In the absence of their acto-myosin system. Based on previous studies that implicate an important role for host cell actin during invasion, we favour the hypothesis that once the TJ is formed, host cell actin plays an important role during invasion of the analysed mutants. In summary, this study leads to three hypotheses. 1. Components of the invasion machinery show multiple redundancies. 2. A compensatory invasion mechanism is in place that can substitute for the loss of a functional actin-myosin A-system. 3. Our current model that predicts a linear motor for the generation of force for motility and invasion needs to be revised. In line with the third possibility, we propose that a gelationsolation osmotic engine could drive the Diacerein parasite propulsion, similar to the model proposed in. Specifically, a gel of actin-like filaments and acidic protein micromolecules secreted by micronemes at the apical end of the cell, would be coated by both immobile heavy and mobile light cations in the cytoplasm, as was shown for actin gels in vitro. The mobile cations generate osmotic pressure that is balanced by tension of the elastic gel. Partial disassembly of the gel caused by degradation of its macromolecules will cause weakening of the gel elastic modulus and gel swelling. This swelling will push the leading edge of the cell forward, providing there are adhesions of the gel to the substrate that are spatially biased to the rear. Subsequent complete disassembly of the older weaker gel and reassembly of the dense gel at the new leading edge Nodakenin completes the protrusion cycle, which can be step-like if the assembly-disassembly is cyclic or smooth if the gel assembles and disassembles continuously. The rear of the cell will be pulled forward by the membrane tension generated by the protrusive force and the cytoplasmic flow from the rear caused by the gradient of the pressure in the cell �C such gradient is made possible by the poroelastic nature of the cell cytoskeleton. Simple estimates in the supplemental material demonstrate that such gelation-solation osmotic engine is physically feasible. The idea of a macroscopic osmotic engine coupled with the gel elasticity was theoretically proposed and experimentally proved a long time ago. Ability of osmotic gradients to propel membrane vesicles at speeds comparable with those reported in this paper was demonstrated experimentally. The proposed hypothesis is also supported by findings that rapid dynamics of adhesions and fast cycles of assembly and disassembly of actin filaments are necessary for the fast motility of the parasite. The fact that high potassium buffer completely blocks gliding motility lends additional support for the model. The high cytoplasmic pressure in the motile parasite is evident from the rounded shape of the GAP45 mutant in our study and from the bleb-like protrusions evident in plasmodium ookinete mutants. In principle, other modes of parasite propulsion driven by the osmotic pressure are possible. First, it was shown theoretically that secretion of charged molecules at one side of the cell, and their degradation at the other side coupled with water flow through permeable cell membrane, can create water intake at the front and outflow at the rear and an accompanying force that can propel the 
cell. Sufficient membrane permeability probably requires aquaporin channels; however, an aquaporin KO in P. berghei is viable and has no defect in gliding motility.