Validation by qPCR revealed that in cell culture induced to overtly express miR-932, sick was unaltered, whereas fitm was downregulated. Akin to ChAT under miR-190 overexpression, Srpk79D was an outlier. The surging of its levels suggest that the gene not only bypasses the canonical mode of miRNA-based regulation, but does so effectively. Compared to miR-190 targets, however, the functionalities of these miR-932 targets are more logically linked to truncated cytoophidia typical in ovaries overexpressing the miRNA. The prospect of there being a relationship between fitm and CTPsyn-cytoophidia localization within the cytosol is consequently very interesting. Through qPCR, CTPsynIsoC levels were determined to not be affected by miR-932 overexpression. Macro-cytoophidia are indeed shorter in nurse cells of such-genotyped females. However, we have additionally noted that often times, this phenotype was accompanied by greater micro-cytoophidia numbers. In other words, the shortening of cytoophidia may be attributed to a failure of monomers to co-localize and tether, rather than a downregulation of CTPsyn expression. As fitm levels saw significant downturns as miR-932 became more available, we hereby hypothesize that the roles its protein play in cellular transport may also include dictating the localization patterns of cytoophidia.Target Prediction Outcomes of miR-975
Mir-975 is a rather novel miRNA. Thus far no concrete reports of any regulatory roles the miRNA may have are available. The miR-975 gene-cluster may also be genus specific: on miRBase, it has been annotated only for six Drosophila species (Kozomara & Griffiths-Jones, 2011). This means that information regarding the probable functions of the small-RNA is scarce, although in silico target prediction (Kheradpour et al., 2007) has indicated that many of the mRNAs it might elicit are involved in membrane-trafficking and organization, cell-cycle regulation and DNA-repair. Endogenously, its activity may be rather limited. Sensor-based assays have shown how, along with miR-932, miR-975 presence in vitro is lowest amongst the for prime candidate miRNAs. Coincidentally, these two also had fewer numbers of predicted targets amongst the four. The phenotypes manifested through its overexpression nonetheless brings about exciting prospects to the possible functions of miR-975. As its cytoophidia-elongating properties expressively suggest that CTPsyn is not directly regulated by the miRNA, it implicates the gene and its protein product in a co-regulatory network of a much grander scale instead.
The three ostensible targets of miR-975 shortlisted for validation are klaroid or koi, mthl1, and split-ends or spen. In 975-ov cell culture, whereas mthl11 remained unchanged, both koi and spen levels were very clearly downregulated. This may explain death-curve assay outcomes of miR-975 overexpression, where cell death was accelerated under either conditions of higher construct dosages or past a certain time threshold of incubation. Presumably, under the exact opposite conditions brought about by sponging i.e. in a miR-975-null environment, the upregulation of koi and spen would occur in its place. Dysregulation of either gene is detrimental to the developing fly (Kracklauer et al., 2007; Mace & Tugores, 2004). The Spen family of proteins is known to have especially vital roles in development. It directs temporally critical events within Notch and Wnt-signalling pathways. As a transcriptional repressor, Spen activity prevents uncontrolled proliferation. However, as these proteins continue to accumulate disproportionately within the 975_sp cell, the negative effects they would have upon growth might have eventually compromised cellular survival as a whole (Chang et al., 2008; Hazegh et al., 2017; Jin et al., 2009).
It is important to distinguish here that the concurrent downregulation of koi and spen does not appear to be causal to the lengthening of cytoophidia in ovaries. In fact, quite the opposite is true. Amongst the candidates, only in the instance of miR-975 overexpression were such polarizing differences observed in vivo to in vitro. Where they were very much reduced in cell culture, koi levels remained steady in Act5c>miR-975 ovarian tissue. The disparity in spen levels was most stark, as its downturn in cells is displaced by a modest upregulation in ovaries. CTPsynIsoC expression was also stimulated in these tissues, where it was unaffected by miR-975 activities in vitro. Together, these results effectively excludes koi as an effector of CTPsyn and its pathways. We can also conclude that miR-975 may not directly elicit spen after all; its downregulation in cell culture may instead be a consequence of upstream events taking place. Many questions remain. One is whether CTPsynIsoC upregulation directly translates into greater protein production, and therefore longer cytoophidia. If so, was CTPsynIsoC hyper-expression triggered by the nucleic-acid binding, signal-transducing properties of Spen proteins? Or is the phenotype orchestrated by completely unknown factors, which only come into effect when miR-975 is overexpressed in a gonadal micro-environment? Further investigation is hence required, as we certainly cannot provide answers to these complex problems within the current information capacity.
Target Prediction Outcomes of miR-1014
Like miR-932 and miR-975, miR-1014 orthologues are yet to be found in organisms outside of the insect class (Kozomara & Griffiths-Jones, 2011). As a result, it is presently understudied, and its functions are thus far unknown. However, it shows bearings of a potentially active miRNA. MiR-1014 is predicted to target over 380 genes within the D. melanogaster transcriptome, with affinities to over 500 unique binding sites. Sensor assays have also demonstrated that miR-1014 is naturally expressed at greater levels than miR-932 and miR-975 both, though to a lesser extent to that of miR-190.
