On 2019-12-17 01:27:50, user Alex Terzibachian wrote:
BI 598 Group 5: Stephanie Yemane, Alex Terzibachian & Gabriela A. Rodríguez-Morales
Review written by undergraduate and graduate students from Boston University as requirement from the BI598 class
Summary
Microglia are cells derived from the mesoderm that function as the macrophages of the central nervous system. Even though their function has historically been linked to the immune system, recent studies suggest that microglia might play an important role in regulating synapse development during early developmental stages through synapses pruning, mediated by the complement pathway. However, it is still unknown if microglia perform the same regulating role in synapses of adult-born cells in the olfactory bulb. To answer this question, the authors of this paper ablated microglia in the olfactory bulb of mice and looked at the functional development of abGCs.
In figure 1, the authors observed microglia interaction with abGCs with the help of in vivo two-photon imaging. They observed that microglia interact more frequently, but not for a longer period of time, with mushroom spines of abGCs compared to filopodial spines. They also saw that the percentage of spines covered by microglial processes during interaction, both filopodial and mushroom were not different. This made them realize that microglia preferentially interact with mushroom spines on developing abGCs. The curves for both the data and offset points also seem to be very similar to one another in all the graphs, with similar cumulative probabilities for the different factors studied.
In figure 2, researchers next wanted to determine if these microglial interactions that were defined in Figure 1, are necessarily involved for the functionality of abGCs during development. To do this, they ablated microglia using chow formulated with a CSF1R inhibitor, PLX622. This inhibitor prevents the signaling required for microglia survival. Mice were started on chow, three weeks later researchers administered lentiviral tdTomato for abGC imaging, and about five weeks post-injection, mouse brains were analyzed via two-photon microscopy where they found that this induced ablation didn’t affect overall number of adult-born neurons in the olfactory bulb, nor the overall gross morphology when PLX mice were compared to controls. They next used calcium response recordings in the dendrites of abGCs to identify differences in responses across control and PLX groups. Researchers found that PLX mice had decreased responsiveness compared to control, they found that PLX mice had dendrites that responded to less odors and that the median lifetime sparseness was decreased as well. The data from these experiments suggest that ablation of microglia induce sparser representation of odors within PLX-treated mice.
In figure 3 they used 2-photon to image abGC dendrites in awake control and PLX-treated mice in order to compare the responsiveness, lifetime sparseness and median number of odors that elicited a response in the cells, to the results found in anesthetized mice. Results showed responses to a lower median of odors as well as lover overall responsiveness and lower lifetime sparseness. Principal component analysis also showed a significant change in response timecourse between awake and anesthetized mice. These results support the idea of an overall decrease in the proportion of responsive dendrites.
They then wanted to look at whether microglia ablation was specific to only developing abGCs, or whether they also affect mature abGCs. To test this, they used similar techniques in figure 4, as seen in figures 2 and 3, on three-month-old labeled abGCs. The Ca2+ heat map traces of the microglia treated with PLX5622 for the 16 different odors was graphed and seen to not have a significant difference in the distribution of responses. They also observed no difference in the number of effective odors, nor in the lifetime sparseness. Hence, they concluded that microglia ablation after development has no effect on odor-evoked response.
At this point, researchers had found that abGCs had reduced responses in the setting of ablated microglia. Next, they wanted to observe if excitatory synapses on abGCs were altered in PLX-treated mice relative to controls. Results of experiments conducted in figure 1 showed that microglia preferentially interact with the mushroom spines of the external plexiform layer of the olfactory bulb. In figure 5, they first aimed to analyze spines of apical dendrites of abGCs. They found that between PLX-treated and control mice, there was no significant difference in spine density, but they found the difference of mean head volume was somewhat significant.
In figure 6 they looked at the electrophysiological properties of abGCs by looking at sEPSCs and sIPSCs to depict the possible effect of spine head size between control mice and mice treated with PLX. To achieve this, the authors performed whole cell recording of labeled cells 5 to 6 weeks post injection. Results showed no difference in sEPSC frequency between both experimental conditions, however the amplitude of the sEPSCs was significantly decreased in PLX-treated mice. Results from the sIPSCs showed no difference in the inhibitory events between the control mice and the ones treated with PLX. These results suggest that functional differences between both experimental conditions could be due to weaker excitatory inputs onto abGCs.
In figure 7, the authors checked whether synaptic inputs were affected by microglia ablation. They did so by recording sEPSCs in abGCs of mice that underwent microglial ablation for three weeks after 3 months of maturation. After processing the raw traces recorded from abGCs in control and in PLX-treated mice in section B of figure 7, sections C and D show that there was no significant change in the frequency or in the amplitude of sEPSCs. This led them to believe that microglia ablation after abGC development has no effect on excitatory synaptic currents.
Merits
Sections C and H in Figure 1 are a great backbone to include in the figure, as it allows the reader to easily identify and differentiate between filopodial and mushroom spines of abGCs.
Figure 2 was very thorough and the supplementary figures were useful in backing their arguments. In supplementary figures 2.1 and 2.2 they revealed that ablation was successful and sustained with continued delivery of PLX-chow; thus, proving that levels of microglia were reliable and consistent. Providing the results for supplementary figure 2.3, was imperative in showing that the number of neurons was not altered and thus any further analysis comparing responses across control and PLX groups were valid.
