The location of midgut bacteria relative to meconial peritrophic membranes (MPMs) and changes in bacterial numbers during midgut metamorphosis were studied in Anopheles punctipennis (Say), Culex pipiens (L.), and Aedes aegypti (L.) pupae and newly emerged adults. After adult emergence in Aedes,Anopheles, and most Culex, there were few to no bacteria in the midgut. In most newly emerged adult mosquitoes, few bacteria were found in either the lumen or within the MPMs/meconia. In a few Culex specimens, high numbers of bacteria were found in the MPMs/meconia and low numbers in the lumen. In all three species bacterial counts were high in fourth instars, decreased after final larval defecation, increased in young pupae, and increased further in old pupae. A very effective gut sterilization mechanism is operating during mosquito metamorphosis and adult emergence. This mechanism appears to involve the sequestration of remaining larval gut bacteria within the confines of the meconium and one or two MPMs and the possible bactericidal effect of the exuvial (molting) fluid, which is ingested during the process of adult emergence.
As filter-feeders, mosquito larvae ingest a wide variety of microorganisms, including bacteria (Walker et al. 1988, Merritt et al. 1992, Straif et al. 1998). Given such indiscriminate intake of microbes, it is probable that some would be pathogenic to adult mosquitoes. The literature on the subject of bacteria in adult mosquito midguts is reviewed briefly in Pumpuni et al. (1996) and Straif et al. (1998). We have been studying the possible role of noncellular peritrophic membranes found in association with the developing adult midgut in light of their potential role in preventing larval bacteria from infecting adult mosquitoes.
The peritrophic membrane or matrix (PM) is a noncellular, chito-proteinous layer that forms around ingested food in the midgut of most insects (Spence 1991, Jacobs-Lorena and Oo 1996, Lehane 1997). Because it separates the food bolus from the midgut epithelium, it may act as a barrier to potentially invasive organisms ingested with food.
Two types of PM are recognized. Type I PMs are formed by secretions from cells distributed over the entire midgut surface. Type I PMs have been described in several orders and in blood-feeding black flies, mosquitoes, sand flies, and tabanids. Type II PMs are produced by the cardial epithelial cells at the foregut/midgut junction and are found in larval Diptera (Lehane 1991).
In mosquitoes, adult females form a type I PM around a blood meal, and in larvae the food bolus is surrounded by a type II PM. At least two PMs also form during the pupal stage, meconial peritrophic membrane 1 or MPM1, forming with a few hours after pupation, and meconial peritrophic membrane 2 or MPM2, forming around the time of pupal-adult emergence (Romoser et al. 2000). The MPMs surround the meconium, the sloughed, degenerating larval midgut epithelium (Moncayo and Romoser 1994). The meconium/MPM complex disappears from the adult midgut within 24–48 h after emergence (Romoser et al. 2000). Two hypotheses regarding the functions of these membranes have been suggested (Moncayo and Romoser 1994). First, they may serve as barriers to potentially harmful microorganisms found in the larval meconium during metamorphosis. Second, they may protect the developing midgut epithelium from the hydrolytic enzymes associated with the autolytic breakdown of the meconium.
The two objectives of our research reported here were as follows: (1) to test the hypothesis that microorganisms, specifically bacteria, become sequestered within the confines of the MPMs; and (2) to determine changes in bacterial count during midgut metamorphosis.
Materials and Methods
Species Studied and Rearing Procedures.
Representative species of the genera Anopheles,Culex, and Aedes were studied. Larval Anopheles punctipennis (Say) were collected in Athens County, OH, and reared in water from their larval habitat. Specimens of Culex pipiens (L.), and Aedes aegypti (L.), Rockefeller strain, were from laboratory colonies and were raised in distilled water. All larvae were fed a 1:1:1 mixture of liver powder, ground rabbit chow, and brewer’s yeast and were maintained at 27 ± 1°C. Newly emerged adults were held in small cardboard containers.
Meconium and Lumen Preparation.
Mosquitoes were tested for differences in the concentration of bacteria inside the MPM/meconium verses the remainder of the midgut lumen. To ensure a bacteria-rich environment, a 5-ml suspension of Salmonella typhimurium (ATCC 14028 American Type Culture Collection, Rockville, MD) was adjusted to an optical density of 0.2 (=1 × 108 to 3 × 108 bacteria per milliliter) and added to the rearing medium during the first stadium. Before dissection, forceps, slides, and the dissecting area were cleaned with 70% ethyl alcohol to reduce bacterial contamination. Dissections of cold-anesthetized mosquitoes were carried out in sterile, distilled phosphate buffered saline (PBS). For An. punctipennis and Cx. pipiens, once the meconium and the midgut were dissected they were placed in separate tubes, each filled with 200 μl of sterile PBS and put into suspension using sterile, mosquito grinders. MPM 1 was often absent or very fragile in Ae. aegypti and the meconium tended to disintegrate during dissection. For this reason, the meconium in this species was left intact in the lumen, and samples for bacterial counts were taken before and after the meconium had been egested.
