Phosphatidylserine turns the gears of phospholipids in B cell lymphoma

Phosphatidylserine (PS) is a quantitively minor class of phospholipids that contributes to cellular membranes (1). PS is synthesized at the endoplasmic reticulum (ER) by PTDSS1 and PTDSS2, which replace the head-group of phosphatidylcholine (PC) or phosphatidylethanolamine (PE), respectively, with serine. Once synthesized, PS is transported primarily by a non-vesicular manner via membrane contact sites, where the ER is closely apposed to other cellular membranes (2). One such major destination is the plasma membrane (PM) where PS serves as a regulator for a number of cellular processes including but not limited to signaling, membrane trafficking, and apoptosis.

Cancer cells are known to adapt their lipid metabolism for growth. This is a wise strategy, but, to the contrary, can be a vulnerable weakness that we can exploit as a therapeutic target. Previous studies identified PS as such a potential target for several types of cancer (3, 4). Accordingly, Aoki and colleagues (5) screened various cell lines with an inhibitor to PTDSS1, and found B cell lymphoma was the most sensitive cancer type to inhibition of PS synthesis by PTDSS1. In fact, Ramos cells, a type of Burkitt lymphoma cell line that was one of the top hits in the screen, showed a significant growth defect in culture as well as in mice xenografts when the PTDSS1 gene was knocked-out. The question is then how does the loss of function of PTDSS1 lead to cell death in B cell lymphoma? As described below, the answer was found through the dysregulation of another type of phospholipid, phosphoinositides, mediated by a B cell-specific signal transducer on the PM, the B cell receptor (BCR).

B cells express BCR, a membrane-bound antibody, at the PM. Upon binding with antigens (ligands), the BCR transduces signals for survival or death depending on the cell’s status. A major signaling cascade via the BCR is the activation of a phospholipase C (PLC; γ2 isoform in B cells) that hydrolyzes phosphatidylinositol 4,5 bisphosphate (PI(4,5)P₂) into diacylglycerol (DAG) and inositol 1,4,5-trisphosphate (IP₃), followed by Ca²⁺ release to the cytosol and thus Ca²⁺ depletion from the ER. This in turn elicits a robust Ca²⁺ influx from outside via the Orai Ca²⁺ channels activated by the ER Ca²⁺ sensor STIM1/2 at the ER-PM contact sites, a process called store-operated Ca²⁺ entry (SOCE) (6). Omi et al. (5) found that the generation of IP₁, a measurable metabolite of IP₃, was enhanced in PTDSS1 KO Ramos cells. Consistent with this, an increase in Ca²⁺ responses, even spontaneously under normal culture conditions without BCR stimulation, was also observed. These Ca²⁺ responses were sufficient to activate caspase-3, indicative of apoptosis. Functional ablation of BCR components repressed this Ca²⁺ response. These data all support the idea that enhanced BCR-mediated phosphoinositide cascade, leading to increased Ca²⁺ responses, caused cell death in PTDSS1 KO cells.

That said, why does the phosphoinositide signaling pathway go awry upon inhibition of PS synthesis? The lipid transport processes, in particular by lipid transfer proteins at membrane contact sites, connected the loss of PS with phosphoinositide malfunction.

As mentioned earlier, PS is transported from the ER to the PM via ER-PM contact sites. This is mediated by members of oxysterol-binding protein-related proteins (ORPs), ORP5 and ORP8 (reviewed in 2). ORP5/8 function as lipid exchangers, mediating transport of PS from the ER to the PM in exchange for phosphatidylinositol 4-phosphate (PI4P) (7). Intriguingly, PTDSS1 inhibition led to accumulation of PI4P and concomitant reduction of PS at the PM, suggesting possible dysregulation of ORP5/8-mediated lipid exchange under PTDSS1 inhibition. In fact, ORP5/8 double KO cells also showed a higher Ca²⁺ response upon BCR stimulation, confirming they phenocopy PTDSS1 KO cells. Interestingly though, this phenocopy was incomplete, strongly indicating a missing mechanism behind the regulation of phosphoinositide metabolism by PTDSS1.

