least for the majority of pathological states characterized by elevated cell death. The nucleosome positioning-based approach developed in this thesis complements earlier approaches based on analysis of methylation patterns, both of which can help to identify the cell type responsible for generating a particular cfDNA fragment; conditions ranging from diabetes to multiple sclerosis and various cancers have already been so detected. Although the low abundance of cfDNA is likely to always limit the sensitivity of the technique, this issue is less serious in the longitudinal scenario. Meanwhile its exceptional scope, convenience, and specificity suggest that the liquid biopsy will become a crucial component of future diagnostic processes.

Histochemical Markers of Myelin Damage and Impaired Remyelination in the Aging Rhesus Monkey Brain: Relationship to Cognitive Performance

Larissa Estrada, PhD, Boston University

Myelin damage is known to increase in the normal aging brain and to correlate with age-related cognitive decline. While the causes of increased myelin damage are unknown, here we consider whether the brain's innate capacity for remyelination diminishes with age and hence could contribute to myelin damage through slow accumulation of myelin defects. Maintenance and repair of myelin depend on oligodendroglia precursor cells (OPCs), which must differentiate into a sufficient number of healthy mature oligodendroglia (oligos), the myelinating cell of the brain. The extracellular matrix molecule hyaluronic acid (HA) has been shown to inhibit maturation of OPCs into mature myelinating oligos. The present study examined aging changes in myelination using four markers: the damaged myelin basic protein (dMBP) antibody, a histochemical reaction to stain HA, and immunohistochemistry for OPCs and mature oligos. These markers were quantified using cell density (oligos and OPCs), percent area stained (HA and dMBP), and fluorescence intensity (HA and dMBP). Relationships between these markers, age, and behavioral measures of cognitive function were investigated using single and multiple regression analyses. Results showed that in the corpus callosum and cingulum bundle of the rhesus monkey, staining for dMBP as a marker of myelin damage strongly correlated with increases in HA. The increase in HA in the cingulum bundle correlated positively with age. OPC density increased with age in both the cingulum bundle and corpus callosum. Mature oligo density did not change significantly with age, but approached a significant increase in the cingulum and approached a significant decrease in the corpus callosum. The increase in OPC density correlated positively with both HA and dMBP in the cingulum bundle. These data are consistent with the hypothesis that HA accumulation contributes to myelin damage by inhibiting the differentiation of OPCs into mature oligodendrocytes, diminishing the brain's innate capacity for remyelination with age.

Comment: Myelin is a complex lipid and protein-rich material, which ensheaths axons, electrically insulating them from the extracellular fluid; it both greatly increases the speed at which an action potential can propagate, and prevents inappropriate cross-stimulation of adjacent neurons. Although the healthy adult brain is capable of replacing damaged myelin, this process breaks down dramatically in a variety of disease states—most famously multiple sclerosis—leading to severe and usually fatal neuropathology. Absent frank disease myelination peaks in humans around age 40, after which degenerative changes such as “myelin balloons” become histologically apparent at an ever-accelerating rate. The extent of this degeneration correlates very tightly with loss of fine motor control, and also to a significant extent with impaired memory and other symptoms of age-related cognitive decline. Since myelin integrity can be measured accurately via MRI, an unusually good opportunity exists to intervene very early in the disease process, potentially decades before serious symptoms appear. Designing such a therapy requires an understanding of how natural remyelination is impaired in the aged milieu, a knowledge gap that this thesis seeks to address. Hyaluronic acid is a large polysaccharide found throughout the body, with well-studied biomechanical and signaling roles. It is continuously and vigorously recycled, with around a third of the total present being degraded and re-synthesized on a daily basis; nonetheless, it accumulates with age in the central nervous system, primarily as a result of increased production by reactive astrocytes. Depleting the excess HA through careful stimulation of its degradation is an obvious therapeutic avenue, but one that must be pursued with great caution since its anti-inflammatory properties help to buffer the age-related increase in background levels of inflammation—indeed, superabundant HA is believed to be a key player in the exceptional longevity of the naked mole rat—whereas its breakdown products exhibit the opposite effect. This repercussion would be more tolerable in middle-aged brains, especially if anti-inflammatory drugs are coadministered, but for the elderly a more subtle approach that decouples elevated HA from impaired OPC differentiation may be required. Fortunately, good progress is being made in the engineering and transplantation of OPCs for the treatment of recognized demyelinating diseases, which should accelerate their translation to a more generally applicable therapy.

