This web page was produced as an assignment for an undergraduate course at Davidson College.

Review Paper 1

Drosophila Microbiome Modulates Host Development and Metabolic Homeostasis via Insulin Signaling




I found this article to be interesting and, for the most part, thorough. The researchers approached the microbiome subject from a refreshing angle. While I have seen papers where the researcher will take bacteria from a gut and analyze its needs in vitro, this team came at the problem from the opposite angle. They looked at which bacteria are vital to the host rather than what substances are vital to the bacteria. As the paper claims, they established a new model for these kinds of studies by finding a monoassosciation with a bacterium that is necessary and sufficient to essentially restore development for conventionally reared flies. One aspect of the paper that frustrated me (which is further mentioned in some of the figure summaries) was when they would statistically compare the two groups that they thought were important to show a reduced phenotype and restoration but not other groups. I would have preferred a simple ns for "not significant" or something along those lines rather than an absence of analysis. For example, in Figure 3b, the graph shows that the time to puparium formation is greatly increased with the monoassosciated mutant and greatly rescued (lowered) by DILP2. However, the bar for wt and the bar for the DILP2 treated ones look significantly different but are not marked as such. It is okay for the data to not show a 100% rescue, but I would prefer to know that over assuming it. Beyond that gripe, I thought the connection from gut bacteria to diabetes (at least in flies) was intriguing and a good area for further research. The researchers showed a lot of different developmental aspects that were affected in the mutants including some I never would have thought of (cell size). I would be curious to know what implications this paper has on our understanding of diabetes in humans.


This paper looks at the bacteria that inhabit the gut or microbiome of Drosophila melanogaster. The researchers begin by establishing optimal growing conditions and characterizing the bacteria that do exist in the gut of the fly. There are five primary species: Commensalbacter intestini, Acetobacter pomorum, Gluconobacter morbifer, Lactobacillus plantarum, and Lactobacillus brevis. The researchers then grew flies that were germ-free with the use of antibiotics and looked at how well they survived in different conditions. The germ-free bacteria grew more slowly as the amount of yeast in their diets decreased. In addition, none survived on a casamino acid diet when germ-free. The researchers then performed an experiment to see if one or all of the bacteria that were wiped from the midgut were responsible for the lethal phenotype. By transplanting in the missing bacteria one at a time (monoassociation), the researchers found that A. pomorum was sufficient to restore the phenotype of the conventionally reared flies. They then created mutant strains by introduction of transposons to A. pomorum and figured out where the lesions were created. The mutants that were the least beneficial to the hosts had mutations in the pyrroloquinoline quinone-dependent alcohol dehydrogenase (PQQ-ADH). The team also sequenced the genome of the A. pomorum (roughly 2.8Mb). The researchers suspected a similarity between A. pomorum and a bacteria found in mosquitoes that provides the host with vital sugars based on the in vitro morphology of the bacteria. They then performed a series of experiments to determine the effect of the P3G5 mutant (mutation in PQQ-ADH) on the host fly. They found that flies monoassociated with the mutant took longer to form a puparium, weighed less as adults, had smaller wings, fewer cells, smaller cells, elevated sugar levels, and elevated lipid levels. The flies were essentially malnurished. They confirmed the association to the diabetic mosquitos by rescuing the flies infected with the mutants with DILP (Drosophila insulin-like peptide) and acetic acid (.2%). This study showed that microbiota can have a profound effect on host development and morphology. In addition, the researchers created a new model for host/microbiota interactions with the monoassociation of D. melanogaster and A. pomorum.

Figure 1

A)The graph displays the different diets the researchers experimented with in order to establish the best diet for rearing microbiota monoassociated Drosophila. Lowering the yeast percentage in the diet appeared to have a marked effect on growth rate. Namely, the lower the yeast percentage in the diet, the longer it took for the larvae to develop a puparium. However, the display does not show significant differences between diet percentages, only between conventionally reared and germ-free reared flies within those groups. The disparity between those two conditions within each diet does appear to get larger as yeast % decreases. 0.1 % yeast was lethal in germ-free pupae as was casamino acid diet pupae that were germ-free. The microscope images display the reduction in size of the pupae when they are germ free in both 2% and casamino acid diets 120hrs after laying

B)This image shows that only A. pomorum monoassociated flies form a puparium at a rate that is not significantly different from conventionally raised as well as fly guts restored with all five microbiota species. They will use this species in their study. They also show the size of the larvae at 120 hours after egg-laying to show that A. pomorum are the ones most similar to wt.

