The increased fat storage demands imposed on adipose tissue during positive energy balance is a challenge that can be resolved by increasing the size of pre-existing adipocytes (hypertrophy) and/or by generating new adipocytes (hyperplasia)1. The relative contribution of hyperplasia and hypertrophy to the expansion of fat mass depends on the individual’s age and the anatomical location of the fat depot. Children and teenagers rely primarily on hyperplasia (mainly when they are overweight), whereas in adulthood hypertrophy of adipocytes is the most common mechanism. Hypertrophy and hyperplasia adaptations relate to lower and higher adipocyte turnover2, respectively, and are associated with different pathophysiological outcomes, irrespectively of age, gender or bodyweight3,4. For a similar body mass index and fat depot, hypertrophic adipocytes correlate with chronic inflammation, dyslipidaemia, fatty liver, increased insulin resistance and increased risk of developing type 2 diabetes. Hyperplastic adipose tissue results in a metabolically benign profile, which provides a cellular rationale for the ‘healthy obesity’ concept that is applied to individuals who have a high body mass index but are resilient to metabolic obesity-associated comorbidities. Understanding why storing fat through adipocytes growing in size versus increasing their number have very different metabolic outcomes is necessary to understand the molecular mechanisms that induce adipocyte metabolic resilience, vulnerability and early functional failure.

In a recent work published in Nature Medicine, Li et al.5 identified a mechanism in patients with obesity and who are hyperinsulinaemic that confers metabolic vulnerability to the adipocytes of their subcutaneous white adipose tissue (scWAT). In the absence of mitosis, these adipocytes exhibited active endo-reduplication of their nuclear genome and premature activation of the senescence program (Fig. 1). The endo-reduplication observation was unexpected as it refutes the current paradigm that mature adipocytes do not re-enter the cell cycle. However, the evidence of senescence provided a mechanism for why adipocyte hypertrophy is dysfunctional and prone to inflammation.

Fig. 1: Hypertrophied adipocytes are associated with a pro-inflammatory metabolic phenotype.
figure 1

Li and colleagues5 show that obesity-associated hyperinsulinaemia triggers endo-reduplication in subcutaneous hypertrophic adipocytes, leading to senescence and a pro-inflammatory response (the senescence-associated secretory phenotype (SASP)).

Full size image

Endo-reduplication is the duplication of genomic DNA without segregating chromosomes during mitosis6. Endo-reduplication is a common phenomenon in protists, bacteria, fungi, plants and the animal kingdom to enable the growth of active metabolic cells, facilitating regeneration and compensatory cellular hypertrophy. In mammals, hepatocyte endo-reduplication is a component of the response to stress and/or injury. Now, Li and colleagues5 provide the first evidence that endo-reduplication also occurs in fully mature human adipocytes when they are becoming hypertrophic and exposed to high insulin levels.

The first hint of endo-reduplication in adipocyte hypertrophy was the upregulation of cell cycle genes in mature adipocytes from the subcutaneous abdominal adipose tissue of patients with obesity and who are hyperinsulinaemic. Of relevance, the adipocytes of patients who are normoglycaemic did not increase cell cycle gene expression, irrespective of whether these individuals are lean or have obesity. The insulin-driven cell cycle fingerprint was lower G1/S-phase and increased G2 markers without mitotic markers in RNA sequencing or confocal imaging. Insulinaemia and fat mass were the only clinical parameters that were positively correlated with cyclin A2 (S/G2 phase), cyclin D1 and phospho-HH3 cell cycle markers. Insulin was the only biochemical parameter associated with transcriptional enrichment in G2 markers. Moreover, the authors showed that insulin increased DNA synthesis (S phase) and nuclear size in hypertrophic adipocytes, a cellular hallmark of entering endo-reduplication.

