Cellular and molecular features of senescence
Introduction: cellular basis for ageing and preliminary intervention strategies
The phenomenon of physical deterioration with advancing age is a common feature of all multicellular life, with each plant or animal species exhibiting characteristic rates of decline as a normally occurring element of its developmental program. Here, the term ‘apparent’ is used because the existence of a specific genetic mechanism selected by evolution to limit the maximal lifespan of an individual – also termed a ‘longevistat’ – is an ongoing subject of debate .
Ageing is a complex and fascinating process that has been the focus of intense efforts to understand and counteract throughout most of human history. Until the 20th century, such initiatives were entirely in the domain of cultural mythology. The most famous of these in Western tradition was the pursuit of a ‘Fountain of Youth’ by the Spanish conquistador Juan Ponce de Leony Figueroa .
Advances in medical technology and rapid increases in understanding of cellular and molecular biology have stimulated increasingly sophisticated efforts to define ageing at a level that may allow interventions resulting in lifespan extension. Unfortunately, although enormous gains have been made in the development of therapies for age-associated diseases, such as chronic inflammatory disorders, efforts to decrease rates of physical ageing as a whole have been hampered by an inability to fully characterize the underlying mechanisms of the ageing process. Thus, clear characterization of physical ageing has been difficult due to the lack of a clear case definition that may lead to intervention strategies.
Early characterizations of cellular ageing – the ‘Hayflick Limit’
The underlying complexity of age-related physical decline has led many in the scientific community to conclude that the process may represent a general deterioration of cellular function, with insufficient unifying features to allow for definitive interventions. Interestingly, although recent characterization of cellular senescence seems to support such conclusions, ambiguity exists in the interpretation of ongoing studies in this field . This work will be discussed in section I-2 of the present review.
Arguments that ageing is too intricate a process to effectively counteract are diminished to some degree by the implications of studies in the 1960s by Hayflick et al. demonstrating the replicative limits to eukaryotic cells in culture . This phenomenon, known as the Hayflick limit, is defined as the average maximum number of doublings a population of normal cells is capable of undergoing before its replicative potential is lost. This maximum in vitro doubling capacity was observed to vary, depending on the median lifespan of the species from which cells used to establish the culture were taken. For example, cultures of human foetal cells are observed to double 40–60 times before losing proliferative potential , whereas cultures established from mice, a short-lived species, double a maximum of approximately 15 times , and cells from Galapagos tortoises, which live well over a century, demonstrate an upward in vitro doubling limit of around 110 times in culture .
Significantly, it has also been observed that cell cultures derived from patients afflicted with progeroid diseases, in which features of rapidly accelerated ageing are a primary symptom, exhibit far lower Hayflick limits than cells from normal individuals . The strong positive correlation between cell lineage doubling potential (as defined by the Hayflick limit and by longevity of a particular individual) may imply the presence of a normally occurring physiological process acting to limit maximum lifespan within a particular species. This is the definition of a ‘longevistat’ as discussed in a review by Dale Bredesen at the Buck Institute for Age Research, at the University of California in San Francisco , which examines evidence for and against the existence of such a process, particularly the contribution of cellular senescence to physical ageing . One implication of the correlation between the Hayflick limit of cells in culture and the median lifespan of the species from which the cells were taken is that, if genetically determined mechanisms for lifespan limitations exist, it may be possible to characterize their underlying features and then to intervene at some point in their normal function in ways that result in lifespan extension.
Life expectancy versus median lifespan
It is important to distinguish between life expectancy and lifespan. Average life expectancy is the length of time an individual within a population of organisms may be expected to survive when disease, accidents, predation and other environmental stressors are factored in. Conversely, median lifespan is a measure of the time an organism is expected to survive in the complete absence of environmental stressors [8, 9]. The objective of age-intervention initiatives is to make use of cutting-edge technology to extend median lifespan significantly in excess of the normal range for a particular species.
Telomere length: correlation between telomere erosion and Hayflick limit in vitro
Studies of changes in nuclear chromatin organization occurring during propagation of cell lines have revealed what may constitute elements of a longevistatic process in multicellular organisms. The replicative potential of a particular cell lineage has been observed to correlate with the length of chromosome tips, called telomeres. These structures, which contain non-coding DNA, are shortened each time a cell divides, and after a finite number of divisions, become critically shortened, signalling a cell to stop dividing . The Hayflick limit may thus be defined on a cellular morphological basis as the number of divisions necessary to critically deplete telomeres.
Cellular immortality, telomerase activity and telomere length
The aforementioned Hayflick phenomenon may offer insight into strategies for intervention in age-related physical deterioration of an organism based on a major feature of cell division: telomere length. For example, treatment of human fibroblasts in vitro with carnosine, a dipeptide antioxidant occurring naturally in vertebrate brain and muscle, decreases telomere erosion rates during cell division and increases the Hayflick limit of treated cultures [11, 12]. An even more potent approach to preserving telomere integrity and maintaining genomic stability of a cell is offered by specifically amplifying expression of telomerase, an enzyme that normally repairs telomeres . The role of telomerase during progressive cell division activity is shown via a diagram in Figure 1.
Initially, chromosome telomere length and structural integrity in eukaryotic cells derived from foetal tissue are maintained by high levels of telomerase. The activity of this enzyme progressively decreases with each cycle of cell division, resulting in greatly shortened telomeres, decreased organization of chromatin and deleterious effects, including cell cycle arrest and cellular senescence [3, 13, 14].
Sustained up-regulation of telomerase potentially enables unlimited replicative capacity for a cell . Indeed, telomerase activity is significantly diminished in cells nearing their Hayflick limit both in vivo and in vitro [3, 13]. Interestingly, multi-organ deterioration in a telomerase-deficient mouse strain that underwent greatly accelerated physical ageing was completely reversed by induction of high activity levels of the enzyme. In these experiments, mice exhibiting physical age near the end of the typical lifespan for this strain were restored to full youthful vigour .