C)This image just shows the range and spread (box and whisker plots) of the different tronsposon-created A. pomorum mutants with relative time to puparium formation on the y axis. The wt time to puparium formation was designated as 1 and all mutants were compared to that. All mutants took longer to form a puparium than the wt. P3G5 is the mutant they will use in this study as it has a mutation in the PQQ-ADH1 gene (gene of interest) and it colonizes the host gut about as well as A. pomorum (shown in supplemental table).

Figure 2

A)The images are photos of larvae at 120 hours after egg-laying and the graphs display box and whisker plots of the day at which a puparium was formed in each condition. For both casamino acid and 0.1% yeast, the wt and P3G5 mutants differed significantly (indicated by triple astericks) in days to puparium formation. This is to be expected since mutants have deficient PQQ-ADH. The mutants take longer to make a puparium.

B)The trend continues into adult flies (5 days old). This is a similar graph and images of adult flies. There is significant difference in both male and female flies with regards to weght. The mutants way significantly less.

C)They used adult female flies and analyzed wing area, cell number, and cell size. The mutants showed reduced phenotypes in all these categories. It is curious that they didn’t show male data as well.

D)The mutants had elevated sugar levels and lipid levels as well to a significant degree. This is indicative of the mutants’ similarity to a diabetic phenotype associated with insulin-like growth factor mutations.

Figure 3

A)Many of the images in figure 3 are rescue experiments. The researchers are using the mutants (P3G5) which defective PQQ-ADH to show some effect on normal phenotype and then rescuing the phenotype with ectopic Drosophila insulin-like peptides (to show a similarity between a diabetic phenotype and the phenotype caused by the mutant. In this image, the researchers note the mutant’s localiztion of a transcription factor, dFOXO, into the nucleus which is abnormal. The third image shows the wt phenotype rescue with DILP expression as the cells appear to have dFOXO in the cytoplasm rather than the nucleus.

B)The images are of the wt, mutant, and rescue at 120 hours after egg-laying, respectively. The graph quantifies the days it took to make a puparium for these larvae. Again, I would like to see comparison between the wt and the rescue. The days do jump significantly up with the mutant and significantly down with the rescue.

C)The images are adult flies at 5 days for visual comparison of size. Again, the researchers take female weight as the standard and quantify the data graphically, but without any indication of statistical comparison between the wt and the rescue. The P3G5 mutants do have significantly less weight and the rescues have significantly more weight than the mutants.

D)The trend continues with wing area, cell number, and cell size. The mutant has a reduced phenotype in all those categories and a significantly greater phenotype with the DILP rescue (when compared to the P3G5 mutants). The images are of adult female fly wings.

E)This is a box and whisker plot of the sugars and lipids of the three categories. The mutants have significantly higher levels of both when compared to wt (large variance in triglyceride levels) and the rescues had significantly lower levels of both when compared to the mutants.

Figure 4

A)This figure shows that acetic acid rescues the phenotypic changes caused by P3G5. This makes sense since PQQ-ADH has acetic acid as one of its byproducts. These images visually show a rescued phenotype when looking at a membrane bound protein called PH-GFP and the localization of a transcription factor dFOXO within the cell(DAPI just stains DNA).

B)This graph shows the acetic acid rescue in respect to time to puparium formation. They also included a control to show that the acetic acid is sufficient to restore the wt phenotype, A. pomorum must be present. The images show larvae at 120 hours after egg-laying. The mutants without acetic acid take much longer to form a puparium while the ones with acetic acid take much less time (as compared to the mutants). The germ-free larvae die even with acetic acid supplementation.

C)The rescue works for both male and female body weight of adults (5 days).

D)Once again, the acetic acid (.2%) rescue restores the wt phenotype for wing area, cell number, and cell size. The wing images are of adult female flies.

E)Lastly, acetic acid appears to rescue the wt phenotype with regards to sugar and lipid levels.


Work Cited

Seung Chul Shin, Sung-Hee Kim, Hyejin You, Boram Kim, Aeri C. Kim, Kyung-Ah Lee, Joo-Heon Yoon, Ji-Hwan Ryu, Won-Jae Lee. 4 November 2011. Drosophila microbiome modulates host developmental and metabolic homeostasis via insulin signaling. Science. Vol. 334.670-674



Shuman Home Page

Genomics Front Page

Davidson College Biology Department


© Copyright 2012 Department of Biology, Davidson College, Davidson, NC 28036
Send comments, questions, and suggestions to: