Sarcolipin alters SERCA1a interdomain communication by impairing binding of both calcium and ATP

Sarcolipin (SLN), a single-spanning membrane protein, is a regulator of the sarco-endoplasmic reticulum Ca2+-ATPase (SERCA1a). Chemically synthesized SLN, palmitoylated or not (pSLN or SLN), and recombinant wild-type rabbit SERCA1a expressed in S. cerevisiae design experimental conditions that provide a deeper understanding of the functional role of SLN on the regulation of SERCA1a. Our data show that chemically synthesized SLN interacts with recombinant SERCA1a, with calcium-deprived E2 state as well as with calcium-bound E1 state. This interaction hampers the binding of calcium in agreement with published data. Unexpectedly, SLN has also an allosteric effect on SERCA1a transport activity by impairing the binding of ATP. Our results reveal that SLN significantly slows down the E2 to Ca2.E1 transition of SERCA1a while it affects neither phosphorylation nor dephosphorylation. Comparison with chemically synthesized SLN deprived of acylation demonstrates that palmitoylation is not necessary for either inhibition or association with SERCA1a. However, it has a small but statistically significant effect on SERCA1a phosphorylation when various ratios of SLN-SERCA1a or pSLN-SERCA1a are tested.


Supplementary informations
Considering the data provided by the literature (see Supplementary Table 1), one of our main concern was to control as much as possible both protein orientation to get properly assembled SLN:SERCA1a complexes, and SLN to SERCA1a ratio to favour specific interaction.It has been shown that, for membrane proteins not emerging from the membrane, i.e. deprived of large cytosolic domain as is the case for SLN, it can be particularly difficult to control the orientation after co-reconstitution ( 1 and references hereafter).The final orientation of membrane protein in proteoliposomes was shown to really depend on the lipid nature, the detergent, the ionic strength, the pH, the membrane curvature, and the shape of the membrane protein [2][3][4][5] .Although the reconstitution of SERCA1a alone or in the presence of PLB leads to about 80±10% of right side oriented proteins as described by Rigaud and coll.and Young and coll. 6,7, this favoured orientation is most probably due to the large cytosolic headpieces of these proteins.Indeed, these large hydrophilic domains may not enter in the destabilized liposomes during reconstitution because of steric restrictions and/or electrostatic interactions with phospholipids 6 , and for PLB, well-known interactions with the SERCA1a N-domain can undoubtedly help in orienting PLB towards the cytosolic leaflet [8][9][10] .SLN is clearly different from PLB as its N-terminus is very short (Supplementary Figure 2) and involved in a few interactions with SERCA1a compared to PLB [11][12][13][14][15][16] .
Furthermore, in our hands, reconstitution of SERCA following the method of Rigaud and coll.leads to a 50/50 distribution of the SERCA orientation 17 , indicating that fine-tuning of reconstitution parameters is certainly necessary.Taking all these points into account, co-reconstitution of recS1a in the presence of SLN would lead to an unknown amount of complexes properly formed and right-side-out oriented, therefore requiring additional cumbersome fine-tuning as well as additional assay to assess complexes formation and orientation.
We and others have shown over the last forty years that DDM-or C12E8-solubilized native or recombinant S1a are stable for long period from a few hours to days depending on the experimental conditions and that this solubilized enzyme behaves as the native SERCA1a [18][19][20][21][22][23][24][25][26] .Note that in the very last reference, yeast-expressed DDM-solubilized purified enzyme behaves as the recombinant SERCA1a in microsomes prepared from cos-7 cells.It indicates again that the lipid environment is not mandatory for measuring the turn-over rate and following the binding of ligands whether by intrinsic fluorescence or by measuring phosphorylation rates.Thus, we chose to use yeast-expressed and purified SERCA1a (recS1a) and synthesized SLN in the presence of DDM or C12E8.This strategy was already successful for structural analysis by NMR of the SLN:SERCA1a complex in dodecylphosphocholine 27 .
We indeed assume that the dynamics of the protein-detergent complex enable the two partners to interact in the right orientation.The agreement between our experiment and others from the literature confirms that the interaction between SLN and SERCA1a occurs in spite of the presence of the detergent.Our protocol should be suitable for studying the role of other regulatory peptides on SERCA1a.

Use of synthesized Palm-SLN and SLN
The RBM GABA strategy was used to synthesize peptide SLN(1-12,E 2,RBM ,C 9,Palm )-SAL (1).In brief, the peptide is prepared on Rink amide ChemMatrix resin (0.1 mmol, degree of substitution: 0.45 mmol/g) by using the microwave-assisted fluorenylmethyloxycarbonyl (Fmoc)-based solid-phase peptide synthesis (SPPS) method.The general synthetic procedures include reductive amination, reduction of the nitro group on the RBM, introduction of the Lys6 tag, protection of the 2-OH by GABA group on the RBM to tolerate the trifluoroacetic acid (TFA) cleavage, removal of the Mmt group on the Cys and palmitoylation.Please see 28 for details on materials provenance and procedure.After the assembly of the full-length peptide, the resin was treated with a TFA cocktail at room temperature to obtain 21 mg of crude peptide (1') (isolated yield: 11.1%).Finally, the GABA protecting group was removed under pyridine/AcOH (7:6, mol/mol) conditions to give the target peptide (1) (16.1 mg, isolated yield: 76.7%).