For each species, bacterial counts in the meconia and lumens of a total of five specimens were taken in each of four time intervals (0–5, 5–10, 10–15, and 15–20 h) after adult emergence, producing an adult postemergence sample size of 20 mosquitoes.
Changes in Bacterial Counts During Metamorphosis.
For each of the three species, 10 specimens from each of the following stages were tested for differences in bacterial counts: (1) fourth instar before final defecation, (2) fourth instar after final defecation, (3) 0- to 10-h-old pupae, (4) 38- to 48-h-old pupae, and (4) 0- to 10-h-old adults. Larvae, pupae, and adults were removed from the rearing medium using watchmaker’s forceps, and caution was taken not to puncture the cuticle. Specimens were surface sterilized by being placed in 70% ethyl alcohol for 10 s and rinsed in sterile PBS for an additional 10 s. Specimens were placed in 200 μl of fresh sterile PBS and ground with a sterile tissue grinder. Bacterial counts also were made on samples of the larval rearing water.
A 1:10 and 1:100 dilution was made with each of the homogenized samples and 1.0 ml of each dilution was spread on separate trypticase soy agar plates using sterile technique and incubated for 24 h at 37°C. After 24 h, agar plates were examined for individual colonies. Each isolated bacterial colony had arisen from a single bacterium and the total number of colonies therefore was equal to the initial number of bacteria present.
Location of Bacteria in Newly Emerged Adult Midguts.
With few exceptions, bacterial counts taken from the lumen and meconia of the three species revealed that there were few to no bacteria in either of these locations in newly emerged adults (Table 1). Overall, 36/60 (60%) adults contained no bacteria in the midgut after adult emergence. Midguts from five Cx. pipiens contained significant numbers of bacteria in both the meconium and the lumen. However, the average concentration of bacteria within the MPMs/meconia far exceeded that in the lumens; 9,520 versus 217 or an average of 97.4% more bacteria in the MPMs/meconia than in the lumens. The remaining samples (19/60; 31.7%) produced 1–210 bacterial colonies per milliliter, and no major differences were seen in the number of bacteria in the meconium versus the lumen. These 1–210 colonies could have been the result of contamination.
Changes in Bacterial Counts During Metamorphosis.
Counts of larval water samples were in the "too many to count" category in all the dilutions we used, so there was no doubt that larvae had ingested very large amounts of bacteria. Bacterial counts taken as a function of developmental stage (Table 2) displayed a consistent pattern among all three species studied. Counts were very high in fourth instars before final defecation, but decreased dramatically after the final larval defecation and voiding of the peritrophic membrane. The bacterial counts tended to increase during the pupal stadium. With the exception of three Cx. pipiens specimens, bacterial counts decreased to nearly zero in all newly emerged adult specimens examined. The three exceptional Culex specimens contained bacterial counts from 2,040 to 4,960 colonies per milliliter, similar to those found in 38- to 48-h-old pupae. Bacterial counts were essentially zero in the remaining seven specimens. Among the three species studied, the overall average decrease in bacterial counts between the fourth instar before final defecation and newly emerged adults was 99.3%.
Bacterial counts taken from the meconium and lumen of newly emerged adult An. punctipennis,Ae. aegypti, and most Cx. pipiens revealed that in most cases there were few to no bacteria found in either of these locations. Among the remaining specimens, the vast majority produced minimal bacterial growth, which fell within a range that could be caused by contamination. However, in Cx. pipiens, five adult midguts contained bacteria in both the meconium and the lumen, but with an average of 97.7% more bacteria in the meconium than in the lumen. It should be noted that Cx. pipiens produces very well-formed MPMs, both MPM1 and MPM2 (Romoser et al., unpublished data).
The findings in Cx. pipiens support our hypothesis that meconial peritrophic membranes sequester bacteria and other microorganisms in preparation for their egestion from the midgut, and the findings in An. punctipennis and Ae. aegypti are not inconsistent with this hypothesis.
The bacterial counts taken as a function of developmental stage corroborate and extend the midgut dissection results. Because we surface sterilized each specimen before homogenization, it is reasonable to assume that the bacteria detected were from the midgut lumen. In all three species studied, bacterial counts tended to increase from postdefecation fourth instars (pharate pupae), to 0- to 10-h-old pupae, to 38- to 48-h-old pupae, respectively. The overall average increase seen in all samples from postdefecation fourth instars to 38- to 48-h-old pupae was 66.14%. Therefore, the hydrolytic enzymes likely to be involved in the autolysis of the sloughed larval epithelial cells (i.e., the meconium) appeared to have no major effect on bacteria trapped within the meconium. However, the dramatic decrease in bacterial count that occurred between old (38–48 h postpupation) pupae and newly emerged (0–10 h postecdysis) adults indicated that a bactericidal mechanism in the lumen acted around the time of adult emergence, but before the egestion of the meconial peritrophic membrane/meconium complex.
