InteractiveFly: GeneBrief

flower: Biological Overview | References

BIOLOGICAL OVERVIEW

Cell competition promotes the elimination of weaker cells from a growing population. This study investigate how cells of Drosophila wing imaginal discs distinguish 'winners' from 'losers' during cell competition. Using genomic and functional assays, several factors implicated in the process were identified, including Flower (Fwe), a cell membrane protein that is conserved in multicellular animals and proposed to be a Ca2+ channel in neurons (Yao, 2009) conserved in multicellular animals. The results suggest that fwe is a component of the cell competition response that is required and sufficient to label cells as 'winners' or 'losers.' In Drosophila, the fwe locus produces three isoforms, fwe^ubi), fwe^Lose-A, and fwe^Lose-B. Basal levels of fwe^ubi are constantly produced. During competition, the fwe^Lose) isoforms are upregulated in prospective loser cells. Cell-cell comparison of relative fwe^Lose and fwe^ubi levels ultimately determines which cell undergoes apoptosis. This "extracellular code" may constitute an ancient mechanism to terminate competitive conflicts among cells (Rhiner, 2010).

Fwe mediates win/lose decision by means of three differentially expressed isoforms, fwe^ubi, fwe^LoseA, and fwe^LoseB. Cells are identified as losers when relative differences of fwe^ubi or fwe^Lose levels are detected. This system bears the advantage that cells are able to survive general stress conditions that uniformly affect the entire population within a compartment. It is proposed that, in outcompeted cells, the fwe transcript is alternatively spliced and fwe^Lose isoforms are induced at the expense of fwe^ubi. Probably both, the downregulation of fwe^ubi, as well as the upregulation of fwe^Lose contribute to establish the lose/win decision. It is not yet known how the alternative splicing is regulated. The simplest possibility is that when cells competing unsuccessfully for extracellular resources are deprived of survival factors (Diaz Moreno, 2005), they are also depleted from some crucial splicing factors and default splicing will result in the formation of the normally repressed Lose forms. The observation that fwe^Lose upregulation was usually detected throughout the entire loser clone and not just at clone borders could be the consequence of a mechanism that propagates the "loser" state in outcompeted clones (Rhiner, 2010).

Two hypotheses are considered as likely: a cell-to-cell signal that efficiently transmits the "Lose verdict" among outcompeted cells. Alternatively, border cells may transiently increase their uptake of survival factors such as Dpp, for example by generating cytoneme-like extensions, which would further deplete survival factors in the interior of the loser clone (Rhiner, 2010).

fwe shares certain features with proapoptotic or growth promoting genes with respect to cell competition, but overall it behaves differently and seems to stand in a class of its own (Rhiner, 2010).

Genes mediating apoptosis (hid, reaper) show a similar behavior to fwe^Lose in certain aspects, in that they are triggered in loser cells and their elimination inhibits cell competition-induced apoptosis. Likewise, fwe^Lose can trigger cell death in clones in the absence of cell competition. However, such proapoptotic factors induce apoptosis when overexpressed ubiquitously, whereas overexpression of fwe^Lose (or lack of fwe^ubi) throughout the wing imaginal disc or in the entire fly does not interfere with cell viability nor organ size. This context-dependence implicates that fwe does not work as a simple killing signal or some sort of toxic protein acting cell autonomously (Rhiner, 2010). Fwe also shares features with genes known to affect normal tissue growth like Minutes (M/+) or dmyc such as cell-nonautonomous effects on survival in a heterotypic background. Homozygously mutant fwe cells show normal survival when all cells of one compartment are of the same genotype, but they are forced to undergo apoptosis when surrounded by wt cells, a hallmark of cell competition. However, this death does not depend on growth differences: (a) fwe-/- cells are forced to activate caspase-3 in the presence of Minute cells, which have a lower proliferation rate, but do express fwe^ubi; and (b) removal or downregulation of fwe throughout a compartment specifically inhibits cell competition without affecting the growth rate of the whole compartment (Rhiner, 2010).

It has been proposed that fwe is a calcium channel (Yao, 2009). However, during cell competition antagonistic functions are observed for the different isoforms, unlike in synaptic vesicles where the two isoforms that that are called Ubi and Lose A in this study seem to be functionally equivalent (Yao, 2009; Rhiner, 2010 and references therein).

Finally, the data show that fwe^Lose is not just an 'eat me signal' because fwe Lose forms are able to trigger Caspase-3 activation and cause cell death before and in the absence of functional engulfment (Rhiner, 2010).

It is proposed that within a multicellular organ cells are constantly tagged by extracellularly exposed fwe epitopes that function as a code. This extracellular code is composed by different fwe isoforms and allows comparison of relative fitness. During cell competition in Drosophila, the fwe isoforms work as a simple ternary code (fwe^ubi, fwe^Lose-A, and fwe^Lose-B) with a binary output, because fwe^ubi is translated as 'intact cellular fitness' whereas fwe^Lose-A and fwe^Lose-B are redundant and lead to cell elimination. The experiments with the C-terminally truncated fwe form suggest that the presence of the Lose epitopes aid in the labeling of cells as losers, although the lack of the ubi tail may also help in their elimination (Rhiner, 2010).

Other molecules are expected to interact with fwe, which are able to interpret the thresholds or read the extracellular epitopes displayed by the fwe isoforms. It is likely that the signal recognized by neighboring cells includes not only the variable C-terminal epitopes, but also the constant extracellular loop because its sequence is conserved from flies to humans (Rhiner, 2010).

The 'code' composed by the fwe isoforms may have biomedical implications beyond cell competition because imbalances in cell fitness appear during aging, cancer formation, and metastasis (Rhiner, 2010).

Drosophila SPARC is a self-protective signal expressed by loser cells during cell competition

During development and aging, animals suffer insults that modify the fitness of individual cells. In Drosophila, the elimination of viable but suboptimal cells is mediated by cell competition, ensuring that these cells do not accumulate during development. In addition, certain genes such as the Drosophila homolog of human c-myc (dmyc) are able to transform cells into supercompetitors, which eliminate neighboring wild-type cells by apoptosis and overproliferate, leaving total cell numbers unchanged. This study identified Drosophila Sparc as an early marker transcriptionally upregulated in loser cells that provides a transient protection by inhibiting Caspase activation in outcompeted cells. Overall, the unexpected existence of a physiological mechanism is described that counteracts cell competition during development (Portela, 2010).

This study describes the existence of a physiological mechanism that counteracts cell competition. Evidence is provided that dSPARC is a specific marker of cell competition, and not a general marker of apoptosis. Transcriptional activation of dsparc sets a higher threshold for Caspase activation in loser cells, possibly by inactivating an unknown secreted Killing Signal (KS), which is produced upon survival factor withdrawal. dSPARC is not a general inhibitor of apoptosis, despite its potent inhibition of cell competition-induced cell death. dSPARC may allow useful cells to recover from transient and limited damage before they are unnecessarily eliminated by their neighbors. These results show that dSPARC and Flower (Fwe) function in parallel and opposing pathways during cell competition, with dSPARC providing transient protection, whereas the 'Fwe Code' promotes cell elimination by labeling cells as 'losers' (Rhiner, 2010). Therefore, it seems likely that during early stages of cell competition, the decision of whether the potential loser cell will finally undergo apoptosis or not is still reversible. This intermediate state, where dSPARC protects outcompeted cells, may prevent the removal of valid cells that suffer only a temporary fitness deficit. However, if the differences in cellular fitness persist and/or are too ample, cell competition-induced apoptosis is, nevertheless, triggered (Portela, 2010).

One possibility is that secreted dSPARC blocks the unknown KS(s) directly in the extracellular space. dSPARC could bind directly to the KS(s) or just form a matrix that serves as a barrier for the KS(s) to reach the loser cells. The other possibility is that dSPARC could activate a protective pathway in an autocrine way that counteracts the effects of the KS. For example, mammalian SPARC has been shown to protect cells from apoptosis in vitro via activation of integrin-linked kinase and AKT. The identity of the killing cell(s) is not yet known (Portela, 2010).

If cell competition is conserved in mammals, this role of dSPARC specifically repressing cell competition may have important consequences for understanding of mammalian development, homeostasis, stem cell replacement, or cancer. In particular, deregulation of this mechanism is likely to be important in cancer, for example by allowing metastatic cells to survive in a new environment or during the expansion of cancerization fields (Portela, 2010).

'Fitness fingerprints' mediate physiological culling of unwanted neurons in Drosophila

The flower gene has been previously linked to the elimination of slow dividing epithelial cells during development in a process known as 'cell competition.' During cell competition, different isoforms of the Flower protein are displayed at the cell membrane and reveal the reduced fitness of slow proliferating cells, which are therefore recognized, eliminated, and replaced by their normally dividing neighbors. This mechanism acts as a 'cell quality' control in proliferating tissues. This study used the Drosophila eye as a model to study how unwanted neurons are culled during retina development and find that flower is required and sufficient for the recognition and elimination of supernumerary postmitotic neurons, contained within incomplete ommatidia units. This constitutes the first description of the 'Flower Code' functioning as a cell selection mechanism in postmitotic cells and is also the first report of a physiological role for this cell quality control machinery. These results show that the 'Flower Code' is a general system to reveal cell fitness and that it may play similar roles in creating optimal neural networks in higher organisms. The Flower Code seems to be a more general mechanism for cell monitoring and selection than previously recognized (Merino, 2013).

The 'Flower Code' was originally described in the Drosophila imaginal discs as a mechanism used by proliferating epithelial cells to recognize, eliminate, and subsequently replace slow dividing cells during tissue growth, a phenomenon known as 'cell competition.' The flower (fwe) locus gives rise to three isoforms of a cell membrane protein with three transmembrane domains: Fwe^Ubi, Fwe^Lose-A, and Fwe^Lose-B (see Flower Isoform Expression in the Developing Fly Retina). The three isoforms share the N-terminal and transmembrane domains and differ solely in their C-terminal extracellular part. During development, Fwe^Ubi is ubiquitously produced in the wing imaginal disc, but if slow dividing cells appear as a consequence of mutations, the slow dividing cells downregulate Fwe^Ubi and upregulate the two Fwe^Lose isoforms (Merino, 2013).

In Drosophila, the Fwe^Lose isoforms are sufficient and necessary to initiate the recognition and subsequent elimination of slow dividing cells by programmed cell death. Such elimination and replacement is triggered if 'fitter' cells are present to substitute the slow dividing ones. Therefore, it has been proposed that a differential display of extracellular Flower domains indicates cellular fitness. There is evidence that the mechanism may be conserved in mammals in which at least four different Flower isoforms are present (Petrova, 2012), among which two isoforms seem to specify the 'Lose-fate' (Merino, 2013).

This study asked whether the 'Flower Code' is restricted to the elimination of slow dividing cells caused by somatic mutations or if it could constitute a more general mechanism for cell recognition and selection occurring also in postmitotic cells (e.g., neurons) and in the absence of somatic mutations (Merino, 2013).

