TrendsMammalian cells asymmetrically segre-gate cellular components (cargoes)during cell division.

The asymmetric segregation of car-goes can become dysfunctional withage, leading to more symmetric segre-gation of cargoes.

TICB 1289 No. of Pages 11

ReviewCreating Age Asymmetry:Consequences of InheritingDamaged Goods inMammalian CellsDarcie L. Moore1,* and Sebastian Jessberger2,*

Accumulating evidence suggests that mammalian cells asymmetrically segre-gate cellular components ranging from genomic DNA to organelles and dam-aged proteins during cell division. Asymmetric inheritance uponmammalian celldivision may be specifically important to ensure cellular fitness and propagatecellular potency to individual progeny, for example in the context of somaticstem cell division. We review here recent advances in the field and discusspotential effects and underlying mechanisms that mediate asymmetric segre-gation of cellular components during mammalian cell division.

Segregated cargoes range from geno-mic DNA to organelles and damagedproteins.

Inheritance of ‘good’ cargoes canenhance cell health and responsivity,whereas inheritance of ‘bad’ cargoescan decrease proliferation rate andincrease cell death.

1Department of Neuroscience, Schoolof Medicine and Public Health,University of Wisconsin-Madison,Madison, WI 53705, USA2Brain Research Institute, Faculty ofMedicine and Science, University ofZurich, 8057 Zurich, Switzerland

*Correspondence:darcie.moore@wisc.edu (D.L. Moore)and jessberger@hifo.uzh.ch(S. Jessberger).

Segregation of Cellular Components During Asymmetric Cell DivisionAsymmetric cell division has been extensively studied over the past decades using a variety ofcell lines, reductionist model organisms, and mammalian in vivo approaches, with the aim tocharacterize how cell divisions can asymmetrically affect the behavior and properties of the twonewly generated daughter cells [1–6]. Given that asymmetric cell division often results in distinctfates of its progeny, the focus of previous research has largely been on the unequal distributionand segregation of fate determinants [7,8] or segregation of polarity elements in non-mammaliansystems [9]. However, more recently the segregation of cellular components that are not directlyinvolved in fate choices and the mechanisms behind unequal distribution of specific cellularcomponents (hereafter called cargoes, see Glossary) uponmammalian cell division have gainedincreasing attention. This is especially true in the context of stem cell divisions where, at leastunder some conditions, one of the daughter cells represents a self-renewed version of theoriginal mother cell. The question that has arisen over the past years is very simple, althoughobtaining an answer has turned out to be somewhat challenging and current knowledge remainsscant: who inherits which parts of the cellular constituents of the mother cell when she divides?

The mechanisms of inheritance are likely not conserved across modes of cell divisions or acrosscell type. However, it is conceptually important to note that even two daughter cells with thesame fate and cellular potency are not composed of an identical share of maternal proteins,lipids, or organelles. As such, each cell division may be considered asymmetric. Given theseconsiderations, we focus on cell divisions that can, but do not necessarily, lead to daughter cellswith distinct fates.

How cellular components segregate during cell division may be of exceptional importance in thecontext of embryonic and somatic stem cell biology. There is ample evidence that specific cells inour bodies have the potential to generate progeny throughout life [10–15]. This regeneration canlead to rapid turnover of somatic stem cell-derived tissues and organs including those of the

Trends in Cell Biology, Month Year, Vol. xx, No. yy http://dx.doi.org/10.1016/j.tcb.2016.09.007 1© 2016 Elsevier Ltd. All rights reserved.

TICB 1289 No. of Pages 11

GlossaryAggregate: oligomeric cluster ofmisfolded or damaged proteins.Cargoes: those molecules, proteins,organelles, etc. that are segregatedbetween two daughter cells duringmitosis.COS7 cells: immortalized Africangreen monkey kidney fibroblast-likecell line.Diffusion barrier: a structure limitingthe movement of specific proteins ormolecules.Endoplasmic reticulum (ER): anorganelle made of interconnectedtubules and cisternae that producethe proteins and lipids of a cell.HEK 293 cells: human embryonickidney cell line.Huntingtin protein (Htt): necessaryfor some normal cellular functioning,but with glutamine expansionmutations leads to Huntingtondisease.Inclusion body: an intracellular sitewhere aggregates are sequestered;typically one per cell.Insoluble protein deposit (IPOD)compartment: non-membrane-bound compartment where non-ubiquitinated, amyloidogenic proteinsare sequestered and not degraded.Juxtanuclear quality control(JUNQ) compartment: non-membrane-bound compartmentwhere ubiquitinated, misfoldedproteins are targeted for eventualrefolding or degradation bycolocalized proteasomes andchaperones.Microtubule organizing center(MTOC): consists of a pair ofcentrioles and pericentriolar material;anchors microtubules at their minus-ends to give direction to theirnucleation; anchors and organizesthe microtubules that will form thecilium.Neural stem cells (NSCs): self-renewing, multipotent cells of theadult brain; undergo both symmetricand asymmetric divisions; candifferentiate into neurons, astrocytes,and oligodendrocytes.Photoactivation: using a pulse of anactivating wavelength to create achemical reaction that can changethe fluorescence emission of aprotein.Progerin: truncated form of thenuclear protein lamin A, typicallygenerated from a point mutation inthe gene encoding lamin A; mutationcauses Hutchinson–Gilford progeria

intestines or the hematopoietic system [16,17]. In addition, stem cells ensure repair mechanismsafter excessive use, such as those that occur through satellite cells in skeletal muscle or in theskin [18,19]. To assure life-long viability and retention of cell division properties, specificcomponents of the cells can be either preferentially retained in the self-renewed stem celldaughter or asymmetrically segregated to the non-stem cell daughter cell. Asymmetricallysegregated cargoes may range from DNA to organelles to damaged proteins. Partially basedon the technical difficulty to birthdate these types of cargo and to image their segregation overtime, our understanding of how cargoes are distributed during cell division remains poor.However, we summarize here and review recent experiments aiming to characterize thesegregation of specific cargoes during asymmetric cell division, with a focus on mammaliancells, and suggest future approaches that will further our understanding of the mechanismsgoverning asymmetric cell division.

