Membrane lipid order of human red blood cells is altered by physiological levels of hydrostatic pressure

Membrane lipid order of human red blood cells is altered by physiological levels of hydrostatic pressure. Am. J. Physiol. 272 Wear-t Cir-c. Physiol. 41): H538-H543, 1997.-The effect of hydrostatic pressure at levels applied in diving or hyperbaric treatment (thus considered “physiological”) on the order of lipid domains in human red blood cell (RBC) membrane was studied. Membrane order was determined by measuring 1) the fluorescence anisotropy (FAn 1 of lipid probes, 2 > the resonance energy transfer from tryptophan to lipid probes, and 3) spectral shifts in Laurdan fluorescence emission. It was found that the application of mild pressure (~15 atm) 1) increased, selectively, the FAn of lipid probes that monitor the membrane lipid core, 2) in- creased the tryptophan FAn, 3) increased the resonance energy transfer from tryptophan to lipid probes residing in the lipid core, and 4) induced changes in the Laurdan fluorescence spectrum, which corresponded to reduced membrane hydration. It is proposed that the application of pres- sure of several atmospheres increases the phase order of membrane lipid domains, particularly in the proximity of proteins. Because the membrane lipid order (“fluidity”) of RBCs plays an important role in their cellular and rheological functions, the pressure-induced alterations of the RBC mem- brane might be pertinent to microcirculatory d served in human .s subjected to elevated pressure. sorders ob-fluorescence anisotropy; resonance energy transfer; diving; hyperbaric treatment .


41): H538-H543,
1997.-The effect of hydrostatic pressure at levels applied in diving or hyperbaric treatment (thus considered "physiological") on the order of lipid domains in human red blood cell (RBC) membrane was studied. Membrane order was determined by measuring 1) the fluorescence anisotropy (FAn 1 of lipid probes, 2 > the resonance energy transfer from tryptophan to lipid probes, and 3) spectral shifts in Laurdan fluorescence emission. It was found that the application of mild pressure (~15 atm) 1) increased, selectively, the FAn of lipid probes that monitor the membrane lipid core, 2) increased the tryptophan FAn, 3) increased the resonance energy transfer from tryptophan to lipid probes residing in the lipid core, and 4) induced changes in the Laurdan fluorescence spectrum, which corresponded to reduced membrane hydration.
It is proposed that the application of pressure of several atmospheres increases the phase order of membrane lipid domains, particularly in the proximity of proteins. Because the membrane lipid order ("fluidity") of RBCs plays an important role in their cellular and rheological functions, the pressure-induced alterations of the RBC membrane might be pertinent to microcirculatory d served in human .s subjected to elevated pressure. sorders obfluorescence anisotropy; resonance energy transfer; diving; hyperbaric treatment .
PRESSURE OF SEVERAL or even tens of atmospheres is applied in hyperbaric treatment or in diving (in commercial and experimental diving, the diver may reach a depth of 300 m, i.e., 30 atm). This range of pressures is therefore considered "physiological" (12). In the clinic, blood cells are routinely separated by centrifugation, and in research, during experimental procedures, the cells and membranes are often subjected to high-speed centrifugation.
Because the hydrostatic pressure induced in spinning is a function of the angular velocity (11,29), these procedures may exert a pressure of several or tens of atmospheres in the separation of blood cells or hundreds or thousands of atmospheres in various experimental procedures.
Many studies have demonstrated that hydrostatic pressure in the range of hundreds of atmospheres alters cell functions and biochemical processes such as membrane lipid molecular order (17) and phasetransition (29) and protein-receptor dissociation (25 ). Other studies have shown that lower pressure levels, in the range of tens of atmospheres, may affect cellular functions such as platelet aggregation (221, release of neurotransmitters (3 ), and cellular distribution of cytoskeletal and adhesion proteins (12).
Previous studies (12, 26) of the effect of pressure on red blood cells (RBCs) have shown that different R.BC functions may be sensitive to different levels of hydrostatic pressure. For example, ionic regulation in deepsea fish RBCs is modulated by hundreds of atmospheres (261, whereas ion transport and ATP metabolism in human RBCs are affected at tens of atmospheres (9). In previous studies, it was observed that the application of hydrostatic pressure in the range of several atmospheres to RBCs induces changes in the membrane composition, leading to resistance of the cells to hemolysis by phospholipase AZ ( 11) and to enhancement of their aggregability (6).
These pressure-induced changes in the RBC membrane indicate that a mild pressure of several atmospheres may alter its physical properties. This study was undertaken to examine the effect of hydrostatic pressure at physiological levels on lipid order in the human RBC membrane as measured by the fluorescence anisotropy (FAn) of membrane lipid probes and the energy transfer from proteins to the lipid probes. It was found that the application of pressure up to 15 atm induces a substantial increase in membrane lipid order. METHODS Application of pressure.
