The availability of in vitro
models has made it possible to study, at the molecular level, Acanthamoeba
interactions with the host cells, such as corneal epithelial/endothelial cells [17
], brain microvascular endothelial cells [21
], fibroblasts [22
], keratocytes [18
], macrophages [24
], neutrophils [26
], and neurones [28
]. Although these models allow in-depth biochemical and molecular investigations in vitro
, thus further elucidating mechanisms of infection, they cannot model whole organism responses to infection at the physiological level. This is particularly relevant in brain infection due to Acanthamoeba
which involves complex interactions between amoeba and the host.
Both Acanthamoeba genotypes studied here in locusts, reduced faecal output at about 5 days post-injection, and killed all locusts within 11 days. Live Acanthamoeba can be recovered from brain lysates of amoebae-injected locusts, and trophozoites can be seen inside infected brains in histological studies. It is intriguing that amoebae are not found in the CNS of infected locusts on day three, and they invaded the brain after 4 or 5 days, with changes in faecal output and fresh body weight respectively becoming apparent. It is tempting to speculate from these temporal relationships that Acanthamoeba-mediated locust death is, at least in part, associated with the parasite's invasion of the brain. Interestingly, Acanthamoeba did invade other parts of the locust CNS such as the suboesophageal ganglion, but other ganglia (such as in the ventral nerve cord) were not investigated for the presence of amoebae in this study. The suboesophageal ganglion is situated below the crop and is connected to the brain by circumoesophageal connectives, and coordinates movements of the mouthparts, and the activity of the salivary glands. Clearly, invasion of the CNS by Acanthamoeba could affect feeding behaviour, as is suggested by the reduction in faecal output in infected locusts. It seems most likely that the changes in locust physiology and behaviour (reduction in body weight and faeces production, and reduced locomotory activity) are consequent on Acanthamoeba-mediated disruption of the blood brain barrier, which leads to neural dysfunction and reduced sensory output/input.
For the first time, histological examination of infected locusts shows that amoebae invaded deep into tissues such as the fat body and muscle, causing appreciable degenerative changes. Thus the amoebae invade these tissues, and are not isolated from them simply because they adhere to the surface of the tissues which are bathed in the haemolymph of the insect's open circulatory system. These findings suggest that Acanthamoeba produced parasitaemia and survived the onslaught of the innate immune defences of locusts. Acanthamoeba invaded different organs of the infected locusts and did not appear to exhibit any specific preference in relation to the tissues that are invaded, a finding that is consistent with the human form of infection, and confirms the validity of locusts as a useful model in which to study the pathogenesis of Acanthamoeba granulomatous encephalitis in vivo. Furthermore, Acanthamoeba granulomatous encephalitis is mostly limited to immunocompromised populations, and insects have an entirely innate immune defence system, suggesting that it is realistic to use locusts as a tractable model in which to study the pathogenesis of Acanthamoeba granulomatous encephalitis.
Although Acanthamoeba spread to many tissues and were found in the haemolymph throughout the course of the infection, none of the isolates (T1 and T4 genotype) were ever found in locust faeces (unpublished observations). For the first time, histological sectioning revealed the occasional presence of some amoebae in the lumen of the locust foregut, but no damage to the wall of the foregut was evident in any of the locusts subjected to microscopic examination. Indeed, the apical surfaces of the cells lining the foregut have a cuticle, which could represent a barrier to penetration by Acanthamoeba. Unfortunately, infected locusts destined for histological examination were not kept isolated from one another (as was the general case), and food replenishment and removal of dead animals took place only once every 24 h, so cannibalism was possible if locusts died shortly after this daily routine. It is likely therefore that amoebae observed occasionally in the lumen of foregut were simply there because they were consumed by cannibalism of a dead infected locust. This is a novel finding and it is strengthened by the fact that the histological sections never revealed evidence of damage to the wall of the foregut and suggest that amoebae do not infect locusts via the oral route, a finding that is consistent with infection in vertebrates.
Another significant finding was the entry of amoebae into the locust CNS, which appeared to be associated with disruption of the neural lamella and the perineurium/glial cell complex that constitutes the locust blood-brain barrier [29
]. This is consistent with the studies in vitro
showing that amoebae cross human brain microvascular endothelial cells, which constitute the blood-brain barrier, by affecting the integrity of the cell monolayer [32
]. At present, the basis of the damage to the locust blood-brain barrier is not clear, i.e., amoeba and/or host inflammatory response. Recent studies in vitro
show that serine proteases secreted by Acanthamoeba
play an important role in affecting the integrity of the human brain microvascular endothelial cell monolayers [32
], and the role of proteases and additional virulence determinants will be addressed in future studies in vivo
using locusts. In addition, there is a need for a comparative study to test several additional Acanthamoeba
isolates of various genotypes in locusts versus