Unfortunately, cellular immortalization, including sustained high telomerase activity, is an intrinsic element of carcinogenic transformation of a cell [15-17]. Moreover, as a major feature of virtually all tumour cells is the absence of replication-associated telomere shortening and abolition of the Hayflick limit, a major anticancer strategy currently being pursued is the identification of telomerase inhibitors that restore progressive decrease in telomere length with cell division, thus causing tumour cells to die upon reaching their Hayflick limit.
Limits to use of in vitro outcome in design of age-intervention strategies
Attempts to identify functional correlations between the Hayflick limit of cultured cells and whole-organism lifespan reveal major paradoxes in efforts to develop strategies for intervention in the ageing process based on outcome of in vitro studies of cellular ageing. Extension or abolition of the Hayflick limit by sustained high telomerase activity in a particular cell line has value in ongoing characterizations of ageing on the cellular level, particularly because it may be induced in any modern biomedical laboratory. Nevertheless, clinical use of this phenomenon in age-intervention strategies must be approached with extreme caution as sustained high telomerase activity is a key element of carcinogenic transformation [16, 17]. For this reason, strategies that may extend the replicative potential of cells in a living organism carry a high risk of cancer formation unless countermeasures are employed in such procedures to decrease the risk of carcinogenesis. Approaches for doing this are explored in section III-3 of the present review.
Another limitation to fibroblast use of in vitro data in the design of animal or human age-intervention strategies is that the behaviour of cells in culture differs substantially from their function as elements of living tissue. In vitro models also suffer from weak relevance to whole-organism ageing. For instance, although some investigators have reported that cells taken from elderly human beings exhibit lower replicative potential than those taken from younger volunteers , subsequent evaluation of methodologies used to draw these conclusions cast doubt on their validity . An additional pitfall for broad application of in vitro data is that cells exhibiting senescent phenotypes have been detected in vivo, but without the characteristic shortening of telomeres characteristic of cultured cells at their Hayflick limit . Such cellular phenotypes may arise from healthy cells when subjected to stress that damages their DNA in ways that fail to activate cell death programs and potentially result in carcinogenic transformation. Cells sustaining such damage undergo replicative arrest, preventing their progression into cancers, but also develop a senescent phenotype through the process of stress-induced premature senescence (SIPS), that express products, especially mediators of inflammation, which may impair healthy tissue function, promote carcinogenesis in neighbouring cells and generally contribute to disease- and age-associated whole-organism deterioration [3, 21, 22]. Young, undamaged cells express products and perform functions that contribute to healthy homoeostasis. Old cells and damaged cells that have developed a senescent phenotype affect surrounding tissue in ways that contribute to age-associated deterioration, including dysregulated inflammation and cancer [3, 21]. The influences on an organism of young, undamaged cells versus stressed cells that become senescent and aged senescent cells (cells which may have reached their Hayflick limit) are outlined by the illustration shown in Figure 2, below.
If the limitations implied by the aforementioned observations are understood and taken into account in the development of research strategies, cell culture models may nevertheless be used as a very potent tool in designing experiments with expected outcome relevant to whole-organism ageing. Indeed, cell culture models that incorporate SIPS cells and other senescent cellular phenotypes have provided evolving unprecedented understanding of how cellular senescence contributes to age-associated deterioration of tissues and whole organisms .
Effect of eliminating senescent cells on age-associated pathologies in progeroid mice
In November 2011, investigators at the Mayo Clinic College of Medicine in Rochester, Minnesota reported that selective ablation of senescent cellular phenotypes from progeroid (age-accelerated) mice significantly delayed onset of age-related physical deterioration and decreased severity of established degenerative symptoms in aged animals . These studies utilized a sub-strain of the BubR1 progeroid mouse bearing INK-ATTAC, a drug-inducible transgene, designed to respond to drug treatment by elimination of cells expressing the tumour suppressor protein p16Ink4, which is a biomarker for senescent cells and a critical contributor to their emergence through elimination of proliferative capability . When compared with untreated mice, drug-treated animals in which significant attrition of senescent cells had occurred were substantially healthier. Age-related phenotypic parameters that were used as a basis for this observation included: incidence of lordokyphosis and cataracts; mean skeletal muscle fibre diameters measured in 10-month-old mice; physical stamina; body and fat depot weights of 10-month-old mice; fat cell diameters in 10-month-old mice; dermis and subdermal adipose layer thickness of 10-month-old mice; and several other major health-related features that deteriorate with age and at a very high rate in the BubR1 progeroid strain, used as a model in the aforementioned studies .
Inhibition of accelerated ageing in mice by telomerase reactivation – 2011 Dana Farber studies
Another striking example of how insight gained from in vitro studies has contributed to the development of approaches to suppression of age-associated pathologies in animals is provided by the study with age-accelerated mice conducted by Jaskelioff et al., described above . This very intriguing research, recently conducted at the Dana Farber Cancer Institute, tested a hypothesis that physical deterioration resembling accelerated ageing in a mouse strain with short, dysfunctional telomeres could be inhibited by systemic amplification of telomerase expression. In these experiments, a transgenic strain of mice was created from telomerase-deficient animals to contain an expression cassette for the enzyme, inducible by the drug 4-hydroxytamoxifen (4-OHT), an oestrogen receptor agonist/antagonist used in therapy for cancer and endocrine disorders. In comparison with sham-treated mice, transgenic animals administered 4-OHT as time-release pellets to reactivate telomerase expression manifested dramatic reversal of the rapid physical deterioration associated with the parent phenotype. Amplified telomerase expression in the transgenic mice restored physical integrity and function of all tissues examined to levels expected in young, healthy animals. These dramatic improvements in prognosis occurred despite the use of mice that exhibited physical deterioration equivalent to extremely aged animals near the end of lifespans expected for the telomerase-deficient strain. Of particular interest was an observation that neural degeneration was completely reversed in animals engineered to contain a 4-OHT-inducible somatic telomerase reactivation element, along with the restoration of smell, reproductive capacity and liver function . The dramatic effect of telomerase reactivation on apparent physical age of mice used for these experiments is shown below in Figure 3.