Mitochondrial Oxidative Stress and Antioxidant Therapy in Arterial Aging

Rachel Gioscia-Ryan, PhD, University of Colorado at Boulder

Advancing age is a primary risk factor for cardiovascular diseases (CVD), due primarily to development of age-related arterial dysfunction, including arterial endothelial dysfunction and stiffening of large elastic arteries. A key cellular mechanism underlying age-related arterial dysfunction is oxidative stress, a state in which cellular production of reactive oxygen species (ROS) such as superoxide exceeds endogenous antioxidant defense capabilities.

Vascular mitochondria are emerging as critical regulators of arterial function. Vascular mitochondria produce physiological levels of ROS (mtROS) for signaling, but mitochondrial dysfunction is characterized by excessive mtROS production. Thus, vascular mitochondria represent a key potential source of arterial oxidative stress; however, the role of mtROS in age-related arterial dysfunction has been unknown. Accordingly, the purpose of this dissertation was to determine the role of mitochondria-derived oxidative stress in arterial aging and to investigate the therapeutic potential of a mitochondria-targeted antioxidant, MitoQ, to ameliorate age-related arterial dysfunction.

In arteries of old mice, mitochondrial superoxide production was ∼3 times greater than in arteries of young mice, and this was associated with arterial endothelial dysfunction, measured as a reduction in nitric oxide-mediated endothelium-dependent dilation (EDD). Acute, ex vivo application of MitoQ to reduce mitochondrial oxidative stress abolished the age-associated impairment in EDD. Moreover, chronic, in vivo MitoQ supplementation (4 weeks in drinking water) in old mice completely restored EDD to levels similar to those of young mice, accompanied by normalization of age-related alterations in protein markers of mitochondrial health measured in whole arteries.

Arterial aging in mice was also characterized by elevated large elastic artery stiffness, assessed in vivo as aortic pulse-wave velocity (aPWV). MitoQ supplementation in old mice reduced aPWV to levels similar to those of young mice and this was at least partially mediated by attenuation of the age-related reduction in arterial elastin content.

Together, these results indicate that mitochondrial oxidative stress is a key mechanism underlying age-related arterial dysfunction. These studies demonstrate that reducing mitochondrial oxidative stress with the targeted antioxidant MitoQ restores EDD and reduces arterial stiffness in old mice, underscoring the therapeutic potential for mitochondria-targeted strategies to reduce mitochondrial oxidative stress, improve arterial function, and reduce CVD risk in humans.

Comment: The contribution of ROS to human aging has been a subject of vigorous enquiry since Denham Harman's seminal articles in 1954 and 1972—the latter elaborating his initial hypothesis to reflect the central role of mitochondria and thus explain the failure of early efforts to therapeutically exploit dietary antioxidants (which simply do not reach the organelle). An extensive body of evidence has accumulated in subsequent decades to support that role: impairing antioxidant defenses specifically in the mitochondrial matrix has profoundly deleterious effects on health and life span, particularly impacting stem cell self-renewal; while overexpression of existing mitochondrial antioxidants, including catalase and peroxiredoxin, rescues the proatherogenic phenotype of ApoE^−/− mice, reduces postischemic heart damage, and so on. Meanwhile, an increasing understanding of the significance of vascular dysfunction in age-related pathologies throughout the body—Alzheimer's, renal dysfunction, and retinopathy, to name a few—has further underlined the clinical significance of the phenomenon. In the context of the vasculature, elevated ROS enhances the migratory behavior of endothelial cells, leading to reduced barrier function and an increase in infiltration of the vessel wall by leukocytes. Superoxide, in particular, also reacts with nitric oxide to form peroxynitrite; this process both depletes an important natural vasodilator and yields a product that can directly poison the superoxide scavenger MnSOD. Meanwhile, increased oxidation and/or an age-associated decline in production of the cofactor BH₄, required for dimerization of endothelial nitric oxide synthase (eNOS), result in a substantial increase in monomeric (“uncoupled”) eNOS. This, unlike the dimeric form, actively produces superoxide—leading to a vicious cycle, particularly as the shear stress caused by impaired vasodilation drives increased eNOS synthesis. MitoQ, synthesized in the late 1990s, was the first in a series of artificial mitochondrially targeted antioxidants developed in response to the failure of systemic antioxidant therapy. It consists of a ubiquinone molecule attached to a lipophilic cation, a formulation that causes it to accumulate in the negatively charged mitochondrial membrane to concentrations many orders of magnitude higher than are present in the surrounding solution. Crucially, the ubiquinone moiety is rapidly regenerated by Complex II, allowing it to function as a long-lasting defense. Although effective in numerous model systems, including that of the current work, MitoQ has a significant limitation with regard to its use in the clinic; at only slightly higher concentrations (less than twice the effective antioxidant dose) it switches to a pro-oxidant mechanism. This finding particularly motivated the subsequent development of the SkQ series of molecules, where ubiquinone is replaced by the plant-derived antioxidant plastoquinone. For such compounds, the window between therapeutic and detrimental doses is a much safer 32-fold; one formulation, targeting dry eye syndrome, is already entering phase III trials in the United States after promising phase II results. Continual use of MitoQ or any of its derivatives is potentially unwise, given that mitochondrial ROS levels regulate autophagy, may suppress cancer, ¹⁵ and participate in many other physiologically important processes (we note that its recent inclusion in the NIA Interventions Testing Program should serve to clarify the net benefit or cost of long-term usage). It is thus fortunate that the results of this thesis indicate a marked therapeutic effect from relatively short-term treatment, potentially by virtue of interrupting the vicious cycle mentioned earlier.