As previously discussed (11, 291, cells at the bottom of a spinning tube are subjected to hydrostatic pressure that is a function of the angular velocity and the height of the aqueous column in the spinning tube, as expressed by the equation (1) where P,, is the atmospheric pressure, p is the aqueous phase density, 03 is the angular velocity of the spinning rotor, and R and R,, are the distances from the center of rotation to the bottom of the tube and to the air-water meniscus, respectively (R -R,, is thus the height of the aqueous column on the cells). Accordingly, in the present study, pressure was applied to the RBCs by centrifugation in a swinging-bucket rotor at an angular velocity and buffer column corresponding to the desired pressure (for example, a pressure of 10 atm was obtained by spinning at 3,000 rpm under 5.5 cm of buffer in a centrifuge having a radius of 19.5 cm). After the application of pressure, the cells were returned to ambient pressure, and their supernatant was collected. The cells were washed twice with isotonic tris(hydroxymethyl)aminomethane (Tris) buffer at pH 7.4 by centrifugation (10 min) at 300 g under an aqueous column of 0.5 cm (which exerts a pressure of 0.5 atm EFFECT OF HYPERBARIC PRESSURE ON MEMBRANE LIPID ORDER H539 and does not produce a significant pressure effect) and were KCl, and 5 mM TrisHCl) at pH 7.4 and separated by resuspended in fresh Tris buffer. centrifugation, as indicated in Application of pressure. This As shown in Ey. 1, the pressure applied by spinning procedure was repeated three times. depends on the height of the aqueous column on top of the Preparation of RBCghosts. RBCs were subjected to lysis by cells. Therefore, to obtain equal pressure on all the cells, the osmotic shock according to the common procedure. In this amount of RBCs used in these experiments was small enough procedure, the membranes are usually separated from the (0.2 ml of packed RBCs) to form a thin layer at the bottom of intracellular content by high-speed centrifugation. This althe tube, with no significant height difference within the RBC ready exerts a pressure of at least hundreds of atmospheres, layer. To rule out a possible drag or shear effect of the which is obviously undesirable in the present study. To spinning, the RBCs to be pressurized were placed at the circumvent this drawback in the present study, after the bottom of the tube and the buffer was layered on top to the application of osmotic shock, the membranes were separated desired height before spinning.
As in previous studies by by column chromatography. RBCs before or after the applica-Chen et al. (6) and Halle and Yedgar (ll), in the control tion of pressure were lysed in Tris buffer (0.5 mM) containing experiments, the RBCs underwent the same procedure but 5 mM NaCl and 0.15 mM KCl, and the membranes were were centrifuged under an aqueous column of 0.5 cm. This separated from the hemoglobin on a Sephadex G-100 column was not sufficient to induce a significant pressure effect even (1). at the highest speed used for pressure treatment.
This rules Membrane lipid order: Membrane lipid order was deterout the possibility that the effect is due to centrifugation-mined by measuring the FAn of lipid probes inserted into the induced cell-cell contact. To further exclude the effects of drag RBC membranes.
To avoid interference of the hemoglobin or shear or cell-cell contact, in a few experiments, the cells with the fluorescence measurements, the FAn was measured were subjected to hydraulic pressure by applying force di-in ghosts prepared from RBCs before or after the application rectly on a syringe filled with an RBC suspension (free of air of pressure to the intact cells (see Preparation of RBCghosts >. bubbles) with the Indeflator manometer (Advanced Cardiovas-Fhorescence probes. The following fluorescent lipid probes, cular Systems) for the desired time (this system was limited which incorporate into and monitor different regions of the to a maximal pressure of only 10 atm and much less conve-membrane, were used: 1,6-diphenyl-1,3,5-hexatriene (DPH), nient to operate). To separate the cells from the extracellular which incorporates into different apolar regions of the memfluid, the pressure-treated suspension was spun for 10 min at brane (15); the cationic probe l-(4-trimethylammonium)-6a low speed (300 g) and water column (0.5 cm> that do not phenyl-1,3,5-hexatriene (TMA-DPH), where the fluorophore induce a significant effect. These two methods were applied to TMA is located between the upper parts of the fatty acyl the same blood samples and, as previously found (6, ll), chains ( 14); 8-anilino-1-naphthalenesulfonic acid (ANS), which produced the same pressure effect (see Fig. 1).