In the three species studied, representative of three major mosquito genera, and in most of the specimens tested, the midgut was nearly sterile or became nearly sterile in recently emerged adults. It seems clear from our results that a very effective antibacterial mechanism in addition to the sequestration (by the MPMs/meconium) and egestion process is operating in the species studied. It is interesting to note the following comment by Straif et al. (1998): “Even though bacteria could be isolated from each blood-feeding stage … , the overall bacteria prevalence of 50% in wild Anopheles may indicate that either only part of the mosquito population acquires bacteria or that bacteria are not stable midgut residents.”
The results of Walker and Romoser (1987) provide an important clue regarding a possible mechanism that could account for the dramatic decrease in the bacterial population that occurs around the time of adult emergence. Early in the ecdysial process, the pharyngeal and cibarial pumps are activated and there is evidence that exuvial (molting) fluid is ingested through the true mouth during the process of adult emergence. This is possible because the stylets of the proboscis remain in separate sheaths and the ecdysial space is continuous with the true mouth. As the exuvial fluid is ingested, gas from the external environment enters the tracheal system via the respiratory trumpets, inflates the newly formed adult tracheae, and then passes into the exuvial space via the newly formed adult spiracles. This is possible because the adult spiracles develop during the pupal period and open into the exuvial space. As the exuvial fluid is ingested, the gas in the exuvial space follows and eventually is ingested, inflating the midgut.
Ingested exuvial fluid, which includes enzymes that had previously digested the old endocuticle during the molting process, may be active in killing bacteria in the midgut lumen. If MPM1 is found to be permeable to these enzymes, it could help explain the low bacterial counts found in the meconium. It is possible that the formation of MPM2 is involved with the sequestration of these enzymes, after they have reduced the bacterial count in the midgut. However, MPM2 may be permeable to the enzymes. Detra and Romoser (1979) showed that the PM in larval Ae. aegypti is permeable to proteolytic enzymes found in the peritrophic space. However, because MPM2 may not always be present, it would not appear to be needed for this function. At the very least, it would provide an additional layer around any microorganisms associated with the meconium.
In addition to MPM formation and the likely swallowing of exuvial fluid, other processes (egestion of the larval food bolus and the cell sloughing that forms the meconium and eventual egestion of the MPM/meconium complex) occur within the mosquito midgut during the process of metamorphosis. Taken together, these processes probably contribute to the eventual reduction/removal of midgut bacteria. A summary sequence follows: (1) Before pupation, fourth instars egest the food bolus and then the peritrophic membrane, emptying the alimentary canal of most of the bacteria present from the larval environment. (2) The larval midgut epithelial cells slough during the pharate pupal/pupal stadium and the new adult midgut epithelium develops from regenerative cells. This would result in the retention of any microorganisms from the larval midgut lumen within the confines of the sloughed cells that are within the meconium. (3) MPM1 then is secreted and conforms to the shape of the meconium. This would further confine any bacteria found in the meconium and further protect the developing adult midgut epithelium from potentially harmful microorganisms. (4) Exuvial fluid is ingested as part of the ecdysial process and is followed by gas that expands the midgut. The exuvial fluid is a potential source of antimicrobial agents that may clear the adult midgut lumen and the meconium of microorganisms. (5) MPM2 forms around the meconium and may confine the ecdysial fluid thereby protecting the adult epithelial cells from enzymatic breakdown, further confining any remaining miroorganisms, and facilitating the egestion of the meconium and its microbial contents.
In addition to the possible antibacterial action of swallowed ecdysial fluid, it is conceivable that salivary gland secretions, known to contain a bacteriolytic factor (Rossignol and Lueders 1986), are released into the exuvial space and swallowed along with the ecdysial fluid.
Our model of bacterial (microbial) reduction/elimination in the newly emerged adult mosquito midgut appears to apply to Anopheles,Culex, and Aedes species and may apply to all mosquitoes. Further, it is possible that this model applies to many holometabolous insects, particularly those that have larvae that are generalist feeders in an environment highly contaminated with microorganisms, such as decaying leaves, dead animals, and pools of water. We currently are investigating the possible antibacterial effects of components of exuvial fluid as well as extending our studies to other kinds of insects.
The authors acknowledge the dedicated assistance of Kristine Oswald. The research reported here has been supported through grants from the Ohio University Research Committee.