To this end, the Drosophila retina was used as a model to study which genes contribute to the removal of supernumerary sensory neurons. The Drosophila eye consists of 800 ommatidia in which each unit is formed by eight photoreceptor neurons (R1-R8), four cone cells, and is surrounded by secondary and tertiary pigment cells. At the periphery of the retina, incomplete ommatidia are formed, probably because not enough neurons can be recruited to build a complete unit. It is believed that such rudimentary photoreceptors are physiologically eliminated during development at the pupal stage in order to purge nonfunctional connections that could interfere with the correct perception of the environment (Merino, 2013).

In order to test whether the Flower Code was involved in the recognition of such incomplete units, different Flower isoforms were genetically downregulated or overexpressed in the Drosophila retina. This study shows that the Flower isoforms are required and sufficient for the recognition and elimination of peripheral photoreceptors. Interestingly, the function of the isoforms appears to be tissue specific because all photoreceptor neurons constitutively express Fwe^Ubi and Fwe^Lose-A, whereas the Fwe^Lose-B isoform is uniquely expressed in the neurons to be culled. This is the first description in which Flower coding specifies the elimination of postmitotic cells, thereby fulfilling an important physiological role, which is independent of external insults. Given the conservation of the fwe locus, these results suggest that cell-cell interactions based on fwe may play a general role in sculpting neural networks by selecting optimal neurons and culling unwanted cells (Merino, 2013).

The mechanism resembles the function of the Flower proteins during the elimination of slow dividing cells in Drosophila, but the individual role of the isoforms appears to be cell-type specific, because all photoreceptor neurons constitutively express two isoforms, Fwe^Ubi and Fwe^Lose-A, whereas only Fwe^Lose-B acts in neurons as Lose form and specifically marks the neurons to be purged (Merino, 2013).

Since defective photoreceptors normally appear localized to the eye periphery, it is proposed that the 'Flower Code,' which is normally activated in randomly appearing unfit cells, became regulated by a positional cue such as Wingless in the context of eye development. This may provide an explanation to why wingless activation, which is normally a prosurvival pathway in Drosophila and mammals, is transformed into a proapoptotic signal in the periphery of the Drosophila eye (Merino, 2013).

This study also found that direct comparison of Flower^Lose-B levels among neurons was absolutely necessary and sufficient for neuronal culling. One interesting aspect of this comparison is how the interaction occurs, because each ommatidia is isolated from neighboring ommatidia by surrounding nonneural cells (such as the cones and the pigment cells). One possible explanation is that neurons compare their levels of Flower^Lose-B among axons when they bundle into the optic stalk, but a direct test of this hypothesis is lacking at this time. Alternatively, Flower^Lose-B-expressing neurons might respond to a secreted signal emanating from neurons not expressing Flower^Lose-B (Merino, 2013).

Importantly, this is the first description of Flower coding in the elimination of postmitotic neurons revealing a physiological role for the conserved transmembrane protein in the nervous system in the absence of external insults. The results suggest that the Flower Code may have similar roles in sculpting and maintaining optimal neural networks in higher organisms and may have implications for normal neurological function or disease. For example, in mammals, during life-long adult neurogenesis, active neuronal selection is known to occur in the hippocampus. This is believed to be linked to memory storage and, interestingly, the process is competitive in nature since only a few of the newborn neurons survive. However, despite their importance, neural selection mechanisms are poorly understood. It would be interesting to know whether the Flower Code in mammals is implicated in those processes of neuronal selection and plasticity (Merino, 2013).

Most importantly, a picture starts to appear, in which different isoforms of the Flower protein are displayed at the cell membrane of many, if not all, cell types, whereby certain 'Flower fingerprints' can signal suboptimal fitness. Cells expressing those isoforms are therefore recognized and culled from the tissue. By judging the newest finding in the nervous system, the 'Flower Code' seems to emerge as a more general mechanism of cell recognition and selection than previously acknowledged (Merino, 2013).

Brain regeneration in Drosophila involves comparison of neuronal fitness

Darwinian-like cell selection has been studied during development and cancer. Cell selection is often mediated by direct intercellular comparison of cell fitness, using "fitness fingerprints". In Drosophila, cells compare their fitness via several isoforms of the transmembrane protein Flower. This paper reports a study of the role of intercellular fitness comparisons during regeneration. Regeneration-competent organisms are traditionally injured by amputation, whereas in clinically relevant injuries such as local ischemia or traumatic injury, damaged tissue remains within the organ. It was reasoned that 'Darwinian' interactions between old and newly formed tissues may be important in the elimination of damaged cells. A model of adult brain regeneration in Drosophila was used in which mechanical puncture activates regenerative neurogenesis based on damage-responsive stem cells. It was found that apoptosis after brain injury occurs in damage-exposed tissue located adjacent to zones of de novo neurogenesis. Injury-affected neurons start to express isoforms of the Flower cell fitness indicator protein not found on intact neurons. This change in the neuronal fitness fingerprint is required to recognize and eliminate such neurons. Moreover, apoptosis is inhibited if all neurons express 'low-fitness' markers, showing that the availability of new and healthy cells drives tissue replacement. In summary, this study found that elimination of impaired tissue during brain regeneration requires comparison of neuronal fitness and that tissue replacement after brain damage is coordinated by injury-modulated fitness fingerprints. Intercellular fitness comparisons between old and newly formed tissues could be a general mechanism of regenerative tissue replacement (Moreno, 2015).

In many clinically relevant injuries, such as stroke or traumatic brain injury, impaired cells remain within an organ. In order to study how damaged brain tissue interacts and may be replaced by newly generated cells after injury, adult flies were subjected to penetrating traumatic brain injury, by lesioning the optic lobe (OL) unilaterally with a thin metal filament. This local mechanical damage has been previously shown to activate quiescent adult neural stem cells and drive regenerative neurogenesis (Fernandez-Hernandez, 2013), therefore leading to the apposition of injury-exposed and intact neurons, as well as de novo generated neurons. Local recruitment and activation of stem cells is a common strategy to regenerate tissues in many organisms (Moreno, 2015).

Traumatic brain injuries typically cause a variable extent of tissue damage. Neurons can persist in vulnerable states due to axon stretching and tearing, activating secondary injury processes (diffuse neuronal depolarization, glutamate excitotoxicity, disturbed calcium homeostasis, etc.), which are poorly understood. To study the fate of impaired brain tissue, cell death was monitored several days after the primary injury (Moreno, 2015).

Previous studies have shown that neuronal apoptosis is detectable within the first hours after damage (AD) as a direct consequence of the mechanical impact (Fernandez-Hernandez, 2013). Extended analysis revealed a second burst of apoptosis starting at around 24 hr AD, with low numbers of apoptotic cells present in the lesioned area, which increased and peaked around 3 days after injury. To determine whether apoptosis occurred within regenerating or pre-existing tissue, TUNEL staining of injured brains was performed in which proliferating cells upon injury were marked with GFP/RFP based on perma-twin labeling (Fernandez-Hernandez, 2013), a mitotic recombination-dependent tracing method, which is activated before brain damage in adult Drosophila to mark newly generated tissue. Three days after brain injury, numerous apoptotic cells were observed in damage-exposed tissue next to new tissue. Even 6 days AD, cells continued to die in the 'old' tissue neighboring patches of regenerated tissue, whereas undamaged OLs did not show apoptosis associated with newly generated cells derived from physiologic adult neurogenesis (Moreno, 2015).

The newly formed tissue observed 6 days after brain damage consisted mainly of newborn neurons, which expressed the panneuronal marker Elav and persisted up to 11 days AD. Regenerated tissue was usually devoid of glial cells and macrophages (Moreno, 2015).

Most apoptotic cells were found close (within three cell diameters) to newly generated cells 3 days and 6 days AD. In contrast, apoptosis rarely occurred in 'perma-twin-marked' new tissue. Overall, apoptotic counts were highest 3 days AD and dropped to one-third around 6 days after injury, accompanied by a proliferative phase, evident from the expansion of perma-twin-marked tissue (Moreno, 2015).

Thus, a burst of delayed cell death was identified in injury-exposed brain tissue that is not caused by the primary mechanical insult but is associated with the onset of regenerative neurogenesis (Moreno, 2015).

In order to find genes that may regulate cell death at regeneration borders, reporters for pathways such as JNK, Hippo, Wingless, and JAK-STAT were tested, that are important for regeneration of fly epithelial tissues. Among these, only TRE-gfp, a sensitive JNK pathway reporter, was strongly induced after brain damage (Moreno, 2015).

JNK signaling was repressed in neurons during all stages or specifically during adulthood, but no significant reduction of cell death was observed 1 or 2 days after brain injury (Moreno, 2015).

Next, it was hypothesized that 'Darwinian-like' interactions between impaired and newly formed tissues may trigger cell death, since negative selection can drive elimination of less fit cells during development or carcinogenesis (Moreno, 2015).

In Drosophila, different isoforms of the conserved Flower protein form tissue-specific fitness fingerprints at the cell surface that mediate negative selection of suboptimal cells when surrounded by fitter cells (Rhiner, 2010; Moreno, 2013). First, it was asked whether Flower isoforms are expressed in the adult brain and, specifically, in the OLs. To this end, transgenic flies carrying a translational flower reporter in which expression of Flower^ubi, Flower^LoseA, and Flower^LoseB can be visualized as fusion proteins to YFP, GFP, and RFP, respectively. FlowerLoseA::GFP was strongly expressed in the adult brain, including the OLs, whereas FlowerLoseB::RFP was not detectable. Since Flower^ubi::YFP signal was of low intensity, it was verified the expression pattern of Flower^ubi with an ubi-specific antibody. It was found that both Flower^LoseA and Flower^ubi localized to cell membranes, but Flower^ubi levels were lower since immunodetection required signal amplification. Next, adult brains were stained for Elav, which showed that mature neurons in the adult brain display FlowerLoseA and low levels of Flower^ubi at the cell surface. This revealed that Flower proteins are not only expressed during nervous system development, but also form similar fitness fingerprints in the adult nervous system (Moreno, 2015).

Subsequently, OLs were unilaterally injured, and FlowerLoseB::RFP induction was observed specifically in the damaged right OLs compared to the undamaged control side. FlowerLoseB::RFP signal was first detectable in few cells 24 hr AD and was then present in numerous cells along the lesion 48 and 72 hr AD, whereas expression levels of Flower^ubi and FlowerLoseA remained similar. Next, the experiments were repeated with a different flower reporter where the three isoforms carry Flag (Flower^ubi), HA (FlowerLoseA), and Myc (FlowerLoseB) tags. Again, it was found that FlowerLoseB::Myc was upregulated in lesioned OLs compared to uninjured brains, whereas FlowerLoseA::HA was expressed at high levels in damaged and undamaged OLs (Moreno, 2015).

Elav staining of flies carrying the flower reporter revealed that FlowerLoseB::RFP was induced at the cell surface of injury-exposed neurons 48 hr AD or present in dying neurons (Elav+) (Moreno, 2015).

These results show that acute brain injury triggers local and dynamic changes in displayed fitness marks on damage-exposed neurons compared to surrounding, non-affected cells: impaired neurons start to signal low fitness via induction of FlowerLoseB, which is not encountered on healthy neurons, whereas FlowerLoseA and Flower^ubi expression remains comparable on injured versus non-injured cells (Moreno, 2015).

