The conformation of human phospholipid scramblase 1, as studied by infrared spectroscopy. Effects of calcium and detergent

    The conformation of human phospholipid scramblase 1, as studied by infrared spectroscopy. Effects of calcium and detergent Nagore Andraka, Lissete S´anchez-Magraner, Marcos Garc´ıa-Pacios, F´elix M. Go˜ni, Jos´e L.R. Arrondo PII: DOI: Reference:

S0005-2736(17)30069-X doi:10.1016/j.bbamem.2017.02.015 BBAMEM 82434

To appear in:

BBA - Biomembranes

Received date: Revised date: Accepted date:

14 September 2016 18 February 2017 22 February 2017

Please cite this article as: Nagore Andraka, Lissete S´anchez-Magraner, Marcos Garc´ıaPacios, F´elix M. Go˜ ni, Jos´e L.R. Arrondo, The conformation of human phospholipid scramblase 1, as studied by infrared spectroscopy. Effects of calcium and detergent, BBA - Biomembranes (2017), doi:10.1016/j.bbamem.2017.02.015

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

The conformation of human phospholipid scramblase 1, as studied by infrared spectroscopy. Effects of calcium and detergent.

PT

Nagore Andraka+, Lissete Sánchez-Magraner+, Marcos García-Pacios, Félix M. Goñi and José L.R. Arrondo*

SC

RI

Unidad de Biofísica (CSIC, UPV/EHU) and Departamento de Bioquímica, Universidad del País Vasco, P.O. Box 644, 48080 Bilbao, Spain.

NU

*Corresponding author. Fax No. +34 94601 3360. E-mail: [email protected] + These two authors contributed equally to this work.

AC CE P

TE

D

MA

ABSTRACT Human phospholipid scramblase 1 (SCR) is a membrane protein that catalyzes the transmembrane (flip-flop) motion of phospholipids. It can also exist in a non membrane-bound form in the nucleus, where it modulates several aspects of gene expression. Catalysis of phospholipid flip-flop requires the presence of millimolar Ca2+, and occurs in the absence of ATP. Membrane-bound SCR contains a C-terminal α-helical domain embedded in the membrane bilayer. The latter domain can be removed giving rise to a stable truncated mutant SCRΔ that is devoid of scramblase activity. In order to improve our understanding of SCR structure infrared spectra have been recorded of both the native and truncated forms, and the effects of adding Ca2+, or removing detergent, or thermally denaturing the protein have been observed. Under all conditions the main structural component of SCR/SCRΔ is a βsheet. Removing the C-terminal 28 aa residues, which anchor SCR to the membrane, leads to a change in tertiary structure and an increased structural flexibility. The main effect of Ca 2+ is an increase in the α/β ratio of secondary structure components, with a concomitant increase in the proportion of non-periodic structures. At least in SCRΔ, detergent (Zwittergent 3-12) decreases the structural flexibility, an effect somewhat opposite to that of increasing temperature. Thermal denaturation is affected by Ca 2+, detergent, and by the presence or absence of the C-terminal domain, each of them influencing in different ways the denaturation pattern. Keywords: Scramblase, membranes, spectroscopy, 2D IR correlation spectroscopy

1. INTRODUCTION Phospolipid scramblases catalyze the transbilayer (flip-flop) motion of phospholipid molecules in an ATP-independent process. The first member of this family, human phospholipid scramblase 1 (from now on abbreviated as SCR), was described in 1996 (1). This was soon followed by the discovery of three other members of the same family (2). The early results in the field were reviewed by Sahu et al. (3). More recently Suzuki et al. (4) described phospholipid scrambling induced by a transmembrane protein (TMEM16F) in animal cells that has been tentatively associated 1

calcium,

protein

structure,

detergents,

infrared

to the Scott syndrome. Phospholipid scramblases usually require the presence of Ca2+ for their catalytic action. SCR is a 318aa, type II endofacial membrane protein, expressed in a variety of cells. It contains a proline-rich domain at the Nterminus, a DNA-binding region (aa 86-118), a palmitoylation motif (aa 184-189), a nuclear localization signal (257-266), a Ca2+-binding site (aa 273-284), and a transmembrane α-helix (aa 291-309) (Scheme 1). The Ca2+- binding motif has a Ca2+ affinity constant in the millimolar range,

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

(5-7). series of experiments it was shown that a synthetic peptide representing the SCR α-helix could spontaneously insert into lipid bilayers (19) and that it had a specificity for binding cholesterol (20). Some of the previous discoveries have raised questions pertaining to the conformation of SCR and SCRΔ under the various experimental conditions. How does the loss of the C-terminal αhelix modify the overall protein conformation, why does Ca2+ appear to destabilize SCR structure, and how is the protein perturbed by the detergent that is used to keep it in suspension are some of those questions. In this paper we describe our attempts to provide the corresponding answers using infrared spectroscopy. This technique has found wide application in the study of protein conformation and dynamics, particularly in the study of membrane proteins (21). Our results provide a structural basis to the previous biochemical/functional observations.

D

MA

and is essential for the scrambling activity of SCR The actual cellular functions of SCR remain somewhat unclear, because in addition to its putative role in mediating transbilayer movement of phospholipids, SCR may also regulate at the nuclear level processes including signaling, cell proliferation and differentiation, apoptosis, injury, transcription, antibacterial, and antiviral activities (8-11). SCR in its palmitoylated form was found to partition with the EGF receptor in putative lipid rafts. EGF stimulates tyrosine phosphorylation of SCR and promotes protein internalization (8). In the absence of palmitoylation, virtually all of the expressed SCR localizes to the nucleus (12). These data suggest that the post-translational acylation determines the protein localization in the cell and regulates its normal function, either in the nucleus or incorporated to the plasma membrane (12,13). A structural model computed by homology modeling suggests that palmitoylation may represent the principal membrane anchorage for the SCR (6).

