Age reprogramming: cell rejuvenation by partial reprogramming

Abstract

'Age reprogramming' refers to the process by which the molecular and cellular pathways of a cell that are subject to age-related decline are rejuvenated without passage through an embryonic stage. This process differs from the rejuvenation observed in differentiated derivatives of induced pluripotent stem cells, which involves passage through an embryonic stage and loss of cellular identity. Accordingly, the study of age reprogramming can provide an understanding of how ageing can be reversed while retaining cellular identity and the specialised function(s) of a cell, which will be of benefit to regenerative medicine. Here, we highlight recent work that has provided a more nuanced understanding of age reprogramming and point to some open questions in the field that might be explored in the future.

Keywords: Age reprogramming; Cellular identity; Epigenetic rejuvenation; H3K9me3; OSKM; Partial reprogramming.

Fig. 1.

Age reprogramming. (A) Key events in age reprogramming research. Numbered references: (1) Wilmut et al., 1997; (2) Sinclair et al., 2016; (3) Takahashi and Yamanaka, 2006; (4) Liu et al., 2007; (5) Marion et al., 2009; (6) Singh and Zacouto, 2010; (7) Lapasset et al., 2011; (8) Horvath, 2013; (9) Manukyan and Singh, 2014; (10) Ocampo et al., 2016; (11) Olova et al., 2019; (12) de Magalhães and Ocampo, 2022. (B) Age reprogramming as a separable element of nuclear reprogramming. Nuclear reprogramming of an old cell (blue) after introduction of the Oct4, Sox2, Klf4 and c-Myc (OSKM) ‘reprogramming factors’ or by somatic cell nuclear transfer (SCNT) results in de-differentiation to an ESC-like state producing iPSCs and nuclear transfer-derived ESCs (NT-ESCs), respectively (green). Re-differentiation of iPSCs and NT-ESCs results in differentiated cells that are rejuvenated (young; red). Age reprogramming aims to bypass the de-/re-differentiation cycle, retaining the specialised functions of the old cell and simply making it younger. Based on Singh and Zacouto (2010). (C) Partial reprogramming as an experimental approach to show that age and developmental reprogramming are separable. ‘Reprogramming factors’ are introduced into an old cell (blue), which is characterised by the expression of age-related markers. During the trajectory from old cell to iPSC (green), a putative stage exists during which the expression of age-related markers is reduced or lost, indicating rejuvenation (red), while the ‘partially reprogrammed’ cell still possesses its specialised characteristics, i.e. it does not exhibit characteristics of embryonic cells. Based on Singh and Zacouto (2010).

Fig. 2.

Transient de-differentiation and age reprogramming. A ‘burst’ of OSKM expression results in transient de-differentiation that takes an old, specialised cell (blue) through a zone of (epigenetic) instability (green ellipse). Entry into the zone partially reprogrammes the cell, transiently de-differentiating and briefly expanding its developmental potential (x-axis). When OSKM expression is stopped, the partially reprogrammed intermediate returns to its specialised phenotype, with the cell now age reprogrammed and exhibiting a younger ‘age’ (y-axis). Recent work indicates that the time spent in the ‘zone of (epigenetic) instability’ could determine the degree of rejuvenation of an old cell, as measured by methylation and transcription clocks (Gill et al., 2022); the optimum time spent in the zone (trajectory A) results in an age reprogrammed cell that is young (red), whereas a longer time spent in the zone (trajectory B) results in an age reprogrammed cell that is older (less young; given in purple). Based on a model by Manukyan and Singh (2012).

Fig. 3.