In order to relate FlowerLoseB induction to cell death during brain regeneration, TUNEL staining was performed of flies transgenic for the YFP/GFP/RFP translational flower reporter. It was observed that FlowerLoseB expression often correlated with cell death 1 to 3 days after brain injury. At 72 hr after injury, 64% of FlowerLoseB::RFP-expressing cells stained positive for TUNEL, raising the possibility that FlowerLoseB expression could drive negative neuronal selection, as previously described for neuronal culling during retina development (Moreno, 2015).

Interestingly, FlowerLoseB was not detected in apoptotic cells 6-14 hr after mechanical injury, suggesting that immediate cell death after mechanical tissue disruption may be FlowerLoseB independent (Moreno, 2015).

To examine whether Flower is functionally implicated in neuronal cell death linked to flowerLoseB upregulation, UASflowerRNAi constructs were conditionally activated in the adult nervous system using the neuronal driver elav-Gal4 and the thermosensitive Gal4 repressor Gal80ts. Five days after gene activation, OLs were lesioned laterally and brains were processed for TUNEL staining (Moreno, 2015).

Three days after injury, knockdown of all flower isoforms (UASRNAifwe_all) or both flower Lose isoforms (UASRNAifweLA/LB) in adult neurons significantly reduced apoptosis in damaged right OLs compared to control brains, where expression of UASlacz was activated instead. Apoptotic numbers were already reduced 24 hr AD when Flower fitness fingerprints were suppressed by the same RNAi constructs (Moreno, 2015).

Sequence similarities between flowerLoseA and flowerLoseB mRNAs did not allow flowerLoseB-specific targeting. These results show that Flower fitness indicator proteins are functionally required in neurons to signal removal of unfit neurons (Moreno, 2015).

Since Flower isoforms have previously been shown to reveal fitness deficits in a non-cell-autonomous manner, tests were performed to see whether uniform overexpression of low fitness marks (in this case FlowerLoseB) would prevent apoptosis coinciding with brain regeneration. To this end, overexpression of UASflowerLoseB, UASflowerLoseA, and UASFlower^ubi was activated in all neurons in adult flies, and the effect on cell death was examined (Moreno, 2015).

When flowerLoseB was ectopically expressed in adult brains, apoptotic counts in lesioned brains were halved 24 hr AD compared to UASlacz control brains and ten times lower at the third day after traumatic brain injury, whereas uninjured left OLs (UASlacz) showed on average 4 apoptotic cells. In contrast, neuronal overexpression of flowerLoseA and Flower^ubi did not significantly affect the number of TUNEL-positive cells 24 and 72 hr after brain injury (Moreno, 2015).

These results show that the majority of cell death 3 days after traumatic brain injury is actively regulated through comparison of neuronal fitness, leading to elimination of FlowerLoseB-expressing impaired neurons when surrounded by intact or newly formed neurons with more advantageous fitness profiles (Moreno, 2015).

Darwinian-like cell selection plays an important role when constructing tissues during development. This study has addressed how the brain weeds out less functional neurons after injury. Fitness-based cell selection was shown to regulate the elimination of damaged tissue during adult brain regeneration in Drosophila. Based on reporter screening and genetic analyses, this study has found that specific isoforms of the cell fitness indicator protein Flower drive the active elimination of impaired neurons at stages in which regenerative neurogenesis provides new neurons for repair (Moreno, 2015).

Traumatic brain injury was shown to cause fitness deficits in injury-exposed neurons, which start to express FlowerLoseB isoforms that are absent on healthy neurons. This reveals for the first time that fitness indicator proteins operate in the adult nervous system and are able to dynamically reflect changes in neuronal fitness states. Local FlowerLoseB upregulation also mediates the culling of sensory neurons in incomplete photoreceptor units during development and therefore seems to present a common signal to mediate negative neuronal selection. These findings open the possibility that conserved Flower proteins may reflect changes in neuronal fitness in other neuropathological conditions (Moreno, 2015).

For further insight, it will be helpful to determine which unfit neurons are recognized and selected for replacement. FlowerLoseB could be upregulated upon physical damage to the neuronal cell body, axon shearing, or disruption of proper wiring or a combination of insults. Moreover, for damaged brain tissue to be replaced, not only neuronal cell bodies but also their axonal projections need to be removed efficiently. It is therefore possible that flowerLoseB induction and 'axon death' signaling molecules, which trigger Wallerian degeneration to allow fast fine-tuning of the nervous system, may be linked (Moreno, 2015).

Importantly, it was showm that damage-modulated fitness indicator proteins are necessary to identify and cull impaired neurons after brain injury. If all neurons are forced to express 'low-fitness' fingerprints, such unfit neurons are not removed by apoptosis. This analysis has shown that damage-exposed cells are specifically eliminated around proliferating zones, where de novo neurogenesis is taking place. A model is proposed in which newly born cells are favored over unfit damaged neurons to reconstitute the adult brain based on Flower fitness fingerprints. One possibility is that neurons partially damaged and/or displaced by the injury upregulate FlowerLoseB. A non-exclusive alternative is that newborn neurons play an active role in the elimination of less fit neurons (Moreno, 2015).

Based on the data, cell death does not seem to be associated with physiologic adult neurogenesis, but further analysis with higher temporal resolution will be required to corroborate these results (Moreno, 2015).

Is fitness-driven elimination of 'old' cells that do not fit into regenerated tissues important in other regenerating tissue types? Preliminary results show that specific Flower isoforms are induced in regenerating wing imaginal discs after cell ablation and in the adult midgut after irradiation (Moreno, 2015).

Moreover, Darwinian-like cell selection could play a role during liver regeneration in mice. An initial study reported a striking increase in apoptosis of host hepatocytes immediately adjacent to transplanted progenitor cells, which can repopulate the liver. It will be interesting to see if mouse Flower homologs also play a role there (Moreno, 2015).

It is therefore proposed that comparison of cellular fitness between damaged and intact tissue may be a common mechanism during regeneration and relevant for stem cell-based replacement therapies after injury (Moreno, 2015).

Elimination of unfit cells maintains tissue health and prolongs lifespan

Viable yet damaged cells can accumulate during development and aging. Although eliminating those cells may benefit organ function, identification of this less fit cell population remains challenging. Previously, a molecular mechanism, based on 'fitness fingerprints' displayed on cell membranes, was identifed that allows direct fitness comparison among cells in Drosophila. This study reports the physiological consequences of efficient cell selection for the whole organism. The study found that fitness-based cell culling is naturally used to maintain tissue health, delay aging, and extend lifespan in Drosophila. A gene, ahuizotl (azot), was identified that ensures the elimination of less fit cells. Lack of azot increases morphological malformations and susceptibility to random mutations and accelerates tissue degeneration. On the contrary, improving the efficiency of cell selection is beneficial for tissue health and extends lifespan (Merino, 2015).

Individual cells can suffer insults that affect their normal functioning, a situation often aggravated by exposure to external damaging agents. A fraction of damaged cells will critically lose their ability to live, but a different subset of cells may be more difficult to identify and eliminate: viable but suboptimal cells that, if unnoticed, may adversely affect the whole organism (Merino, 2015).

What is the evidence that viable but damaged cells accumulate within tissues? The somatic mutation theory of aging proposes that over time cells suffer insults that affect their fitness, for example, diminishing their proliferation and growth rates, or forming deficient structures and connections. This creates increasingly heterogeneous and dysfunctional cell populations disturbing tissue and organ function. Once organ function falls below a critical threshold, the individual dies. The theory is supported by the experimental finding that clonal mosaicism occurs at unexpectedly high frequency in human tissues as a function of time, not only in adults an embryos (Merino, 2015).

Does the high prevalence of mosaicism in our tissues mean that it is impossible to recognize and eliminate cells with subtle mutations and that suboptimal cells are bound to accumulate within organs? Or, on the contrary, can animal bodies identify and get rid of unfit viable cells (Merino, 2015)?

One indirect mode through which suboptimal cells could be eliminated is proposed by the 'trophic theory,' which suggested that Darwinian-like competition among cells for limiting amounts of surv ead to removal of less fit cells. However, it is apparent from recent work that trophic theories are not sufficient to explain fitness-based cell selection, because there are direct mechanisms that allow cells to exchange 'cell-fitness' information at the local multicellular level (Merino, 2015).

In Drosophila, cells can compare their fitness using different isoforms of the transmembrane protein Flower. The 'fitness fingerprints' are therefore defined as combinations of Flower isoforms present at the cell membrane that reveal optimal or reduced fitness. The isoforms that indicate reduced fitness have been called FlowerLose isoforms, because they are expressed in cells marked to be eliminated by apoptosis called 'Loser cells.' However, the presence of FlowerLose isoforms at the cell membrane of a particular cell does not imply that the cell will be culled, because at least two other parameters are taken into account: (1) the levels of FlowerLose isoforms in neighboring cells: if neighboring cells have similar levels of Lose isoforms, no cell will be killed; (2) the levels of a secreted protein called Sparc, the homolog of the Sparc/Osteonectin protein family, which counteracts the action of the Lose isoforms (Merino, 2015 and references therein).

Remarkably, the levels of Flower isoforms and Sparc can be altered by various insults in several cell types, including: (1) the appearance of slowly proliferating cells due to partial loss of ribosomal proteins, a phenomenon known as cell competition; (2) the interaction between cells with slightly higher levels of d-Myc and normal cells, a process termed supercompetition; (3) mutations in signal transduction pathways like Dpp signaling; or (4) viable neurons forming part of incomplete ommatidia. Intriguingly, the role of Flower isoforms is cell type specific, because certain isoforms acting as Lose marks in epithelial cells are part of the fitness fingerprint of healthy neurons. Therefore, an exciting picture starts to appear, in which varying levels of Sparc and different isoforms of Flower are produced by many cell types, acting as direct molecular determinants of cell fitness. This study aimed to clarify how cells integrate fitness information in order to identify and eliminate suboptimal cells. Subsequently, the physiological consequences were analyzed of efficient cell selection for the whole organism (Merino, 2015).

In order to discover the molecular mechanisms underlying cell selection in Drosophila, this study analyzed genes transcriptionally induced using an assay where WT cells (tub>Gal4) are outcompeted by dMyc-overexpressing supercompetitors (tub>dmyc) due to the increased fitness of these dMyc-overexpressing cells. The expression of CG11165 was strongly induced 24 hr after the peak of flower and sparc expression. In situ hybridization revealed that CG11165 mRNA was specifically detected in Loser cells that were going to be eliminated from wing imaginal discs due to cell competition. The gene, which was named ahuizotl (azot) after a multihanded Aztec creature selectively targeting fishing boats to protect lakes, consists of one exon. azot's single exon encodes for a four EF-hand-containing cytoplasmic protein of the canonical family that is conserved, but uncharacterized, in multicellular animals (Merino, 2015).