AC CE P

TE

SCR has also a predicted inside-to-outside oriented transmembrane helix (aa 291–309), and on this basis SCR is classified as a type II membrane protein. Finally, the remaining chain portion (aa 310–318) is considered as a short exoplasmic tail (14,15). In contrast to this model, Bateman and coworkers provide an alternative SCR model in which the helix instead of being inserted into the membrane is buried within the protein core (6). A number of recent contributions have provided data that support a role for SCR at the membrane level, without ruling out additional roles at the nuclear level, and indicate that the Cterminal α-helix may indeed have a transmembrane location. Francis et al. (16) showed that the C-terminal helix was essential for membrane insertion and scrambling activity and the data received confirmation from our laboratory (17). We prepared a mutant (SCRΔ) in which the last 28 aa residues, including the α-helix that presumably constitutes the transmembrane domain, were lacking. SCRΔ had lost the scramblase activity, and its affinity for Ca2+ was decreased by one order of magnitude (17) although this truncated mutant had still the capacity to bind membranes (18). In a different 2

2. MATERIALS AND METHODS 2.1. Protein purification. Construction of the mutant SCRΔ, cloning of SCR, and purification of SCR and SCRΔ have been described in detail elsewhere (17). Briefly, SCRΔ was extracted from inclusion bodies and it was purified using a HisTrap HP column with a stepwise 0 to 500 mM imidazole gradient. SCR was expressed as a fusion protein with MBP at the N-terminal end and it was purified in two steps, first using an amylose resin and then a DEAESepharose column. Finally, the MBP-SCR was digested with Factor Xa and the SCR was purified using a HisTrap HP column in the presence of 5mM Zwittergent 3-12. The purified proteins were identified in SDS-PAGE gels and stored at –80º C. 2.2. IR measurements. For infrared spectroscopy measurements SCRΔ was dialyzed against 150 mM NaCl, 20 mM Tris-HCl pH 7.5 buffer (TC buffer) + 1 mM EDTA at 4 ºC overnight, in the presence of 5 mM Zwittergent 3–12 when required. The protein solution was then concentrated to 2 mg ml-1 and next dialyzed against TC buffer in the presence or absence of 5 mM Zwittergent 3–12 at 4 ºC for 6 h in order to remove urea and EDTA.

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

To generate the 2D IR correlation maps, heating was used as the perturbation to induce time-dependent spectral fluctuations and to detect dynamic spectral variations on the secondary structure of SCR. Obtention of the twodimensional synchronous and asynchronous maps has been described previously (22).

3. RESULTS

AC CE P

TE

D

MA

SCR was dialyzed against TC-5 mM Zwittergent 3–12 buffer +1 mM EDTA and then against TC-5 mM Zwittergent 3–12. Next the protein was concentrated to 2 mg ml-1. The proteins were centrifuged at 25,000 x g for 40 min at 4 ºC to avoid protein aggregates and then protein aliquots (0.1 ml) were freeze-dried. Finally they were resuspended in the appropriate volume of D2O-based buffer, in the presence and absence of calcium, and the infrared spectra were recorded. The spectra were retrieved in a Bruker Tensor 27 (Bruker Optik GmbH, Ettlingen, Deutschland) spectrometer equipped with a liquid nitrogen-refrigerated mercury-cadmium-telluride detector. Samples were measured using a demountable Peltier liquid cell (Biotools Inc., USA) with excavated calcium fluoride BioCell windows (BioTools), and 25 µm optical path. A total of 143 interferograms min-1 were generated at 2 cm-1 resolution and averaged over one minute intervals. Opus 5.0 software from Bruker Optics was used for data acquisition. The samples were heated from 4–80 ºC at a rate of 1 ºC min-1. The quantitative results shown are averages from two closely similar traces, examples of which are shown in Fig. S1.

3.1. Native (SCR) and truncated (SCRΔ) scramblase. Ca2+ effects. The amide I bands of the IR spectra of SCR are shown in Fig. 1 in the presence and absence of 5 mM Ca2+, at three temperatures. The bands have been fitted to a number of components corresponding to secondary structure elements. Component band assignment was performed as discussed by Arrondo et al. (21), and was in most cases straightforward. At 20 ºC in the absence of Ca2+ (Fig. 1A) the largest spectral component corresponds to βsheet, indicated by the bands at 1630 cm-1 and 1678 cm-1, the so-called high- and low-frequency components of the β-sheet spectrum (Table 1). The band component at 1647 cm-1 includes probably αhelix and random (non-periodic) structures (see Discussion) and the 1661 cm-1 signal has been assigned to turns, although more recently it has 3

also been associated to intrinsically disordered structures (23). The small band components at 1610-1613 cm-1 correspond to amino acid side chains (24), presumably, for this protein, to the asymmetric guanidinum stretching vibration of the arginines present in the central portion of the SCR sequence (Scheme 1). The fractional areas of the different secondary structure components are summarized in Table 1. The same preparation in the presence of 5 mM Ca2+ (Fig. 1D) exhibits a different amide I band shape, which is not due to major changes in the wavenumber (position) of the components, but in their relative areas, the random + helix fraction increasing at the expense of the β-sheet, turns remaining constant on average (Table 1). This would be compatible with the idea of Ca2+ inducing a partial β-sheet unfolding. In turn, this would explain previous data according to which Ca2+ increased binding of the hydrophobic probe ANS to SCR (18). 5 mM is a Ca2+ concentration that allows full scramblase activity (transbilayer lipid motion) (17).

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

A preliminary analysis of T effects was carried out by examining the changes in the amide I band full width at half-height (FWHH) (Fig. 3). This parameter usually increases with T and particularly so during thermal protein denaturation (21). For SCR (in the presence of detergent and in the absence of Ca2+) (Fig. 3) the width at halfheight increases slowly and gradually between 4 and ≈ 40 ºC. A transition period follows, in which the slope of the curve increases, until it reaches a maximum value above 65 ºC. Then the bandwidth increases steeply with T, while the protein is probably undergoing thermal denaturation. For the truncated mutant lacking the putative transmembrane domain SCRΔ, under the same conditions, i.e. presence of detergent and absence of Ca2+, FWHH at the lower temperatures is narrower than for the native SCR. However, WHH increases steeply above ≈ 25 ºC and reaches a maximum at ≈ 65 ºC. This is probably showing a gradual process of thermal denaturation along that wide T range. In turn this could be due to an intrinsic low stability of the truncated protein. When SCRΔ is examined in the absence of detergent (Fig. 3) FWHH values are higher than those measured with detergent at all temperatures, in agreement with the spectra described above (compare Fig. 2A and 2D). However FWHH increases only slightly with T in the 4-80 ºC range (Fig. 3). These preliminary observations are confirmed and amplified when the individual spectra are examined. The spectra for SCR (detergent, no Ca2+) (Fig. 1 A-C and Table 1) show little if any difference between 20 and 37 ºC, while the spectrum at 80 ºC is compatible with a protein structure in the process of thermal denaturation. At 80ºC the major features are (i) the shifts in the components related to extended structures (1627 and 1680 cm-1), in turn responsible for the increased FWHH, (ii) the increase in the proportion of unordered components (1643 cm-1), and (iii) in particular a marked decrease in the proportion of the 1630/1627 β-sheet component (Table 1). When the SCRΔ spectra are considered, gradually increasing temperature in the absence of Ca2+ has some effects on the amide I band between 20º C and 37º C (Fig. 2 D, E), with a rearrangement of components in the 1640-1675 cm-1 region that modifies the high-frequency arm