The Immortal Strand HypothesisOriginally formulated by John Cairns, the idea that DNA strands asymmetrically segregate todaughter cells with the aim to protect long-lived cells from DNA damage and replication errorshas been tested experimentally over the past four decades [20]. This hypothesis, known as theimmortal strand hypothesis, proposes that, upon DNA replication, sister chromosomes non-randomly segregate to daughter cells, and that this may represent a mechanism of how stemcells protect their genomic information from errors of DNA duplication by inheriting the olderparental DNA strand (i.e., immortal), whereas the second daughter cell receives the newer (i.e.,mortal) sister chromatid [20–22]. Over the past decades the immortal strand hypothesis hasremained highly controversial, with data convincingly supporting and others clearly refuting theevidence of non-random inheritance of DNA strands during stem cell division [23–32].

There are numerous reasons why the proof or final rejection of the immortal strand hypothesishas turned out to be difficult. First of all, there is clear evidence that non-random segregation ofsister chromatids is highly dependent on the (stem) cell type analyzed and the specific chro-mosomes studied. Whereas the evidence for non-random segregation of sister chromatids inmuscle stem cells, called satellite cells, is strong, findings in other types of somatic stem cells donot always support the immortal strand hypothesis [30]. In addition, the experimental approachto characterize segregation of DNA strands has been largely based on stable labeling of DNAbefore cell division using, for example, thymidine analogs such as 5-bromo-2’-deoxyuridine(BrdU) [23,31]. Apart from the requirement to fix cells and tissues to analyze segregation ofBrdU-labeled chromatids, the saturation and extensive labeling of thymidine analogs itself maycause changes in genomic structure and DNA replication.

In this context, the modified hypothesis of non-random sister chromatid segregation that wasrecently proposed is intriguing because this suggests that non-random segregation of DNAstrands is associated with DNA replication stress [33]. Slowed or stalled replication forks causeDNA damage during replication in the newer strand, and segregation of that newly acquired DNAdamage to the non-stem cell daughter may ensure genomic stability of the long-lived stem celldaughter [33]. Thus, the modified immortal strand hypothesis proposes a context-dependentmechanism for when and why sister chromatids segregate non-randomly. Such a model mayexplain conflicting results and provides a testable hypothesis using existing but also novel toolsto study non-random segregation of DNA strands. Indeed, a recent technique, CO-FISH(chromosome orientation fluorescence in situ hybridization), does not depend on nucleosideanalog labeling, and was recently used inDrosophila germline stem cells. CO-FISH revealed thatsister chromatids of the X and Y sex-chromosomes appear to be non-randomly segregated, incontrast to autosomes [32]. This study, together with others, also revealed that non-randomDNA segregation at least partially occurs through SUN/KASH protein-dependent mechanisms[32,34], providing mechanistic details for the biased segregation. However, at the same time

2 Trends in Cell Biology, Month Year, Vol. xx, No. yy

TICB 1289 No. of Pages 11

syndrome, a disease of prematureaging.Spinocerebellar ataxia type 3:disease caused by CAG expansionsin the gene encoding ataxin 3; leadsto an abnormal build-up of proteinaggregates that cannot be degraded,and eventual cell deathStress foci: non-membrane-bound,small cytoplasmic protein aggregateSUN/KASH: proteins containingSUN (inner nuclear membrane; e.g.,SUN1, SUN2) and KASH (outernuclear membrane; nesprins)domains that physically interact in theperinuclear space.Vimentin: type III intermediatefilament that is part of thecytoskeleton; can bind to thenucleus, ER, and mitochondria, aswell as to many other potentialproteins and organelles.

novel techniques based on high-throughput genomic sequencing or stable isotope labeling havenot found any evidence that human stem cells in diverse parts of the body are protected fromnormally occurring rates of genomic mutations [28,30].

In summary, the hypothesis that some cells, and more specifically stem cells, show asymmetricsegregation of chromatids is neither fully proven nor finally refuted. Instead, it appears that non-random inheritance of DNA strands shows a high degree of cell type specificity and might becontext-dependent (i.e., high replication stress may bias towards non-random segregation ofsister chromatids). Thus, the immortal strand may not exist for all mammalian cell types or underall cellular conditions, but the findings that asymmetry in DNA strand segregation can occurseem plausible. Further, non-random sister chromatid segregation may not only represent amechanism to protect the DNA of the long-lived daughter cell but may also be important totransduce diversity in non-stem daughter cells that appears to occur in neural stem cells(NSCs) during replication [35].

Asymmetric Segregation of OrganellesThere is accumulating evidence that organelles are also non-randomly distributed betweendaughter cells. While numerous organelles have been widely studied for their asymmetricsegregation in non-mammalian systems [36,37], we focus here on select organelles thatspecifically have been demonstrated to be asymmetrically distributed during mammalian celldivision.

MitochondriaMitochondria are the powerhouses of cells, producing ATP that allows the cell to perform itsmany functions. Changes in mitochondrial health can thus affect the ability of a cell to functionproperly, prompting the need to manage this heterogeneity. In yeast, there is a selectiveinheritance of mitochondria between the mother and bud that favors the bud [38,39]. Recently,asymmetric segregation of mitochondria has been interrogated in mammalian stem cells and celllines with different outcomes [40–42].