To monitor Azot expression, a translational reporter was designed resulting in the expression of Azot::dsRed under the control of the endogenous azot promoter in transgenic flies. Azot expression was not detectable in most wing imaginal discs under physiological conditions in the absence of competition. Mosaic tissue was generated of two clonal populations, which are known to trigger competitive interactions resulting in elimination of otherwise viable cells. Cells with lower fitness were created by confronting WT cells with dMyc-overexpressing cells, by downregulating Dpp signaling, by overexpressing Flower^Lose isoforms, in cells with reduced Wg signaling, by suppressing Jak-Stat signaling or by generating Minute clones. Azot expression was not detectable in nonmosaic tissue of identical genotype, nor in control clones overexpressing UASlacZ. On the contrary, Azot was specifically activated in all tested scenarios of cell competition, specifically in the cells undergoing negative selection. Azot expression was not repressed by the caspase inhibitor protein P35 (Merino, 2015).

Because Flower proteins are conserved in mammals, tests were made to see if they are also able to regulate azot. Mouse Flower isoform 3 (mFlower³) has been shown to act as a 'classical' Lose isoform, driving cell elimination when expressed in scattered groups of cells, a situation where azot was induced in Loser cells but is not inducing cell selection when expressed ubiquitously a scenario where azot was not expressed. This shows that the mouse Flower^Lose isoforms function in Drosophila similarly to their fly homologs (Merino, 2015).

Interestingly, azot is not a general apoptosis-activated gene because its expression is not induced upon eiger, hid, or bax activation, which trigger cell death. Azot was also not expressed during elimination of cells with defects in apicobasal polarity or undergoing epithelial exclusion-mediated apoptosis (dCsk) (Merino, 2015).

azot expression was analyzed during the elimination of peripheral photoreceptors in the pupal retina, a process mediated by Flower-encoded fitness fingerprints. Thirty-six to 38hr after pupal formation (APF), when Flower^Lose-B expression begins in peripheral neurons, no Azot expression was detected in the peripheral edge. At later time points (40 and 44hr APF), Azot expression is visible and restricted to the peripheral edge where photoreceptor neurons are eliminated. This expression was confirmed with another reporter line, azot{KO; gfp}, where gfp was directly inserted at the azot locus using genomic engineering techniques (Merino, 2015).

From these results, it is concluded that Azot expression is activated in several contexts where suboptimal and viable cells are normally recognized and eliminated (Merino, 2015).

To understand Azot function in cell elimination, azot knockout (KO) flies were generated by deleting the entire azot gene. Next, Azot function was analyzed using dmyc-induced competition. In the absence of Azot function, loser cells were no longer eliminated, showing a dramatic 100-fold increase in the number of surviving clones. Loser cells occupied more than 20% of the tissue 72hr after clone induction (ACI). Moreover, using azot{KO; gfp} homozygous flies (that express GFP under the azot promoter but lack Azot protein), it was found that loser cells survived and showed accumulation of GFP. From these results, it is concluded that azot is expressed by loser cells and is essential for their elimination (Merino, 2015).

In addition, clone removal was delayed in an azot heterozygous background (50-fold increase, 15%), compared to control flies with normal levels of Azot. Cell elimination capacity was fully restored by crossing two copies of Azot::dsRed into the azot^-/- background demonstrating the functionality of the fusion protein. Silencing azot with two different RNAis was similarly able to halt selection during dmyc-induced competition. Next, in order to determine the role of Azot's EF hands, a mutated isoform of Azot (Pm4Q12) was generated and overexpressed, that carried, in each EF hand, a point mutation known to abolish Ca²⁺ binding. Although overexpression of wild-type azot in negatively selected cells did not rescue the elimination, overexpression of the mutant AzotPm4Q12 reduced cell selection, functioning as a dominant-negative mutant. This shows that Ca²⁺ binding is important for Azot function. Finally, staining for apoptotic cells corroborated that the lack of Azot prevents cell elimination, because cell death was reduced 8-fold in mosaic epithelia containing loser cells (Merino, 2015).

The role of azot in elimination of peripheral photoreceptor neurons in the pupal retina was examined using homozygous azot KO flies. Pupal retinas undergoing photoreceptor culling (44hr APF) of azot^+/+ and azot^-/- flies were stained for the cell death marker and the proapoptotic factor. Consistent with the expression pattern of Azot, the number of Hid and TUNEL-positive cells was dramatically decreased in azot^-/- retinas compared to azot^+/+ retinas (Merino, 2015).

Those results show that Azot is required to induce cell death and Hid expression during neuronal culling. Therefore, tests were performed to see that was also the case in the wing epithelia during dmyc-induced competition. Hid was found to be expressed in loser cells and the expression was found to be strongly reduced in the absence of Azot function (Merino, 2015).

Finally, forced overexpression of Flower^Lose isoforms from Drosophila were unable to mediate WT cell elimination when Azot function was impaired by mutation or silenced by RNAi (Merino, 2015).

These results suggested that azot function is dose sensitive, because heterozygous azot mutant flies display delayed elimination of loser cells when compared with azot WT flies. Therefore advantage was taken of the functional reporter Azot::dsRed to test whether cell elimination could be enhanced by increasing the number of genomic copies of azot. Tissues with three functional copies of azot were more efficient eliminating loser cells during dmyc-induced competition and most of the clones were culled 48hr ACI. From these results, it is concluded that azot expression is required for the elimination of Loser cells and unwanted neurons (Merino, 2015).

Next, it was asked what could be the consequences of decreased cell selection at the tissue and organismal level. To this end, advantage was taken of the viability of homozygous azot KO flies. An increase of several developmental aberrations was observed. Focus was placed on the wings, where cell competition is best studied and, because aberrations, including melanotic areas, blisters, and wing margin nicks, were quantified. Wing defects of azot mutant flies could be rescued by introducing two copies of azot::dsRed, showing that the phenotypes are specifically caused by loss of Azot function (Merino, 2015).

Next, it was reasoned that mild tissue stress should increase the need for fitness-based cell selection after damage. First, in order to generate multicellular tissues scattered with suboptimal cells, larvae were exposed to UV light and Azot expression was monitored in wing discs of UV-irradiated WT larvae that were stained for cleaved caspase-3, 24hr after treatment. Under such conditions, Azot was found to be expressed in cleaved caspase-3-positive cells. All Azot-positive cells showed caspase activation and 17% of cleaved caspase-positive cells expressed Azot. This suggested that Azot-expressing cells are culled from the tissue. To confirm this, later time points (3 days after irradiation) were examined; the increase in Azot-positive cells was no longer detectable. The elimination of azot-expressing cells after UV irradiation required azot function, because cells revealed by reporter azot{KO; gfp}, that express GFP instead of Azot, persisted in wing imaginal discs from azot-null larvae. Tests were performeed to see if lack of azot leads to a faster accumulation of tissue defects during organ development upon external damage. azot^-/- pupae 0 stage were irradiated, and the number of morphological defects in adult wings was compared to those in nonirradiated azot KO flies. It was found that aberrations increased more than 2-fold when compared to nonirradiated azot^-/- flies (Merino, 2015).

In order to functionally discriminate whether azot belongs to genes regulating apoptosis in general or is dedicated to fitness-based cell selection, whether azot silencing prevents Eiger/TNF-induced cell death was exanubed. Inhibiting apoptosis (UASp35) or eiger (UASRNAieiger) rescued eye ablation, whereas azot silencing and overexpression of AzotPm4Q12 did not. Furthermore, azot silencing did not impair apoptosis during genitalia rotation or cell death of epithelial precursors in the retina. These results highlight the consequences of nonfunctional cell-quality control within developing tissues (Merino, 2015).

The next part of the analysis demonstrated that the azot promoter computes relative Flower^Lose and Sparc Levels. Epistasis analyses were performed to understand at which level azot is transcriptionally regulated. For this purpose, the assay where WT cells are outcompeted by dMyc-overexpressing supercompetitors was used. It was previously observed that azot induction is triggered upstream of caspase-3 activation and accumulates in outcompeted cells unable to die. Then, upstream events of cell selection were genetically modified. Silencing fwe^Lose transcripts by RNAi or overexpressing Sparc both blocked the induction of Azot::dsRed in WT loser cells. In contrast, when outcompeted WT cells were additionally 'weakened' by Sparc downregulation using RNAi, Azot is detected in almost all loser cells compared to its more limited induction in the presence of endogenous Sparc. Inhibiting JNK signaling with UASpuc did not suppress Azot expression (Merino, 2015).

The activation of Azot upon irradiation was examined. Strikingly, it was found that all Azot expression after irradiation was eliminated when Flower Lose was silenced and also when relative differences of Flower Lose where diminished by overexpressing high levels of Lose isoforms ubiquitously. On the contrary, Azot was not suppressed after irradiation by expressing the prosurvival factor Bcl-2 or a p53 dominant negative. These results show that Azot expression during competition and upon irradiation requires differences in Flower Lose relative levels (Merino, 2015).

Finally, the regulation of Azot expression in neurons was analyzed. Silencing fwe transcripts by RNAi blocked the induction of Azot::dsRed in peripheral photoreceptors. Because Wingless signaling induces Flower^Lose-B expression in peripheral photoreceptors, tests were performed to see if overexpression of Daxin, a negative regulator of the pathway, affected Azot levels. Axin overespression completely inhibited Azot expression. Similarly, overexpression of the cell competition inhibitor Sparc also fully blocked Azot endogenous expression in the retina. Finally, ectopic overexpression of Flower^Lose-B in scattered cells of the retina was sufficient to trigger ectopic Azot activation. These results show that photoreceptor cells also can monitor the levels of Sparc and the relative levels of Flower^Lose-B before triggering Azot expression (Merino, 2015).

These results suggest that the azot promoter integrates fitness information from neighboring cells, acting as a relative 'cell-fitness checkpoint.'

To test if fitness-based cell selection is a mechanism active not only during development, but also during adult stages, WT adult flies were exposed to UV light and monitor Azot and Flower expression were monitored in adult tissues. UV irradiation of adult flies triggered cytoplasmic Azot expression in several adult tissues including the gut and the adult brain. Likewise, UV irradiation of adult flies triggered Flower Lose expression in the gut and in the brain. Irradiation-induced Azot expression was unaffected by Bcl-2 but was eliminated when Flower Lose was silenced or when relative differences of Flower Lose where diminished in the gut. This suggests that the process of cell selection is active throughout the life history of the animal. Further confirming this conclusion, Azot function was essential for survival after irradiation, because more than 99% of azot mutant adults died 6 days after irradiation, whereas only 62.4% of WT flies died after the same treatment. The percentage of survival correlated with the dose of azot because adults with three functional copies of azot had higher median survival and maximum lifespan than WT flies, or null mutant flies rescued with two functional azot transgenes (Merino, 2015).

The next part of the study addressed the role of cell selection during aging. Lack of cell selection could affect the whole organism by two nonexclusive mechanisms. First, the failure to detect precancerous cells, which could lead to cancer formation and death of the individual. Second, the time-dependent accumulation of unfit but viable cells could lead to accelerated tissue and organ decay. We therefore tested both hypotheses (Merino, 2015).

It has been previously shown that cells with reduced levels for cell polarity genes like scrib or dlg are eliminated but can give rise to tumors when surviving. Therefore this study checked if azot functions as a tumor suppressing mechanism in those cells. Elimination of dlg and scrib mutant cells was not affected by RNAi against azot or when Azot function was impaired by mutation, in agreement with the absence of azot induction in these mutant cells. However, azot RNAi or the same azot mutant background efficiently rescued the elimination of clones with reduced Wg signaling (Merino, 2015).

