Effects of Skp and LPS on the folding of OmpA

I have first investigated how Skp and LPS affect the folding of OmpA using an electrophoretic mobility assay. Previously, different electrophoretic mobilities of folded (30 kDa) and denatured OmpA (35 kDa) were observed in sodium dodeclysulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) if the samples were not boiled prior electrophoresis (Schweizer et al.1978).

In folding experiments, appearance of the 30 kDa band was correlated to the formation of native structure in OmpA (Dornmair et al.1990; Surrey and Jähnig 1992; Arora et al.2000).

Here I have used this simple assay and monitored the time course of OmpA folding in different experiments at various Skp, LPS, and lipid concentrations. Five representative experiments are shown in Figure 3.1.

When reacted with Skp in aqueous solution, OmpA remained unfolded over a time course of 180 min. Only one band was observed at 35 kDa indicating unfolded OmpA (U). For comparison, the migration of native OmpA (F) at 30 kDa is shown in the first lane of gel 1.

When OmpA was reacted with a 5-fold molar excess of LPS, about 45% of OmpA folded over a time course of 180 min. I have determined the mole fraction of folded OmpA by densitometry of the two bands representing the folded and the unfolded forms.

In contrast, when OmpA was reacted with preformed phospholipid bilayers in absence of Skp and LPS, about 75% of OmpA folded at pH 7 and the kinetics were slower (Figure 3.1A, Gel 4 and Figure 3.1B, ●).

The densitometric analysis of gel 2 (Figure 3.1B ▲) showed that the yield of folded OmpA quickly levelled off within an hour after initiation of folding. However, when OmpA was reacted with both, first Skp and then LPS, only very little folding was observed (Figure 3.1A, gel 3 and Figure 3.1B, ♦). The conclusion was that Skp inhibits the folding of OmpA into LPS. I was then interested to investigate whether OmpA that was first reacted with Skp and LPS would insert and fold into preformed phospholipid bilayers. Gel 5 in Figure 3.1A indicates that OmpA quickly folds under these conditions. Up to 90% of OmpA folded with fast kinetics, as analyzed by densitometry (Figure 3.1B, ●). Additional experiments indicated

Figure 3.1 Membrane insertion and folding of OmpA is affected by LPS and Skp.

A, SDS-polyacrylamide gels showing the time courses of OmpA folding at 30 °C. The arrows indicate the migration of unfolded (U) and folded (F) OmpA. The folding of OmpA was monitored after dilution of the denaturant in the presence of a 4-fold molar excess of Skp (gel 1), a 5-fold molar excess of LPS (gel 2), a 4-fold molar excess of Skp followed by a 5-fold molar excess of LPS (gel 3), a 200-fold molar excess of phospholipids (gel 4), and a 4-fold molar excess of Skp followed by a 5-fold molar excess of LPS followed by the immediate addition of a 200-fold molar excess of phospholipids (gel 5). For comparison, the migration of native OmpA (F) at 30 kDa is shown in the first lane of gel 1. Where present, the migration of Skp (17 kDa) is also indicated. The experiments 2–5 were done in parallel. In all of the five experiments, OmpA, LPS, Skp, and lipid vesicles were from the same stock solutions. B, densitometric analysis of the fraction of folded OmpA as determined from SDS gels as shown panel A: Skp/OmpA = 4:1 mol/mol (■); LPS/OmpA = 5:1 mol/mol (▲); Skp/LPS/OmpA = 4/5/1 mol/mol/mol (♦); phospholipid bilayers (●); and Skp/LPS/OmpA = 4/5/1 mol/mol/mol in presence of phospholipid bilayers (●).

that the sequence in which Skp and LPS were added to OmpA in these experiments had only a small effect on the time course of OmpA folding when the reactants were mixed quickly without any delays (data not shown).

The yields obtained in this experiments were very similar to the yields reported previously for the folding of OmpA into preformed lipid bilayers of dimyristoylphosphatidylcholine at the same temperature and pH (Surrey and Jähnig 1995).

The faster rate of folding and the increased folding yields in presence of Skp and LPS were not simply an additive effect of the folding of OmpA in presence of Skp and LPS and the folding of OmpA into lipid bilayers (Gels 3 and 4) since the sum of the folding yields (Figure 3.1B, ●) of these two reactions is still smaller than the yields obtained in presence of all three components (Gel 5, Figure 3.1B, ●).

My observation that OmpA remained unfolded in solution when both, Skp and LPS, were present but inserted and folded with high yields when preformed lipid bilayers were added, led to investigation of the kinetics of OmpA folding into bilayers with only one periplasmic component, either Skp or LPS, present.

Figure 3.2 Efficient folding of OmpA into lipid bilayers requires the simultaneous presence of Skp and LPS. The OmpA-folding kinetics were determined by electrophoresis and densitometry.

OmpA, Skp, LPS, and lipid were reacted as described under “Material and methods” at 30 °C. The Skp/OmpA ratio was 4 mol/mol. The LPS/OmpA ratio was 5 mol/mol, and the lipid/OmpA ratio was 200 mol/mol.A, folding was initiated after urea dilution by immediate addition of lipid bilayers in absence of Skp and LPS (●), in presence of Skp (♦), in presence of LPS (▲), and in presence of both Skp and LPS (●).B, the corresponding experiments were performed but with a 30-min delay between urea dilution in presence or absence of Skp and the addition of lipid or addition of first LPS

Figure 3.2 shows the densitometric analysis of the formation of folded OmpA as a function of time after addition of lipid bilayers. In panel A, Skp, LPS and lipid were added to OmpA immediately after strong dilution of urea. In panel B, LPS, and lipid were added 30 min after urea dilution in presence or in absence of Skp. In all experiments, a 30 min delay time resulted in slower OmpA folding kinetics.

The folding kinetics into lipid bilayers were fastest and folding was most complete, when Skp and LPS were both present (●). When Skp and LPS were both absent, folding into lipid bilayers was less efficient when lipid bilayers were added after 30 min compared to immediate lipid addition (●).

However, when only one of the periplasmic components, either Skp (♦) or LPS (▲) was added, OmpA folding into lipid bilayers was strongly reduced.

Even slower folding kinetics and lower yields of folded OmpA were observed when the delay time was increased to 1 hour for all folding reactions (data not show). This decrease in folding efficiency was smallest and up to 90% folded OmpA were still detected when Skp and LPS were both present, while in absence of both, LPS and Skp, final yields of folded OmpA in lipid bilayers dropped below 60%.