Telomerase-deficient mice engineered to contain a 4-OHT-inducible telomerase reverse transcriptase-oestrogen receptor (TERT-OR) allele respond to stimulation with 4-OHT by reactivation of telomerase activity and maintenance of telomere length along with healthy function of tissues and organs. Mice not treated with 4-OHT exhibit low telomerase activity and rapid deterioration of multiple cellular and organ functions expected as an outcome of defective telomeres. Conversely, mice implanted with 4-OHT-releasing pellets experienced somatic telomerase reactivation and significant delay of progeric phenotype onset. Dramatic results of these evaluations are shown below. Here, the 4-OHT-treated, telomerase-enhanced, 48-week-old mouse on the left exhibits a physical age apparently younger than the 35-week-old non-4-OHT-treated progeric animal .
The relevance of these findings to vertebrate ageing remains to be determined. Despite the enhanced cancer risk likely to be engendered by telomerase-based ‘rejuvenation therapies’, it is tempting to predict that results of the work at Dana Farber reveal a solid mechanistic strategy for intervention in the normal process of ageing. However, at the time of this writing, such a conclusion would be premature, especially since Jaskelioff et al. did not report a favourable impact on lifespan as an outcome of their study. Moreover, the extremely high carcinogenic potential of telomerase augmentation makes it unlikely that such an approach to age intervention will become practical for human beings unless and until more effective means of suppressing and/or treating cancer have been developed.
Finally, observations reported by the Dana Farber group were based on a highly specialized progeric mouse strain bred for short telomere length and short lifespan. Hence, their findings should not be extrapolated to other vertebrate models, including human beings, until ongoing studies of telomerase amplification on normal ageing begin to produce definitive data. The status of telomere integrity and telomerase expression in a particular cell species influences its pattern of gene expression (epigenomic profile), which is manifested as cellular phenotype. Here, the term ‘epigenome’ refers to gene expression profiles for any particular cell type, which define its phenotype and homoeostatic role. Epigenomic profiles of cells within a tissue develop in response to interaction between the internal makeup of a cell and its external environment. This, and many other influences, create mosaics of cell phenotype within tissue that ultimately act as major determinants of lifespan, as discussed in section on shifting epigenomic profiles, below.
Major epigenomic profiles represented in tissue: functional cell categories
Lifespan definition and major contributing factors
‘Lifespan’ for any particular animal, plant or cell, may be broadly defined as the average length of time that the organism may resist succumbing to environmental and internal stressors – in the absence of disease, accident or other catastrophic events. The ability of an individual to remain alive and functional is dependent at a fundamental level on the number and distribution of five major functional cell categories, each defined by a unique epigenomic profile. These are described briefly as follows:
- Regenerative mitotic (stem) cells: These are cells capable of maintaining tissue integrity by proliferative response to damage, including ‘progenitor’ stem cells, which are specialized types of stem cell differentiated to a degree that allows them to replace a specific type of tissue such as muscle cells or neurons .
- Quiescent post-mitotic cells: These cells are healthy and terminally differentiated to functional roles in which cell division is not necessary. Examples include nerves , heart muscle and other specialized tissue.
- Transformed cells (cancers): Cells that have undergone cancerous transformation represent the greatest drawback to regenerative capacity of a particular type of tissue. A capacity for tissue regeneration has given organisms that have it, such as vertebrates, an enormous adaptive advantage over multicellular life forms in which non-dividing cells constitute the predominant tissue . However, proliferation-capable cells that sustain non-lethal genomic damage in the form of somatic mutations, often propagate this damage by dysregulated clonal expansion . In many cases, such damaged cells exhibit destructive phenotypic characteristics that damage or kill the host. This is the definition of cancer .
- Senescent post-mitotic cells: Senescent cells are those which are damaged in ways that disable their proliferative ability, but do not cause carcinogenic transformation or cell death. The first stage of cellular senescence is damage-induced cell cycle arrest, also called replicative senescence (RS). Typically, this growth arrest is triggered by stress-induced DNA damage responses (DDR), and also may be induced as a result of effects of telomere erosion – a phenomenon that produces stress-related metabolites . These influences activate tumour suppressor genes, in particular the p53 and pRB pathways, in which damaged DNA and other by-products of cellular stress interact with constitutively expressed cytoplasmic precursors of transcription factors, to form complexes that negatively regulate cell division cycle (CDC) elements and halt proliferation . These cells may remain functional to some degree, depending on the level of damage responsible for abolishing proliferative potential. However, damaged cells in sustained replicative arrest typically undergo a process called ‘geroconversion’, in which hallmark phenotypic characteristics emerge that makes their presence detrimental for the host [3, 27]. Geroconversion is a damage-induced alteration in cellular homoeostasis that causes the emergence of cell forms that adversely impact tissue function. DDR activation resulting from oxidative damage to a cell hyperactivates growth-promoting cellular process, especially the nutrient-sensing complexes mTORC1 and mTORC2, which both include the serine/threonine protein kinase mTOR. These complexes respond to oxidative stress, systemic nutrient level elevation and other cues to maintain high levels of cell growth-related gene expression, especially in the presence of nutrients . As this input of gene products is not counterbalanced by cell division, senescent cells assume a characteristic large, flat or distorted morphology, and often accumulate high levels of pH 6-active beta galactosidase, a lysosomal hydrolase up-regulated by an overload of a cell's degradative machinery as a result of high levels of stress-associated detritus [3, 27, 28]. Senescent cells may also express proteases and inflammatory mediators, such as interleukin-6 (IL-6) that trigger unregulated and pathological inflammatory processes. Accumulation of senescent cells adversely affects their tissue microenvironment, contributing to metabolic syndrome/type 2 diabetes, cancer and many other age-associated disorders . Conversely, deletion of senescent cells delays the onset and reduced severity of age-related pathologies in mice . Further demonstration of rapamycin-mediated mTOR inhibition to enhance autophagic clearance of toxic aggregates and mitigate severity of age-associated disease was reported in October 2011 by researchers at the Institute of Clinical Medicine at the University of Eastern Finland. These investigators demonstrated that the restoration of autophagy by rapamycin-mediated suppression of mTOR activity enhances the viability of human retinal epithelial cells through clearance of lipofuscin deposits [29, 30]. These results further underscore the clinical potential of senescent cells as therapeutic targets. Indeed, as suggested by the authors of the aforementioned report, the results presented suggest the clinical value of a rapamycin-based ophthalmic solution (eye drop) for the treatment for age-associated macular degeneration .