Moreover, the high number of suboptimal cells produced by UV treatment did not lead to tumoral growth in azot-null background. Thus, tumor suppression mechanisms are not impaired in azot mutant backgrounds, and tumors are not more likely to arise in azot-null mutants (Merino, 2015).

Also tests were performed to see whether the absence of azot accelerates tissue fitness decay in adult tissues. Focused was placed on the adult brain, where neurodegenerative vacuoles develop over time and can be used as a marker of aging. The number was compared of vacuoles appearing in the brain of flies lacking azot (azot^-/-), WT flies (azot^+/+), flies with one extra genomic copy of the gene (azot^+/+; azot⁺), and mutant flies rescued with two genomic copies of azot (azot^-/-;azot^+/+). For all the genotypes analyzed, a progressive increase was observed in the number and size of vacuoles in the brain over time. Interestingly, azot^-/- brains showed higher number of vacuoles compared to control flies (azot^+/+ and azot^-/-;azot^+/+) and a higher rate of vacuole accumulation developing over time. In the case of flies with three genomic copies of the gene (azot^+/+; azot⁺), vacuole number tended to be the lowest (Merino, 2015).

The cumulative expression of azot was analyzed during aging of the adult brain. Positive cells were detected as revealed by reporter azot{KO; gfp}, in homozygosis, that express GFP instead of Azot. A time-dependent accumulation of azot-positive cells was observed (Merino, 2015).

From this, it is concluded that azot is required to prevent tissue degeneration in the adult brain and lack of azot showed signs of accelerated aging. This suggested that azot could affect the longevity of adult flies. Flies lacking azot (azot^-/-) had a shortened lifespan with a median survival of 7.8 days, which represented a 52% decrease when compared to WT flies (azot^+/+), and a maximum lifespan of 18 days, 25% less than WT flies (azot^+/+). This effect on lifespan was azot dependent because it was completely rescued by introducing two functional copies of azot. On the contrary, flies with three functional copies of the gene (azot^+/+; azot⁺) showed an increase in median survival and maximum lifespan of 54% and 17%, respectively (Merino, 2015).

In conclusion, azot is necessary and sufficient to slow down aging, and active selection of viable cells is critical for a long lifespan in multicellular animals (Merino, 2015).

The next part of the study demonstrates that death of unfit cells is sufficient and required for multicellular fitness maintenance. The results cited above show the genetic mechanism through which cell selection mediates elimination of suboptimal but viable cells. However, using flip-out clones and MARCM, this study found that Azot overexpression was not sufficient to induce cell death in wing imaginal discs. Because Hid is downstream of Azot, it was wondered whether expressing Hid under the control of the azot regulatory regions could substitute for Azot function (Merino, 2015).

In order to test this hypothesis, the whole endogenous azot protein-coding sequence was replaced by the cDNA of the proapoptotic gene hid (azot{KO; hid}) flies. In a second strategy, the whole endogenous azot protein-coding sequence was replaced by the cDNA of transcription factor Gal4, so that the azot promoter can activate any UAS driven transgene (azot{KO; Gal4} flies. The number of morphological aberrations was compared in the adult wings of six genotypes: first, homozygous azot{KO; Gal4} flies that lacked Azot; second, azot{KO; hid} homozygous flies that express Hid with the azot pattern in complete absence of Azot; third, azot^+/+ WT flies as a control; and finally three genotypes where the azot{KO; Gal4} flies were crossed with UAShid, UASsickle, another proapoptotic gene, or UASp35, an apoptosis inhibitor. In the case of UASsickle flies, a second azot mutation was introduced to eliminate azot function. Interestingly, the number of morphological aberrations was brought back to WT levels in all the situations where the azot promoter was driving proapoptotic genes (azot{KO; hid}, azot{KO; Gal4} × UAShid, azot{KO; Gal4} × UASsickle with or without irradiation. On the contrary, expressing p35 with the azot promoter was sufficient to produce morphological aberrations despite the presence of one functional copy of azot. Likewise, p35-expressing flies (azot{KO; Gal4}/azot⁺; UASp35) did not survive UV treatments, whereas a percentage of the flies expressing hid (26%) or sickle (28%) in azot-positive cells were able to survive (Merino, 2015).

From this, it is concluded that specifically killing those cells selected by the azot promoter is sufficient and required to prevent morphological malformations and provide resistance to UV irradiation (Merino, 2015).

The next part of the study demonstrated that death of unfit cells extends lifespan It was asked whether the shortened longevity observed in azot^-/- flies could be also rescued by killing azot-expressing cells with hid in the absence of Azot protein. It was found that azot{KO; hid} homozygous flies had dramatically improved lifespan with a median survival of 27 days at 29°C, which represented a 125% increase when compared to azot^-/- flies, and a maximum lifespan of 34 days, 41% more than mutant flies (Merino, 2015).

Similar results were obtained at 25°C. It was found that flies lacking azot (azot^-/-) had a shortened lifespan with a median survival of 25days, which represented a 24% decrease when compared to WT flies (azot^+/+), and a maximum lifespan of 40 days, 31% less than WT flies (azot^+/+). On the contrary, flies with three functional copies of the gene (azot^+/+; azot⁺) or flies where azot is replaced by hid (azot{KO; hid} homozygous flies) showed an increase in median survival of 54% and 63% and maximum lifespan of 12% and 24%, respectively (Merino, 2015).

Finally, the effects of dietary restriction on longevity of those flies was tested. It was found that dietary restriction could extend both the median survival and the maximum lifespan of all genotypes. Interestingly, dietary restricted flies with three copies of the gene azot showed a further increase in maximum lifespan of 35%. This shows that dietary restriction and elimination of unfit cells can be combined to maximize lifespan (Merino, 2015).

In conclusion, eliminating unfit cells is sufficient to increase longevity, showing that cell selection is critical for a long lifespan in Drosophila (Merino, 2015).

This study has shown that active elimination of unfit cells is required to maintain tissue health during development and adulthood. The gene (azot), whose expression is confined to suboptimal or misspecified but morphologically normal and viable cells. When tissues become scattered with suboptimal cells, lack of azot increases morphological malformations and susceptibility to random mutations and accelerates age-dependent tissue degeneration. On the contrary, experimental stimulation of azot function is beneficial for tissue health and extends lifespan. Therefore, elimination of less fit cells fulfils the criteria for a hallmark of aging (Merino, 2015).

Although cancer and aging can both be considered consequences of cellular damage, no evidence was found for fitness-based cell selection having a role as a tumor suppressor in Drosophila. The results rather support that accumulation of unfit cells affect organ integrity and that, once organ function falls below a critical threshold, the individual dies (Merino, 2015).

Azot expression in a wide range of 'less fit' cells, such as WT cells challenged by the presence of 'supercompetitors,' slow proliferating cells confronted with normal proliferating cells, cells with mutations in several signaling pathways (i.e., Wingless, JAK/STAT, Dpp), or photoreceptor neurons forming incomplete ommatidia. In order to be expressed specifically in 'less fit' cells, the transcriptional regulation of azot integrates fitness information from at least three levels: (1) the cell's own levels of Flower^Lose isoforms, (2) the levels of Sparc, and (3) the levels of Lose isoforms in neighboring cells. Therefore, Azot ON/OFF regulation acts as a cell-fitness checkpoint deciding which viable cells are eliminated. It is proposed that by implementing a cell-fitness checkpoint, multicellular communities became more robust and less sensitive to several mutations that create viable but potentially harmful cells. Moreover, azot is not involved in other types of apoptosis, suggesting a dedicated function, and - given the evolutionary conservation of Azot - pointing to the existence of central cell selection pathways in multicellular animals (Merino, 2015).

Cell mixing induced by myc is required for competitive tissue invasion and destruction

Cell-cell intercalation is used in several developmental processes to shape the normal body plan. There is no clear evidence that intercalation is involved in pathologies. This study used the proto-oncogene myc to study a process analogous to early phase of tumour expansion: myc-induced cell competition. Cell competition is a conserved mechanism driving the elimination of slow-proliferating cells (so-called 'losers') by faster-proliferating neighbours (so-called 'winners') through apoptosis and is important in preventing developmental malformations and maintain tissue fitness. Using long-term live imaging of myc-driven competition in the Drosophila pupal notum and in the wing imaginal disc, this study showed that the probability of elimination of loser cells correlates with the surface of contact shared with winners. As such, modifying loser-winner interface morphology can modulate the strength of competition. Elimination of loser clones requires winner-loser cell mixing through cell-cell intercalation. Cell mixing is driven by differential growth and the high tension at winner-winner interfaces relative to winner-loser and loser-loser interfaces, which leads to a preferential stabilization of winner-loser contacts and reduction of clone compactness over time. Differences in tension are generated by a relative difference in F-actin levels between loser and winner junctions, induced by differential levels of the membrane lipid phosphatidylinositol (3,4,5)-trisphosphate. These results establish the first link between cell-cell intercalation induced by a proto-oncogene and how it promotes invasiveness and destruction of healthy tissues (Levayer, 2015).

To analyse quantitatively loser cell elimination, long-term live imaging was performed of clones showing a relative decrease of the proto-oncogene myc in the Drosophila pupal notum, a condition known to induce cell competition in the wing disc. Every loser cell delamination was counted over 10 h, and the probability of cell elimination was calculated for a given surface of contact shared with winner cells. A significant increase was observed of the proportion of delamination with winner-loser shared contact, whereas this proportion remained constant for control clones. The same correlation was observed in ex vivo culture of larval wing disc. Cell delamination in the notum was apoptosis dependent and expression of flower^lose (fwe^lose), a competition-specific marker for loser fate, was necessary and sufficient to drive contact-dependent delamination. Moreover it was confirmed that contact-dependent death is based on the computation of relative differences of fwe^lose between loser cells and their neighbours. Thus, cell delamination in the notum recapitulates features of cell competition (Levayer, 2015).

This suggests that winner-loser interface morphology could modulate the probability of eliminating loser clones. Using the wing imaginal disc, winner-loser contact was reduced by inducing adhesion- or tension-dependent cell sorting and observed a significant reduction of loser clone elimination. This rescue was not driven by a cell-autonomous effect of E-cadherin (E-cad) or active myosin II regulatory light chain (MRLC) on growth, death or cell fitness but rather by a general diminution of winner-loser contact. Competition is ineffective across the antero-posterior compartment boundary, a frontier that prevents cell mixing through high line tension. Accordingly, there was no increase in death at the antero-posterior boundary in wing discs overexpressing fwe^loseA in the anterior compartment. However, reducing tension by reducing levels of myosin II heavy chains was sufficient to increase the shared surface of contact between cells of the anterior and posterior compartments, and induced fwe^lose death at the boundary. Altogether, it is concluded that the reduction in surface contact between winners and losers is sufficient to block competition, which explains how compartment boundaries prevent competition (Levayer, 2015).