Rational Drug Combinations Design Against Intratumoral Heterogeneity and Clonal Evolution

Boyang Zhao, PhD, Massachusetts Institute of Technology

Cancer is a clonal evolutionary process. This results in complex clonal architecture and intratumoral heterogeneity in each patient. This also presents challenges for effective therapeutic intervention—with constant selective pressure to induce or select pre-existing resistant subclones toward drug resistance. Mathematical/computational modeling from population genetics, evolutionary dynamics, and engineering are being utilized to a greater extent in recent times to study tumor progression, intratumoral heterogeneity, drug resistance, and rational drug scheduling/combinations design. In this thesis, we present several joint quantitative and experimental approaches for the rational design of drug combinations to tackle the issue of intratumoral heterogeneity and clonal evolution.

Using a tractable experimental system with predefined tumor compositions, we derived computational approaches to rationally design drug combinations with the goal of minimizing a given heterogeneous tumor. We found that the best drug combinations can oftentimes be nonintuitive as they do not contain component drugs most effective for the individual subpopulations. This was the result of a need for combinatorial considerations on the effects of each drug on all subpopulations, hence at times leading to nonintuitive drug regimens. We validated our computational model predictions in vitro and in vivo in a preclinical model of Burkitt's lymphoma, with predictable evolutionary trajectories on treatment. Next, we extended this methodology to study the effects of more complex tumor heterogeneity on combinatorial drug design, with similar conclusions. Sampling and statistical analyses over a range of tumor compositions can further inform effective drug combinations under some uncertainty in initial tumor heterogeneity. Moving beyond a model where we have control of initial tumor composition, we sought to examine collateral resistance and sensitivity during clonal evolution. Using a murine model of Ph+ acute lymphoblastic leukemia, we performed drug selection and pharmacological screen experiments. We observed important evolutionary processes of selection and drift in giving rise to resistance to clinically used BCR-ABL1 inhibitors. Remarkably, the resistant population also became hypersensitized to nonclassical BCR-ABL1 inhibitors at intermediate stages of the clonal evolution, in this so-called temporally collateral sensitivity. Mathematical modeling and experimentation brought additional insight into the evolutionary dynamics and mechanism of action, with demonstrated in vivo efficacy.

These quantitative approaches, complemented with extensive experimentation, facilitate a principled approach to our understanding of making forward predictions that directly inform therapeutic drug regimen designs.