Defining a ‘critical window’ for age reprogramming and designing reprogramming regimes. (A) The classical virally-transduced OSKM expression paradigm is shown at the top. The ‘critical window’ (highlighted in yellow) extends from day 7 to 15. This window was chosen, inter alia, on the basis of a plateau of expression for three clusters of fibroblast-specific genes (F1, F2 and F3) in the face of eAge falling from 50 to 20 years; cellular identity (defined by expression of F1, F2 and F3) is stable while eAge falls. Also shown is expression of the (early) pluripotency gene cluster (green) and the cell-cycle inhibitor p21 (CDKN1A; blue) that initially falls but then exhibits a rising plateau of expression during the ‘critical window’. Accumulation of senescence-associated gene products towards the end of the ‘critical window’ may provide an explanation for why transient OSKM expression beyond day 13 results in reduced age reprogramming. Based on a model by Simpson et al. (2021) and Singh et al. (2019). HDF, human dermal fibroblast. (B) Cyclic OSKM expression paradigms can be used to develop cell-type specific age reprogramming regimes. Based on the ‘critical window’ in A, a ‘zone of optimal age reprogramming’ can be defined whereby maintaining eAge between 20 and 50 enables age reprogramming to proceed without suppression of somatic identity. The ‘plasticity’ of cellular phenotype determines the type of cyclic expression paradigm to be used. Expression paradigm 1 (eAge¹; orange) has a cyclic regime in which OSKM is expressed for 1 day followed by no expression for 6 days. This would be appropriate for hepatocytes, for which expression for more than 1 day is known to be deleterious. Expression paradigm 2 (eAge²; purple), using 2 days of OSKM expression and 5 days of no expression, could be used for cardiomyocytes that are more refractory to OSKM reprogramming and require a longer ‘burst’ of OSKM expression. For both paradigms, the aim is to keep eAge between 20 and 50, where there is low level expression of the early pluripotency gene cluster (green) that follows the cyclic expression paradigm, while levels of somatic cell gene expression (grey) remain elevated. Once eAge is in the ‘zone of optimal age reprogramming’ it can be kept there if the ‘bursts’ of OSKM are less frequent, as shown. Based on a model by Singh et al. (2019).

Fig. 4.

Age reprogramming during pre-/early-post implantation development in the mouse. The hallmarks of ageing, telomere attrition, mitochondrial dysfunction, DNA repair and epigenetic drift are rejuvenated during early development: telomeres are extended in vivo during pre-implantation development; there is in vitro evidence that mitochondrial function is rejuvenated in differentiated iPSC derivates and via partial reprogramming; DNA repair is elevated in ESCs; and there is epigenetic rejuvenation of histone and DNA modifications during pre-/early-post implantation development that results in a global repressive epigenetic landscape in or around gastrulation, before entry into the phylotypic progression (note that the embryos depicted at E8.5 and older give a rough idea of when phylotypic progression occurs, although an exact phylotypic ‘stage’ for vertebrates has been difficult to identify). H3K9me3 increases and becomes enriched over promoters, gene bodies and termination sites in mesodermal and endodermal cells at E8.25 (Nicetto et al., 2019; Wang et al., 2018). Global H3K27me3 histone modification and CpG methylation levels reach maximal levels at around E6.5 (Auclair et al., 2014; Zheng et al., 2016), although many PRC2-targeted CGIs are demethylated through the PRC2-mediated recruitment of Tet dioxygenases (Li et al., 2018; Zhang et al., 2018). Remarkably, resetting of H3K9me3, H3K27me3 and CpG methylation broadly coincides with the eAge ground state at around E6.5-E7.5. Ect, ectoderm; End, endoderm; Epi, epiblast; Mes, mesoderm; PS, primitive streak.

Fig. 5.

Heterochromatin-like complexes safeguard cellular identity. Small heterochromatin-like complexes can ‘nucleate’ the assembly of larger H3K9me3-marked heterochromatin-like domains. Targeting and formation of the complexes occurs in several steps. (1) KRAB-ZNF binds to its DNA binding site through its zinc-fingers (Zn). (2) The KRAB domain of KRAB-ZNF interacts with KAP1. One molecule of KAP1 binds to an HP1 dimer, which in turn binds to H3K9me3 (red circles). (3) The PHD domain of KAP1 is an E3 ligase that cooperates with UBE2i to sumoylate (SUMO2) the KAP1 bromodomain. (4) The sumoylated bromodomain is bound by the NuRD complex that deacetylates acetylated histones in preparation for histone methylation. (5) SETDB1 H3K9 HMTase interacts with the sumoylated (SUMO2) bromodomain and generates H3K9me3 (red circles). (6) The ATRX/DAXX complex is bound to KAP1, HP1 and H3K9me3. ATRX/DAXX incorporates replacement histone H3.3 into chromatin, thereby reinforcing nucleation. (7) HP1 recruits an H3K36me3 HMTase and generates H3K36me3 (orange circles). (8) HP1 recruits an H4K20 HMTase that generates H4K20me3 (green circles). (9) KAP1 binds to the maintenance DNA methylase DNMT1 and its co-factor Np95. DNMT1 maintains cytosine methylation at the site of assembly. Modified from Singh and Newman (2020). The interaction of HP1 with H3K36me3 HMTase has been described (Zaidan and Sridharan, 2020) and it is known that the heterochromatin-like complex can generate the H3K36me3 modification (Sripathy et al., 2006). Several components of heterochromatin-like complexes have been identified in screens as being important in safeguarding cellular identity (see Table 1).