In human mammary stem-like cells, which undergo asymmetric division, photoactivation ofmitochondrial proteins to label the age of mitochondria and subsequent imaging through mitosisrevealed asymmetric segregation of older mitochondria (>10 h following photoactivation) to themore flat, adherent epithelial-type daughter cell, leaving the more round, stem-like daughter withless ‘old’ mitochondria, and more stem-cell traits, including the ability to form mammospheres[40]. This asymmetry was not seen when cell division occurred earlier than 10 h followingphotoactivation, nor when only bulk mitochondria were observed, suggesting that older proteinsspecifically are targeted for biased inheritance. Older mitochondrial proteins localized perinu-clearly through a Drp1-dependent mechanism, and disruption of either the localization orfunction of Drp1 resulted in a decrease in stem-like functions. Interestingly, decoupling allthe mitochondria in the cell did not change the asymmetric segregation of the proteins,suggesting that a decreased mitochondrial membrane potential was not the signal for asym-metric segregation [40]. Thus, the localization of the older proteins may be the driving forcebetween the inheritance bias in mammary stem-like cells, although the mechanism is unclear.The functional cost of inheriting these ‘older’mitochondria appears to be a loss of ‘stemness’ inthe targeted daughter cell.

In non-stem mammalian cells, similar types of experiments have found no evidence for asym-metric segregation of mitochondria in mouse embryonic fibroblasts [40], or in the bulk mito-chondrial movement in human embryonic kidney (HEK 293) cells [42], suggesting that thisasymmetric segregation may be more specific to cells with ‘stemness’ properties which mayhave different requirements for their behavior. Future experiments to determine the mechanisms

Trends in Cell Biology, Month Year, Vol. xx, No. yy 3

TICB 1289 No. of Pages 11

behind this asymmetric sorting of older mitochondria and its generalization to other stem cellsshould further reveal the importance of this segregation to stem cell function.

Centriole Inheritance and the Primary CiliumDuring the G1 stage of the cell cycle, cells possess a centrosome containing a mother and adaughter centriole. In S phase, the original centrioles semi-conservatively duplicate and grownew daughters. When these newly created pairs separate between daughter cells duringmitosis, one daughter cell inherits the older mother centriole [2], resulting in asymmetricinheritance. Is there a cost or benefit to the inheritance of the older centrosome? In theembryonic neocortex, radial glial stem cells, the neurogenic precursors in the developing cortex,retain the older mother centriole, while the daughter cells inherit the younger mother centrioleand differentiate, suggesting a correlation between ‘stemness’ and inheritance of the oldercentriole [43]. Depletion of ninein, a protein important for the maturation of the centrosome andenriched at the old mother centrosome, disrupts asymmetric inheritance of centrosomes. This inturn results in the reduction of radial glial progenitors in the mouse embryonic cortex, suggestingthat asymmetric inheritance of the old mother centriole is necessary for the proliferation of radialglia and thus proper development of the cortex [43].

Similarly, biased inheritance of centrosome-associated structures also confers stem cell traits.The older mother centriole along with its associated appendage proteins (together called thebasal body), nucleates the primary cilium, an organelle that projects from the cell and respondsto external cues, coordinating signaling pathways during development and in tissue homeosta-sis. Before mitosis, the cilium partially disassembles, leaving a vesicle containing a remnant ofciliary membrane with a distinct lipid composition, and an enrichment of ciliary signaling proteins.This ciliary membrane remnant remains attached to the basal body during mitosis and ispreferentially asymmetrically inherited by the daughter cell that maintains ‘stemness’ [44].The daughter cell that has inherited the older mother centriole and the ciliary membrane remnantis the first to generate the primary cilium, giving a selective advantage to the daughter cellbecause it is the first to respond to extracellular signaling cues that may regulate proliferation[44–46]. Whether this advantageous inheritance confers ‘stemness’ remains unclear.

Specific centrosomal proteins also have been shown to asymmetrically associate with the oldcentriole, such as cenexin/outer dense fiber protein 2 (ODF2), and centrosomal protein 164 kD(Cep164), whereas centrobin is enriched at the daughter centriole [47]. Whether these proteinscontribute to cell function following inheritance is not well understood. Taken together, thesestudies suggest that asymmetric inheritance of the older mother centriole and the primary ciliumremnant favors increased ‘stemness’.

MidbodyWhen a cell undergoes mitosis, the two resulting daughter cells remain connected by anintercellular bridge containing the midbody, an electron-dense structure that is composed ofhundreds of proteins [48,49]. The non-mitotic functions of this structure have revealed importantfindings related to midbody inheritance, accumulation, degradation, and release, and theirfunctional consequences [48].

The question of midbody inheritance or release has been a highly debated subject. At the end ofcytokinesis, abscission of the intercellular bridge can occur asymmetrically on one side of themidbody, and this allows internalization of the midbody remnant by one of the daughter cells,specifically to that which inherited the old centrosome (midbody accumulation or inheritance)[50]. By contrast, when abscission occurs on both sides of the midbody it is released into theextracellular space (midbody clearance) [48]. These findings suggest that cells can regulatewhether they utilize midbody accumulation or midbody clearance.

4 Trends in Cell Biology, Month Year, Vol. xx, No. yy

TICB 1289 No. of Pages 11

Midbody accumulation correlates with enhanced ‘stemness’ or increased pluripotency similar tothe inheritance of the old centrosome [50,51]. Increasing midbody accumulation throughprevention of autophagy also increases reprogramming efficiency of induced pluripotent stemcells [50]. Autophagy has been identified to play a central role in midbody degradation andclearance despite the suggestion that stem cells possess low autophagic activity [50,51].

By contrast, midbody loss appears to correlate with stem cell differentiation [50,51]. In neuralprogenitor cells in vivo, midbody release is enhanced during the early stages of neurogenesis,corresponding to an increase in differentiation [52,53]. Once released into the extracellularspace, themidbody can be engulfed by neighboring cells, where it can influence cellular behavior[54]. Because themidbody possesses many components of major signaling platforms, includingWnts, MAPK, and chemokines, the inheritance of this remnant may increase stemness in thereceiving cells [55], allowing long-distance passage of enriched signaling platforms betweencells.