An increase in the levels of phosphatidylinositol (PI), the precursor of PI4P and PI(4,5)P₂, in PTDSS1 KO cells motivated the authors to test another lipid exchanger, Nir2/3, which transport PI from the ER to the PM and phosphatidic acid (PA) in the opposite direction. Nir2/3, as PI/PA exchangers, facilitate the “PI cycle,” in which PA derived from PLC-mediated hydrolysis of PI(4,5)P₂ is recycled to re-synthesize PI (8, 9). Thus, the Nir2/3-mediated process might be aberrantly accelerated to fuel PI under PTDSS1 inhibition. As expected, accumulation of PI, and consequent enhancement of Ca²⁺ response under PTDSS1 inhibition were both suppressed in Nir2/3 double KO cells, suggesting a contribution of Nir2/3 in accumulation of PI and hence its metabolic products PI4P and PI(4,5)P₂.

Collectively, the current study highlights a unique feature of B cell lymphoma in that their survival particularly depends on PS synthesis by PTDSS1. The study also reveals a pathophysiological crosstalk between non-vesicular lipid transport at ER-PM contacts and Ca²⁺ signaling (10). Mechanistically, inhibition of PS synthesis by PTDSS1 resulted in down-regulation of ORP5/8-mediated PS/PI4P exchange and up-regulation of PI cycling by Nir2/3-mediated PI/PA exchange. Dysfunction of these processes both contribute to the build-up of phosphoinositide lipids in the PM, thereby leading to BCR-mediated hyper-activation of phosphoinositide signaling and ultimately death (Fig. 1).

In B cell lymphoma, PS inhibition provoked the dysregulation of several other phospholipids. First, an insufficient PS synthesis at the ER prevented lipid counter-flow between the ER and the PM, leading to the build-up of phosphoinositides, in particular PI4P, in the PM. In this situation, ORP5/8 should be able to form ER-PM contacts in theory (as demonstrated, albeit by exogenously expressed ORP8 (5)), but could apparently not accomplish PI4P transport to the ER. This might be due to the lower availability of PS in the ER, which could hamper efficient lipid exchange with PI4P, but also due possibly to the less active status of Sac1, a PI4P phosphatase in the ER whose enzymatic activity can be upregulated by PS (11). Note that a similar cause-and-effect relationship, albeit in the opposite direction, has been observed in a PS metabolic disease, Lenz-Majewski syndrome, where the gain-of-function mutations in PTDSS1 lead to accumulation of PS at the ER (12). In this situation, PI4P levels decreased in the PM due to dysregulation of ORP8-mediated PI4P/PS exchange along with activation of Sac1 by PS. These observations indicate a new mode of inter-relationship where the metabolic states of two different lipids are reciprocally influenced via their lipid countertransport. However, regulatory mechanisms including how ORP5/8 and metabolic enzymes sense and control these lipid levels merits further investigation. Second, PI levels increased upon PS inhibition via acceleration of the PI cycle by Nir2/3-mediated PI/PA exchange. However, this PI increase was not completely suppressed by Nir2/3 double KO as the authors mentioned. The underlying mechanism remains unclear. Last but not least, PE was also affected by PS inhibition. Given that exogenously supplied PE rescued the growth defect of B cell lymphoma with PS inhibition (5), the PS-PE axis is another layer of survival/death regulation.

In sum, this study sheds light on an essential role of PS in B cell lymphoma. PS keeps a balance of the tightly inter-connected phospholipidome as if it turns the interlocked gears of phospholipids. Once the gear of PS stops working, the others also malfunction. Understanding how each gear (phospholipid) turns properly in sync will not only uncover the biological significance of phospholipid inter-relationships, but also aid in the development of new strategies for treating diseases including B cell lymphoma.