TI - DISCUSSION . AB - In this study , we assessed to what extent Cav1 expression contributes to the formation of biochemically identifiable rafts , such DRMs , NDRs , CO-sensitive domains , and physically ordered domains . Since Cav1 associates with specific membrane lipids and thus may be involved in lipid metabolism ( 17 ) , we started our study with a complete PL analysis of MEFs from WT or Cav1-/- animals . We found a higher proportion of SM and increased abundance of smaller , more saturated PLs in Cav1-/- cells but no difference in membrane-associated cholesterol . Our data conclude that Cav1 expression is not required for the targeting of cholesterol to DRMs , NDRs or CO-sensitive domains , despite the contribution of Cav1 to ordered domains in adherent cells . In summary , our data suggest that Cav1 is involved in the organization of cell architecture -dependent lipid rafts . Genetic deletion of Cav1 increases the proportion of esterified cholesterol without altering total cholesterol levels ( 39 , 47 ) and increases relative SM levels and the incorporation of saturated fatty acids into PLs . Ether PLs levels and the rate of PC and SM synthesis are not altered by caveolin deficiency . The mechanisms that regulate PL levels or the fatty acid incorporation into PLs are poorly understood ; thus , it is unclear precisely how Cav1 achieves such modulation in cellular lipid homeostasis . Similar effects were not observed when Cav1 expression was induced in a cell type that otherwise did not form caveolae ( 24 ) , suggesting that the role of Cav1 in lipid homeostasis in vivo is more complex than the formation of caveolae . Recently , it was reported that Cav-/- mice on a normal chow diet have increased saturated and monounsaturated cholesterol esters in their circulating lipoproteins compared with their WT controls ( 48 ) . Taken together , it appears that Cav1 plays a role in maintaining the balance of saturated to unsaturated lipids , but further studies are required to understand the mechanism of this regulation . The isolation and biochemical characterization of lipid rafts by detergent extraction ( as DRMs ) or sonication ( as light membranes ) have attracted criticism , as they may not reflect rafts in intact cells ( 6 , 9 ) . We have recently shown that isolation protocols can be adjusted so that DRMs and light membranes can represent similar domains , which , in macrophages , have similar structural properties to those found in intact cells ( 11 ) . Surprisingly , we found no difference in the cholesterol content of isolated lipid rafts from Cav1-/- and WT MEFs , suggesting that caveolar cholesterol contributes a negligible or minor amount of cholesterol to DRMs and light membranes , respectively , although Cav1 itself fractionated almost completely into DRMs . Taken alone , the biochemical data suggest that both cell types have the same amount of total raft lipids ; when Cav1 is present , it associates with raft fractions and generates caveolae , but Cav1 is not required for the formation of DRMs , NDRs or CO-sensitive domains . In contrast , the data obtained by Laurdan microscopy in intact , adhered cells suggest otherwise , with Cav1 expression resulting in more ordered and more abundant raft domains . It should be kept in mind , however , that the biochemical and microscopy analysis assess rafts in two different experimental systems . The microscopy images record the structure of the adherent plasma membrane adjacent to the substratum , while in DRM and NDR assays , whole detached cells are fragmented containing both plasma membranes as well as intracellular membranes . We have previously shown that cell shape and integrin engagement significantly contribute to membrane order and , further , that Cav1 expression and PHOSphorylation regulate membrane order at focal adhesions ( 34 ) . Hence , the microscopy analysis of membrane order also detects cell architecture -dependent membrane domains but is limited to the plasma membrane , whereas DRM and NDR analysis suggest that there is per se no biochemical difference in cholesterol-enriched membrane domains between detached WT and Cav1-/- cells . Taken together , it appears that Cav1 expression has little effect on the cholesterol distribution within membranes but does contribute to anchorage -dependent membrane order at the cell surface . In this context , it is noteworthy that we found no difference in total levels between the two cell types , while the proportional increase in SM or saturated PLs in Cav1-null cells does not contribute to higher cholesterol levels in isolated raft domains . The inverse is observed when cholesterol or sphingomylin levels are manipulated acutely ; treatment with cyclodextrin to remove membrane cholesterol lowers SM levels in DRMs and NDRs , while treatment of membranes with sphingomyelinase reduces DRM and NDR cholesterol ( 11 , 20 ) . One interesting hypothesis is that the higher levels of SM and saturated PLs in Cav1-/- cells compensate for the loss of Cav1 , resulting in the formation of noncaveolae raft domains that isolate as DRMs and NDRs . Whether subtle shifts in lipid profile , such as changes in fatty acid chain length in specific PLs subclasses , affect raft formation in general will only be resolved when lipid profiles and raft abundance are compared across cell types . Our observation that Cav1-/- cells had identical levels of CO-sensitive cholesterol to WT cells refutes the previous suggestion that CO specifically targets caveolar-cholesterol ( 25 ) . Interestingly , CO-sensitive depletion of cholesterol was largely limited to DRM in both cell types , while the CO product ( cholestenone ) was recovered predominantly in detergent-soluble fractions . This could be explained if lipid rafts with their high SUBstrate concentration are the main target for CO , but that the cholestenone generated redistributes to nonraft domains . The absence of the 3beta-hydroxyl group in cholestenone may cause such redistribution as it is important for the interaction with sphingolipids ( 49 , 50 ) . By visualizing CO action on lipid monolayers , Slotte ( 51 ) has not only shown that the oxidation reaction converts the condensed phase ( cholesterol-rich phase ) into an expanded phase ( cholestenone-rich phase ) but also located CO activity in the expanded phase or at the boundary between expanded and condensed phases . Furthermore , the catalytic rate of CO decreases with increasing membrane order of artificial lipid bilayers ( 52 ) . In agreement , we only observed a small shift of raft proteins toward the nonraft fractions after CO treatment , but not the complete disorganization of DRMs observed in other studies ( 53 , 54 ) that used different CO treatment conditions and cell type ( 55 ) . Overall , our and other data strongly suggest that CO activity is complex and not an appropriate tool for simple characterization of caveolae . In summary , we identified that Cav1 does not contribute to DRMs and NDR but contributes to cell architecture -dependent raft domains in adherent cells . Our complete lipid analysis of WT and Cav1-/- MEF further suggests that Cav1 plays a role in lipid homeostasis by regulating the balance in cholesterol esterification and between saturated and unsaturated fatty acids in PLs .