Given the emerging importance of stem cells in regulating cancer progression, the relationship ofmidbody inheritance and cancer stem cell-like characteristics has been investigated. Cancerstem cells have either moremidbody accumulation or undergo lessmidbody release, suggestingthat midbody inheritance may be an important regulator of tumorigenesis [50,51]. How exactlythe midbody can regulate these behaviors remains unclear.

Taken together, these studies suggest that the inheritance and accumulation of the midbodycorresponds with stem cell behavior, while midbody loss parallels with more differentiation-likebehaviors [48]. Whether the fate of the cell leads to a specific midbody behavior, or whether themidbody behavior confers cell fate, is not yet clear.

Asymmetric Segregation of Damaged or Misfolded ProteinsThe asymmetric inheritance of damaged proteins has only recently been approached inmammalian systems but has been extensively studied in other organisms, such as Drosophilaand budding yeast [9,56]. In the mammalian system, the majority of studies on the asymmetricsegregation of damaged proteins have focused on induced protein aggregates, the over-expression of specific aggregation-prone proteins, with the aim to associate these findings withcellular events occurring in different degenerative diseases. These experiments were based onthe first description of the mammalian aggresome, a structure containing aggregated, misfoldedproteins that are transported on microtubules to themicrotubule organizing center (MTOC),and caged in the intermediate filament vimentin [57]. The revelation that a cell aggregates thesemisfolded proteins into a single structure led to questions of how this might function in a dynamicsystem: specifically, how are aggregates inherited, what is the cost of that inheritance, and whatis the mechanism governing the asymmetric segregation of aggregates?

Induced AggregatesStudies investigating the segregation of aggregates in mammalian cells during asymmetricdivision have suggested that the inheritance of aggregates to the ‘other’ daughter cell mayimprove cellular fitness for the proliferating, stem, or longest-lived daughters [42,58,59].

Original studies in mammalian cell culture lines revealed that cells that overexpressed ahuntingtin (Htt) fragment that formed aggregates could undergo normal mitosis [59]. Thisaggregate structure did not break down during mitosis, and was selectively inherited by only oneof the daughter cells, leaving the other cell free of damage. The aggregates were in closeproximity to the MTOC, likely as a result of a dynein-dependent process [60], and weresurrounded by a cage-like vimentin structure [59]. Thus, this study implicates the pericentro-somal matrix, vimentin, microtubules, and dynein in asymmetric segregation of protein

Trends in Cell Biology, Month Year, Vol. xx, No. yy 5

TICB 1289 No. of Pages 11

aggregates in dividing mammalian cells, similar to the machinery observed in initial earlyaggresome studies [57,60]. These in vitro findings were also supported by observations in ahuman neurodegenerative disease where protein aggregates are formed in cells. In intestines ofpatients with spinocerebellar ataxia type 3, the more-differentiated cells of the intestinalcrypts contained aggregates, while the stem cells remained clean [59]. This suggests thatasymmetric segregation of aggregates may also occur in stem cells in vivo, potentially to benefitthe stem daughter cell.

What is the cost of inheriting protein aggregates? Live-cell imaging of Htt overexpression in ratPC12 cells revealed that the time from one round of mitosis to the next was longer in thedaughter cell that inherited the Htt+ inclusion body and that this inheritance led to an increase incell death [61]. These effects could be mediated by impairment of the proteasome, a proteincomplex important for degradation of damaged, misfolded, or targeted proteins. In only onereport has the inheritance of the inclusion body resulted in positive effects in the daughter cell,such as increased resistance to stress, as well as longer neurite outgrowth in cells forced todifferentiate [61]. The underlying mechanisms behind these findings remain currently unclear.

Misfolded proteins have been shown to be sorted into two distinct inclusion body-like structuresin mammalian cells: a juxtanuclear quality control (JUNQ) compartment, and an insolubleprotein deposit (IPOD) compartment [62]. Soluble misfolded proteins together with activeproteasomes are localized to the JUNQ compartment, and are confined by a vimentin cage [42].By contrast, insoluble aggregates, such as Htt protein, are moved to the IPOD and are notdegraded [42]. However, it remained unclear how inclusion bodies are segregated duringmitosis and how accumulation in these compartments affected the fitness of future cell progeny.

Recent studies in mammalian cell lines showed that the intermediate filament vimentin confinesthe JUNQ and is inherited asymmetrically by daughter cells during mitosis [42]. The IPODinclusion body is also inherited asymmetrically, although its segregation is not regulated togetherwith JUNQ inheritance. Thus, in mammalian cells, the same daughter does not always inheritboth inclusions [42], while in yeast, for example, the mother always inherits both the IPOD andthe JUNQ [62]. Furthermore, only the inclusion bodies and not small stress foci are retainedasymmetrically, assuming that aggregation of these proteins is required before asymmetricsegregation can take place during mitosis. The inheritance of the JUNQ results in an increase inthe amount of time until the next division, thus describing a functional cost to this inclusioninheritance [42].

Themechanism behind the inheritance of JUNQ appears to involve the association of JUNQwiththe cytoskeleton and, specifically, the intermediate filament vimentin. The formation of a vimentincage occurs before the accumulation of substrates to the JUNQ, suggesting that vimentin mayinform the location of the aggregate [42]. Recently, reduced mitochondria, P-body markers, andstress granule markers have been reported to colocalize with collapsed vimentin, suggestingthat vimentin could mediate the functional asymmetric segregation of these molecules as well asof misfolded proteins [63].

Taken together, these studies suggest that mammalian cells undergo cellular rejuvenation duringmitosis by sorting specific aggregates between daughter cells through cytoskeletal mechanismsto increase the cellular fitness and resilience of one of the progeny.