MA

Spectra of the truncated mutant SCRΔ obtained under similar conditions are shown in Fig. 2, and the corresponding quantitative data are collected in Table 2. In the absence of Ca2+ (Fig. 2D) the mutant spectrum is more complex than the wild-type (Fig. 1A). The main structural component is still β-sheet, unordered structures and β-turns are also detected, but in addition two new component bands appear at 1640 cm-1 and 1658 cm-1. They can be assigned to flexible open loops (25), and to α-helix respectively. The IR spectrum of SCRΔ in the presence of Ca2+ was examined next (Fig. 2G). In this case 20 mM Ca2+ was used because the affinity of the mutant for this cation was greatly diminished (17). As seen with the native SCR, Ca2+ does not change the number or position of the secondary structure components but their relative proportions are modified (Table 2), the non-periodic/random components increasing at the expense of β-sheet.

AC CE P

TE

D

3.2. Detergent effects on SCRΔ. Native SCR requires the presence of detergent to remain in solution, as expected from an integral membrane protein. In our study, detergent was 5 mM Zwittergent 3-12. The surfactant effect on SCR structure can only be guessed at, since SCR is not stable in solution in its absence. However the truncated mutant SCRΔ can exist homogeneously dispersed both in the presence and absence of 5 mM Zwittergent 3-12. Detergent effects can be deduced by comparing Figs. 2A and 2D (Table 2). At 20 ºC the overall spectra are clearly different. This is mainly due to the shifts in bands corresponding to extended structures. The 1630 cm-1 band in the presence of detergent, assigned to a “classical” β-sheet is shifted in its absence to 1623 cm-1, a frequency commonly found in either amorphous (26) or ordered (27) protein aggregates. The increased high-frequency component at 1679/1680 cm-1 in the absence of detergent is also typically more prominent in protein aggregates.

3.3. Temperature effects. In order to assay the effects of temperature on the SCR and SCRΔ structures, spectra were recorded in the 4-80 ºC range. Representative spectra at 20, 37, and 80 ºC are shown in Figs. 1, 2, and the corresponding numerical values are summarized in Tables 1, 2. 4

ACCEPTED MANUSCRIPT

SC

RI

PT

observed in the intensities, but not positions, of the band components. The 1649/1651 cm-1 (α-helix) signal area decreases from 26 to 16 %, while the 1642/1643 cm-1 (flexible open loops) increases from 14 % to 18 %, and the predominant β-sheet at 1631 cm-1 goes up from 32 % to 37 %. Between 37 ºC and 80 ºC (Fig. 2 I) further changes in percent areas are observed: the 1639/1643 cm-1 signal almost vanishes, while the β-sheet component at 1630/1631 cm-1 increases its width and relative area from 37 % to 49 %.

NU

3.5. 2D IR correlation spectroscopy The use of 2D-IR correlation spectroscopy has been proposed recently for the study of proteins. In this procedure an external perturbation is correlated with the dynamic response of the protein‚ to increase the amount of information obtained from the infrared spectrum (28). Proteins are a good target for this method‚ since changes induced by an external perturbation can be studied in more detail than with the conventional infrared. This approach‚ essentially similar to 2D-NMR spectroscopy‚ uses correlation analysis of the dynamic fluctuations caused by an external perturbation to enhance spectral resolution without assuming any line-shape models for the bands. The perturbation can be achieved through changes in temperature, or by the presence of lipids, or other external ligands. The power of the 2D correlation approach results primarily in an increase of the spectral resolution by a dispersal of the peaks along a second dimension that also reveals the time-course of the events induced by the perturbation.

TE

D

MA

of the overall amide I band, the 1630 cm-1 component remaining unaltered. Between 37º C and 80º C however (Fig. 2 E, F) the proportion of α-helix (1658/1654 cm-1) decreases markedly and correspondingly the fraction of non-periodic elements increases, as a direct consequence of the thermal stress. The large β-structure band remains constant in position and relative area. A shift and an increase in the 1664-1668 cm-1 component contribute to the increase in amide I bandwidth (Fig. 3). The spectrum of SCRΔ at 20 ºC in the absence of detergent was described above as reminiscent of an aggregated protein. Hardly any changes are detected between 20 and 37 ºC (Fig. 2A-B). In the 37-80 ºC range (Fig. 2B-C) most of the band components remain in position, except for a shift of the β-sheet signal (1624 to 1621 cm1 ), accompanied by a decreased area. In addition, the proportion of α-helix (1653/1654 cm-1) decreases and correspondingly the fractions of non-periodic elements (1640/1641 cm-1) and flexible loops (1666/1667 cm-1) are increased. These changes are compatible with the observed small thermal increase in FWHH (Fig. 3).