Expressional changes of the three chosen targets of the miRNA show that predictions were on the nose. Whilst NetB levels were unperturbed, both CrebA and CdsA were dramatically downregulated in cell culture. Unlike with miR-975 targets, this trend was mirrored in Act5c>miR-1014 ovarian tissues as well. For CTPsyn, it was such an encouragement to see that not only does this miRNA have the potential to bind to sites on CdsA in particular, its overexpression does indeed bring about a downturn in the gene’s expression. This is because of the roles CdsA play as an enzyme with direct connections to CTP through the phospholipid synthesis pathway (Liu et al., 2014). Specifically, it catalyses the condensation of phosphatidic acid and CTP into the starting molecule within phospholipid production i.e. cytidine diphosphate diacylglycerol (CDP-DAG). We hypothesize that as cellular CdsA is depleted, accumulation of under-utilized CTP confers a negative feedback mechanism upon CTPsyn (Aronow & Ullman, 1987). These proteins becomes inactivated and polymerize into filaments to avoid degradation, ultimately producing the elongated cytoophidia phenotype characteristic of miR-1014-overexpressing ovaries.
The Low Probability of a Direct-binding Inhibitory Relationship Between CTPsynIsoC and Any of the Prime Candidate miRNAs
Apart from miRNA target prediction, we also had CTPsynIsoC analysed by a variation of the TargetScanFly algorithm for miRNA-binding sites. The ORF-scanner reads the entirety of a gene of interest, identifies its 3’UTR, and assigns potential miRNAs to regions within the sequence in the case of seed complementarity (Schnall-Levin et al., 2010). A conserved site is described by Schnall-Levin et al. as those which are conserved beyond amino-acid coding. This means that seed sequences found on a single gene is the same or nearly identical even cross-species and extensively diverged lineages. Another determinant to conservation is branch-length scoring, based on the sum of phylogenetic branch lengths between species containing a specific site (Kheradpour et al., 2007). Under this rule, a site is only deemed conserved if it has a branch length score of above 60% or 3.16. Although two of the four prime candidate miRNAs (miR-975 and miR-932) were each predicted to have a binding site directly on the 3’UTR of CTPsynIsoC, neither was conserved outside of the Sophophora subgenus. In actuality, several of the outcomes put forth in this chapter discourages the notion that either of these miRNAs comply with this prediction.
One of the assays conducted was designed to assess whether prime candidates directly bind to 3’UTR region of CTPsynIsoC. Double-transfection involves a miRNA-overexpression construct, introduced together with a plasmid encoding for an eGFP gene modified to end with the 282-bp 3’UTR sequence of CTPsynIsoC. Significant signal attenuation of eGFP away from controls shows that a miRNA might indeed have a compatible seed sequence along this region, and therefore could be a direct regulator of CTPsynIsoC. The results of this assay demonstrated that miR-975 and miR-1014, and not miR-932, are the ones are most likely to have this relationship with CTPsynIsoC. And yet signal reductions were subtle at best. Furthermore, CTPsynIsoC levels were unaffected in all miRX-ov cell cultures. These little clues allude to the possibility that even if any of these miRNAs were to truly regulate the gene in vivo, effects were expected to be modest and non-extreme. Alas, subsequent qPCR analysis of miR-975-overexpressing ovarian tissues quickly debunked its case as a canonical regulator of CTPsynIsoC. Its expression levels not only saw an upturn under such conditions, but the upregulation was significantly drastic. We are hereby more inclined to believe that though CTPsyn and cytoophidia formation truly do respond to miRNA-overexpression events, the gene itself is not directly regulated by the candidates under evaluation here. The naturally low expression levels of both miR-975 and miR-932 – as the two most likely to target CTPsynIsoC – only provides more support to this assumption. As a matter of fact, if this is to be viewed in lieu to its poor targeting by known D. melanogaster miRNA families, it is not a stretch to say that CTPsyn may not be implicitly bound by any miRNAs at all.
Overall, the stringent algorithm of TargetScanFly listed only the potential interactions between five miRNAs and specific sequences within the 3’UTR region of CTPsynIsoC to be of any merit. It is important to note that in this case, only the canonical aspects of miRNA-based regulation are considered. For one, the traditionally accepted view has always been that miRNA action would often negatively affect the levels of its mRNA targets. However, this has been challenged, as several miRNA species are found to induce, rather than reduce genetic expression (Catalanotto, Cogoni, & Zardo, 2016; Ørom, Nielsen, & Lund, 2008; Place et al., 2008). Additionally, emerging evidence over the past years has dictated that these small RNAs may assert its regulatory functions by binding to intronic, coding regions, and 5’UTR targets (Lytle, Yario, & Steitz, 2007; Qian et al., 2016). These outlying binding sites were found to be distinctively enriched in D. melanogaster, too (Schnall-Levin et al., 2010). Consequently, with the whole open-reading frame of CTPsynIsoC included for target prediction, thirteen miRNAs were identified as potential regulators of the gene. In the future, exploring these routes of in silico target prediction as well may be in the best of interests, so that no avenues are left unaddressed.