Endogenous Protein CargoesWhile the previously described experiments suggest that asymmetric inheritance of damagedproteins is specific to the overexpression and induction of misfolded or mutant proteins, multiplereports also have characterized the segregation of endogenously expressed proteins. These

6 Trends in Cell Biology, Month Year, Vol. xx, No. yy

TICB 1289 No. of Pages 11

studies have primarily been performed in stem cells, showing the importance of asymmetricsegregation in these cell types.

Total polyubiquitinated proteins, as well as specific forms of phospho-Smad1 and phospho-b-catenin, both reported to be targeted for degradation with these modifications, have been foundto be asymmetrically segregated in mitotic human embryonic stem cells (ESCs) and COS7 cells[64]. This protein accumulation was found in nuclear bays colocalizing with the centrosome, andrequired microtubule transport for movement to the centrosome [64]. It is not clear, however,whether this inheritance is specific to the older or younger centrosome, or which mechanismsmight be controlling this segregation.

Recently, asymmetric segregation of ubiquitinated proteins and vimentin was also found inmammalian NSCs during mitosis [65]. The segregation of these cargoes was dependent on theage of the cell: young NSCs predominantly segregated ubiquitinated proteins and vimentinasymmetrically, while NSCs from old mice segregated these cargoes more symmetrically [65].Similarly to budding yeast [66] and Caenorhabditis elegans [67], a diffusion barrier has beenidentified in the endoplasmic reticulum (ER) membrane of NSCs that may potentially regulatethe asymmetric segregation of cargoes. In NSCs, this diffusion barrier weakens with chronologi-cal age, correlating with a more symmetric segregation of cargoes during mitosis [65].

Because the nuclear envelope (NE) breaks down and becomes part of the ER duringmammalianmitosis, it has been suggested that proteins of the NE may be important in diffusion barrierfunction and the asymmetric distribution of cargoes. Disruption of NE proteins through over-expression of the premature-aging mutant protein progerin leads to a weakened diffusionbarrier, and more symmetric segregation of cargoes such as ubiquitinated proteins [65]. Thus, inNSCs, mitotically aggregated ubiquitinated proteins may be segregated to one daughter cell bya diffusion barrier mediated by interactions of the nuclear envelope with the cytoskeleton. Afunctional consequence of the asymmetric segregation of these cargoes also has been reportedthrough time-lapse imaging, demonstrating decreased cell proliferation for the daughter cell thatinherits these cargoes [65].

In both the embryonic and adult brain in vivo, the non-stem daughter cell inherits ubiquitinatedproteins and vimentin, leaving the stem cell clean, whereas in the old brain there is a moresymmetric inheritance of ubiquitinated proteins, further supporting the in vitro evidence for aweakened diffusion barrier and symmetric segregation of molecules with age [65]. Thesefindings confirm studies performed in Drosophila stem cell compartments showing that asym-metric segregation of damaged proteins led to the longest-lived progeny remaining clean [61].

Because these studies focus on endogenously produced proteins, bundled only during mitosis,and dispersed throughout the cell during interphase, they suggest that this inheritance might benormal practice in proliferating cells. Furthermore, that this segregation becomes dysfunctionalwith age implies that this process is important for maintaining cellular fitness and preventingcellular aging.

Potential Mechanisms and ConsequencesMany potential mechanisms have been suggested for the asymmetric segregation of damagedproteins, both induced and endogenous, including cytoskeleton interactions either through theMTOC, microtubules, and vimentin, as well as the presence of a diffusion barrier in the cleavagefurrow of dividing cells. The details of how these proteins perform the segregation, or how theymake the choice for which daughter cell receives the cargo, remain unknown. Proteasomefunction also may be significant in the asymmetric segregation of cargoes. For example, it hasbeen reported in T lymphocytes that asymmetric inheritance of proteasomes results in an

Trends in Cell Biology, Month Year, Vol. xx, No. yy 7

TICB 1289 No. of Pages 11

Outstanding QuestionsWhat is the mechanism behind theasymmetric segregation of cargoes?Is it different for each cargo?

How does the cell know which daugh-ter needs to be renewed?

Does the asymmetric segregation ofcargoes influence the cell to adopt aparticular fate?

What is the influence of the cellularcontext (e.g., replication stress) onasymmetric inheritance of cargoes?

What other cargoes are asymmetricallysegregated between daughter cells?

Does chronological aging affect thesegregation of all cargoes?

What does the more differentiateddaughter do with the inheritance ofdamaged proteins?

Is this effect generalizable to all asym-metric cell divisions?

Are all ‘good’ proteins segregated withthe old centrosome and all ‘bad’ pro-teins with the young centrosome?

asymmetric degradation of the transcription factor T-bet [68]. Because both the centrosome [69]and the JUNQ compartment [42] have been suggested to have localized proteasome activity,their identified asymmetric segregation places them as interesting candidates to mediate a rolefor proteasomes in the asymmetry of cargoes. Future experiments will need to characterize themechanisms behind these asymmetric segregations and determine if they function similarly withall cell cargoes, and in all cell types.

The functional outcome of the asymmetric inheritance of damaged proteins is also somewhatinconsistent between studies. The majority suggest that the inheritance of these molecules by aproliferating cell leads to a reduction in stem-cell proliferation rate or increased cell death[42,58,65]; however, it has also been reported that inheritance by a differentiating cell can leadto enhanced differentiation [58], suggesting a positive role for this asymmetry. Understanding thebasal differences in the proliferating versus differentiating intracellular milieu may reveal the basisfor these different responses.