AC CE P

3.4. Combined T and Ca2+ effects. Thermal effects are in general influenced by the presence of Ca2+. For native SCR (in detergent) in the presence of 5 mM Ca2+ (Fig. 1 D, E) important changes are observed when the protein is heated from 20 to 37 ºC (in contrast with the invariance of spectra in Fig. 1 A, B in the absence of Ca2+): the fractional area of the unordered components (1648 cm-1) increases, and the β-sheet area decreases in about the same proportion (Table 1). When T is increased to 80 ºC (Fig. 1F) a novel band component appears at 1669 cm-1, attributed to β-turns. The high-frequency extended sheet signal at ≈ 1679 cm-1 increases gradually with temperature, contributing to the increased band width. When the 80 ºC SCR spectra with and without Ca2+ are compared (Fig. 1 C, F) the same spectral components are seen, but in the presence of 5 mM Ca2+ the α-helix signal (1658/1655 cm-1) is more prominent than in its absence, and the reciprocal is true of the 1669/1667 cm-1 turns. When SCRΔ, in the presence of detergent and of 20 mM Ca2+, is gradually heated between 20 and 37 ºC (Fig. 2 G, H) marked changes are 5

Correlations between bands are found through the so-called synchronous and asynchronous contour maps that correspond to the real and imaginary parts of the cross-correlation of spectral intensity at two wavenumbers. In a synchronous 2D map, the peaks located along the diagonal (autopeaks) correspond to changes in intensity induced (in this case) by temperature and they are always positive. The cross peaks indicate an in-phase relationship between the two bands involved, i.e. that two vibrations of the protein characterized by two different wavenumbers and are being affected simultaneously. Asynchronous maps show out-ofphase correlations between cross peaks and give an idea of the sequential order of events produced

ACCEPTED MANUSCRIPT

PT

4. DISCUSSION

SC

RI

The above results increase our understanding of the influence of Ca2+, detergent, and of the putative C-terminal transmembrane domain on the overall SCR structure and its secondary structure components. In addition, certain aspects of spectral interpretation deserve some discussion. 4.1. Band assignments. Band assignment in IR spectroscopy is not always a straightforward procedure, because the secondary structure components found in real proteins do not always correspond to the ideal canonical ones (21). Assignment is sometimes helped by comparing the spectra of a given protein under different conditions. In the spectrum of native SCR at 20 ºC we find a band at 1647 cm-1 borderline between α-helix and unordered components. It can be either a pure band or a mixture of two. A band component arising purely from unordered (non-periodic) structures is not expected to change in position with temperature, but the opposite would happen if a structured component was present. What we observe (Fig. 1 A-C) is that the 1647/1646 cm-1 component at 20 and 37 ºC is split at 80 ºC between a large 1643 cm-1 and a smaller 1655 cm-1 band, positions that are commonly assigned to unordered and α-helix respectively. It appears as if these two rather large bands were not resolved at the lower temperatures. In the SCRΔ mutant though (in the presence of detergent) the 1649 and 1658 cm-1 band maxima are separate enough to allow resolution even at 20 ºC (Fig. 2 D). Note that these two bands are narrower than the apparently composite band at 1647 cm-1 in Fig. 1 A. Thus in this case the Tinduced splitting helps in assigning the 1647 cm-1 band to amalgamated α-helix and non-periodic structures. A further assignment that merits discussion is that of the 1627/1628 cm-1 signal of SCR at 80 ºC, both with (Fig. 1 F) and without

AC CE P

TE

D

MA

Fig. 4 shows the different synchronous maps, and the corresponding asynchronous maps are shown in the Supplementary Material (Fig. S2). In the present study, perturbation has been achieved by temperature. Figures 4A and 4B show respectively native SRC and SRCΔ. It can be seen from the autopeaks that the frequencies of the main peaks are very similar in both cases, as observed in Figs. 1 and 2. Changes in intensity of the major peaks indicate that the two proteins exhibit differences in their sensitivity to temperature, indicating a different conformation. Quantitation of peak intensity in 2D spectra has not been reliably achieved yet, but the overall spectral shape suggests that both proteins have a similar flexibility, with a higher intensity of the peak around 1620 cm-1 in the native protein. The asynchronous spectra also show these differences and indicate that the process of thermal unfolding is different in both proteins.

conformation it could be assumed that in the absence of detergent, the truncated mutant protein is less stable.

NU

by the perturbation, i.e. an asynchronous peak is produced if the vibrations of the functional groups corresponding to the varying wavenumbers change each at a different time. The asynchronous peak will be positive if the change in one of two bands occurs earlier than in the second band, and negative in the opposite case.

In the presence of calcium the observed behavior is again close to what is seen by curve fitting, and the autopeaks are similar in number, although in this case the peak positions in the maps exhibit some shifts. At first glance, it can be seen that in the native protein the changes in autopeak intensity are lower in the 1680 cm-1 peak, this would point to a more compact protein in the presence of Ca2+. In the case of the mutant, a similar pattern is obtained pointing also to a more stable system. In the case of the mutant with or without detergent, a different pattern is obtained in the 2D IR maps in the absence of detergent when compared to the presence of detergent, since in the absence of detergent, a simpler map is observed. This is probably due to the width of the components and would correlate with the bandshape that is shown in Fig. 2. In terms of 6

ACCEPTED MANUSCRIPT

NU

SC

RI

PT

SCRΔ (Fig. 2D) is the same as the 1661 cm-1 in SCR (Fig. 1A). The spectral (Figs. 1, 2) and functional (17) data in combination point to a rearrangement of the protein tertiary structure with a decreased overall stability upon removal of the C-terminal 28 aa residues containing the transmembrane peptide. A further interesting observation is that the SCRΔ mutant at 20 ºC presents already the six components that are visible for SCR only at 80 ºC (Fig. 2 D and 1 C respectively). The doublets 1640/1649 and 1658/1669 (SCRΔ, 20 ºC) correspond to the doublets 1643/1655 and 1667/1680 (SCR, 80 ºC). The general shift towards higher wavenumbers at 80 ºC is indicative of a higher thermal mobility. The effect induced by T on SCR is very similar to that caused by deleting the transmembrane domain. This is probably an indication that in both cases the protein is being partially denatured/destabilized, although perhaps through different pathways.

TE

D

MA

Ca2+ (Fig. 1 C). Protein aggregation is detected by IR spectroscopy in the form of extended structure signals at wavenumbers below 1625 cm-1 (26, 27). In proteins containing a major α-helical component under native conditions the appearance of these extended structures at 1617-1625 cm-1 upon denaturation is clearly visible (29). However in those proteins, as in SCR, containing a large fraction of β-structure when native, aggregation may give rise either to re-ordered extended structures, or to components shifted in position due to changes in their environment (18). An example of intermolecular reorganization is found in amyloid fiber formation (30). For native SCR the overall change in Amide I shape would point to a structural reorganization. This would be in agreement with a thermal denaturation process leading ultimately to exposure of internal hydrophobic regions and extensive intramolecular rearrangement. A more common form of extended β-structures found in aggregating proteins is seen at 1621/1624 cm-1 for SCRΔ in the absence of detergent (Fig. 2 A-C).