Loser clones have been reported to fragment more often than controls, whereas winner clones show convoluted morphology, suggesting that winner-loser mixing is increased during competition. This could affect the outcome of cell competition by increasing the surface shared between losers and winners. Clone splitting was used as a readout for loser–winner mixing. Two non-exclusive mechanisms can drive clone splitting: cell death followed by junction rearrangement, or junction remodelling and cell–cell intercalation independent of death. To assess the contribution of each phenomenon, the proportion of clones fragmented 48 h after clone induction (ACI) was systematically counted. A twofold increase was observed in the frequency of split clones in losers (wild type (WT) in tub-dmyc) versus WT in WT controls. Overexpressing E-cad or active myosin II was sufficient to prevent loser clone splitting, whereas blocking apoptosis or blocking loser fate by silencing fwe^lose did not reduce splitting. Finally, the proportion of split clones was also increased for winner clones either during myc-driven competition or during Minute-dependent competition. Altogether, this suggested that winner–loser mixing is increased independently of loser cell death or clone size by a factor upstream of fwe, and could be driven by cell–cell intercalation. Accordingly, junction remodelling events leading to disappearance of a loser–loser junction were three times more frequent at loser clone boundaries than control clone boundaries in the pupal notum. The rate of junction remodelling was higher in loser–loser junctions and in winner–winner junctions than in winner–loser junctions. The preferential stabilization of winner–loser interfaces should increase the surface of contact between winner and loser cells over time. Accordingly, loser clone compactness in the notum decreased over time whereas it remains constant on average for WT clones in WT background. Similarly, the compactness of clones in the notum also decreased over time for conditions showing high frequency of clone splitting in the wing disc, whereas clone compactness remained constant for conditions rescuing clone splitting. Altogether, it is concluded that both Minute- and myc-dependent competition increase loser–winner mixing through cell–cell intercalation (Levayer, 2015).

It was then asked what could modulate the rate of junction remodelling during competition. The rate of junction remodelling can be cell-autonomously increased by myc. Interestingly, downregulation of the tumour suppressor PTEN is also sufficient to increase the rate of junction remodelling through the upregulation of phosphatidylinositol (3,4,5)-trisphosphate (PIP3). It was reasoned that differences in PIP3 levels could also modulate junction remodelling during competition. Using a live reporter of PIP3 that could detect modulations of PIP3 in the notum, a significant increase of PIP3 was observed in the apico-lateral membrane of tub-dmyc–tub-dmyc interfaces compared with WT–WT and WT–tub-dmyc interfaces (Fig. 3a, b). Moreover, increasing/reducing Myc levels in a full compartment of the wing disc was sufficient to increase/decrease the levels of phospho-Akt (a downstream target of PIP3, whereas fwe^loseA overexpression had no effect. Similarly, levels of phospho-Akt were relatively higher in WT clones than in the surrounding M^−/+ cells. Thus differences in PIP3 levels might be responsible for winner–loser mixing. Accordingly, reducing PIP3 levels by overexpressing a PI3 kinase dominant negative (PI3K-DN) or increasing PIP3 levels by knocking down PTEN (UAS-pten RNAi) were both sufficient to induce a high proportion of fragmented clones and to reduce clone compactness over time in the notum , whereas increasing PIP3 in loser clones was sufficient to prevent cell mixing. Moreover, abolishing winner–loser PIP3 differences through larval starvation prevented loser clone fragmentation, the reduction of clone compactness over time in the notum and could rescue WT clone elimination in tub-dmyc background. It is therefore concluded that differences in PIP3 levels are necessary and sufficient for loser–winner mixing and required for loser cell elimination (Levayer, 2015).

It was then asked which downstream effectors of PIP3 could affect junction stability. A relative growth decrease can generate mechanical stress that can be released by cell-cell intercalation. Accordingly, growth reduction through Akt downregulation is sufficient to increase clone splitting and could contribute to loser clone splitting. However, Akt is not sufficient to explain winner-loser mixing because, unlike PIP3, increasing Akt had no effect on clone splitting. PIP3 could also modulate junction remodelling through its effect on cytoskeleton and the modulation of intercellular adhesion or tension. No obvious modifications of E-cad, MRLC or Dachs (another regulator of tension) was detected in loser cells. However, a significant reduction of F-actin levels and a reduction of actin turnover/polymerization rate were observed in loser-loser and loser-winner junctions in the notum. Similarly, modifying Myc levels in a full wing disc compartment was sufficient to modify actin levels, and F-actin levels were higher in WT clones than M-/+ cells. This prompted a test of the role of actin organization in winner-loser mixing. Downregulating the formin Diaphanous (Dia, a filamentous actin polymerization factor) by RNA interference (RNAi) or by using a hypomorphic mutant was sufficient to obtain a high proportion of fragmented clones and to reduce clone compactness over time, whereas overexpressing Dia in loser clones prevented clone splitting (UAS-dia::GFP) and compactness reduction. This effect was specific to Dia as modulating Arp2/3 complex (a regulator of dendritic actin network) had no effect on clone splitting. Thus, impaired filamentous actin organization was necessary and sufficient to drive loser-winner mixing. These actin defects were driven by the differences in PIP3 levels between losers and winners. Thus Dia could be an important regulator of competition through its effect on cell mixing. Overexpression of Dia was indeed sufficient to reduce loser clone elimination significantly (Levayer, 2015).

Filamentous actin has been associated with tension regulation. It was therefore asked whether junction tension was modified in winner and loser junctions. The maximum speed of relaxation of junction after laser nanoablation (which is proportional to tension) was significantly reduced in loser-loser and winner-loser junctions compared with winner-winner junctions. This distribution of tension has been proposed to promote cell mixing. Accordingly, decreasing PIP3 in clones reduced tension both in low-PIP3-low-PIP3 and low-PIP3-normal-PIP3 junctions, whereas overexpressing Dia in loser clones or starvation were both sufficient to abolish differences in tension, in agreement with their effect on winner-loser mixing and the distribution of F-actin. Thus the lower tension at winner-loser and loser-loser junctions is responsible for winner-loser mixing. Altogether, it is concluded that the relative PIP3 decrease in losers increases winner-loser mixing through Akt-dependent differential growth and the modulation of tension through F-actin downregulation in winner-loser and loser-loser junctions (Levayer, 2015).

Several modes of tissue invasion by cancer cells have been described, most of them relying on the departure of the tumour cells from the epithelial layer. This study suggests that some oncogenes may also drive tissue destruction and invasion by inducing ectopic cell intercalation between cancerous and healthy cells, and subsequent healthy cell elimination. myc-dependent invasion could be enhanced by other mutations further promoting intercalation (such as PTEN). Stiffness is increased in many tumours, suggesting that healthy cell-cancer cell mixing by intercalation might be a general process (Levayer, 2015).

A Ca2+ channel differentially regulates Clathrin-mediated and activity-dependent bulk endocytosis

Clathrin-mediated endocytosis (CME) and activity-dependent bulk endocytosis (ADBE) are two predominant forms of synaptic vesicle (SV) endocytosis, elicited by moderate and strong stimuli, respectively. They are tightly coupled with exocytosis for sustained neurotransmission. However, the underlying mechanisms are ill defined. Previous work has shown that the Flower (Fwe) Ca2+ channel present in SVs is incorporated into the periactive zone upon SV fusion, where it triggers CME, thus coupling exocytosis to CME. This study shows that Fwe also promotes ADBE. Intriguingly, the effects of Fwe on CME and ADBE depend on the strength of the stimulus. Upon mild stimulation, Fwe controls CME independently of Ca2+ channeling. However, upon strong stimulation, Fwe triggers a Ca2+ influx that initiates ADBE. Moreover, knockout of rodent fwe in cultured rat hippocampal neurons impairs but does not completely abolish CME, similar to the loss of Drosophila fwe at the neuromuscular junction, suggesting that Fwe plays a regulatory role in regulating CME across species. In addition, the function of Fwe in ADBE is conserved at mammalian central synapses. Hence, Fwe exerts different effects in response to different stimulus strengths to control two major modes of endocytosis (Yao, 2017).

A tight coupling of exocytosis and endocytosis is critical for supporting continuous exocytosis of neurotransmitters. CME and ADBE are well-characterized forms of SV endocytosis triggered by moderate and strong nerve stimuli, respectively. However, how they are coupled with exocytosis under distinct stimulation paradigms remains less explored. A model is proposed based on the present data. When presynaptic terminals are mildly stimulated, SV release leads to neurotransmitter release and the transfer of Fwe channel from SVs to the periactive zone where CME and ADBE occur actively. The data suggest that this channel does not supply Ca²⁺ for CME to proceed. However, intense activity promotes Fwe to elevate presynaptic Ca²⁺ levels near endocytic zones where ADBE is subsequently triggered. Thus, Fwe exerts different activities and properties in response to different stimuli to couple exocytosis to different modes of endocytosis (Yao, 2017).

It has been previously concluded that Fwe-dependent Ca²⁺ influx triggers CME (Yao, 2009). However, the current results suggest alternative explanations. First, the presynaptic Ca²⁺ concentrations elicited by moderate activity conditions, i.e., 1-min 90 mM K⁺/0.5 mM Ca²⁺ or 20-s 10-20 Hz electric stimulation, are not dependent on Fwe. Second, expression of 4% FweE79Q, a condition that abolishes Ca²⁺ influx via Fwe, rescues the CME defects associated with fwe mutants, including decreased FM1-43 dye uptake, a reduced number of SVs, and enlarged SVs. Third, raising the presynaptic Ca²⁺ level has no beneficial impact on the reduced number of SVs observed in fwe mutants. These data are consistent with the observations that a Ca²⁺ influx dependent on VGCCs triggers CME at a mammalian synapse. Hence, Fwe acts in parallel with or downstream to VGCC-mediated Ca²⁺ influx during CME (Yao, 2017).

ADBE is triggered by intracellular Ca²⁺ elevation, which has been assumed to be driven by VGCCs that are located at the active zones. However, the data strongly support a role for Fwe as an important Ca²⁺ channel for ADBE. First, following exocytosis, Fwe is enriched at the periactive zone where ADBE predominates. Second, Fwe selectively supplies Ca²⁺ to the presynaptic compartment during intense activity stimulation, which is highly correlated with the rapid formation of ADBE upon stimulation. Third, 4% FweE79Q expression, which induces very subtle or no Ca²⁺ upon strong stimulation, fails to rescue the ADBE defect associated with loss of fwe. Fourth, treatment with a low concentration of La³⁺ solution that specifically blocks the Ca²⁺ conductance of Fwe significantly abolishes ADBE. Lastly, the role of Fwe-derived Ca²⁺ influx in the initiation of ADBE mimics the effect of Ca²⁺ on ADBE at the rat Calyx of Held. As loss of fwe does not completely eliminate ADBE, the results do not exclude the possibility that VGCC may function in parallel with Fwe to promote ADBE following intense stimulation (Yao, 2017).

