Direct visualization of membrane-spanning pores formed by a Leishmania amazonensis pore-forming cytolysin, as probed by atomic force microscopy

We have previously shown that Leishmania amazonensis produces and secretes a cytolysin that lyses membranes of mammalian cells, including macrophages, its host cell. Using the patch-clamp technique, we have previously demonstrated that the mechanism by which this cytolysin rupture macrophages plasma membrane is by pore formation, which lead us to name it leishporin. While we have characterized leishporin in several aspects, its molecular identity is still unknown. Its behavior suggests that leishporin is, or depend on, a protein, but recent results also suggests that a non-protein molecule is involved in cell lysis. Although the patch-clamp has undeniably revealed that L. amazonensis extracts generates pores in macrophages, these structures have not been spotted on cell membranes, which prompted us to several questions: 1) What is the appearance of leishporin-induced pores? Is it similar to that of other described pores? 2) Do these pores physically span lipid bilayers? 4) Are their directly-measured sizes compatible with those previously suggested by patch-clamp? 5) Do these pores fuse with one another, enlarging in size, as suggested by our previous reports? In the present work, we have used two membrane models, erythrocytes and liposomes, to visualize pores induced by the cytolysin on parasite extracts. Leishporin-mediated lysed erythrocytes or liposomes were analyzed by atomic force microscopy (AFM), which allowed us to visualize multiple membrane-spanning pores of variable diameters, ranging from 25 to 230 nm. They do not resemble to protein-formed pores, but rather, to pores made by small molecules such as lipids or peptides, as also visualized by AFM. Our results suggest that the maximum size for individual pores formed by leishporin is around 32 nm, but indicate that they are prone to coalesce, originating large membrane damages that leads to cell collapse, what seems to be a unique property among pore-forming cytolysins. Author summary One of the mechanisms whereby a cell can be destroyed is by punching holes into their membranes. Through these holes, due to differences in osmolarity between the outside and the inside of a cell, water flows towards the cytoplasm causing plasma membrane ruptures, which damages or lyses cells. We have previously described in the protozoan parasite Leishmania amazonensis one of such activities. Using an electrophysiology technique, we have found that parasite extracts lyse cells by making pores on their membranes. However these pores were not directly visualized so far. In this report, using a high-resolution-type scanning microscopy, the atomic force microscopy, we showed in red blood cells membranes and artificial lipid membranes (liposomes) the physical aspect of the pores we described earlier. We observed that these pores are circular-shaped structures with variable diameters, ranging from 25 to 230 nm that span the whole thickness of both types of membranes. We verified that L. amazonensis extracts-mediated pores resemble to pores formed by lipids or peptides and not by pores formed by proteins and that they may fuse with one another forming larger holes.


67
Protozoans of the genus Leishmania are intracellular macrophage-dwelling 68 parasites that infect man and other vertebrates. They cause a multifaceted disease 69 known as leishmaniasis, whose different outcomes depend on the more than 20 parasite directly causes host cell rupture [4][5][6][7] or that amastigotes are released by exocytosis  In previous studies, we have described a cytolytic activity present in Leishmania 88 amazonensis amastigostes and promastigostes that functions optimally at 37°C and pH 89 5.5 (conditions that are found inside the PV), but also at lower temperatures and/or 90 neutral pH [10,11]. This cytolytic activity has mostly been characterized in L. 6 116 Pore-formation by promastigote extracts was first suggested by an indirect 117 method, osmotic protection with polyethileneglycol in erythrocytes [10,11], and first 118 demonstrated by the patch-clamp technique in macrophages [12]. However, we had not 119 yet directly visualized the pores and/or the membrane damage caused by its action. As 120 known from previous works, human erythrocytes [11,12] and liposomes made from 121 dipalmitoilphosphatidilcholine (DPPC) [13] are lysed after incubation with cytolytically 122 active L. amazonensis membrane extracts. Atomic force microscopy (AFM) has been 123 commonly employed to visualize pore-formation produced in biological samples by 124 different pore-forming molecules [19][20][21][22][23][24][25]. In the present study, we used the AFM 125 tapping-mode technique in an attempt to visualize cell and artificial membranes damage 126 caused by leishporin. In fact, we were able to visualize the pores we have previously 127 described on the surfaces of lysed erythrocytes and also on lipid films derived from 128 disrupted liposomes that had been treated with L. amazonensis membrane extracts. This 129 is the first visual confirmation that the membrane damage caused by L. amazonensis 130 pore-forming cytolysin has indeed the appearance of a pore that spans the entire width 131 of the membranes studied. Moreover, a unique property of the cytolysin to form 132 coalescent pores was confirmed. The cytolytic activity of parasite mExt was assessed by a hemolytic assay using 161 human erythrocytes (HuE) as targets, as previously described [11]. Briefly, in 96 round-          arrows) are also shown ( Fig. 5A and C (panel 2)). Figures 5B and D (panel 2) show the 305 depths measurements of some of the pores. In these particular measurements the depths 306 were around 5.2 nm, 5.8 nm, 6.6 nm (Fig. 5B (panel 2)), 4.9 nm, 6.8 nm and 7.3 nm 307 (Fig. 5D (panel 2)). Figure 5E   Many of the observed pores were measured to obtain their diameter and depth.

313
Diameter was assumed as the horizontal distance between two points at the half of the 314 vertical distance. Depth was assumed as the vertical distance between the surface line 315 and the deepest point of the valley. Figure 6 shows the frequency of these two   [41], lipids [42,43], peptides [40], and lipopeptides [34]. In  This type of difference in measurements using different techniques has been already 403 reported. For instance, for EspB and EspD, two hemolytic proteins produced by the 404 type III Secretion System from E.coli, osmoprotection assays revealed a minimal pore-405 size of 3-5 nm whilst AFM showed pores ranging between 55-65 nm [45].

406
It was quite conspicuous that in liposomes we could visualize pores much larger 407 in diameters than in erythrocytes. One possibility is that, liposomes were more 408 overloaded with mExt than erythrocytes, an approach used in order to maximize the 19 409 probability to visualize the pores. Another possibility is that in liposomes, pores could 410 more easily fuse with each other giving rise to larger pores. About the variation on the 411 depths of the pores, we can speculate that the shallowest pores (4 nm) might not be 412 completely traversed the membrane and the deepest ones could be due to the fact that 413 the films are not totally straight (in fact, some sort of undulations appears in the 414 images) and the AFM probe could be measuring more than the thickness of the 415 membrane (for instance, in cases where the pore would be in the top of the undulation).

416
A remarkable finding was that, in liposomes, we could observe a pattern in diameters of 417 pores. Many pores had invariably the same diameter, all multiples of ~32.5 nm (Fig. 6), 418 suggesting that individual pores reach a maximum of ~32.5 nm in diameter, but can 419 fuse together, creating larger pores with multiples of this diameter. This maximum size 420 could due to some steric hindrance of molecules that constitute the pore, which 421 hampers further pore growth. Why these quantic increments in pore diameters were not

436
In Leishmania life cycle, host cell death/rupture passively caused by amastigote 437 overpopulation has been the customary assumption for release of parasites [4][5][6][7]. Lately, 438 due to leishporin activity at both acidic and neutral pH, we have proposed that, it may Listeria monocytogenes [50,51] Legionella pneumophila [52], Toxoplasma gondii [53] 444 and some Plasmodium species [48,49,54]. 445 The capacity to cause cell damage on their host cells has been reported as an T.cruzi, and pore formation has been shown to facilitate its entry in host cells [55,56].