Concluding RemarksHow cellular cargoes become segregated into two daughter cells during mammalian cell divisionhas gained increasing attention over the past several years (Table 1). For some organelles (suchas centrosomes and centrioles) there is extensive evidence for asymmetric segregation,whereas for others (such as sister chromatid distribution and mitochondrial inheritance) thepicture is much less clear. Conceptually it is important to note that the inheritance of cargoes isnot necessarily only good or bad (see Outstanding Questions). Some instances of asymmetricsegregation appear to be beneficial for one of the daughter cells where inheritance may mediatea degree of cellular memory: for example, the stem cell that inherits the mother centrosome andits associated proteins is the first daughter cell to remake a primary cilium allowing quickresponses to its environment. Similarly, inheriting the midbody remnant and thus large numbersof proteins involved in signaling can affect proliferation and defer stemness specifically to the

Is the fate of the cell inheriting this cargodetermined by the inheritance, or doesthe cell type decide the inheritance?

Key Figure

Asymmetric Segregation of Cargoes During Mammalian Cell Division

[Non-stem][Stem]

Ubiqui�natedproteins

Oldmitochondria

Aggregate

Youngcentrosome

Vimen�n

‘Immortal’DNA strand

Midbodyremnant

Ciliarymembrane

Oldcentrosome

Youngmitochondria

‘Mortal’DNA strand

Figure 1. [1_TD$DIFF]Shown is a schematic of a dividing cell and cellular components that have been described previously to beasymmetrically inherited by the two daughter cells upon cell division of the mother cell. Note that this depiction cannot begeneralized to all cell types so far analyzed, given that non-random segregation of cargoes presumably shows high cell type-and context-dependent specificity. In addition, these cargoes were typically not studied at the same time to confirm theircolocalization. The left side of the cell depicted in anaphase represents the part of the daughter cell that will retain stem cellproperties, whereas the non-stem daughter cell (in this example two distinct fates of the daughters will occur) is depicted onthe right.

8 Trends in Cell Biology, Month Year, Vol. xx, No. yy

TICB1289

No.of

Pages

11

Table 1. Studies Focused on the Asymmetric Segregation of Misfolded or Damaged Proteins During Mammalian Cell Division

Cell Type Cargo Localization Functional Consequenceof Cargo Inheritance

Mechanism Asymmetric/SymmetricFate of Cells

If Asymmetric,Who Inherits?

Refs

Hamster O23 andHEK293 cells

Htt fragment (either HDQ74 orHDQ119 + XFP)Vimentin

Perinuclear – MTOC;vimentin

Symmetric fate (cell lines);asymmetric fate (humanintestine)

Non-stem progeny(differentiated orcommitted)

[59]

Human ES, COS7,L cells

Endogenous proteins targetedfor degradation (phospho-Smad1,phospho-b-catenin, andpolyubiquitinated proteins)

Pericentrosomal(nuclear bay);colocalized withmicrotubules duringinterphase

Both daughtersrespond to BMPtreatment similarly

SuggestsMTOC andcentrosomeinheritance

Symmetric fate [64]

PC12 cells Htt103Q–eGFP Proliferating conditions:longer cell cycle length,increased cell death,increased resistanceto stress;differentiatingconditions: enhanceddifferentiation (longerneurite)

Suggestsproteasomeimpairment

Symmetric fate [58]

HEK293, N2a,CHO cells

HttQ119–eYFPVHL–XFPVimentin

VHL (JUNQ)–proteasomes,vimentin, centrosome,MTOC;Htt (IPOD) – noassociation

Longer cell cycle length SuggestsMTOC,cytoskeleton,specificallyvimentin

Symmetric fate [42]

Mouse and ratNSCs

Endogenous ubiquitinatedproteinsVimentin

Longer cell-cycle length Diffusionbarrier(correlation)

Symmetric fate (culturedcells); asymmetric fate(embryonic and adultbrain)

Non-stemdaughter

[65]

Trendsin

CellB

iology,Month

Year,V

ol.xx,No.yy

9

TICB 1289 No. of Pages 11

Table 2. Current Generalizable Positive and Negative Consequences of Inheritance of Reported SegregatedComponents

‘Positive’ Inheritancea ‘Negative’ Inheritancea

‘Younger’ mitochondria Misfolded/aggregated proteins

Old centrosome High levels of ubiquitinated proteins

Ciliary membrane ‘Older’ mitochondria

Midbody remnant

‘Immortal’ DNA strand

aThese have been shown in at least one example to have ‘positive’ or ‘negative’ effects, however, they may be cell type- orcontext-dependent.

stem cell daughter. At the same time there are more negative outcomes associated with cargoinheritance that mediate cellular and replicative aging: the non-stem daughter cell may inheritprotein aggregates, more damaged proteins, and the older mitochondria (Figure 1 and Table 2).As such, asymmetric segregation of cargoes may represent a way to rejuvenate proliferatingcells to ensure a lifelong capacity of cell genesis.

Without a doubt there will not be a single mode of asymmetric inheritance that holds true for allcell types under all cellular conditions. Instead, it seems plausible that there is a high cellularspecificity (for instance depending on the cellular longevity of daughter cells) and context-dependency that may change under physiological conditions (e.g., replicative stress) and thatmay also largely influence results obtained using in vitro approaches. Recent technologicaladvances allowing birthdating of proteins and organelles, using, for example, approaches basedon photoconversion of tagged proteins in combination with improved imaging systems thatenable the longitudinal observation of single clones over extended times, will help to characterizethe presence of the potential underlying mechanisms of asymmetric segregation of cellularcomponents during mammalian cell division.

AcknowledgmentsSpecial thanks to S. Rinehart for artwork. The laboratory of S.J. is supported by grants from the Swiss National Science

Foundation (31003A_156943 and BSCGI0_157859), the European Molecular Biology Organization (EMBO) Young

Investigator program, and the European Research Council (ERC). D.L.M. was supported by a Long-Term Fellowship

of the Human Frontier Science Program. D.L.M. and S.J. are members of the Zurich Neuroscience Center (ZNZ). We

apologize to all colleagues whose work could not be cited owing to space limitations.