AC CE P

4.2. Effect of removing the C-terminal end. The mutant SCRΔ lacks the C-terminal 28 aa residues (including the transmembrane domain) and is unable to catalyze the transbilayer lipid flipflop motion (16,17). From the point of view of structure it appears that some of the bands that putatively overlap in SCR (Fig. 1 A, see Discussion above) are resolved in SCRΔ (Fig. 2 D). This is the case of the wide 1647 cm-1 band in SCR that gives rise to 1640 cm-1 + 1649 cm-1 bands in SCRΔ (Fig. 2D), and the component at 1661 cm-1 in SCR (Fig. 1A) that splits into narrower signals at 1658 + 1669 cm-1 in SCRΔ (Fig. 2D). Perhaps in the mutant protein some peptides can develop intramolecular interactions that decrease their motion, thus decreasing the widths of the corresponding band components. It is also possible that in the new environment SCRΔ band components are shifted with respect to their counterparts in SCR and this improves separation of components. Concomitantly the percent area of the β-sheet component decreases from ≈ 50 % (native) to ≈ 43 % (mutant). This may have been compensated by the increase in the 1640 + 1649 cm-1 bands in SCRΔ (≈ 30 % together) (Fig. 2D) as compared with the 1647 cm-1 band in SCR (≈ 20 %). The sum area of 1658 + 1669 cm-1 band in 7

4.3. Effects of Ca2+. SCR exists usually in the cytoplasm i.e. with Ca2+ concentrations in the micromolar range, and only becomes active when Ca2+ levels are increased by 1000-fold, e.g. in response to cell injury, apoptosis, or cell activation (1,3,7,17). Thus the 0 and 5 mM Ca2+ used in the present study correspond to the two physiological SCR conformations, respectively inactive and active. The main effect of Ca2+ on native SCR at 20 ºC appears to be a decrease in the proportion of βsheet (Fig. 1 A, D, Table 1), accompanied by an increase in the composite band at 1647 cm-1, other spectral components remaining more or less invariant. The Ca2+ effect on native SCR can be clarified by examining the spectra at high temperature, specifically at 80 ºC. As discussed above, T induces the splitting of the 1647 cm-1 signal into its two components, namely nonperiodic structures (1643 cm-1) and α-helix (1655 cm-1). Addition of Ca2+ to SCR at 80 ºC (Fig. 1 C, F), hardly modifies the extended β-structures (probably not canonical β-sheet any longer, see above) but increases the α-helical component, leaving unchanged the proportion of unstructured protein. The idea that Ca2+ increases the proportion of α-helical components is reinforced by the observations on SCRΔ (Fig. 2 D, G, Table 2). In

ACCEPTED MANUSCRIPT

AC CE P

TE

D

PT

RI

SC

MA

4.4. Effects of detergent. Zwittergent 3-12, or lauryl sulfobetaine, is N – dodecyl – N , N – dimethyl – 3 – ammonio – 1 -propanesulfonate, a surfactant that finds extensive use in biological and non-biological applications (31, 32). SCR could only be kept in a homogeneous dispersion in the presence of detergents, previous experiments had shown that 5 mM Zwittergent 3-12 was optimal for that purpose (5). The spectra in Fig. 1 were all taken in the presence of detergent, and most studies on SCRΔ were also performed in Zwittergent 3-12 for a better comparison of the native and truncated proteins (17). For this reason the spectra in Fig. 2 D-I were recorded in detergent. However SCRΔ can be homogeneously dispersed in the absence of detergent, and in fact the spectra shown in Fig. 2A-C were recorded with no detergent and can be used to observe the detergent effects on SCRΔ (compare Fig. 2 A and D, or B and E, or C and F). A similar comparison could not be carried out with native SCR because it cannot be isolated nor kept stably in dispersions without added surfactants. Detergent appears to have two main effects on the SCRΔ spectrum (Fig. 2 A, D). One is the large (7 cm-1) shift of the β-structure signal at 1623/1630 cm-1. β-structures below 1625 cm-1 are usually found in aggregated proteins, however this does not seem to be the case in SCRΔ because the position of that large band is sensitive to both detergent and temperature, and this is usually not the case with protein denaturation signals (26, 33). A second effect is the splitting of the 1653 cm-1 band into two components at 1649 cm-1 and 1658 cm-1. The phenomenon of band splitting and overlapping in this protein has been discussed above. The effects of temperature on SCRΔ are also different with and without detergent, the

change in spectral width being clearly smaller in the absence of detergent, in which case the amide I width is considerably larger even at 20 ºC. Note that the spectral components in Fig. 2 A (no detergent, 20 ºC) and 2 E (detergent, 37 ºC) are very similar. This suggests that the same effect, i.e. structural relaxation is occurring when SCRΔ in the presence of detergent at 20 ºC (Fig. 2 B) is subjected to either heating to 37 ºC or to detergent removal. In general SCRΔ + detergent has features in common with the native SCR, e.g. overall bandwidth, position of the main β-band, or percent area occupied by the α + unordered structures (Figs. 1 A, 2 D). This is probably an indication that the detergent is providing a substitute amphipathic environment to those hydrophobic regions in the protein that have been left exposed to the water medium as a result of the C-end peptide removal. A similar phenomenon was observed for another type II protein, TrwB, when the membrane anchor was deleted (34, 35).

NU

the truncated mutant non-periodic and α-helical components are resolved even at 20 ºC, as discussed above. In this case, as with native SCR, Ca2+ causes a decrease in β-sheet that, at least for SCR, is accompanied by an increase in the αhelical signal at 1658 and 1649 cm-1, respectively in the absence and presence of Ca2+. Various evidences concur in showing that, in the presence of Ca2+, SCRΔ at 20 ºC (Fig. 2G) exhibits several properties of the thermally denatured native SCR at 80 ºC (Fig. 1F), namely the proportions of αhelix and β-sheet structures.