monitors the interface between the apolar tail and the polar In all experiments, after the application of pressure, the head of phospholipids and may be an indicator of changes in cells were separated from the extracellular medium and surface charge (24); l-acyl-2-[12-(9-anthryl)-ll-trans-dodecsuspended in buffer by gentle tilting of the suspension tube, enoyl] -sn-glycero-3-phosphocholine (APC ), the fluorophore of and all spectroscopic measurements were carried out at which is located in the center of the bilayer perpendicular to ambient pressure. the fatty acyl chains (18); N-1 12-(9-anthryl)-ll-trans-dodec-Preparation of RBC suspension. Blood was drawn from enoyl] -sphingosine-1-phosphocholine (ASM), which is similar human volunteers into EDTA-containing test tubes and to APC but prefers the proximity of membrane proteins (18); centrifuged at low speed (equivalent to 300 g) under an and the fatty acid analog 12-(9-anthryl)-11-trans-dodecanoic aqueous column of 0.5 cm., i.e., under conditions that were acid (AA). Taking into account the extremely slow phosphonot sufficient to induce significant pressure (~0.5 atm). where I,.,. and I\+ are the fluorescence intensities measured with a vertical polarizer (denoted by the first indexes) and an analyzer (denoted by the second indexes) mounted vertically and horizontally, respectively, and G = I&Vh is the correction factor. IhV is intensity measured with horizontal polarizer and vertical analyzer. For each sample, fluorescence was corrected for the scattering by unlabeled ghost membranes.
Resonance energy transfer. Resonance energy transfer (RET) from tryptophan (excited at 280 nm> to theYlipid probes in RBC membranes was determined by the decrease in the fluorescence intensity of the tryptophan emission (330 nm> induced by the presence of the lipid probes, which absorb light in the tryptophan emission range. RET is a measure of the proximity of the emitting tryptophan to the absorbing lipid. Generalized polarization. Generalized polarization (GP> of Laurdan was calculated according to the equation GP = (B -R)/(B + R), where B (blue) and R (red) are the fluorescence maxima characteristic of the gel and liquid-crystalline lipid phases, respectively. This value is a measure of the relative content of these two phases in the membrane (20,21). Figure 1 demonstrates the effect of hydrostatic pressure on the FAn of DPH in RBC membranes isolated after the application of pressure to intact RBCs. As shown in Fig. 1, the anisotropy was increased substantially as a function of both the pressure applied (Fig.  1A) and the duration of treatment (Fig. 1B) and reached a plateau of 50% increase after the application of 15 atm for 1 h. This experiment demonstrates that application of a pressure of several atmospheres to RBCs exerts a considerable increase in membrane lipid order.

RESULTS
To examine a possible effect of pressure on RBC membrane proteins, we determined the FAn of trypto-phan in RBC membranes after the application of pressure up to 15 atm to the intact RBCs. As shown in Fig. lB, the tryptophan FAn is also increased, although to a lesser extent than the FAn of DPH.
To learn about a possible differential pressure effect on the order of different lipid domains in the RBC membranes, we measured the FAn of the lipid probes listed in METHODS, which reside in different lipid regions of the RBC membrane, as well as the energy transfer from tryptophan to these probes. The results, presented in Table 1, show a selective effect. The application of pressure increased the FAn of DPH, which incorporates, with low selectivity, into different apolar regions of the membrane (15); of AA, which distributes between both halves of the bilayer; and of ASM (although to a lesser extent), which incorporates predominantly into the outer leaflet of the membrane in the proximity of proteins (18). This treatment did not significantly affect the FAn of TMA-DPH? ANS, and APC. Table 1 also shows that the pressure treatment induced a significant increase in the tryptophan FAn and increased the RET from tryptophan to those lipid probes in which the FAn values were affected by pressure. This may suggest that the application of pressure preferentially alters the physical state of lipids in the neighborhood of proteins, and these changes are most expressed in the central region of the bilayer.
As noted in the introduction, the application of pressure reduces RBC susceptibility to hemolysis by phospholipase A2 (11) and enhances RBC aggregation (6). This is associated with the release of membrane components into the extracellular fluid and can be reversed by reincubation of pressure-treated cells with the extracellular fluid collected after the application of pressure ("conditioned medium"). In accordance with this, Table 2 shows that incubation of pressure-treated  Values are means t-SD from 3 independent experiments. RBC, red blood cell; RET, resonance energy transfer; DPH, 1,6-diphenyl-1,3,5hexatriene; AA, 12-(9-anthrylk11-~IYZIL+dodecenoic acid; ASM, N-112-(9-anthryl )-ll-fl-n~zs-dodecenoyl]-sphingosine-l-phosphocholine; TMA-DPH, 1-~4-trin~ethylammonium~-6-phenyl-1,3,5-hexatriene; ANS, %anilino-1-naphthalenesulfonic acid; APC, 1-acyl-2-[ 12-(9-anthrylb 1 l-trtrlzs-dodecenoyl] -sr2.-glycero-3-phosphocholine. RBC ghosts were prepared from RBCs that were subjected to a hydrostatic pressure of 15 atm or from control RBCs that were prepared by same procedure but without significant pressure. RBC ghosts were fluorescent labeled as described in AIETIIODS. RET   treated cells with the extracellular fluid collected after the appli cation of pressure (the conditi oned medium). It s eems that the primary effect of the press ure treatment is to induce constitutive changes in the membrane, and these changes are then sensed by the different fluorescent probes.