References

1. Henderson, K.A. and Gottschling, D.E. (2008) A mother's sacri-

fice: what is she keeping for herself?Curr. Opin. Cell Biol. 20, 723–728

2. Pelletier, L. and Yamashita, Y.M. (2012) Centrosome asymmetryand inheritance during animal development. Curr. Opin. Cell Biol.24, 541–546

3. Inaba, M. and Yamashita, Y.M. (2012) Asymmetric stem cell divi-sion: precision for robustness. Cell Stem Cell 11, 461–469

4. Homem, C.C. and Knoblich, J.A. (2012) Drosophila neuroblasts: amodel for stem cell biology. Development 139, 4297–4310

5. Morrison, S.J. and Kimble, J. (2006) Asymmetric and symmetricstem-cell divisions in development and cancer.Nature 441, 1068–1074

6. Gotz, M. and Huttner, W.B. (2005) The cell biology of neuro-genesis. Nat. Rev. Mol. Cell Biol. 6, 777–788

7. Bonifacino, J.S. (2014) Adaptor proteins involved in polarizedsorting. J. Cell Biol. 204, 7–17

8. Kusek, G. et al. (2012) Asymmetric segregation of the double-stranded RNA binding protein Staufen2 during mammalian neuralstem cell divisions promotes lineage progression. Cell Stem Cell11, 505–516

10 Trends in Cell Biology, Month Year, Vol. xx, No. yy

9. Knoblich, J.A. (2010) Asymmetric cell division: recent develop-ments and their implications for tumour biology.Nat. Rev. Mol. CellBiol. 11, 849–860

10. Chao, M.P. et al. (2008) Establishment of a normal hematopoieticand leukemia stem cell hierarchy.Cold Spring Harb. Symp. Quant.Biol. 73, 439–449

11. Watt, F.M. et al. (2006) Epidermal stem cells: an update. Curr.Opin. Genet. Dev. 16, 518–524

12. Brack, A.S. and Rando, T.A. (2012) Tissue-specific stem cells:lessons from the skeletal muscle satellite cell. Cell Stem Cell 10,504–514

13. He, S. et al. (2009) Mechanisms of stem cell self-renewal. Annu.Rev. Cell Dev. Biol. 25, 377–406

14. Zhao, C. et al. (2008) Mechanisms and functional implications ofadult neurogenesis. Cell 132, 645–660

15. Tetteh, P.W. et al. (2015) Plasticity within stem cell hierarchies inmammalian epithelia. Trends Cell Biol. 25, 100–108

16. Barker, N. et al. (2007) Identification of stem cells in small intestineand colon by marker gene Lgr5. Nature 449, 1003–1007

17. Doulatov, S. et al. (2012) Hematopoiesis: a human perspective.Cell Stem Cell 10, 120–136

TICB 1289 No. of Pages 11

18. Charge, S.B. and Rudnicki, M.A. (2004) Cellular and molecularregulation of muscle regeneration. Physiol. Rev. 84, 209–238

19. Tumbar, T. et al. (2004) Defining the epithelial stem cell niche inskin. Science 303, 359–363

20. Cairns, J. (1975) Mutation selection and the natural history ofcancer. Nature 255, 197–200

21. Tajbakhsh, S. (2008) Stem cell identity and template DNA strandsegregation. Curr. Opin. Cell Biol. 20, 716–722

22. Rando, T.A. (2007) The immortal strand hypothesis: segregationand reconstruction. Cell 129, 1239–1243

23. Rocheteau, P. et al. (2012) A subpopulation of adult skeletalmuscle stem cells retains all template DNA strands after celldivision. Cell 148, 112–125

24. Shinin, V. et al. (2006) Asymmetric division and cosegregation oftemplate DNA strands in adult muscle satellite cells. Nat. Cell Biol.8, 677–687

25. Karpowicz, P. et al. (2005) Support for the immortal strand hypoth-esis: neural stem cells partition DNA asymmetrically in vitro. J. CellBiol. 170, 721–732

26. Potten, C.S. et al. (2002) Intestinal stem cells protect their genomeby selective segregation of template DNA strands. J. Cell Sci. 115,2381–2388

27. Conboy, M.J. et al. (2007) High incidence of non-random templatestrand segregation and asymmetric fate determination in dividingstem cells and their progeny. PLoS Biol. 5, e102

28. Tomasetti, C. and Bozic, I. (2015) The (not so) immortal strandhypothesis. Stem Cell Res. 14, 238–241

29. Potten, C.S. et al. (1978) The segregation of DNA in epithelial stemcells. Cell 15, 899–906

30. Steinhauser, M.L. et al. (2012) Multi-isotope imaging mass spec-trometry quantifies stem cell division and metabolism. Nature 481,516–519

31. Kiel, M.J. et al. (2007) Haematopoietic stem cells do not asym-metrically segregate chromosomes or retain BrdU. Nature 449,238–242

32. Yadlapalli, S. and Yamashita, Y.M. (2013) Chromosome-specificnonrandom sister chromatid segregation during stem-cell division.Nature 498, 251–254

33. Charville, G.W. and Rando, T.A. (2013) The mortal strand hypoth-esis: non-random chromosome inheritance and the biased seg-regation of damaged DNA. Semin. Cell Dev. Biol. 24, 653–660

34. Rambhatla, L. et al. (2001)Cellular Senescence: ExVivo p53-Depen-dent Asymmetric Cell Kinetics. J. Biomed. Biotechnol. 1, 28–37

35. Wei, P.C. et al. (2016) Long neural genes harbor recurrent DNAbreak clusters in neural stem/progenitor cells. Cell 164, 644–655

36. Ouellet, J. and Barral, Y. (2012) Organelle segregation duringmitosis: lessons from asymmetrically dividing cells. J. Cell Biol.196, 305–313