- Cells executing a death program: Damage to a cell activates several processes that may result in its death. Four major types of cell death have been characterized at the time of this writing :
- Type I cell death/apoptosis: This is a tightly regulated process activated by specific biochemical signalling cues, in which a damaged cell is neatly dismantled in ways that avoid random release of cell contents, thus avoiding tissue toxicity [31, 32].
- Type II cell death/autophagy: Autophagy (or ‘self-eating’) is a process in which the internal components of the cell, in particular, organelles and proteins, are enzymatically degraded. Activation of the autophagic program may not result in cell death, particularly if damage to the cell has not been severe. In fact, under many conditions, it is an adaptive response to stress, particularly starvation, in which degraded cell components may be re-utilized to promote cell survival and restoration of healthy function .
- Type III cell death/necrosis: This pathway typically occurs in response to severe stressors imposed on tissue, such as mechanical, chemical or thermal trauma, dysregulated inflammation, microbial challenge and cancer. A major feature of necrosis is the uncontrolled release of cell contents, which may subsequently initiate pro-inflammatory pathways in ways that may become pathological .
- Type IV cell death/‘necroptosis’: This process exhibits cell morphological features very similar to necrosis, but also is characterized by biochemical processes that are shared with apoptosis, along with a unique temporal progression distinct from both apoptosis and necrosis [31, 33].
Progressive shift in epigenomic profiles: a major unifying feature of multicellular ageing
Integrating cellular and whole-organism features
Despite the complexity and seeming randomness of the ageing process, studies of its cellular and molecular features have begun to reveal a comprehensive picture of age-associated physical decline that takes into account findings drawn both from work with cells in culture and whole organisms. The health and viability of any multicellular organism at all developmental stages, from embryonic to senescent, is a function of the epigenomic profiles of its constituent cells [34, 35].
The overall physical status of an organism, broadly defined as health and viability, is completely defined by the relative representation of cells in the five major functional cell status categories described above in Section I-2. Each of these categories represents a characteristic pattern of gene expression constituting the epigenome of a particular cell, establishing its phenotype and activity within a particular tissue. Collectively, interacting cellular epigenomic profiles are manifested as tissue and whole-organism health status. Ageing may therefore be defined as the progressive shift in epigenomic profiles of cells that collectively make up an individual, and age-related physical deterioration occurs as a function of how far these shifts move an individual away from steady-state homoeostatic equilibrium (representing ideal health).
Individuals with an elevated fraction of cells in a healthy quiescent post-mitotic state (Category II) are likely to exhibit balanced homoeostasis and good overall health. Likewise, organisms undergoing significant growth, in which large numbers of dividing cells (Category I) are adding healthy tissue, may exhibit robust viability, particularly if the cell growth is part of an existing developmental program representing the transitional stages of early life. The viability of an organism with high representation of Category I cells may nevertheless be low if high rates of cell proliferation are occurring in response to trauma. In such cases, elevated mitotic activity may be an adaptive response, with the potential to restore a physiologically compromised individual to its optimally healthy state.
The presence of category III cells (cancer) is an obvious and typically lethal disadvantage, even when the percentage representation of such cell species is low. Signalling pathways that result in geroconversion into senescent cellular forms (Category IV) effectively block carcinogenic transformation. Nevertheless, age- or trauma-related tissue build-up of senescent cellular phenotypes creates progressively toxic tissue microenvironments that favour transition to cancer and other degenerative processes . The overall deleterious influence of senescent cells on whole-organism viability is significant in normal ageing. Cells that have undergone replicative arrest as result of both stress-induced damage and telomere erosion (which constitutes a special kind of stress, as will be discussed below), develop into senescent forms that accumulate over time and become major contributors to age-associated pathologies such as atherosclerosis, stroke and other chronic syndromes characterized by dysregulated inflammation [3, 13, 27].
Role of cell death programs in tissue cell phenotype representation
Of major interest in the context of the present review is the relative representation of Category V – that is, cells that have been activated to execute signalling pathways leading to death via apoptosis, autophagy, necrosis or necroptosis. Each of these four major modes of cell death is potentially useful in counteracting normal age-associated physical decline and maintaining the representation of cellular phenotypes that maximize an individual's retention of youthful characteristics. The rate and magnitude of cell death occurring in a particular tissue establishes the distribution of cellular phenotypes and effectively defines virtually all aspects of an individual's physical makeup at any given developmental stage and environment. It is therefore likely that strategies for intervention in normal ageing will become increasingly focused on methods for skewing representation of cells towards increased percentages of healthy, quiescent forms, and where needed, healthy replicative or regenerative types: cell phenotype categories I and II described in section I-2 above) respectively. At the time of this writing, therapeutic methods that skew cells in a particular tissue towards desired phenotypic representation patterns have not been used specifically to target ageing in human beings. Phenotypic skewing by selective induction of death, versus survival programs within a tissue may nevertheless become an increasingly important tool in life extension strategies with increased understanding of molecular mechanisms leading to cell death. An exploration of how oxidative stress contributes to activation of the apoptotic death program is provided in Part II of the present review.