Reverse-phase chromatography profiles and ESI-MS analysis are available in Supplementary
Figure 3 for peptides 1-3, respectively.Final products were lyophilized for long-term storage until being used.
Extreme care has to be taken for resuspension of hydrophobic peptides in detergent solution to prevent aggregation of peptides by hydrophobic, non-specific interactions in an aqueous environment.
For Palm-SLN and SLN, we followed the procedure described in 29 .Briefly, SLN powder was initially resuspended in a large volume of pure water and lyophilized to evaporate most of the TFA.The rinsed powder was then suspended in presence of 10 mM HCl to exchange the anionic residual TFA that may interact with the positive charges of the peptide for chloride.This suspension was lyophilized again.
Three additional cycles of -dissolving in pure water / lyophilisation -will allow to eliminate traces of TFA and to get back to a pH compatible with functional tests.For final solubilisation of SLN in detergent, powder was then dissolved in 50% TFE and an amount of detergent to reach a 100:1 mole:mole detergent:SLN ratio.Here, we used C12E8 as it is fully compatible with functional characterization of SERCA1a 23 .This suspension was then lyophilized again to evaporate TFE.Sarcolipins (Palm-SLN or SLN) were finally resuspended at 0.8 mg/mL (about 200 µM) in a 10 mM potassium phosphate buffer pH 7.4, in the presence of 2.5 mg/mL C12E8.

Expression and Purification of recombinant SERCA1a (recS1a)
The yeast expression plasmid pYeDP60-SERCA1a-BAD (WT) and expression in Saccharomyces cerevisiae was described previously 30 .Yeast membranes were solubilized by DDM for subsequent purification of recombinant Ca 2+ -ATPase by Streptavidin affinity chromatography 31 .Purified recombinant SERCA1a was recovered in a buffer containing 50 mM MOPS-Tris pH7, 100 mM KCl, 5 mM MgCl2, 2.1 mM CaCl2, 40 % glycerol (v/v) and 0.5 mg/mL DDM, together with some thrombin remaining from the elution procedure.The protein concentration in the purified fraction was in the 0.05 to 0.15 mg/mL range depending on the batch.

DDM-solubilized native SERCA1a purification
SR vesicles were suspended at a concentration of 1 mg/mL in a buffer containing 50 mM MOPS-Tris pH7, 100 mM KCl, 5 mM MgCl2, 0.1 mM CaCl2, 20 % glycerol (v/v) and 10 mg/mL DDM (detergent:protein ratio 10:1 w/w), and incubated for 10 minutes at 20°C.Solubilized material was recovered after centrifugation for 10 minutes at 120,000 x g on a TLA100.3rotor (Beckman Coulter TL100 ultracentrifuge).Supernatant was injected on a size exclusion chromatography column (Superdex 200 10/300 GL on an Akta Purifier FPLC system, GE Healthcare) equilibrated with the same buffer but now containing only 0.5 mg/mL DDM.Chromatography was performed at 4°C and at a flow rate of 0.5 mL/min.Fractions corresponding to the main peak were collected, pooled and stored at -80°C.For further details, see 32 .

Steady-state ATPase activity measurements
Measurements were performed using classical enzyme coupled assay.We measured the rate of NADH oxidation in the presence of 0.1 mg/mL lactate deshydrogenase, 0.05 mg/mL pyruvate kinase, 1 mM phosphoenolpyruvate, 0.25 mM NADH (protocol was adapted from 22 ) and in the additional presence of 2 mg/mL C12E8 to limit time-dependent inactivation of the purified SERCA1a 20,23 .a "r" referred to rabbit isoforms, "h" to human isoform.b HEK-293 cells and myocytes were used for co-expression and microsomes preparation.DOPC or EYPC-EYPA bilayers obtained from co-reconstitution of native SR with recombinant or synthetic SLN.c molar ratio does not take into account the possible presence of endogenous SLN. "n.d." for not determined.As a comparison, Lipids:SLN:SERCA1a ratio is about 70:1:1 to 70:2:1 mole:mole:mole in native SR.d Ca1/2 corresponds to the amount of calcium necessary to attain half-maximal activity (calcium uptake or turn-over depending on the system used).e Aggregation number for DPC is 54.Proteins are distributed in about 6 mM DPC micelles (1000 micelles:50 SLN:1 SERCAa mole:mole:mole).f Aggregation number for DDM is 80-120.Proteins are distributed in about 7-11 µM DDM micelles (130 micelles:10 SLN:1 SERCA1a mole:mole:mole; see Methods section for a detailed calculation).