Images in figure 5 were a very clear and concise depiction of the spines in question, scale bars and insets are clear.
Major Criticisms
The paper states that recordings performed in awake mice suggested that PLX-tretaed mice showed lower responsiveness, a lower median of odors and lower lifetime sparseness. However, only the change in fluorescence elicited by the odor was significantly different between both groups, the median of odors and lifetime sparseness was not significantly different. This begs the question if in fact there is a decrease in the number of responses dendrites or a change in the effect elicited by the odor in the responsive dendrites.
Another major criticism for this paper would be that the authors did not identify the 16 odors used to produce the heat maps seen in figures 2 and 4, nor mention why they were used. There could be other odors that could elicit different types of activity, and not having an explanation for using these doesn’t allow the reader to be for certain that there aren’t other patterns of activity. In giving an explanation the their choices, they can eliminate more doubt.
In the experiment in figure 5, researchers changed the experimental timeline – they administered the lentiviral injection before treatment with PLX chow, and analyzed mice brains four weeks after lentiviral administration and PLX treatment. This was odd as they suddenly flipped the protocol they were using before. Not much information was given regarding reasoning for timeline structure; could the injection have agitated microglia prior to PLX treatment? Researchers do mention that the lentiviral injection did not only target the under/developed abGCs; however, if the same protocol was used across experiments, they would have had the same baseline difference throughout – at least permitting for consistent data collection. For panels C and D in figure 5, analysis of spine head volume was presented as both averaged bar graphs and as a cumulative distribution; whereas analysis of spine density was only done via analysis of averaged spine density. The cumulative probability had reached statistical significance where the bar graph did not. Because of the statistical significance of the cumulative probability, researchers deemed there was an observed trend in increased spine density in PLX-treated mice. It is questionable as to why there was no cumulative distribution was done for differences in spine density across testing groups.
In addition to the electrophysiology shown in figure 6, the authors could have added a figure showing the fluorophore+ cells that they recorded from.
Overall, there should be more methodology included for the different protocol timelines used, in addition to including the motivation for the experiments conducted, and for changes in initial procedural timeline.
Minor Criticisms
The paper states that the animals were presented with a panel of 16 monomolecular odors, however, the data from figure 3 only includes responses to 15 odors. This should be revised to include the 16th odors or update the total number of odors included in the panel to 15.
A minor criticism for the first figure would be to include the Wilcoxon rank sum test values next to the graphs in the figure as well, in order to make the statistical evidence clearer when presenting the data.
In figure 2F-I, all data presented were deemed significant, although these images were depicted differently than 2H and 2I. For 2H and I, the individual data points were plotted in addition to the curve/bar graph which allows for readers to see the distribution of the individual data points in addition to the curve. However, for figures 2F and 2G the curves for control and PLX do not appear to be significantly shifted. Presenting these figures in the same manner would, or simply including the individual data points on the curve would make this figure more convincing in its efforts to depict the significance of the observed changes.
N sizes for supplementary figures 2.1 and 2.2 were rather small. It was useful to see the sustained ablation of microglia with continued PLX delivery – however, where a maximum of three mice were observed throughout development (up to nine weeks), what is the ratio of failed/successful attempts of ablation.
The n size for figure 5 could be larger to emphasize the differences across control and PLX groups. Researchers analyzed almost 1,000 spines in about twelve cells from each group. Although researchers confirmed the ablation of microglia by immunostaining, and confirmed the sustained depletion of microglia for up to nine weeks with continued PLX delivery, they had only confirmed doing so in 3-4 mice through up to nine weeks of constant PLX delivery (Fig2.1, Fig 2.2). If they could confirm this maintained ablation in more mice, could a larger n be used in this figure to demonstrate whether there are significant differences in spine density/volume that are outside of inherent compensatory mechanisms.
For the electrophysiology experiments in figure 6, the experimental group should be blind to the experimenter and data taken at different times post slicing should be shown in order to neglect any contribution of recording time to electrophysiological properties of the cells.
Another minor criticism would be to include error bars on bar graphs in all figures. This would allow the reviewer to look at the sparsity of all the raw data and visually see on the graphs how precise those averaged values are.
Future Directions
Researchers characterized the effects of PLX induced microglia ablation; however, it would be interesting and more revealing to observe the remaining microglia. Is there functionality the same? Does this induced depletion of microglia promote the signaling of another response pathway that could have an effect on the connectivity of these neurons in the olfactory bulb that they were observing?
It would be interesting to see the effects of stopping ablation by discontinuing chow, and further observe the induced changes of discontinuing this ablation. This would allow for a more full picture of the role of microglia in the integration of abGCs in neuronal circuitry of the olfactory bulb during development. If chow was stopped after the developmental stages; would the deficits observed still remain? Would the re-introduction of microglia contribute to more efficient and specific responses, meaning cells could still be integrated into the circuitry and strengthen certain connections?
Another future direction could be to look further downstream at the mitral cell level or even Piriform cortex and see what possible effects microglia ablation could have in odor representation.