Developmental shifting of cellular phenotypic representation in tissue
Eukaryotic multicellular organisms begin as an undifferentiated pluripotent stem cell, normally in the form of a single cell embryo, which may proliferate with little differentiation to form a peri-implantation embryo, or blastocyst. The component cells then undergo further proliferation and differentiation to form an organism's three classes of progenitor tissue: endoderm, mesoderm and ectoderm. During early life, each tissue contains proliferation-competent progenitor cells (functional category Type I) in percentages that are elevated relative to later developmental stages. These ‘oligopotent’ stem cells are capable of terminal differentiation into specific types of tissue and serve as a pool of reserve tissue to replace quiescent post-mitotic cells (functional category II) as they become depleted through stress-induced attrition. Cell damage which fails to move a cell into functional category V through activation of a death program may cause genetic changes, causing transition to functional category III (cancer). Vertebrate eukaryotes have evolved a major anticancer adaptive program followed by damaged cells, which causes their conversion into senescent phenotypes (category IV), which are incapable of proliferation, but remain alive. Progressively decreasing percentages of quiescent and regenerative cells in tissues during adulthood through old age are paralleled by increased representation of senescent cells . At the time of this writing, although no direct mechanistic correlations between cellular senescence and whole-organism ageing have been established, the known effects of senescent cells on resident tissues provide the basis for a compelling hypothesis. Specifically, a possible explanation for physical decline that accompanies the chronological age of an organism is that progressively elevated numbers of senescent cells in each tissue adversely affect tissue integrity and function in ways that are manifested as senescent whole-organism phenotypes. Approaches to testing of this hypothesis in ways that may lead to strategies for counteracting normal ageing are discussed in section III-3 of the present review. A conceptual picture of an age-related shift in representation of functional cell categories is shown below, in Figure 4.
Tissue regeneration and cell death: tools for modulation of age-related physical decline
Stem cell-mediated tissue regeneration strategies
Approaches that may be used in skewing modes of cell death in ways that favour youthful representation of functional cell categories in tissue will be discussed in section III-3 of this review. Below in the present section, some of the approaches for augmenting a representation of Category I, healthy replication-competent (stem) cells will be considered. One treatment that relies on such a process, bone marrow transplantation (BMT), has a well-established track record in clinical medicine. This method, which regenerates haematopoietic tissue by supplementation with allogeneic stem cells, is discussed below.
BMT – an early, prototypic stem-cell therapy
The most widely used form of cellular regenerative therapy that relies on repopulation of a tissue niche with Category I (mitotic) cells is BMT. BMT, for reasons discussed below, has features that allow it to serve as a very early prototypical model for stem cell-based rejuvenation therapy. Nevertheless, BMT has never been used specifically as an anti-ageing measure, and in its present form, it is not likely to have direct application in any age-intervention strategies. BMT is a drastic, invasive and painful treatment, typically used for patients afflicted with neoplastic lymphoproliferative disorders (including leukaemias) who have proven refractory to chemotherapy, radiation and other commonly used anticancer regimens. Nevertheless, observations of the behaviour of transplanted haematopoietic stem cells (HSC) help establish guidelines for transplantation of a wide range of other progenitor cell types. For example, recent studies demonstrating the ability of HSC from young mice to suppress cell senescence-related biomarkers and rejuvenate renal tissue in aged mice , suggest that BMT may provide a therapeutic paradigm of potential use in the design of clinical protocols for counteracting physical ageing.
Preparation of a patient for BMT may involve complete ablation of bone marrow, using whole-body radiation or chemotherapeutic agents such as cyclophosphamide and busulfan. Although on a case-specific basis, a ‘non-myeloablative’ protocol that leaves recipient bone marrow in place may be used. The reduced level of trauma as a result of treatment with this latter procedure is offset somewhat by an increased risk of cancer relapse. A patient is then administered either autologous or allogeneic cell preparations, which have high percentages of HSC. These may be obtained from donor bone marrow or from the peripheral blood of either a donor or the patient by a process called apheresis, which extracts HSC from blood. The therapeutic effect depends upon the ability of transplanted HSC to replace a patient's original haematopoietic tissue and provide the individual with a self-renewing source of progenitors for each of the haematopoietic lineages.
Refinements in the clinical use of BMT have resulted in steady improvements in its efficacy during the late 20th and early 21st centuries. Nevertheless, the adverse effects associated with this procedure underscore the dangers associated with the use of stem-cell transplantation in broad clinical venues. The major complications of BMT include several life-threatening conditions that develop as a result of the method of applying this treatment, and also as a consequence of stem-cell behaviour. The general immunosuppression that accompanies the destruction of recipient bone marrow, along with the effects of immunosuppressive drugs, leaves a patient vulnerable to opportunistic infections and sepsis. BMT recipients are also highly susceptible to new malignancies that may arise from carcinogenic influences on host tissue of marrow-ablative treatments. BMT carries an additional risk of tumour evolution from transplanted HSC, as placement in their new environment may stimulate dysregulated proliferative responses, including carcinogenesis.
Another major obstacle to successful BMT currently faced by healthcare providers is the requirement for a close human leucocyte antigen (HLA) tissue match between donor and recipient to avoid graft-versus-host disease (GVHD), an often fatal syndrome resulting from immune attack on host tissue by transplanted donor leucocytes . Typically, a prospective BMT recipient's siblings are screened for suitability of donors and failing this, the search is widened to unrelated persons. Nevertheless, at the time of this writing, a large percentage of patients requiring allogeneic HSC are unable to identify matching donors .