Influence of Fat Metabolism Pathways in Cytoophidia Formation
It became increasingly clear as we proceeded through in silico target prediction and qPCR that there are strong connections between fat metabolism and the cytoophidium. This should not come as a surprise, as CTP is known to be a required substrate in phospholipid production (Kennedy & Weiss, 1956). Two very significant expressional changes observed in regards to miR-932 and miR-1014 could be related back to this pathway and each miRNA’s shortening and elongating effects upon cytoophidia, respectively, and we believe that these may provide new insights into the regulation of cytoophidia formation.
Overexpression of miR-932 induces a severe downturn in fitm mRNA levels in both in vitro and in vivo tissue. The fat-storage inducing transmembrane protein gene of D. melanogaster (dme-fitm) is a homologue of the human FIT2 gene. The latter is a part of a large family of proteins found primarily along the endoplasmic reticulum (ER) membrane (Kadereit et al., 2008). Its name suits its fat-safekeeping function: FIT2 mediates the partitioning of triglyceride (TAG) fatty acids from the ER into lipid-rich cellular organelles called lipid-droplets (LD) when fat is abound within the cell (Guo et al., 2009). Where its overexpression increases LD numbers, FIT2-knockdown, though non-lethal, reduces the accumulation of LDs rapidly. Meanwhile, overexpression of miR-1014 downregulates CdsA. The relationship between its enzyme product, CTP and initiation of phospholipid synthesis has already been discussed in the previous section (Liu et al., 2014). We have also related how this encourages filamentation of CTPsyn into cytoophidia. Nevertheless, the most crucial aspect of the discussion here is how this situation undermines the ability of the cell to generate phospholipids, and what that ultimately means to other cellular processes.
Carbohydrates are often first in line to be consumed for energy in a majority of organisms including D. melanogaster. However, upon a prolonged period of starvation, TAGs usually evolve to displace carbohydrate reserves as the main energy source (Owen et al., 1998). To replace metabolised TAGs, parts of the plasma membrane break off to form autolysosomes containing phospholipids, which are promptly rerouted and stored within LDs. These events explain why, in starved cells, not only are LDs larger, but greater numbers of them are found (Rambold, Cohen, & Lippincott-Schwartz, 2015).
It is widely acknowledged that whilst CTPsyn forms the bulk of its cytoophidia, the protein is not the sole component of the filament (Liu, 2011). In itself the structure’s discovery was rooted to the utilization of an anti-Cup, and not anti-CTPsyn, primary antibody, demonstrating that at the very least, there is one non-CTPsyn constituent of cytoophidia (Liu, 2010). Discernible gaps are also seen along the nurse cell macro-cytoophidia in greater resolution images, in yet another sign of its non-homogenous nature.
This is where starvation and subsequent fat metabolism may bear links to cytoophidia formation. Under miR-1014 overexpression, the downregulation of CdsA eventually depletes ovarian cells of their phospholipid reserves. This creates an environment mimicking starvation which, as aforementioned, culminates in the depletion of phospholipids to compensate for the loss of other fat resources. A previous study has found that elongated cytoophidia were formed in the brains of starved larvae, and at greater frequencies (Aughey et al., 2014). True enough, miR-1014-overexpressing egg chambers hold much longer macro-cytoophidia, with the occasional increment in numbers of micro-cytoophidia observed in follicle cells. Subsequently, the same study showed that once larvae has been refed, cytoophidia disassembling occurred quickly. It can be said that this is mirrored in the truncation of cytoophidia under miR-932 excess: Act5c>miR-932 tissues see much reduced FITM protein levels, thus mimicking conditions of both high-fat food consumption and FIT2-knockdown (Miranda et al., 2014).
CTPsynIsoC levels in either case of miRNA overexpression were already revealed to be undisturbed. This means that modifications to cytoophidia length are not due to greater or reduced CTPsyn-protein presence, but are artefacts of the polymerization and tethering process. Starvation dynamics explained above dictate that the factor most likely to be changing in parallel to these opposing displays of cytoophidia behaviour are LDs. Apart from lipid storage, these membrane-enclosed organelles have a demonstrated ability to contain and interact with numerous proteins such as histones and membrane-trafficking enzymes (Cermelli et al., 2006). Droplet-recruitment also protects a protein from degradation (Welte, 2007); which is suspected to be one of the main objectives for CTPsyn-filamentation in the first place. Taken together, these are the reasons why we believe LDs to be a literal missing piece of the cytoophidium puzzle. Co-immunoprecipitation, TEM-visualization of tagged-TAG localization patterns, as well as generation and subsequent CTPsyn-centric immunostaining of tissues from fitm-knockout mutants, could be worthwhile endeavours towards figuring out whether this hypothesis holds true.