Chain order profile in lipid HII phases.

Dear Sir: In an interesting recent paper, Sternin et al. (1988) presented data on the orientational acyl chain order in the hexagonal HI, phase. Deuterium nuclear magnetic resonance (2H NMR) measurements were made on perdeuterated tetradecanol in a mixture with l-palmitoyl-2-oleoyl phosphatidylethanolamine in water. The data were analyzed by a novel method which involved first obtaining the oriented spectra from the 2H NMR powder pattern (dePaking) and, second, assigning the resonances from the perdeuterated chain by uniformly dividing the intensity under the dePaked spectrum and then monotonically attributing the corresponding mean quadrupole splittings of these spectral segments to the methylene groups of the chain. Relatively few order parameter measurements on hexagonal lipid phases are available, and therefore it was not possible to assess the fidelity of the new method of spectral analysis by comparison with other data, for example from measurements on specifically deuterated systems. In view of the importance of the results to the dynamic properties of inverted lipid phases, it is particularly useful to know to what extent the order parameters obtained are influenced by the method of spectral analysis. In our continuing studies on the lamellar, La, to inverted hexagonal, HI,, phase transition in distearoyl phosphatidylethanolamine (DSPE) (Seddon et al., 1983, 1984) we have recently performed 2H NMR measurements on both perdeuterated and specifically deuterated DSPE. Many of the single methylene resonances were resolved in the dePaked spectra from DSPE perdeuterated in the sn-2 chain in the HI, phase, and comparison with the results from specifically deuterated samples has permitted assignment of these resonances. This work therefore allows a direct comparison of the order parameters obtained by conventional means with the new method of analysis developed by Sternin et al. (1988). The dePaked spectrum of DSPE perdeuterated in the sn-2 chain, dispersed in 2.4 M NaCl at 970C is given in Fig. 1. At this temperature, the lipid is in the HI, phase (Marsh and Seddon, 1972; Seddon et al., 1983), as also evidenced by the low values of the outer quadrupole splittings. Many of the resonances are resolved as separate peaks, arising most probably from single methylene groups. The spectrum has been analyzed by the method of Sternin et al. (1988), as is indicated by the stick diagram given below the dePaked spectrum. An exact correspondence between the resolved peaks and the stick diagram is not obtained, although near coincidence is obtained in several cases. The resolved resonances in the dePaked spectrum have been assigned by comparison with the spectra from DSPE specifically deuterated in the sn-2 chain (Sankaram, M. B., and D. Marsh, manuscript in preparation). Plots of the quadrupole splitting versus chain segment position are compared with those given by the stick plot in Fig. 2. The individual quadrupole splittings from the specifically deuterated samples are also included for comparison. These do not fit exactly with those from the perdeuterated sample because the hexagonal transition temperature is somewhat different for the perdeuterated sample, but the fit suffices fully for the purposes of assignment. Although the overall values of the quadrupole splittings given by the two methods in Fig. 2 are of rather similar magnitude, there are significant differences in the shape of the segmental profile. In particular, the method of Sternin et al. (1988) removes the order parameter plateau between the C-3 and C-6 positions and replaces it by a profile of steadily decreasing quadrupole splittings. On the other hand, it reproduces quite well the quadrupole splittings of the segments closer to the terminal methyl end of the chain. The large differences in the two assignments for the quadrupole splittings at the C-2 position arise from the special conformation of this group in the sn-2 chain, which is well known for the La-phase (Seelig and Seelig, 1975) and is preserved in the HI, phase. Such effects are difficult to allow for a priori in the method of analysis used by Sternin et al. (1988). If this is done on the basis of the specifically deuterated assignments, the fit between the two methods is improved somewhat, but the quadrupole splittings for the C-6