Mesenchymal stromal cell influence on differentiation and immunogenicity in bone marrow cultures
A rapidly evolving and very exciting strategy for increasing safety and efficacy of BMT, with profound implications for sustained regenerative capacity of all tissues, is the use of mesenchymal stromal cells (MSC), which were first identified as a subpopulation within the bone marrow that may act as a ‘feeder’ cell layer to sustain HSC function in vitro . A major unique property of MSC is ‘multi-potency’. That is, their interaction with tissue culture surfaces stimulates differentiation into a wide variety of functional cellular phenotypes, exhibiting ectodermal, endodermal and mesodermal epigenomic expression profiles [39, 40]. The potential for therapeutic use of such phenotypic ‘plasticity’ by MSCs is augmented by their negligible immunogenicity . This property enables MSCs to evade alloreactive cell recognition, thereby allowing their clinical use without the requirement for donor–recipient HLA tissue match . MSC are also observed to down-regulate T lymphocyte immune activity , a property which offers enormous potential for their application in the prevention and management of tissue damage resulting from hyperactive inflammatory processes. Indeed, host MSC are observed to reduce the risk and severity of GVHD in BMT patients through the inhibition of T cell-dependent pro-inflammatory processes , the recruitment of donor HSC to host bone marrow, enhancing their engraftment and promoting haematopoiesis .
The foregoing observations reveal host-derived MSC to be a critical component for bone marrow regeneration and maintenance. However, the evolution of MSC-based preventive medicine and therapies will require sustainable sources of allogeneic MSCs. Fortunately, MSC lines derived from umbilical cord blood may be functionally expanded in vitro to promote engraftment of progenitor stem cells capable of regenerating a wide variety of tissues .
Placenta-expanded MSC product-1 (PLX-1) aids in engraftment of HSC derived from umbilical cord blood (UCB). The first mainstream medical application of UCB MSC was made by Pleuristem Therapeutics of Haifa, Israel, which in 2008 obtained U.S. F.D.A. clearance for Phase I clinical trials of a product called PLX-1 cells, which are placental-derived MSCs expanded ex vivo and stored frozen before being administered to patients by i.v. injection .
Relevance of PLX-I and related MSC to age intervention
A major objective of Pleuristem and other parties developing allogeneic cell lines capable of enhancing representation of regenerative cells in a particular tissue is to broaden the range of tissues into which products such as PLX-I may be engrafted. Progressive refinement of this technology may ultimately allow a sustained regimen of treatments to continually regenerate a comprehensive approach that facilitate achievement of this objective. A major outcome may include maintenance of a healthy and balanced tissue homoeostasis for very long time periods – which could be manifested as a prolonged lifespan. Such an outcome would also require simultaneous protection against cancer and deletion of senescent cells. Consideration of factors contributing to these objectives is discussed below.
Activity of mTOR in bone marrow
BMT offers a particularly good model for initial efforts to restore the regenerative capacity of a representative tissue and prolong the time period that it is stable and functional. Elevated representation of senescent cells in the bone marrow of a recipient can adversely affect the capacity of stem cells introduced via BMT to restore function and viability of haematopoietic . As previously described, a major contributor to the development of the senescent cellular phenotype is hyperactivation of nutrient sensor and growth pathways, in particular mTOR and its derivative complexes mTORC1 and mTORC2[3, 27, 28]. These findings suggest that age-intervention strategies are likely to include pharmacological inhibition of mTOR activity, an approach that will be discussed in Part 3 of the present review.
In healthy bone marrow, or marrow successfully restored to functional capacity by BMT, a pool of self-renewing ‘progenitor’ HSC (Category I in section I-2 above) are distributed throughout the tissue and maintained in a metabolically active quiescent state which, in response to physiological cues, is capable of progression into the cell cycle and production of blood components. During healthy HSC proliferation, cell mass increases due to up-regulation of mTORC1 signalling keeps pace with cell division and cells of normal size and function are produced. However, in a tissue environment which favours potentially pathological levels of cell growth signalling, mTORC activity becomes hyperactivated as a result of the critically decreased availability of the major negative regulators of this enzyme complex, resulting in the exhaustion and depletion of HSCs. Dysregulation of mTOR is also observed to modulate the activity of the enzyme in ways that promote enhanced mitochondrial biogenesis, with resulting increases in oxidative stress  and decreased differentiation of HSC into cell forms that contribute to normal homoeostatic function. By contrast, the capacity of HSC to regenerate functional blood cell lineages is dramatically increased by inhibition of mTOR. As mTOR is a negative regulator of autophagic processes, inhibition of this enzyme increases autophagy, resulting in clearance of toxic debris and increasing autophagy and suppression of the emergence of senescent cellular phenotypes . These aforementioned outcome taken in context with demonstration of the rejuvenating effects of BMT from young mouse donors on aged mouse kidney , underscore the value of BMT as an example of how stem cells might be applied as a countermeasure to age-associated physical decline.
Contribution of cell growth-promoting mechanisms to cellular senescence
Hyperactivity of growth-promoting processes such as mTOR activity is a major contributor to the toxicity of tissue environments that contain a high representation of senescent cell types [3, 44]. This, and other features of senescent cells, underscores the requirement to counteract the influence of these cells – and if possible eliminate them as an integral element of age-intervention strategies. Any approach to rejuvenation-oriented skewing of functional cell types in a tissue must consider how dysregulated mTOR hyperactivity and other adverse influences of senescent cellular phenotypes, such as their pro-inflammatory and carcinogenic effects , may affect healthy cells.
Influence of senescent tissue environment on stem-cell function
The effect of the tissue environment on stem-cell function is explored in an elegant study by Irina Conboy et al. in an investigation of progenitor stem cells for muscle, called myogenic satellite cells (2005). This investigation revealed that myosatellite regenerative function was significantly impaired in tissue from young healthy mice that had been grafted into the senescent environment of aged animals. This investigation further revealed that the myosatellite regenerative and functional capacity was restored when tissue taken from elderly donors was placed in younger recipients .