8

4.5. Structure-function relationship. Our results provide new insight on the disputed matter of the physiological role of PLSRC1. A major observation is that deletion of the C-terminal helix promotes a decrease in protein stability as can be seen in Fig. 2. This suggests that the helix favors a more compact protein conformation that could be involved in the binding of the calcium ions, bringing these residues closer together. This structural loss due to the lack of the helix hampers the adoption of a functional conformation by PLSCR1 resulting in hindrance of the flip-flop activity. These results are in agreement with previous reports that demonstrate the increased affinity of the mutant for ANS as a consequence of the exposure to the aqueous environment of otherwise inner hydrophobic regions (17). Our results are in agreement with Arashiki et al. (36) who showed that in human RBC, PLSCR1 is the predominant scramblase, and that its activity resides in the transmembrane domain. The authors also show that PLSCR1 formed an oligomer, suggesting that it functions by producing a cluster with an integral

ACCEPTED MANUSCRIPT

TE

D

RI

SC NU

MA

A further interesting observation in this context is that the SCRΔ mutant at 20 ºC presents already the six secondary-structure components that are visible for SCR only at 80 ºC (Fig. 2 D and 1 C respectively). The doublets 1640/1649 and 1658/1669 (SCRΔ, 20 ºC) correspond to the doublets 1643/1655 and 1667/1680 (SCR, 80 ºC). The general shift towards higher wavenumbers at 80 ºC is indicative of a higher thermal mobility. The effect induced by temperature (T) on SCR is very similar to that caused by deleting the transmembrane domain. This is probably an indication that in both cases the protein is being partially denatured/destabilized, although perhaps through different pathways.

AC CE P

5. CONCLUSIONS. 1. The main secondary structure component of SCR is β-sheet.

9

2. Removing the C-terminal 28 aa residues, i.e. the putative transmembrane domain, leads to an overall tertiary structure change with increased structural flexibility, and only quantitative changes in the secondary structure components. 3. Ca2+ causes a variation in the equilibrium between α- and β-structures, and an increase in the proportion of nonperiodic/unordered components. 4. Adding detergent has the effect of reducing the degree of structural flexibility in the protein, somewhat opposite to the effect of increasing temperature. 5. Thermal stress (heating to 80 ºC) gives rise to marked changes in the protein conformation, whose nature depends in turn on the presence of detergent, Ca2+, or the C-terminal domain. 6. The secondary structure of the mutant at 20°C is rather similar to that of the wild type found at 80°C, i.e. upon thermal denaturation. This implies that the C terminus stabilizes somehow the secondary structure.

PT

“hole” that allows the passage of phospholipid hydrophilic head regions.

ACCEPTED MANUSCRIPT

PT

Acknowledgments: LSM was supported by the JAEdoc grant from the Spanish National Research Council (CSIC), NA was a predoctoral student supported by the Basque Government. This work was supported by grant IT 849-13 from the Basque Government and by FEDER/MINECO grant Nº BFU 2015- 66306- P.

RI

Conflict of interest: The authors declare that they have no conflicts of interest with the contents of this article.

AC CE P

TE

D

MA

NU

SC

Author contribution: NA and LSM conducted the experiments, analyzed the results and took part in the writing of the manuscript. MGP helped with IR measurements. FMG and JLRA conceived the idea and wrote the paper.

10

ACCEPTED MANUSCRIPT REFERENCES

PT

1. Basse, F., Stout, J. G., Sims, P. J., and Wiedmer, T. (1996) Isolation of an erythrocyte membrane protein that mediates Ca2+-dependent transbilayer movement of phospholipid. J. Biol. Chem. 271, 17205-17210

RI

2. Wiedmer, T., Zhou, Q., Kwoh, D. Y., and Sims, P. J. (2000) Identification of three new members of the phospholipid scramblase gene family. Biochim. Biophys. Acta 1467, 244-253

SC

3. Sahu, S. K., Gummadi, S. N., Manoj, N., and Aradhyam, G. K. (2007) Phospholipid scramblases: an overview. Arch. Biochem. Biophys. 462, 103-114

NU

4. Suzuki, J., Umeda, M., Sims, P. J., and Nagata, S. (2010) Calcium-dependent phospholipid scrambling by TMEM16F. Nature 468, 834-838

MA

5. Stout, J. G., Zhou, Q., Wiedmer, T., and Sims, P. J. (1998) Change in conformation of plasma membrane phospholipid scramblase induced by occupancy of its Ca2+ binding site. Biochemistry 37, 14860-14866

TE

D

6. Bateman, A., Finn, R. D., Sims, P. J., Wiedmer, T., Biegert, A., and Soding, J. (2009) Phospholipid scramblases and Tubby-like proteins belong to a new superfamily of membrane tethered transcription factors. Bioinformatics. 25, 159-162

AC CE P

7. Sahu, S. K., Aradhyam, G. K., and Gummadi, S. N. (2009) Calcium binding studies of peptides of human phospholipid scramblases 1 to 4 suggest that scramblases are new class of calcium binding proteins in the cell. Biochim. Biophys. Acta 1790, 1274-1281 8. Sun, J., Nanjundan, M., Pike, L. J., Wiedmer, T., and Sims, P. J. (2002) Plasma membrane phospholipid scramblase 1 is enriched in lipid rafts and interacts with the epidermal growth factor receptor. Biochemistry 41, 6338-6345 9. Zhou, Q., Ben-Efraim, I., Bigcas, J. L., Junqueira, D., Wiedmer, T., and Sims, P. J. (2005) Phospholipid scramblase 1 binds to the promoter region of the inositol 1,4,5-triphosphate receptor type 1 gene to enhance its expression. J. Biol. Chem. 280, 35062-35068 10. Kuo, Y. B., Chan, C. C., Chang, C. A., Fan, C. W., Hung, R. P., Hung, Y. S., Chen, K. T., Yu, J. S., Chang, Y. S., and Chan, E. C. (2011) Identification of phospholipid scramblase 1 as a biomarker and determination of its prognostic value for colorectal cancer. Mol. Med. 17, 41-47 11. Kusano, S. and Eizuru, Y. (2013) Interaction of the phospholipid scramblase 1 with HIV-1 Tat results in the repression of Tat-dependent transcription. Biochem. Biophys. Res. Commun. 433, 438-444 12.

Wiedmer, T., Zhao, J., Nanjundan, M., Sims, P.J. (2003) Palmitoylation of phospholipid scramblase 1 controls its distribution between nucleus and plasma membrane. Biochemistry 42,1227-1233.

13. Zhao, J., Zhou, Q., Wiedmer, T., Sims, P.J. (1998) Palmitoylation of phospholipid scramblase is required for normal function in promoting Ca2+-activated transbilayer movement of membrane phospholipids. Biochemistry 37, 6361-6366.

ACCEPTED MANUSCRIPT 14. Zhou, Q., Zhao, J., Stout, J.G., Luhm, R.A., Wiedmer, T., Sims, P.J. (1997) Molecular cloning of human plasma membrane phospholipid scramblase. A protein mediating transbilayer movement of plasma membrane phospholipids. J Biol Chem. 272, 18240-18244.

PT

15. Sims, P.J., Wiedmer, T. (2001) Unraveling the mysteries of phospholipid scrambling. Thromb. Haemost. 86, 266-275.

RI

16. Francis, V. G., Mohammed, A. M., Aradhyam, G. K., and Gummadi, S. N. (2013) The single Cterminal helix of human phospholipid scramblase 1 is required for membrane insertion and scrambling activity. FEBS J. 280, 2855-2869

NU

SC

17. Sanchez-Magraner, L., Posada, I. M., Andraka, N., Contreras, F. X., Viguera, A. R., Guerin, D. M., Arrondo, J. L., Monaco, H. L., and Goni, F. M. (2014) The C-terminal transmembrane domain of human phospholipid scramblase 1 is essential for the protein flip-flop activity and Ca²⁺ -binding. J. Membr. Biol. 247, 155-165

Posada, I.M., Busto, J.V., Goñi, F.M. and Alonso, A. (2014) Membrane binding and insertion of the predicted transmembrane domain of human scramblase 1. Biochim. Biophys. Acta 1838, 388397.

D

19.

MA

18. Posada, I. M., Sanchez-Magraner, L., Hervas, J. H., Alonso, A., Monaco, H. L., and Goni, F. M. (2014) Membrane binding of human phospholipid scramblase 1 cytoplasmic domain. Biochim. Biophys. Acta 1838, 1785-1792

TE

20. Posada, I. M., Fantini, J., Contreras, F. X., Barrantes, F., Alonso, A., and Goni, F. M. (2014) A cholesterol recognition motif in human phospholipid scramblase 1. Biophys. J. 107, 1383-1392

22.

AC CE P

21. Arrondo, J. L. R. and Goñi, F. M. (1999) Structure and dynamics of membrane proteins as studied by infrared spectroscopy. Prog. Biophys. Mol. Biol. 72, 367-405. Turnay, J., Olmo, N., Gasset, M., Iloro, I., Arrondo, J.L., Lizarbe, M.A. (2002) Calciumdependent conformational rearrangements and protein stability in chicken annexin A5. Biophys J. 83, 2280-2291.

23. Sanchez-Magraner, L., Cortajarena, A. L., Garcia-Pacios, M., Arrondo, J. L., Agirre, J., Guerin, D. M., Goni, F. M., and Ostolaza, H. (2010) Interdomain Ca(2+) effects in Escherichia coli alphahaemolysin: Ca(2+) binding to the C-terminal domain stabilizes both C- and N-terminal domains. Biochim. Biophys. Acta 1798, 1225-1233 24. Barth, A. (2000) The infrared absorption of amino acid side chains. Prog. Biophys. Mol. Biol. 74, 141-173 25. Fabian, H., Naumann, D., Misselwitz, R., Ristau, O., Gerlach, D., and Welfle, H. (1992) Secondary structure of streptokinase in aqueous solution: a Fourier transform infrared spectroscopic study. Biochemistry 31, 6532-6538 26. Garcia-Pacios, M., Fernandez-Higuero, J. A., de la Arada, I., and Arrondo, J. L. R. (2008) Protein stability studied by Infrared Spectroscopy. Biotechnol. &Biotechnol. Eq. 22, 625-628 27. de la Arada, I., Andraka, N., Pacios, M. G., and Arrondo, J. L. (2011) A conventional an 2DCOS infrared approach to the kinetics of protein misfolding. Curr. Protein Pept. Sci. 12, 181-187 28. Noda, I., Ozaki, Y. (2004) Two-dimensional correlation spectroscopy. Wiley, Chichester.

ACCEPTED MANUSCRIPT 29. Iloro, I., Goni, F. M., and Arrondo, J. L. R. (2005) A 2D-IR study of heat- and [(13)C]ureainduced denaturation of sarcoplasmic reticulum Ca(2+)-ATPase. Acta Biochimica Polonica 52, 477-483

PT

30. de la Arada, I. and Arrondo, J. L. R. (2007) An Infrared Study of Fibril Formation in Insulin from Different Sources. In Pifat-Mrzljak, G., editor. Supramolecular Structure and Function, Springer, Dordrecht

RI

31. Chowdhury, E. H. and Akaike, T. (2005) Rapid isolation of high quality, multimeric plasmid DNA using zwitterionic detergent. J. Biotechnol. 119, 343-347

NU

SC

32. Martin-Garcia, B., Velazquez, M. M., Rossella, F., Bellani, V., Diez, E., Garcia Fierro, J. L., Perez-Hernandez, J. A., Hernandez-Toro, J., Claramunt, S., and Cirera, A. (2012) Functionalization of reduced graphite oxide sheets with a zwitterionic surfactant. Chemphyschem. 13, 3682-3690

AC CE P

TE

D

MA

33. Lefevre, T., Arseneault, K., and Pezolet, M. (2004) Study of protein aggregation using twodimensional correlation infrared spectroscopy and spectral simulations. Biopolymers 73, 705-715 34. Segura, R. L., Aguila-Arcos, S., Ugarte-Uribe, B., Vecino, A. J., de la Cruz, F., Goni, F. M., and Alkorta, I. (2013) The transmembrane domain of the T4SS coupling protein TrwB and its role in protein-protein interactions. Biochim. Biophys. Acta 1828, 2015-2025 35. Segura, R. L., Aguila-Arcos, S., Ugarte-Uribe, B., Vecino, A. J., de la Cruz, F., Goni, F. M., and Alkorta, I. (2014) Subcellular location of the coupling protein TrwB and the role of its transmembrane domain. Biochim. Biophys. Acta 1838, 223-230. 36. Arashiki, N., Saito, M., Koshino, I., Kamata, K., Hale, J., Mohandas, N., Manno, S., Takakuwa, Y. (2016) An Unrecognized Function of Cholesterol: Regulating the Mechanism Controlling Membrane Phospholipid Asymmetry. Biochemistry 55, 3504-3513.

Table 1. Secondary structure components of native SCR solubilized in detergent in the presence or absence of 5 mM calcium. Data derived from the amide I band of infrared spectra recorded in D 2O medium. The spectra are shown in Fig. 1. Average values of two closely similar experiments. Table 2. Secondary structure components of SCRΔ in buffer or in detergent and in the presence or absence of 20 mM calcium. Data derived from the amide I band of infrared spectra recorded in D 2O medium. The spectra are shown in Fig. 2. Average values of two closely similar experiments.

PT

ACCEPTED MANUSCRIPT

Native SCR, no Ca

1647

α-helix Non-periodic structures

1647

1630 1613

β-sheet (high frequency components) Aminoacids side chains

1630 1613

1646

18.8

50

1630 1612

45.1

MA

19.7

PT ED

Native SCR, 5mM Ca

Mean Position 1681 1674 1663

Assignment β-sheet (low frequency components)

Turns, intrinsically disordered structures

α-helix Non-periodic structures

1633 1612

β-sheet (high frequency components) Aminoacid side chains

AC

Table 1

CE

1648

37ºC Position % Area 1678 2.3 1660 33.7

% Area 2 28.2

SC

Position 1678 1661

NU

Assignment β-sheet (low frequency components) Turns, intrinsically disordered structures

RI

20ºC Mean Position 1678 1661

2+

20ºC

Position 1681 1674 1663

% Area 1.9 1.5 27.4

1648 1633

80ºC Position 1680 1667 1656 1643 1628

% Area 7.8 14.8 15.4 26.9 30.2

2+

37ºC Position % Area 1677 7.1 1662

26.2

30.7

1647

36.3

36.9

1632 1612

30.4

80ºC Position 1680

% Area 8.2

1670 1659 1645 1629

6.4 25.1 26.4 33

ACCEPTED MANUSCRIPT

2+

1679

β-sheet (low-frequency components)

1679

1667 1653

1667 1653

1623

Turns, intrinsically disordered structures α-helix loops β-sheet (high-frequency components)

1604

Aminoacid side chains

Mean Position

1640 1623

Position

% Area

4.6

1679

6.4

1681

3.9

11.6 20.8 17.7 45.4

1666 1653 1640 1624

12.3 23 19.3 39.5

1667 1654 1641 1621

17.8 11.7 29.1 37.5

-

1605

-

1604

-

Position

% Area

Position

% Area

0.3 7.1

1679

1 20.9

NU

1604

2+

SCRΔ, detergent, no Ca 20ºC Position % Area

Assignment

80ºC % Area

MA

1640

37ºC Position

RI

Assignment

SC

Mean Position

PT

SCRΔ, no detergent, no Ca 20ºC Position % Area

37ºC

80ºC

1679

0.6

1683 1675

Turns, intrinsically disordered structures

1669

13.5

1665

10.9

1668

1659

α-helix

1659

12.6

1654

19.3

1659

6.3

1649

Non-periodic structures

1649

19.2

1644

17.1

1649

23.8

1641

10.8

-

-

1639

2.7

1630

43.2

1630

45.3

1630

45.3

β-sheet (low-frequency components)

1669

PT ED

1679

Loops

1630

β-sheet (high-frequency components)

AC

CE

1641

2+

SCRΔ, detergent, 20 mM Ca 20ºC

37ºC

80ºC

Mean Position

Assignment

Position

% Area

Position

% Area

Position

% Area

1680

β-sheet (low-frequency components)

1680

1.5

1679

2.9

1680

2.6

1661

Turns, intrinsically disordered structures

1672

9.6

1670

11.2

1661

17.6

1660

14.8

1669

14.2

-

-

-

-

-

-

1659

10.6

1650

α-helix

1650

25.6

1652

16.3

-

-

-

-

-

-

-

-

1649

21.2

1641

loops

1641

13.7

1644

18.3

1639

2.4

1631

β-sheet (high-frequency components)

1631

32

1631

36.5

1631

48.

Table 2

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

SC

RI

PT

FIGURE LEGENDS SCHEME 1. Sequence alignment of SCR (NCBI Accession Number ABM81805.1) and the truncated form SCRΔ. Below, a diagram summarizing the known domains within the protein. The stars indicate Trp residues. FIGURE 1. Amide I region of the infrared spectra of native SCR solubilized in detergent, in the presence or absence of 5 mM calcium. SCR in detergent without Ca+2, (A) 20ºC, (B) 37ºC, (C) 80ºC. SCR in detergent with 5 mM Ca+2, (D) 20ºC, (E) 37ºC, (F) 80ºC. The spectra were curve-fitted to show the different secondary structure components as detailed in Table 1. FIGURE 2. Amide I region of the infrared spectra of SCRΔ. (A) SCRΔ, no detergent, no Ca+2, (A) 20ºC, (B) 37ºC, (C) 80ºC. SCRΔ in detergent, no Ca+2, (D) 20ºC, (E) 37ºC, (F) 80ºC. SCRΔ in detergent + 20 mM Ca+2, (G) 20ºC, (H) 37ºC, (I) 80ºC. The spectra were curve-fitted to show the different secondary structure components as detailed in Table 2 FIGURE 3. Thermal denaturation of SRC and SRCΔ. The full widths at half-height (FWHH) of the amide I bands are plotted as a function of temperature for SCR in detergent (▼), for SCRΔ without detergent (●) and for SCRΔ with detergent (о). No Ca+2 in these samples. FIGURE 4. Synchronous (Φ) 2D-IR correlation map of SRC and SRCΔ. Correlation maps for native and mutant scramblase in the presence/absence of Ca2+/detergent were recorded in the temperature interval 37–80°C. The contour maps do not show small peaks below the established threshold.

16

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Scheme 1

17

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 1

18

AC

CE

PT ED

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 2

19

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 3

20

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 4

21

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

D

GRAPHICAL ABSTRACT

AC CE P

TE

2D-IR correlation map of human phospholipid scramblase 1 showing the predominant β-sheet structures.

22

ACCEPTED MANUSCRIPT Highlights

AC CE P

TE

D

MA

NU

SC

RI

PT

1. The structure of phospholipid scramblase 1 (SCR) has been studied by IR and 2DIR correlation spectroscopies. 2. The main secondary structure component of SCR is β-sheet. 3. Removing the SRC transmembrane domain causes only quantitative changes in the protein secondary structure. 4. Ca2+ causes a variation in the equilibrium between α- and β-structures. 5. Adding detergent has the effect of reducing the degree of structural flexibility in the protein.

23