© Schattauer 2015 Thrombosis and Haemostasis 114.1/2015 208 Instability of cytosolic phospholipase A 2 α variant upon cellular expression as a basis for its clinical presentation Aida Zulueta 1 ; Cristina Razzari 2 ; Gessica Fontana 2 ; Eti A. Femia 2 ; Elena M. Faioni 2 ; Marco Cattaneo 2 ; Marco Trinchera 3 1 Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milano, Italy; 2 Medicina 3, Azienda Ospedaliera San Paolo, and Dipartimento di Scienze della Salute, Università degli Studi di Milano, Milano, Italy; 3 Dipartimento di Medicina Clinica e Sperimentale, Università dell’Insubria, Varese, Italy Correspondence to: Marco Trinchera Dipartimento di Medicina Clinica e Sperimentale Università dell’Insubria, Varese, Italy Tel.: +39 0332 39 7160, Fax: +39 0332 39 7119 E-mail: marco.trinchera@uninsubria.it Received: November 10, 2014 Accepted after minor revision: February 24, 2015 Epub ahead of print: April 23, 2015 http://dx.doi.org/10.1160/TH14-11-0926 Thromb Haemost 2015; 114: 208–210 Dear Sirs, We recently described two siblings suffering bleeding diathesis and recurrent gastroin- testinal ulcers, associated with homozygous 1723G>C transition in PLA2G4A gene, which changed Asp575 to His in cytosolic phospholipase A 2 α (cPLA 2 α) (1). Although recurrent gastrointestinal ulcers were also described in a patient with heterozygous nonsynonymous amino acid substitutions Ser111Pro, Arg485His, and Lys651Arg (2) and in two patients with a homozygous 4 bp deletion (g.155574_77delGTAA) determin- ing a frameshift of 10 amino acids before a premature stop codon (p.Val707fsX10) (3), the bleeding diathesis was a typical feature of our patients. Interestingly, the impair- ment in arachidonic acid metabolism as- sociated to the Asp575His mutation was ex- ceptionally severe, as shown by the very low levels of serum thromboxane B 2 , which were over ten-fold lower than those found in the other cases of inherited cPLA 2 α defi- ciency and in subjects on chronic treatment with aspirin, an irreversible cyclo-oxyge- nase-1 inhibitor. To characterise the defect further, wild-type (WT) and variant cPLA 2 α were permanently expressed in human embryonic kidney HEK-293 cells as fusion proteins with the HaloTag sequence in the pFN21 vector (see Suppl. Methods, available online at www.thrombosis-online. com). Cell transfection and selection were performed following a previously reported procedure (4). Four clones from the WT and 4 from the variant cPLA 2 α-transfected HEK-293 cells were selected because of Ha- loTag expression, as determined by fluor- escence microscopy upon vital staining with TMR-ligand. Mean fluorescence inten- sity was 14.10 ± 8.87 (mean value ± stan- dard deviation) in the WT clones and 3.77 ± 3.25 in the Asp575His clones, respect- ively, as assessed by flow cytometry (▶Fig- ure 1a). These values suggest a much lower expression of the mutant construct, al- though an exception is apparent in the case of clone 4. However, the crude fluorescence intensity could be also affected by the amount of tagged fragments or free tag po- tentially formed. Expression levels of the HaloTag/cPLA 2 α transcript in the clones, quantitated by competitive RT-PCR (see Suppl. Methods, available online at www. thrombosis-online.com), were rather vari- able, as expected in this expression system, and did not differ between WT and mutant clones (▶Figure 1b). In particular, mutant clone 1 displayed the highest value measured among all clones, reaching an amount that was just one-half of β-actin transcript. Conversely, the WT HaloTag/ cPLA 2 α fusion protein was detectable by western blotting with both the anti-HaloTag and the anti-cPLA 2 α antibodies as a very prominent band of about 140 kDa in all 4 WT clones, while it appeared as a faint spot in mutant clone 1 only, and was undetect- able in the others (▶Figure 1 c). Note- worthy, in mutant clone 4, a spot of much lower molecular weight (close to that of the HaloTag alone) was detected with the anti- HaloTag antibody only, indicating the ex- pression of an aberrant protein lacking the bulk of cPLA 2 α sequence (Suppl. Figure 1, available online at www.thrombosis-online. com). The fusion protein from WT clone 2 and the one from Asp576His clone 1 were both purified through a HaloTag binding resin and eluted by the action of a specific protease (see Suppl. Methods, available on- line at www.thrombosis-online.com), ob- taining the proteins free from the HaloTag peptide. Western blot analysis of the puri- fied proteins showed a good recovery of WT cPLA 2 α and a substantial loss of the mutated one (▶Figure 1d). It is not possi- ble at this stage to establish whether the protein was actually lost or misfolded to such a degree that could not be recognised by the antibody. These results indicate that the WT protein is stable in vitro, while the variant is not. Thus, we were able to repro- duce in a cell system the main features re- ported in the patient’s platelets, where the mRNA transcribed from the mutated cPLA 2 α was expressed at the same levels as in other family members, while the protein was undetectable (1). The mutant protein, when detectable, appears only in the cyto- sol and not in a particulate fraction (see Suppl. Figure 2, available online at www. thrombosis-online.com), suggesting that the mutation does not impair solubility. In conclusion, the 1723G>C transition in PLA2G4A gene allows normal mRNA tran- scription and processing, but gives rise to an unstable protein. This mechanism of disease was already reported for other in- herited haematologic (5) and systemic de- fects (6). According to a recently proposed model (7), the interaction of some molecu- lar chaperones with mutated proteins de- tects potentially functional but excessively unstable proteins and directs them towards degradation instead of folding. In our ex- pression system, which is able to force ex- pression via both transcription through the strong cytomegalovirus promoter and translation through the HaloTag sequence, we found all mutant clones expressing the fusion transcript, 2 clones not expressing the fusion protein at all, one expressing the HaloTag only, and one expressing a mini- mal amount of fusion protein, which was lost/misfolded during purification. Since such a clone was the one expressing an enormous amount of transcript, we specu- lated that subsequent translation saturated the degradation mechanism, providing the minimal amount of fusion-protein Letters to the Editor For personal or educational use only. No other uses without permission. All rights reserved. Downloaded from www.thrombosis-online.com on 2018-04-27 | ID: 1001066444 | IP: 54.70.40.11