A fascinating corollary experiment by these investigators demonstrated that the rejuvenating effect of younger tissue on stem cells from elderly individuals could be mediated by humoural factors. These procedures made use of a technique called ‘parabiotic pairing’, in which the circulatory systems of two syngenic mice were linked, exposing the aged animal to the environment of the younger one and vice versa.
Restoration of youthful competence to myosatellite cells in aged mice was determined to have occurred due to reactivation of a signalling pathway called ‘Notch’, which declines in strength with age, but is restored in the elderly member of a parabiotic pair through mechanisms yet to be characterized. These findings, and analysis of the effect of senescent tissue environment on HSC, clearly suggest that the tissues of elderly individuals cannot be restored to youthful vigour by simply introducing a new supply of progenitor stem cells. It is likely that the functionality of these cells would be rapidly degraded by the influence of resident senescent cellular phenotypes. A possible countermeasure to senescent cell influence might be the incorporation into genomes of progenitor stem cells, to be administered therapeutically, of expression cassettes engineered to produce activators for critical pathways such as notch. The expression of such a signalling molecule might be made drug-inducible, as was accomplished by the team at Dana Farber, who constructed a 4-OHT-inducible cassette to increase telomerase production in mice .
At present, however, such a strategy is highly speculative. Moreover, safety issues make such an approach extremely challenging. For example, inflammatory cytokines (including IL-6) which may be up-regulated in senescent cells by an anti-apoptotic protein HMGB1 are known to stimulate Notch signalling in ways that augment breast cancers . These observations underscore the need to consider how a senescent environment may affect stem cells introduced as an element of rejuvenation therapy – and how some future novel class of stem cell programmed to express signalling pathway mediators may endanger resident tissue.
Altering tissue cellular phenotypic ratios by varying oxidative stress dosage: induction of apoptotic responses versus senescence
Future strategies that are able to impart sustained youthful vigour to whole organisms will almost certainly include methods for selectively eliminating cell types that detract from the functional efficiency of component tissues. In a broad sense, external stress applied to cells, both in vivo and in vitro, achieves this purpose. Varying the intensity of a wide range of stressors to which a tissue or organism is exposed has been observed to reproducibly shift the percentage representation of senescent cells, with those executing each of the four major pathways to death (apoptosis, necrosis, necroptosis and autophagy). Relatively low stress levels favour predominance of senescent forms, followed by the induction of apoptotic programs at higher levels, transitioning to necrosis when stress levels become extreme .
Induction of autophagy and necroptosis
Cells which do not respond to oxidative stressors by senescence or apoptosis may die via ‘necroptosis’, or undergo autophagy, which can result in either cell death or survival under improved conditions of cellular homoeostasis . Cells are influenced to respond to various stress conditions by a particular death or survival pathway depending on the presence of certain pharmacological agents, or as part of unique pathologies that may favour skewing of cellular activity patterns into one of the aforementioned stress-response modes . Cells may sustain a very wide variety of stressors, all forms of which cause activation of DDR mechanisms, resulting in cell cycle arrest and either senescence, repair and survival, activation of a death program or carcinogenic transformation [3, 27]. Typically, DDR activation occurs as a result of a reaction between DNA and reactive oxygen molecules that cause strand breaks and production of DNA degradation products which trigger cell cycle arrest through induction of the ‘Gatekeeper’ tumour suppressive proteins p53 and pRB1/p16. This either directly inhibits the transcription of genes needed for mitosis, or promotes the expression of other protein inhibitors such as p21, the expression of which is driven by an interaction of p53 with the p21 element's promoter . DDR activation hyperstimulates cell growth mechanisms such as mTORC, which in turn promote the development of senescent phenotypes and toxic tissue microenvironments .
Cell stress: an epigenomic rheostat
Cellular consequences of DDR activation
The major common feature in virtually all influences that cause stress-induced epigenomic changes is an interaction between a cell's DNA and reactive oxygen molecules produced by the stressor – with resulting activation of DDR processes and conversion of the cell from one of the phenotypic categories described in Section I-2 above into another [3, 27, 31]. Some stressors activate DDR without DNA damage . Therefore, within certain limits, manipulation of external stress levels (including mild thermal stress, cold and other influences that avoid significant cellular damage) potentially provides a general approach for shifting epigenomic profiles in ways that promote tissue and whole-organism vitality. By the foregoing analysis, if stress levels delivered to a particular tissue can be varied with some degree of precision, then differential application of such stressors constitutes a biological rheostat. Ideally, such an approach might be used to reproducibly vary the percentage representation of certain cell phenotypes, much as an electrical rheostat is used to adjust the level of analogue signals on (for instance) a light dimmer, or audio volume control.
Oxidative stress-mediated influence on components of the nuclear cytoskeleton: relationship with progeria and normal ageing
An example of how altering tissue oxidative stress levels may be used to manipulate ageing-related cellular signalling is provided by the outcome of the study that demonstrated how up-regulation of oxidative stress by the drug 6-hydroxydopamine alters phosphorylation patterns of the nuclear cytoskeletal proteins lamin A/C. Altered phosphorylation changes the interaction between this lamin and heat shock protein-90 (hsp-90) in human neuroblastoma cells in ways that correlate with the role of lamin A/C in age-related cellular and tissue changes . These results are particularly significant in the context of the role of progerin, a mutant variant of lamin A, which adversely alters nuclear scaffolding, causing disruption of cellular homoeostasis, cell cycle arrest and rapid geroconversion into senescent cellular phenotypes .
