Hepatocytes Deficient in Nuclear Envelope Protein Lamina-associated Polypeptide 1 are an Ideal Mammalian System to Study Intranuclear Lipid Droplets

Behind the stoNE wall: A fervent activity for nuclear lipids

The four main types of biomolecules are nucleic acids, proteins, carbohydrates and lipids. The knowledge about their respective interactions is as important as the individual understanding of each of them. However, while, for example, the interaction of proteins with the other three groups is extensively studied, that of nucleic acids and lipids is, in comparison, very poorly explored. An iconic paradigm of physical (and likely functional) proximity between DNA and lipids is the case of the genomic DNA in eukaryotes: enclosed within the nucleus by two concentric lipid bilayers, the wealth of implications of this interaction, for example in genome stability, remains underassessed. Nuclear lipid-related phenotypes have been observed for 50 years, yet in most cases kept as mere anecdotical descriptions. In this review, we will bring together the evidence connecting lipids with both the nuclear envelope and the nucleoplasm, and will make critical analyses of these descriptions. Our exploration establishes a scenario in which lipids irrefutably play a role in nuclear homeostasis.

Nuclear lipid droplet: Guardian of nuclear membrane lipid homeostasis?

Lipid droplets (LDs) are cytoplasmic organelles, but they are also found within the nucleus in small numbers. Nuclear LDs that form at the inner nuclear membrane (INM) often increase in response to perturbation in phosphatidic acid (PA) and/or diacylglycerol (DAG), both implicated in various INM functions. Nuclear LDs also increase upon downregulation of seipin, a protein that can trap PA and DAG in the endoplasmic reticulum. Notably, both PA and DAG appear to be more densely distributed on the surface of nuclear LDs than in the INM. I propose that nuclear LDs play a role in regulating the PA and DAG level in the INM, thereby contributing to the lipid homeostasis in this compartment.

Nuclear lipid droplets in Caco2 cells originate from nascent precursors and in situ at the nuclear envelope

Intestinal epithelial cells convert excess fatty acids into triglyceride (TAG) for storage in cytoplasmic lipid droplets and secretion in chylomicrons. Nuclear lipid droplets (nLDs) are present in intestinal cells but their origin and relationship to cytoplasmic TAG synthesis and secretion is unknown. nLDs and related lipid-associated promyelocytic leukemia structures (LAPS) were abundant in oleate-treated Caco2 but less frequent in other human colorectal cancer cell lines and mouse intestinal organoids. nLDs and LAPS in undifferentiated oleate-treated Caco2 cells harbored the phosphatidate phosphatase Lipin1, its product diacylglycerol, and CTP:phosphocholine cytidylyltransferase (CCT)α. CCTα knockout Caco2 cells had fewer but larger nLDs, indicating a reliance on de novo PC synthesis for assembly. Differentiation of Caco2 cells caused large nLDs and LAPS to form regardless of oleate treatment or CCTα expression. nLDs and LAPS in Caco2 cells did not associate with apoCIII and apoAI and formed dependently of microsomal triglyceride transfer protein expression and activity, indicating they are not derived from endoplasmic reticulum luminal LDs precursors. Instead, undifferentiated Caco2 cells harbored a constitutive pool of nLDs and LAPS in proximity to the nuclear envelope that expanded in size and number with oleate treatment. Inhibition of TAG synthesis did affect the number of nascent nLDs and LAPS but prevented their association with promyelocytic leukemia protein, Lipin1α, and diacylglycerol, which instead accumulated on the nuclear membranes. Thus, nLD and LAPS biogenesis in Caco2 cells is not linked to lipoprotein secretion but involves biogenesis and/or expansion of nascent nLDs by de novo lipid synthesis.

Dynamic regulation of hepatic lipid metabolism by torsinA and its activators

Depletion of torsinA from hepatocytes leads to reduced liver triglyceride secretion and marked hepatic steatosis. TorsinA is an atypical ATPase that lacks intrinsic activity unless it is bound to its activator, lamina-associated polypeptide 1 (LAP1) or luminal domain-like LAP1 (LULL1). We previously demonstrated that depletion of LAP1 from hepatocytes has more modest effects on liver triglyceride secretion and steatosis development than depletion of torsinA. We now show that depletion of LULL1 alone does not significantly decrease triglyceride secretion or cause steatosis. However, simultaneous depletion of both LAP1 and LULL1 leads to defective triglyceride secretion and marked steatosis similar to that observed with depletion of torsinA. Depletion of both LAP1 and torsinA from hepatocytes generated phenotypes similar to those observed with only torsinA depletion, implying that the 2 proteins act in the same pathway in liver lipid metabolism. Our results demonstrate that torsinA and its activators dynamically regulate hepatic lipid metabolism.

Functional interaction of torsinA and its activators in liver lipid metabolism

TorsinA is an atypical ATPase that lacks intrinsic activity unless it is bound to its activators lamina-associated polypeptide 1 (LAP1) in the perinuclear space or luminal domain-like LAP1 (LULL1) throughout the endoplasmic reticulum. However, the interaction of torsinA with LAP1 and LULL1 has not yet been shown to modulate a defined physiological process in mammals in vivo . We previously demonstrated that depletion of torsinA from mouse hepatocytes leads to reduced liver triglyceride secretion and marked steatosis, whereas depletion of LAP1 had more modest similar effects. We now show that depletion of LULL1 alone does not significantly decrease liver triglyceride secretion or cause steatosis. However, simultaneous depletion of both LAP1 and LULL1 from hepatocytes leads to defective triglyceride secretion and marked steatosis similar to that observed with depletion of torsinA. Our results demonstrate that torsinA and its activators dynamically regulate a physiological process in mammals in vivo .