as large as that produced by the ShB peptide (pKa *7.7), which seems consistent with the experiments using fluorescently-labeled peptides in which the surface partition coefficients estimated for the interaction of the ShB peptide with PA vesicles, were always higher than those corresponding to the ShB-L7E peptide.

Differential Scanning Calorimetric studies of peptide insertion into lipid bilayers:
    The possible implication of the hydrophobic domains of the phospholipid bilayers in the association of the peptides to the vesicles was investigated by differential scanning calorimetry (DSC) using synthetic dimirystoyl derivatives of the different phospholipids, which have convenient phase transition temperatures at 23.5, 23.3 and 49.6 C, for DMPC, DMPG and DMPA, respectively. In these experiments, peptide insertion into the lipid bilayer and therefore, its interaction with the phospholipid acyl chains, is expected to prevent part of the phospholipid molecules from undergoing the phase transition characteristic of the "pure" phospholipid species and thus, decrease the phase transition enthalpy in a peptide concentration-dependent manner (Papahadjopoulos et al., 1975). We observed that neither the ShB nor the ShB-L7E peptides cause any effects on the transition temperature or the transition enthalpy of DMPC vesicles (data not shown). On the other hand, confronting the ShB peptide to anionic phospholipid vesicles (Figure 2A and B) causes a large, concentration-dependent decrease in the transition enthalpy, which is slightly more pronounced in DMPG than in DMPA vesicles, without modifying the phase transition temperature. A plot of the observed changes in the transition enthalpies versus the peptide/phospholipid molar ratios used in these studies (Figure 2C), predicts that each ShB peptide molecule prevents an average of 3 to 4 anionic phospholipid molecules from undergoing the phase transition. As to the mutant ShB-L7E peptide, its effects on the DMPA or DMPG phase transition enthalpies are negligible compared to those of the ShB peptide (Figure 2A and B).
    DMPG vesicles were also used to study whether the observed peptide insertion into the phospholipid bilayer is pH-dependent. DMPG was chosen for these studies because it lacks chemical groups whose titration within the pH region of interest could interfere with the peptide insertion phenomena determined by DSC. Figure 3 shows that indeed, the ability of the ShB peptide to insert into the hydrophobic domains of the anionic bilayer is strongly pH-dependent. A maximum decrease in the DMPG phase transition enthalpy in the presence of the ShB peptide can be observed at pH 6.0, while increasing the pH to 8.5 or beyond, causes that the ShB peptide is no longer capable to insert into the phospholipid bilayer. In spite of the observed lack of peptide insertion under the latter conditions, peptide binding experiments carried out at pH 8.5 using NBD-labeled peptides as described above, demonstrates that the ShB peptide still binds to anionic phospholipid vesicles, although with a much lower affinity (Table I). As to the mutant ShB-L7E peptide, its effects on the DMPG phase transition enthalpies are negligible at all the pH's explored (Figure 3).

FTIR studies of peptide conformation:
    Figure 4 shows the infrared amide I region of the spectra corresponding to the ShB and ShB-L7E peptides in the presence of different phospholipid vesicles. Either peptide in the presence of PC vesicles (panels A and B) (as well as in plain buffers in the absence of lipids), exhibit a bell-shape amide I band with a maximum centered at 1645 cm-1 characteristic of non-ordered protein structures. The similarities observed between the spectra of the peptides in solution and in the presence of PC vesicles is consistent with the lack of binding of either peptide to the zwitterionic PC vesicles. On the other hand, the spectra of the ShB peptide in the presence of either PA or PG vesicles exhibit a very prominent amide I component at 1623 cm-1 and a smaller one at 1689 cm-1 (lower traces in panels C and E, Figure 4), which have been related to the adoption of a strongly hydrogen-bonded * structure (Demel et al., 1990). Such 1623 cm-1 component in the ShB spectra appears readily and has approximately the same relative importance regardless of the peptide concentration or the peptide to phospholipid molar ratio used in the FTIR experiments (ranging 1.2 to 10 mg/ml and 2 to 60 (by mole), respectively) (Fernandez-Ballester et al., 1995). Even though they are similar, the ShB spectra taken in the presence of PG or PA vesicles differ in that the characteristic 1623 cm-1 component is more heat-stable in the former, since at temperatures as high as 70 C, the absorbance at 1623 cm-1 seen in the ShB/PG samples is still maintained at no less than 65% of that observed at room temperature, whereas that seen in the presence of PA vesicles decreases more 



Pie de Table 1 y datos

Table I: Fluorescence emission maxima (l max) and surface partitioning constants (Kp*) exhibited by NBD-labeled peptides in the presence of phospholipid vesicles.

l max (nm)		Kp*x10-4 (M-1)

peptide designation		pH		buffer	PC	PA	PG		PC	PA	PG
ShB-NBD		7		555	551	531	531		no binding	77.1 (±11.6)	2.80 (±1.10)
ShBL7E-NBD		7		554	551	533	534		no binding	3.05 (±1.45)	0.72 (±0.20)
											
ShB-NBD		8.5		552	---	530	536		---	1.95 (±0.65)	0.64 (±0.09)
ShBL7E-NBD		8.5		553	---	536	552		---	0.62 (±0.025)	£0.05