High expression of progerin has been identified as the cause of Hutchinson-Gilford progeria syndrome (HGPS) . HGPS-afflicted human beings exhibit rapidly advanced rates of physical ageing, dying at approximately 13 years of age from the same spectrum of disorders, such as cardiovascular syndromes, which are typically the cause of death in elderly human beings . Sporadic expression of a cryptic splice variant for lamin A also results in progressive accumulation of the defective protein in cells of healthy individuals, thus allowing the use of progerin as a reliable biomarker of ageing [51, 52]. Therefore, the aforementioned studies by Nakamura et al., demonstrating the capacity of oxidative stress to act through nuclear lamin to affect age-related cellular biomarkers , increase the scope of potential contribution of oxidative cellular processes to progerin-related cellular and whole-organism senescence. Cells of the cardiovascular system are, for reasons not completely understood at the time of this writing, particularly prone to disruption of cellular homoeostasis due to build-up of progerin and other toxic debris . This phenomenon may at least partly explain the observation that cardiovascular disease risk factors, including hypertension, hyperlipidaemia, diabetes, insulin resistance, atherosclerosis and heart failure, increase dramatically with age as contributors to mortality among the elderly . The effect of these influences may be mitigated somewhat by short ischaemic periods interspersed with brief periods of reflow just prior to reperfusion of cardiac tissue, which is a process called ‘ischaemic post-conditioning’ . Characterization of mechanisms by which cellular debris disrupts cardiovascular cell function in ways that exacerbate risk factors and lead to disease is an area of ongoing intensive research.
mTOR, autophagy and oxidative stress-related effects on geroconversion: progerin-mediated
Research ongoing at the time of this writing has shown that the progerin-mediated geroconversion of both HGPS and normal human cells may be prevented and under some conditions reversed by induction of autophagy through suppression of mTOR with rapamycin, an mTOR-inhibitory drug . In normal cells, mTOR activity suppresses autophagy thus allowing progerin and other cellular debris to create oxidative damage resulting in cell cycle arrest and geroconversion to senescent cell phenotypes . For this reason, inhibition of mTOR activity allows autophagic processes to occur at elevated rates, causing ‘cellular housecleaning’ that clears progerin and other mis-functional metabolic products to an extent that enhances healthy cell function . mTOR activity is strongly stimulated by oxidative stress and inhibited by antioxidants [56, 57]. The ability of rapamycin to inhibit mTOR and thus facilitate autophagic clearance of senescence-inducing protein aggregates is potentially a very powerful tool in age intervention. Nevertheless, the high toxicity of the drug makes it impractical for its administration at dosages needed for sustained suppression of senescent cell accumulation in a particular animal . Despite that, a strategy that makes use of a principle called ‘phytochemical synergy’ may allow for substantial reduction in its effective dosage – and thus toxicity. These studies demonstrated that non-toxic phytochemicals extracted from Ginkgo biloba were capable of modulating calcium metabolism in ways that allowed the immunosuppressant FK506 to mediate powerful antiarrhythmic effects in a rodent heart model at significantly lower dosage than FK506 alone . Similar strategies pursued in ongoing studies at University of Debrecen may allow for reduction in the effective dosage and thus toxicity of rapamycin.
Cardiovascular ageing: why age-related physical decline most often involves cardiovascular tissue
The effect of progerin and related protein aggregates in various cell types yield insight into the reason cardiovascular pathologies are a widely prevalent common feature of age-related physical deterioration. Mechanistic insight is provided by an observation that human vascular cells express progerin. These cells are particularly susceptible to disruption of cellular homoeostasis and deterioration in healthy function . Thus, in normal ageing, it is likely that increased representation of cells bearing toxic protein aggregates by all tissues will impact the cardiovascular system earliest and most severely.
Recent studies by Rakesh Kukreja and colleagues at Virginia Commonwealth University revealed that rapamycin constitutes a particularly powerful countermeasure to cardiovascular ageing. Using a murine model of myocardial ischaemia-reperfusion (I/R) injury with isolated hearts in a Langendorff perfusion system, along with parallel in vitro experiments, these investigators demonstrated the ability of rapamycin to significantly protect whole hearts against I/R injury, and further showed that the drug suppressed simulated ischaemia-reoxygenation (SI/RO) damage to primary cardiac myocytes. These studies revealed that the underlying mechanism included activation of JAK2-STAT3 transcription factors, with a subsequent cascade of cardioprotective signalling events, including phosphorylation of ERK, STAT3, eNOS and glycogen synthase kinase-3ß in concert with increased pro-survival Bcl-2 to Bax ratio . These outcome provide a basis for improved strategies for prevention and treatment of cardiovascular disease in ways that directly counteract degenerative changes occurring during the normal process of ageing.
The foregoing reports describe how oxidative stress may promote accumulation of senescent cells through increased mTOR activity and by adverse effects on cytoskeletal components such as lamins. A picture emerges from the aforementioned, suggesting a strategy for inhibition of age-related physical decline by promoting levels of autophagy sufficient to enable ongoing disposal of toxic cellular products such as progerin, but not high enough to cause significant autophagic cell death. A major mediator in achieving such a balance is mTOR, the variation in oxidative stress in particular, as a primary tool for manipulating mTOR activity.
It is important at this point to qualify the above analysis by emphasizing that such an approach could almost certainly never be finely tuned enough to use it as a primary mechanism of rejuvenating tissue – or a whole organism. The use of dietary antioxidants represents a popular, but poorly understood application of redox status adjustment that is occasionally touted as ‘anti-ageing’ therapy. This notwithstanding, selective use of antioxidants and possibly other pharmacological agents will almost certainly have value in optimizing conditions for the activity of stem cells that might be used for sustained tissue regeneration approaches. As the purpose of this review was to establish a solid molecular biological and cellular basis for developing a general life extension strategy, Part II of the paper is a comprehensive review of cellular redox sources of reactive oxygen and how these molecular species interact with cell components to either promote survival or activate a death pathway, with a focus on apoptosis.