Comment: Forty-five years after the “War on Cancer” began, and despite annual research expenditures now well into the tens of billions of dollars, it remains one of the leading causes of death both worldwide and in the United States. The situation is further aggravated by the various degenerative changes that predispose an aging population to carcinogenesis. ¹⁶ Our investments have not been wasted—an array of drugs now exist that can efficiently destroy tumor cells with particular biological abnormalities—but almost all forms of the disease continue to resist any lasting medical solution, evolving orders of magnitude faster than healthy cells to dispense with each such vulnerability as it is exploited. Of course, this problem is well known: chemotherapy is now routinely delivered as combinations of two to five drugs, each with a distinct mechanism of action, designed in theory to eradicate the entire population of neoplastic cells before any one of them can acquire resistance to the full cocktail. However, such combinations are typically assembled on the basis of synergistic cytotoxicity in preclinical studies of model tumor lines, an approach that neglects the complexity of each patient's disease—and indeed has yielded only incremental improvements in long-term survival over single molecules. To truly maximize the potential of our existing arsenal, we must learn to adapt our therapeutic strategies far more precisely and dynamically. The volume and variety of relevant data now available for each patient suggest that doing so will require a shift from human decision-making to fully computational methods. This thesis makes excellent progress in demonstrating both that requirement and the value of the latter approach. In a mature form, the automated design of combinations will have the additional benefit of minimizing the need for skilled labor—thus improving both accessibility and overall costs.

Technology Development in Mouse Genetics and Epigenetics

Chikdu Shivalila, PhD, Massachusetts Institute of Technology

The importance and significance of a model organism in biological research cannot be overstated. The mouse, in particular, has been very useful in understanding questions in many areas of research such as developmental biology, cancer biology, neuroscience, and genetics. However, even though the methods to make transgenic mice and gene knockin and knockouts have been successful, they are very inefficient, labor intensive, and costly. Therefore, in this thesis, we developed a novel methodology to rapidly and efficiently modify the mouse genome. Using CRISPR/Cas9, a novel genome-engineering technology developed from bacteria, we were able to genetically modify mouse embryonic stem cells and make mice that carried genetic modification by zygotic injections. Using CRISPR/Cas9, we were able to make mice in as little as three weeks that contained multiple gene knockouts, single-nucleotide modifications, GFP and mCherry reporter alleles, epitope-tagged alleles, and conditional alleles.

Another interesting area of research in mouse genetics is epigenetic regulation, specifically how DNA methylation regulates development, gene expression, and cell state. Multiple studies have shown that this epigenetic modification plays an important regulatory role in these processes; however, the technology that has existed so far to investigate DNA methylation has only been able to look at snapshots of methylation patterns in fixed cell populations. In this thesis, we have developed a novel technology named Reporter of Genomic Methylation (RGM), which allows for the investigation of methylation dynamics at a single-cell resolution in vivo. The RGM technology was developed using a minimal synthetic secondary DMR promoter that drives the expression of a florescent protein. Using CRISPR/Cas9, the RGM reporter can be integrated into any genomic locus, where it can report on the methylation state of its surroundings. We further show that the RGM reporter activity reflects the methylation state of noncoding regulatory elements such as promoters and enhancers. Furthermore, we show that the RGM technology allows for the dynamics of methylation and demethylation to be observed at these noncoding loci, as cells transition between a pluripotent and differentiated state.

Comment: This thesis reports on recent work conducted at MIT in the laboratory of Rudolf Jaenisch, whose unintentional creation of the first transgenic mice in 1973 is generally considered to mark the beginning of the modern era of animal genetic engineering. It was not until 1989 that a (subsequently Nobel-winning) procedure was developed for knocking out existing genes, often a more useful manipulation for research purposes than the introduction of new DNA. Unfortunately, that protocol is laborious, frequently taking years to complete: extract, culture, and modify embryonic stem cells; reinsert the engineered cells into blastocysts, implant those into a surrogate mother; and then cross-breed the initially chimeric offspring over several generations to reach full homozygosity. In the last decade, much faster techniques have been developed that use site-specific nucleases at the single-cell stage to introduce specific modifications, moreover, enabling the engineering of model species with poorly characterized ES cells. Unfortunately, the design of such enzymes is itself nontrivial, frequently taking months of work, and in practice has been limited to a minority of laboratories. CRISPR/Cas9, by contrast, is a cheap and straightforward technique readily implementable by most research groups. The impact of being able to produce models with complex genetic alterations in just threeweeks—the length of a murine pregnancy—can hardly be overstated. Meanwhile, it is well established that the methylation status of many genomic sites changes predictably in tight correlation with chronological age, and indeed resets to a “youthful” configuration during induction of pluripotency. The second technology described here, which enables the continual monitoring of DNA methylation at single-cell resolution, will be of great value in establishing the nature of any causal relationship between those changes and age-related degeneration. ¹⁷