37. Lerit, D.A. et al. (2013) Organelle asymmetry for proper fitness,function, and fate. Chromosome Res. 21, 271–286

38. Vevea, J.D. et al. (2014) Inheritance of the fittest mitochondria inyeast. Trends Cell Biol. 24, 53–60

39. McFaline-Figueroa, J.R. et al. (2011) Mitochondrial quality controlduring inheritance is associated with lifespan and mother-daugh-ter age asymmetry in budding yeast. Aging Cell 10, 885–895

40. Katajisto, P. et al. (2015) Asymmetric apportioning of aged mito-chondria between daughter cells is required for stemness.Science348, 340–343

41. Dalton, C.M. and Carroll, J. (2013) Biased inheritance of mito-chondria during asymmetric cell division in the mouse oocyte.J. Cell Sci. 126, 2955–2964

42. Ogrodnik, M. et al. (2014) Dynamic JUNQ inclusion bodies areasymmetrically inherited in mammalian cell lines through the asym-metric partitioning of vimentin. Proc. Natl. Acad. Sci. U.S.A. 111,8049–8054

43. Wang, X. et al. (2009) Asymmetric centrosome inheritance main-tains neural progenitors in the neocortex. Nature 461, 947–955

44. Paridaen, J.T. et al. (2013) Asymmetric inheritance of centrosome-associated primary cilium membrane directs ciliogenesis after celldivision. Cell 155, 333–344

45. Anderson, C.T. and Stearns, T. (2009) Centriole age underliesasynchronous primary cilium growth in mammalian cells. Curr.Biol. 19, 1498–1502

46. Piotrowska-Nitsche, K. and Caspary, T. (2012) Live imaging ofindividual cell divisions in mouse neuroepithelium shows asymme-try in cilium formation and Sonic hedgehog response. Cilia 1, 6

47. Reina, J. and Gonzalez, C. (2014) When fate follows age: unequalcentrosomes in asymmetric cell division. Philos. Trans. R. Soc.Lond. 369

48. Chen, C.T. et al. (2013) Resurrecting remnants: the lives of post-mitotic midbodies. Trends Cell Biol. 23, 118–128

49. Skop, A.R. et al. (2004) Dissection of the mammalian midbodyproteome reveals conserved cytokinesis mechanisms. Science305, 61–66

50. Kuo, T.C. et al. (2011) Midbody accumulation through evasion ofautophagy contributes to cellular reprogramming and tumorige-nicity. Nat. Cell Biol. 13, 1214–1223

51. Ettinger, A.W. et al. (2011) Proliferating versus differentiating stemand cancer cells exhibit distinct midbody-release behaviour. Nat.Commun. 2, 503

52. Dubreuil, V. et al. (2007) Midbody and primary cilium of neuralprogenitors release extracellular membrane particles enriched inthe stem cell marker prominin-1. J. Cell Biol. 176, 483–495

53. Marzesco, A.M. et al. (2005) Release of extracellular membraneparticles carrying the stem cell marker prominin-1 (CD133) fromneural progenitors and other epithelial cells. J. Cell Sci. 118, 2849–2858

54. Crowell, E.F. et al. (2014) Engulfment of the midbody remnant aftercytokinesis in mammalian cells. J. Cell Sci. 127, 3840–3851

55. Dionne, L.K. et al. (2015) Midbody: from cellular junk to regulator ofcell polarity and cell fate. Curr. Opin. Cell Biol. 35, 51–58

56. Denoth Lippuner, A. et al. (2014) Budding yeast as a modelorganism to study the effects of age. FEMS Microbiol. Rev. 38,300–325

57. Johnston, J.A. et al. (1998) Aggresomes: a cellular response tomisfolded proteins. J. Cell Biol. 143, 1883–1898

58. Bufalino, M.R. and van der Kooy, D. (2014) The aggregation andinheritance of damaged proteins determines cell fate during mito-sis. Cell Cycle 13, 1201–1207

59. Rujano, M.A. et al. (2006) Polarised asymmetric inheritance ofaccumulated protein damage in higher eukaryotes. PLoS Biol.4, e417

60. Johnston, J.A. et al. (2002) Cytoplasmic dynein/dynactin mediatesthe assembly of aggresomes. Cell Motil. Cytoskeleton 53, 26–38

61. Bufalino, M.R. et al. (2013) The asymmetric segregation of dam-aged proteins is stem cell-type dependent. J. Cell Biol. 201, 523–530

62. Kaganovich, D. et al. (2008) Misfolded proteins partition betweentwo distinct quality control compartments. Nature 454, 1088–1095

63. Pattabiraman, S. and Kaganovich, D. (2014) Imperfect asymme-try: the mechanism governing asymmetric partitioning of dam-aged cellular components during mitosis. Bioarchitecture 4,203–209

64. Fuentealba, L.C. et al. (2008) Asymmetric mitosis: unequal segre-gation of proteins destined for degradation. Proc. Natl. Acad. Sci.U.S.A. 105, 7732–7737

65. Moore, D.L. et al. (2015) Amechanism for the segregation of age inmammalian neural stem cells. Science 349, 1334–1338

66. Shcheprova, Z. et al. (2008) A mechanism for asymmetric segre-gation of age during yeast budding. Nature 454, 728–734

67. Lee, D.A. et al. (2012) Tanycytes of the hypothalamic medianeminence form a diet-responsive neurogenic niche.Nat. Neurosci.15, 700–702

68. Chang, J.T. et al. (2011) Asymmetric proteasome segregation asa mechanism for unequal partitioning of the transcription factor T-bet during T lymphocyte division. Immunity 34, 492–504

69. Wigley, W.C. et al. (1999) Dynamic association of proteasomalmachinery with the centrosome. J. Cell Biol. 145, 481–490

Trends in Cell Biology, Month Year, Vol. xx, No. yy 11