Ideally, entomopathogenic fungi should be virulent to their insect hosts independent of host age, sex and physiological state. However, as seen here, nutritional state can significantly affect host susceptibility. Female A. aegypti that had ingested a blood meal rapidly became less susceptible to infection by M. anisopliae than sucrose fed females. This reduced susceptibility persisted for a duration of approximately 96 h, at which time survival rates returned to levels seen for sucrose fed females exposed to fungi.
These results are similar to those published by Mnyone et al.  when comparing blood-fed and glucose-fed An. gambiae , although no alterations in survival times were observed when exposing mosquitoes to fungi 3 h post-blood feeding. An alteration in survival rates (increased survival) was observed at 12 h and 36 h post-blood feeding. According to these authors, there were no differences in the survival of sugar-fed and blood-fed females 72 h post-feeding.
Mnyone et al.  suggested that reduced host susceptibility could be due to a putative increase in nutrient reserves of blood fed insects when compared to sugar-fed insects, thus delaying the detrimental effects of the fungus on the host, which in turn resulted in increased survival rates. However, up-regulation of the immune system following blood feeding could be responsible for the alteration in susceptibility.
In order to identify the possible causes for the reduced susceptibility of blood-fed female mosquitoes it is necessary to understand the fungal infection process. Entomopathogenic fungi normally infect their host by penetrating the integument, a well studied process that involves both physical and chemical (enzymatic) mechanisms .
The infection process of entomopathogenic fungi can be separated into three phases: (a) adhesion and germination of the conidia on the insect's cuticle; (b) penetration of integument by the germ tube, in order to invade the haemocoel and (c) development of the fungus within the haemolymph (colonization) resulting in host death . There is a time lag between initial contact with the fungus and haemolymph colonization. Conidial germinating can be seen from 8 h under laboratory conditions (unpublished data: RIS) and as the earliest host-death can be seen after 24 h, cuticle penetration thus takes up to 16 h in highly virulent isolates.
Even before reaching the haemolymph, the insect will have already activated an immune response to the microbial invasion. Insect defenses are (a) physical barriers, such as the cuticle and peritrophic membrane, epithelial barriers, some of which carry specific immune mechanisms; (b) protease cascades leading to coagulation and melanization; (c) cellular responses such as phagocytosis and encapsulation, and (d) production of antimicrobial peptides and reactive oxygen species .
It is known that blood feeding in combination with sucrose significantly increases the melanization cascade in An. stephens, whereas insect fed on a sucrose only diet displayed a low immuno-competence . In the current study, A. aegypti females were fed blood once, and subsequently offered sucrose ad libitum , which may have increased immune activity of the mosquitoes. It is common to provide sugar solutions for maintaining mosquitoes in laboratory studies; however, field captured A. aegypti females were shown not to have fed on sugars, indicating that this energy source is not necessary for survival and reproduction [21–23].
Studies of the effects of blood feeding on the longevity of mosquitoes have been contradictory. One hypothesis is that mosquitoes that feed on blood have lower survival rates due to a diversion of nutrients to egg production . Other experiments reported that blood feeding increased the survival of mosquitoes, compared with those feeding only on sugar [15, 24, 25]. In the present study a slight increase in survival of blood-fed females was seen when compared to sucrose only fed females.
Blood fed An. gambiae suffered a series of physiological changes such as expression of genes related to formation of the peritrophic matrix, digestion, immunity and egg development . Insect antimicrobial peptides and proteins take 1-3 hours to produce and 12 to 48 hours to reach their peak . In the current study females became susceptible again to the fungus 96 hours after blood feeding, possibly when due to down-regulation of the immune system returning to the pre-blood meal levels.
The selection of highly virulent fungi is important for the control of A. aegypti especially as the dengue virus develops rapidly. On the seventh day, the females become infective  and transmit the virus following each blood meal . The lethal effects of the entomopathogenic fungus should therefore occur before the parasite develops infective forms.
In the current study, on the seventh day of evaluation, there was significant mortality of A. aegypti fed on sucrose or blood, emphasizing the potential of the fungus for mosquito control. The highest value for S50 was 6 days, resulting in 50% vector mortality before the final development of the Dengue virus . There were no differences in the values of S50 for LPP133 when compared to isolate ESALQ 818, except for zero and 72 hours post-blood feeding.
Blood fed wild strain females exposed to ESALQ 818, had significantly higher survival rates compared with females fed with sucrose except 120 h post-blood feeding. The value of S50 of wild strain females fed with sucrose was higher compared with mosquitoes reared in the laboratory (Rockefeller strain), 5 and 3 days respectively. It is probable that wild strain females are subject to selection pressure which results in greater genetic diversity and increased immunity, compared with females reared in the laboratory.
Reduction in mosquito susceptibility to fungi following a blood meal is of concern, thus females resting on fungus impregnated surfaces immediately after blood feeding will have a reduced risk infection. However, host seeking female A. aegypti remain highly susceptible to fungal infection. It should be remembered that reduced susceptibility is only a temporary phenomenon and may not alter the long term effects of fungi on the mosquito population. These results highlight the need for further studies on the insect-fungus interaction at a cellular and molecular level.