Adipocytes from patients with obesity and who are hyperinsulinaemic showed an increase in the expression of G2 markers and decreased expression of proliferation markers (Ki67 and PCNA), evidencing cell cycle exit (while cells arrested in G2), as well as a marked increase in cyclin D1. The unexpected increase in cyclin D1 and its correlation with senescence-related genes was the critical insight revealing that patients with obesity and who are hyperinsulinaemic had senescent mature hypertrophic adipocytes in their subcutaneous adipose depots. Moreover, hyperinsulinaemia (but not obesity status, gender or age) accounted for the blockage of the cell cycle and activation of the canonical senescence program.

The senescence-associated secretory phenotype (SASP)7 reflects metabolic activity and increased chemokine and cytokine secretion. The RNA-sequencing profiling of hypertrophic adipocytes from patients with obesity and who are hyperinsulinaemic confirmed the upregulation of senescence-mediated, adipocyte-derived pro-inflammatory factors, including IL6, CXCL8, IL1b, CXCL10 and CXCL2, providing further evidence that senescence induced by insulin contributes to chronic low-grade inflammation in patients with obesity.

Li et al.5 subsequently confirmed that hyperinsulinaemia was the driver that initiated cell cycle progression and triggered the subsequent cell cycle block, senescence and inflammation. They used palbociclib, an inhibitor of CDK4 and CDK6, and metformin to block mTOR-mediated mitogenic effects upstream of cyclin D1. Palbociclib (in combination with insulin) blocked cell cycle progression (with negligible effects on cyclin D1), and increased senescence markers in adipocytes from patients with obesity and who are hyperinsulinaemic without triggering a pro-inflammatory profile (SASP). Metformin also reduced cell cycle progression, slashed mTOR signalling and decreased cyclin D1, but decreased the senescence genetic program. Instead, metformin induced the expression of genes linked to quiescence and limited inflammation in cultured adipocytes independently of the hyperinsulinaemia status of the patient. Thus, Li et al.5 concluded that inhibition of senescence upstream of cell cycle entry could be an anti-inflammatory strategy in hypertrophic adipose tissue.

A fundamental question, with pathogenic implications, was whether the endo-reduplication mechanism was restricted to subcutaneous adipose tissue or whether it was also a feature of intra-abdominal adipose tissue, which is known to be more pro-inflammatory. The authors concluded that endo-reduplication was not a primary mechanism triggering the typical inflammation of the intra-abdominal depot of individuals with obesity. Given the pathogenic relevance of visceral adipocytes, it was unexpected that they exhibited only marginal cell cycle activation and senescence markers compared to subcutaneous adipocytes and responded weakly to the mitogenic action of insulin. The fact that the effect was restricted to scWAT may indicate that this depot might be more susceptible to dysfunction and/or that the initial dysfunction of scWAT might initiate a cascade knock-on effect that leads to the dysfunction of visceral adipose tissue. Moreover, the results of Li et al.5 could be interpreted as insulin triggering endo-reduplication in scWAT in a desperate attempt to activate the cell cycle as an expansion strategy to accommodate the excess of fat. However, this adaptive mechanism failed when a robust compensatory hyperinsulinaemia response triggered endo-reduplication, senescence and inflammation under severe metabolic stress.

These findings are exciting as they identify insulin-triggered senescence of subcutaneous adipocytes as a critical event that prompts a flow of events leading to irreversible adipose tissue dysfunction. This mechanism provides a rationale for therapeutic approaches that allow fat mass expansion and simultaneously prevent inflammatory, metabolic complications. Similarly, these data identify endo-reduplication as a biomarker of imminent failure of subcutaneous adipose tissue and likely subsequent expansion and inflammation of visceral adipose tissue. Given the regional differences in endo-reduplication and premature senescence in adipose tissue, it is unclear whether preventing the senescence of subcutaneous adipose tissue might, indirectly, be beneficial for visceral adipose tissue if it decreases requirements for fat storage and inflammation. It is also unknown whether preventing the pre-senescence program in scWAT might lead to more-hypertrophied adipocytes and cell death. The evidence provided by Li et al.5 fully justifies investigating these and other questions, such as how changes in lipid turnover during weight loss or in lipodystrophy affect the endo-reduplication and senescence program in adipose tissue.