Interestingly, Ca²⁺ influx via Fwe does not control SV exocytosis during mild and intense stimulations. How do VGCC and Fwe selectively regulate SV exocytosis and ADBE, respectively? One potential mechanism is that VGCC triggers a high, transient Ca²⁺ influx around the active zone that elicits SV exocytosis. In contrast, Fwe is activated at the periactive zone to create a spatially and temporally distinct Ca²⁺ microdomain. A selective failure to increase the presynaptic Ca²⁺ level during strong stimulation is evident upon loss of fwe. This pinpoints to an activity-dependent gating of the Fwe channel. Consistent with this finding, an increase in the level of Fwe in the plasma membrane does not lead to presynaptic Ca²⁺ elevation at the Calyx of Held when the presynaptic terminals are at rest or subject to mild stimulation. However, previous studies showed that, in shi^ts terminals, blocking CME results in the accumulation of the Fwe channel in the plasma membrane, elevating Ca²⁺ levels. It is possible that Dynamin is also involved in regulating the channel activity of Fwe or that the effects other than Fwe accumulation associated with shi^ts mutants may affect intracellular Ca²⁺ handling. Further investigation of how neuronal activity gates the channel function of Fwe should advance knowledge on the activity-dependent exo-endo coupling (Yao, 2017).

Although a proteomic analysis did not identify ratFwe2 in SVs purified from rat brain, biochemical analyses show that ratFwe2 is indeed associated with the membrane of SVs. The data show that 4% of the total endogenous Fwe channels efficiently promotes CME and ADBE at the Drosophila NMJ. If a single SV needs at least one functional Fwe channel complex during exo-endo coupling, and one functional Fwe complex comprises at least four monomers, similar to VGCCs, transient receptor potential cation channel subfamily V members (TRPV) 5 and 6, and calcium release-activated channel (CRAC)/Orai1, then it is anticipated that each SV contains ~100 Fwe proteins (4 monomers x 25). This suggests that Fwe is highly abundant on the SVs. It is unlikely that many SVs do not have the Fwe, as a 25-fold reduction of the protein is enough to ensure functional integrity during repetitive neurotransmission. Finally, the results for the SypHy and dextran uptake assays at mammalian central synapses indicate the functional conservation of the Fwe channel in promoting different modes of SV retrieval. In summary, the Fwe-mediated exo-endo coupling seems to be of broad importance for sustained synaptic transmission across species. (Yao, 2017).

A synaptic vesicle-associated Ca2+ channel promotes endocytosis and couples exocytosis to endocytosis

Synaptic vesicle (SV) exo- and endocytosis are tightly coupled to sustain neurotransmission in presynaptic terminals, and both are regulated by Ca(2+). Ca(2+) influx triggered by voltage-gated Ca(2+) channels is necessary for SV fusion. However, extracellular Ca(2+) has also been shown to be required for endocytosis. The intracellular Ca(2+) levels (<1 microM) that trigger endocytosis are typically much lower than those (>10 microM) needed to induce exocytosis, and endocytosis is inhibited when the Ca(2+) level exceeds 1 microM. This study identified and characterized a transmembrane protein associated with SVs that, upon SV fusion, localizes at periactive zones. Loss of Flower results in impaired intracellular resting Ca(2+) levels and impaired endocytosis. Flower multimerizes and is able to form a channel to control Ca(2+) influx. It is proposed that Flower functions as a Ca(2+) channel to regulate synaptic endocytosis and hence couples exo- with endocytosis (Yao, 2009).

Ca2+ influx triggers both SV exo- and endocytosis. Since SV retrieval requires much lower Ca2+ levels than those that elicit release, it was proposed that the Ca2+ levels needed to initiate endocytosis derived from diffusion of VGCC-dependent Ca2+ influxes from active zones. However, several reports have documented a requirement for extracellular Ca2+ during endocytosis. Furthermore, it has been proposed that a specific Ca2+ channel is required for SV endocytosis in Drosophila. The present study identified and characterized a SV- and presynaptic membrane-associated protein with three or four transmembrane domains that is evolutionarily conserved but has not been previously characterized in any species. Animals that lack Flower display the typical features of endocytic mutants. These include supernumerary boutons, a low number of SVs in boutons at rest, a severe depletion of SVs upon stimulation (except at active zones), enlarged SVs, a decrease in FM1-43 uptake, a rundown in neurotransmitter release upon repetitive stimulation, and an accumulation of endocytic intermediates (Yao, 2009).

flower mutants exhibit impaired Ca2+ handling, even when the boutons are at rest. This argues that Flower plays a role in Ca2+ homeostasis at rest as well as during endocytosis. Since Flower is associated with SVs and the presynaptic membrane, a reduction in Flower levels may cause lower resting Ca2+ levels because Ca2+ efflux from SVs or Ca2+ influx from the extracellular compartment is impaired. However, a role for Flower in SVs is unlikely. First, experimental evidence suggests that SVs have no or a very limited role in the sequestration of Ca2+ at NMJs in Drosophila. Second, although single SV fusions with the plasma membrane in flower mutants elicit larger amplitudes than in wild-type animals, our data suggest that this is due to the fact that the SVs in flower mutants are larger than wild-type SVs. Indeed, there is a near perfect correlation between the SV size and the size of the mEJPs, as observed in some other endocytic mutants. This suggests that the Flower protein present in the presynaptic membrane affects the resting Ca2+ levels, but does not exclude a role for the protein in SVs (Yao, 2009).

The second TM domain of Flower contains a 9 aa motif similar to the selectivity filters identified in Ca2+-gating TRPV channels. TRPV5 and 6 channels are homo- or multimeric channels that have been shown to form a pore lined by four negatively charged amino acids, similar to VGCCs. This study shows that Flower can form tetramers and higher order multimers and that its heterologous expression in salivary gland cells promotes Ca2+ influx. In addition, substitution of a conserved, negatively charged glutamate (E) residue in the second TM domain with a neutral amino acid (Q) abolishes this Ca2+ influx. Furthermore, the purified Flower protein in proteoliposomes can form a Ca2+-permeable cationic channel. A simple model is proposed to account for the data. SV exocytosis is triggered by VGCCs located at active zones. Subsequently, SVs, and hence Flower proteins are integrated in the presynaptic membrane, where they mostly localize to periactive zones, known sites for endocytosis). A homomultimeric Flower complex then promotes Ca2+ influx which triggers clathrin-mediated endocytosis. This is also supported by the observation that higher extracellular Ca2+ alleviates the endocytic defect. It is proposed that endocytosis removes most, but not all, of the Flower protein, thereby removing a key trigger for endocytosis. Thus, Flower may perform a simple autoregulatory role for itself during endocytosis. This model also addresses how exo-endocytosis coupling may be mediated at presynaptic terminals. The remainder of the Flower protein that is not endocytosed may help to regulate basal Ca2+ levels (Yao, 2009).

Cold temperature improves mobility and survival in Drosophila models of autosomal-dominant hereditary spastic paraplegia (AD-HSP)

Autosomal-dominant hereditary spastic paraplegia (AD-HSP) is a crippling neurodegenerative disease for which effective treatment or cure remains unknown. Victims experience progressive mobility loss due to degeneration of the longest axons in the spinal cord. Over half of AD-HSP cases arise from loss-of-function mutations in spastin, which encodes a microtubule-severing AAA ATPase. In Drosophila models of AD-HSP, larvae lacking Spastin exhibit abnormal motor neuron morphology and function, and most die as pupae. Adult survivors display impaired mobility, reminiscent of the human disease. This study shows that rearing pupae or adults at reduced temperature (18°C), compared with the standard temperature of 24°C, improves the survival and mobility of adult spastin mutants but leaves wild-type flies unaffected. Flies expressing human spastin with pathogenic mutations are similarly rescued. Additionally, larval cooling partially rescues the larval synaptic phenotype. Cooling thus alleviates known spastin phenotypes for each developmental stage at which it is administered and, notably, is effective even in mature adults. Further, cold treatment rescues larval synaptic defects in flies with mutations in Flower (a protein with no known relation to Spastin) and mobility defects in flies lacking Kat60-L1, another microtubule-severing protein enriched in the CNS. Together, these data support the hypothesis that the beneficial effects of cold extend beyond specific alleviation of Spastin dysfunction, to at least a subset of cellular and behavioral neuronal defects. Mild hypothermia, a common neuroprotective technique in clinical treatment of acute anoxia, might thus hold additional promise as a therapeutic approach for AD-HSP and, potentially, for other neurodegenerative diseases (Baxter, 2014).

This study demonstrates that cold temperature alleviates reduced mobility and survival caused by loss of Spastin function in Drosophila. This is the case for flies lacking endogenous spastin, as well as those expressing pathogenic human Spastin. Cold treatment during the pupal stage of development is sufficient to enhance the eclosion rate, climbing ability and lifespan of spastin mutant adults. Furthermore, cold administered only after pupal development, to fully developed adults, also improves mutant mobility. The timing of these two effective periods is consistent with the idea that cold alleviates spastin mutant phenotypes by acting on the developing adult nervous system during pupal metamorphosis, but is also potent after the nervous system has matured. This is extremely promising from a clinical viewpoint, suggesting that the therapeutic window in AD-HSP includes both developing and mature nervous systems (Baxter, 2014).

Although wild-type levels of mobility and survival were not often achieved, the temperature shift to 18°C confers considerable improvement. Some cold-treated flies are able to jump and even fly briefly, behaviors not observed in untreated mutants. Cooling can match or exceed the efficacy of rescue by the microtubule destabilizing drug vinblastine, which has been proposed as a therapeutic approach for AD-HSP. It has been shown that vinblastine doubles the ~12% eclosion rate of spastin5.75 null mutants; however, this study found the drug to be ineffective for null and HL44,HR388 eclosion, but improved eclosion by 65% for HWT,HR388, which is a more common, representative AD-HSP genotype associated with milder pathogenesis. In comparison, pupal cooling of spastin5.75 null mutants increases eclosion by 70% (Baxter, 2014).

Importantly, cooling during the pupal and adult stages does not affect eclosion or motor behavior in wild-type flies. This suggests that cooling not only compensates for defects in neuronal function caused by lack of Spastin (or other mutations), but is also innocuous to properly functioning neurons. Although cooling administered at the larval stage is ultimately deleterious to both control and spastin mutant adults, mutant larval synapses are effectively restored to wild-type morphologies. This suggests that cold is beneficial for some spastin-mediated defects at this stage, but also has nonspecific, toxic effects on a cell population required later, in adults (Baxter, 2014).

What is the mechanism(s) underlying the rescuing effect of cold? The demonstration that cold alleviates not just spastin mutant phenotypes, but also mutant phenotypes in fwe and kat-60L1, indicates that that rescuing effects of cold on nervous system function might be quite broad. All three genes are important in synapse formation, although kat-60L1 has been shown to act post-rather than pre-synaptically at larval and pupal stages. Reduced temperature could thus be generally beneficial to synaptic dysfunction, perhaps by reducing activity or metabolic load. Alternatively, fwe, spastin and kat-60L1 might share a common pathway component(s), as yet undiscovered, that is directly affected by cold. For example, cold itself is well known to destabilize microtubules, particularly at temperatures below 20°C, and is often used in experiments to depolymerize microtubules. Cold could thus substitute directly for the microtubule-severing function of Spastin by promoting microtubule destabilization. Cold-mediated rescue of Kat-60L1 mutants support this idea; however, obvious differences in stable microtubule distribution at cold-treated spastin5.75 synapses or in Drosophila S2R+ cells were not observed; fwe mutants, which have not been implicated in microtubule dysregulation, were also rescued by cooling (Baxter, 2014).

