Mevalonate Pathway-mediated ER Homeostasis Is Required for Haploid Stability in Human Somatic Cells

The somatic haploidy is unstable in diplontic animals, but cellular processes determining haploid stability remain elusive. Here, we found that inhibition of mevalonate pathway by pitavastatin, a widely used cholesterol-lowering drug, drastically destabilized the haploid state in HAP1 cells. Interestingly, cholesterol supplementation did not restore haploid stability in pitavastatin-treated cells, and cholesterol inhibitor U18666A did not phenocopy haploid destabilization. These results ruled out the involvement of cholesterol in haploid stability. Besides cholesterol perturbation, pitavastatin induced endoplasmic reticulum (ER) stress, the suppression of which by a chemical chaperon significantly restored haploid stability in pitavastatin-treated cells. Our data demonstrate the involvement of the mevalonate pathway in the stability of the haploid state in human somatic cells through managing ER stress, highlighting a novel link between ploidy and ER homeostatic control.

The halving of genome copy from the normal diploid state potentially has pleiotropic effects on cellular homeostasis in haploid cells. An apparent feature of haploid cells is their halved cellular volume to diploids with the halving of total protein content (Yaguchi et al., 2018). Though these features possibly have profound influence on intracellular processes -such as the metabolic control -in haploid state, it remains elusive what aspects of the metabolism alter and characterize cellular phenotypes of haploid cells.
The mevalonate pathway metabolizes acetyl-CoA to produce sterol isoprenoids, and non-sterol isoprenoids that mediate diverse biosynthetic processes essential for cell construction and proliferation (Buhaescu and Izzedine, 2007;Mullen et al., 2016). Among mevalonate-derived metabolites, cholesterol serves as a structural component of cell membranes and a precursor of fundamental biomolecules, such as steroid hormones.
Mevalonate-derived polyisoprenols, such as dolichol phosphates are essential components of glycoprotein synthesis and endoplasmic reticulum (ER) homeostasis participating in protein N-glycosylation, C-and O-mannosylation, and GPI-anchor production (Carlberg et al., 1996;Chojnacki and Dallner, 1988;Doucey et al., 1998). Mevalonatederived isoprenoids are also used for the prenylation of small GTPases, which mediates their association to the cell membrane and signal transduction for dynamic processes such as cytoskeletal reorganization and vesicular trafficking (Leung et al., 2006;Wang and Casey, 2016). Recent studies have revealed that the mevalonate pathway controls cell size by optimizing mitochondrial functionality through protein prenylation (Miettinen and Björklund, 2015;Miettinen and Björklund, 2016;Miettinen et al., 2014). Inhibition of the rate-limiting enzyme of mevalonate pathway, 3-hydroxy-3methylglutaryl-coenzyme A reductase (HMGCR) by statins perturbs this homeostatic control, leading to increased cell size in different types of cultured cells (Miettinen and Björklund, 2015).
In this study, in an attempt to modulate cell size by an HMGCR inhibitor pitavastatin in human haploid HAP1 cells (Carette et al., 2011), we found that the inhibitor compromised the stability of the haploid state in these cells. Interestingly, a recent chemical screen searching for compounds that stabilize haploid state has also identified statins leading to the selective loss of haploid cells (Olbrich et al., 2019).
However, whether the mevalonate pathway is indeed involved in promoting haploid stability, and how the inhibition of the pathway may lead to destabilization of haploid state is still unknown. Using a pharmacological approach, we specified that statininduced ER stress as a process responsible for the destabilization of the haploid state.

Materials and methods
Cell culture and flow cytometry Laboratories and used at a dilution of 1:1000.
Immunoblotting was performed as previously described (Yaguchi et al., 2018). We used the ezWestLumi plus ECL Substrate (ATTO) and a LuminoGraph II chemiluminescent imaging system (ATTO) for signal detection. We quantified immunoblotting signals using the Gels tool in ImageJ software (National Institutes of Health).

Mevalonate production is required for the stability of the haploid state in HAP1 cells
To test the effects of inhibition of the mevalonate pathway on cell size control, we treated human haploid cell HAP1 with a competitive HMGCR inhibitor pitavastatin, which has been reported to increase the size of human cell lines such as Jurkat (Miettinen and Björklund, 2015). Although we did On the other hand, treatment with 2 M GGTI-298 mildly arrested haploid HAP1 cells at the G1 phase within 24 h (Fig. 3C), consistent with a previous report in several cell types (Sun et al., 1999). In prolonged culture for 20 d in the presence of 2 M GGTI-298, the haploid-to-diploid conversion was considerably slowed down compared to non-treated control, presumably because of the moderate G1 arrest (Fig. 3D). Therefore, the suppression of either protein farnesylation or geranylgeranylation did not phenocopy the pitavastatin-induced haploid destabilization in our long-term experiment.

Pitavastatin destabilizes the haploid state by evoking ER stress
Since statins potentially induce ER stress by suppressing dolichol phosphates biosynthesis and inhibiting protein N-glycosylation (Chojnacki and Dallner, 1988), we next tested the possibility that pitavastatin destabilizes the haploid state through perturbing ER homeostasis. For this, we tested the effect of pitavastatin on ER stress in HAP1 cells using immunoblot analysis of ATF4 or CHOP/GADD153/DDIT3, the unfolded protein response (UPR) components whose expression increases upon the induction of ER stress (Iurlaro and Muñoz-Pinedo, 2016;Marciniak et al., 2004;Wang et al., 2019). Treatment with 0.5 M pitavastatin for 3 d significantly increased the expression of both ATF4 and CHOP ( Fig. 4A and B).
Mevalonate supplementation canceled the ATF4 and CHOP upregulation in pitavastatin-treated cells ( Fig.   4A and B), demonstrating that pitavastatin evoked ER stress specifically through blocking mevalonate metabolism.
Finally, we determined whether ER stress induction is the cause of pitavastatin-mediated destabilization of haploid state in HAP1 cells. For this, we tested the effect of an ER stress-reducing chemical chaperone, tauroursodeoxycholic acid (TUDCA) (Ozcan et al., 2006;Yoon et al., 2016), on the haploid stability of HAP1 cells. Co-treatment with TUDCA did not affect ATF4 expression, but substantially blocked CHOP upregulation in pitavastatin-treated cells ( Fig.   4A and B), presumably reflecting the complex effects of chemical chaperones on different factors in the UPR pathways (Uppala et al., 2017). In contrast, TUDCA did not change the cholesterol level in pitavastatin-treated cells assessed by filipin staining ( Fig. 2A-C). In long-term passages, co-treatment of TUDCA significantly slowed down haploid-todiploid conversion in pitavastatin-treated cells ( Fig.   4C and D). Therefore, restoration of ER homeostasis by TUDCA substantially improved the stability of the haploid state in the presence of pitavastatin, demonstrating that haploid destabilization by pitavastatin is caused, at least in part, through the induction of ER stress.

Discussion
It is assumed that ploidy differences have pleiotropic effects on intracellular biosynthetic processes and that the altered biosynthesis, in turn, affects cellular physiology at different ploidy states. However, it remains mostly elusive what biosynthetic processes have influences on ploidy-linked cellular phenotypes.
The lower availability of intracellular space may limit stress-responding ER expansion in haploid cells, hence lower tolerance to unfolded protein accumulation.
Mevalonate metabolism is an essential process that supports diverse biosynthetic pathways.