RBCs in their conditioned medium reduced the FAn of DPH considerably (although not completely). A similar trend was observed with the tryptophan FAn, but because the pressure effect on the tryptophan FAn was considerably smaller than that on the DPH FAn (Fig.  l), the reduction in its FAn was close to the experimental error and not sufficient for quantitative derivation.
The fluorescence spectrum of Laurdan has been shown to be useful for learning about the changes in the relative content of the gel and liquid-crystalline lipid phases in membranes (21), as well as the changes in the membrane cholesterol content rel ative to th e phospholipids (20, 21), which result in reduced membrane hydration (16). In view of the changes in the membrane lipid order, described above, we examined the effect of pressure treatment on the Laurdan spectrum. As shown in Fig. 2, the application of pressure increased the Laurdan peak at the gel phase relative to that at the liquid-crystalline phase. In addition, we found that the Laurdan GP (21) is increased as well after this treatment. This increase in GP indicates that the pressure treatment reduces membrane hydration, and the changes in the excit ation spectru m are suggestive of a relative increase in the membrane cholesterol content, but a full detailed characterization of the changes in the membrane composition is required to confirm this. The changes in the Laurdan spectrum are in agreement with the increased lipid order observed with the other lipid probes (21).
In the present study, we used fluorescent molecules that differentially probe various areas of the membrane. Of special interest are the differences in the fluorescence parameters among different lipid probes carrying the same fluorophore, i.e., between DPH and TMA-DPH (with a rod-like fluorophore), and among AA, APC, and ASM, which have a discoid fluorophore. The latter three probes are oriented in the membrane with their long axis parallel to that of the surrounding fatty acyl chains. Moreover, the anthrylvinyl fluorophore was shown to induce only a minor disturbance in the surrounding lipids when attached to the end of the fatty acyl chain (5). In a homogenous environment, anthrylvinyl-labeled phospholipids with different polar head groups show very close or even identical fluorescence parameters (8). Therefore, the relatively large differences in the RET values of lipid probes with different head groups (APC and ASM) suggest that they distribute differently in the membrane and thus indirectly confirm the existence of lipid domains or microdomains in the membrane.
Noteworthy is the finding that the application of pressure affected the fluorescence parameters of different probes to a different degree. As shown in Table 1, such treatment resulted in a significant increase in both the tryptophan RET to and the FAn of DPH, AA  was used (19), the cells at the bottom layer are subthese parameters remained unchanged. It follows that jetted to relatively higher pressures, which might be the application of pressure at this level may have a sufficient to induce changes in the membrane fluidity. different effect on the various regions of the RBC We suggest that when a biological system is subjected membrane.
This seems to be accompanied by a de-to centrifugation, the possible effects of hydrostatic crease in the an increase membrane water content an in the rel ative cholesterol d possibly by content, as indicated by the changes in the Laurdan fluorescence spectrum observed after the pressure treatment.
The mechanism of this phenomenon may be elucidated by the finding that the pressure induces changes in the membrane composition and in the tryptophan FAn. As noted above, the pressure induces the release of lipid(s) and protein into the extracellular medium. Because membrane proteins do not rotate during the excited-state lifetime of the lipid probes used, it is possible that the application of pressure induces the release of membrane proteins with a lower tryptophan FAn and thus increases the relative content of proteins with a higher tryptophan FAn in the membrane.
induced by a pressure of several atmospheres. This It has been previously shown (29) (6,11) have shown that the pressure-induced changes in RBCs may be reversed to a large extent when the pressuri zed cells are incubated W ith th .eir own conditioned medi urn for a interface are more susceptible to pressure. Moreover, Institute (Bethesda, MD), we examined the effect of considering the shedding of membrane proteins dissaturation diving on the RBCs of human volunteers. It cussed above, it is not unlikely that the primary effect was found that, after 1 h in <2 atm, RBC aggregation of the pressure is on the membrane proteins, and this is membranes by column chromatography, as in the present study, enables the detection of the initial effects of followed by the changes in the lipid composition and pressure when the membrane is most susceptible to pressure. phase change.
Previous studies (4,lO) used fluorescence lipid probes to study RBC ghost lipid order and reported a higher basal level and smaller range of change in FAn than those found in the present study. However, in those studies, RBC ghosts were prepared by high-speed centrifugation.
which alreadv exerts a pressure of over 100 atmispheres.
Thus, under those experimental procedures, the effects of a relatively lower pressure are overlooked.
On the other hand, the separation of RBC