In humans, cooling has been shown to be generally neuroprotective, and mild or moderate therapeutic hypothermia (e.g. 33–35°C) has long had clinical applications, including reducing neurological injury in patients following cardiac arrest, traumatic brain injury, epilepsy and stroke. Furthermore, exposure to even near-freezing temperatures results in minimal neuropathology in rat and cat neocortex and hippocampus. Although commonly administered in situations involving acute brain injury, the mechanism by which cooling confers neuroprotection or therapeutic improvement is unknown, multifactorial and context-dependent (Baxter, 2014).

It will be important to characterize the in vivo effects of cold in mouse models of AD-HSP. The specificity of the effect of cold on mutant and not wild-type animals, together with the spatially localized neurodegeneration in AD-HSP, suggest that moderate hypothermia could be applied in a highly targeted manner in this disease context, with minimal negative effects. Future studies should furthermore elucidate the underlying cellular mechanisms and potentially broader applications of cold in alleviating neuronal dysfunction in neurodegeneration. Because Drosophila are ectothermic, with body temperatures that vary with their environment, they provide a straightforward system in which the cell biological effects of temperature change can be studied in vivo (Baxter, 2014).

Flower lose, a cell fitness marker, predicts COVID-19 prognosis

Risk stratification of COVID-19 patients is essential for pandemic management. Changes in the cell fitness marker, hFwe-Lose, can precede the host immune response to infection, potentially making such a biomarker an earlier triage tool. This study evaluated whether hFwe-Lose gene (homolog of Drosophila Flower) expression can outperform conventional methods in predicting outcomes (e.g., death and hospitalization) in COVID-19 patients. A post-mortem examination of infected lung tissue was performed in deceased COVID-19 patients to determine hFwe-Lose's biological role in acute lung injury. Then an observational study (n = 283) was performed to evaluate whether hFwe-Lose expression (in nasopharyngeal samples) could accurately predict hospitalization or death in COVID-19 patients. In COVID-19 patients with acute lung injury, hFwe-Lose is highly expressed in the lower respiratory tract and is co-localized to areas of cell death. In patients presenting in the early phase of COVID-19 illness, hFwe-Lose expression accurately predicts subsequent hospitalization or death with positive predictive values of 87.8-100% and a negative predictive value of 64.1-93.2%. hFwe-Lose outperforms conventional inflammatory biomarkers and patient age and comorbidities, with an area under the receiver operating characteristic curve (AUROC) 0.93-0.97 in predicting hospitalization/death. Specifically, this is significantly higher than the prognostic value of combining biomarkers (serum ferritin, D-dimer, C-reactive protein, and neutrophil-lymphocyte ratio), patient age and comorbidities (AUROC of 0.67-0.92). The cell fitness marker, hFwe-Lose, accurately predicts outcomes in COVID-19 patients. This finding demonstrates how tissue fitness pathways dictate the response to infection and disease and their utility in managing the current COVID-19 pandemic (Yekelchyk, 2021).

C9ORF7 (also known as CACFD1; referred to as hFWE) encodes human Flower (hFWE) and generates four protein-coding splice variants (hFWE1-hFWE4). Mining of the Encode database showed high levels of acetylation at the 27th lysine residue of the histone H3 protein (H3K27Ac) and hypersensitivity to DNAase-I around the +1 open reading frame (ORF), suggesting that the hFWE locus is transcriptionally active. The hFWE exons are conserved, and alternative splicing produces four hFWE isoforms of transmembrane proteins, with membrane-spanning domains as predicted. To identify potential hFWEWin or hFWELose isoforms (as observed in Drosophila), CRISPR-assisted MCF-7 hFWEKO cells were used that expressed individual hFWE isoforms, and they were used for a co-culture protocol. MCF-7 hFWEKO cells lacked expression of hFWE mRNA, but did not differ from wild-type cells in terms of proliferation, cell-cycle, reactive oxygen species (ROS), annexin-V, and clonogenic assays. Lentivirus-assisted overexpression of the four hFWE isoforms had no effect on cell proliferation, cell death, cell volume, cellular sphericity, or annexin-V staining when compared to MCF-7 hFWEKO cells. From these data, it is concluded that, as in Drosophila, human cells homogenously expressing single hFWE isoforms do not undergo apoptosis (Yekelchyk, 2021).

To identify potential hFWEWin and hFWELose isoforms, live-cell imaging was performed, which showed that cells expressing hFWE1 or hFWE3 undergo cell death only when co-cultured with cells expressing either hFWE2 or hFWE4; cell death was accompanied by increased proliferation of cells expressing hFWE2 or hFWE4. These findings suggest that hFWE2 and hFWE4 function as hFWEWin isoforms whereas hFWE1 and hFWE3 function as hFWELose isoforms. Furthermore, elimination of hFWELose-expressing cells occurred through caspase-dependent apoptosis, because the addition of the caspase inhibitor Z-VAD-FMK inhibited this process and annexin-V staining. Cells expressing individual isoforms (hFWE1-hFWE4) did not induce loss of co-cultured hFWEKO cells, indicating that hFWE-mediated cell elimination requires cells that express both hFWEWin and hFWELose. Elimination of hFWELose cells was dependent on contact with hFWEWin cells, because no apoptosis was observed during imaging of low-density-plated hFWE1-GFP+ and hFWE2-RFP+ cells. Tests were performed to see whether factors secreted by hFWEWin-expressing cells were involved in cell elimination by transferring conditioned medium from hFWE2-expressing cells onto hFWE1-expressing cells. This had no significant effect on cell death. Live-cell calcium imaging using Fluo-4AM dye showed that expression of individual hFWE isoforms in MCF-7 hFWEKO cells had no effect on cellular calcium levels, and the calcium chelator BAPTA-AM did not interfere with the elimination of hFWELose-expressing cells, suggesting that hFWE isoforms do not function as calcium transporters and their function is calcium-independent. It is concluded that human Flower proteins work as canonical fitness fingerprints (Yekelchyk, 2021).

To study the role of Flower proteins in human cancer, quantitative PCR (qPCR) was used to measure the expression of hFWE isoforms in laser-captured formalin-fixed paraffin-embedded (FFPE) samples of cancerous and adjacent stromal tissue from individuals with malignant tumours (25 breast, 21 colon), benign tumours (10 breast, 10 colon), and normal tissue (10 breast, 10 colon, from matched patients). All hFWE isoforms showed poor expression in normal tissue. The tumour tissue showed increased expression of hFWEWin isoforms, and stromal tissue showed significant upregulation of hFWElose isoforms (more in malignant tumours than in benign tumours and their stroma). Sequential laser capture of breast and colon stromal tissue showed a significant increase in expression of hFWELose isoforms in the first 400 μm from the edge of the tumour tissue, where genes involved in the regulation of apoptosis were concurrently upregulated. Fluorescence in situ hybridization (FISH) showed increased expression of hFWEWin isoforms exclusively in breast cancer cells; further qPCR showed that, unlike tumour epithelial cells, normal epithelial cells do not exhibit high hFWEWin expression. MCF-7 hFWEKO cells expressing hFWE2 induced apoptosis of co-cultured primary epithelial cells or fibroblasts (Yekelchyk, 2021).

Two anti-hFWE antibodies were generated (one against the N terminus that is common to all four isoforms, and the other specific to the hFWEWin isoforms (encoded by exon 3)). Immunohistochemistry was used to validate the antibody specificity in sections of hFWEKO cells overexpressing hFWE1, hFWE2, hFWE3 or hFWE4. Immunostaining with anti-hFWEWin and anti-hFWE-N-term antibodies produced positive membrane staining in breast cancer cells. Immunohistochemistry with the anti-hFWEWin antibody showed that hFWEWin was expressed exclusively in breast, colon, and squamous cell carcinoma (SCC) cancers. Immunohistochemistry of the same breast, colon, and SCC cancers with anti-hFWE-N-term antibodies-which recognize both hFWEWin and hFWELose isoforms-showed staining of both cancer and stromal cells. Both antibodies showed poor staining in normal tissue. The main observations from this analysis were that both types of hFWE isoform tend to be low in healthy tissue, hFWEWin isoforms are upregulated in cancer tissue (more in malignant than in benign cancers), and levels of hFWELose isoforms are unusually high in the cancer-adjacent stroma (Yekelchyk, 2021).

These results suggested that tumour cells exhibit high hFWEWin expression and tend to grow in areas of stroma with high hFWELose expression. To investigate whether overexpression of hFWELose in host tissue and hFWEWin in the tumour might result in aggressive tumour growth, FweKO mice were generated bearing orthotopic breast tumours derived from MCF-7 cells. Individual hFWE isoforms were overexpressed in the mammary tissue of FweKO mice to study the effects of host cells, and hFWEKO MCF-7 cells overexpressing GFP-tagged individual hFWE isoforms were used to generate the tumours. When all possible combinations were analyzed, it was found that MCF-7 hFWEKO cells expressing hFWE2-GFP or hFWE4-GFP generated aggressive tumours when implanted into mammary tissue expressing hFWE1 or hFWE3 . The results suggest that interactions between expression of hFWEWin isoforms in tumours and hFWELose isoforms in the stroma are important for tumour growth (Yekelchyk, 2021).

The role of host tissue hFWE isoforms was investigated by using hFWE2-overexpressing hFWEKO HCT-116 and MCF-7 cells to generate tumours in FweWT and FweKO mice. Both HCT-116 hFWE2+ and MCF-7 hFWE2+ tumours were larger with higher incidences of metastasis in FweWT mice than in FweKO mice. The expression of hFWELose isoforms was upregulated in wild-type MCF-7 cells when they were co-cultured with hFWE2-expressing hFWEKO MCF-7 cells. Similarly, expression of hFWELose isoforms was increased in tumour-adjacent tissue from FweWT mice after implantation of hFWE2-expressing MCF-7 tumours when compared with before. From these experiments, it is concluded that the expression of Win isoforms in the tumour is sufficient to induce increased expression of Lose isoforms non-autonomously in the surrounding cells. These data show that stromal Lose and tumoral Win isoforms cooperate to promote aggressive cancer growth (Yekelchyk, 2021).

Next, whether inhibition of Flower could provide anti-cancer benefits was investigated. Amongst ectopic tumours derived from hFWEWT and hFWEKO HCT-116 cells, those derived from hFWEWT cells were larger. Knockdown of hFWE expression by short hairpin RNA (shRNA) in hFWEWT HCT-116 cells before xenografting reduced both tumour volume and probability of metastasis. The growth and metastatic potential of HCT-116 tumours treated with shRNA targeting hFWE was rescued by ectopic expression of codon-optimized hFWE2 cDNA (which is unaffected by anti-hFWE shRNA). In a gain-of-function experiment, hFWE2 was overexpressed in CCL-218 (colon cancer) cells, which express low levels of hFWEWin isoforms; this resulted in a nearly threefold increase in their tumour growth potential (Yekelchyk, 2021).

A panel of 18 metastatic cell lines of multiple origins (breast, colon, prostate, pancreatic, and ovarian) was screened for hFWEWin expression. Cells with higher (HCT-116 colon and DU-145 prostate) or lower (CCL-218 colon and OVCAR-8 ovarian) expression of hFWEWin were selected, and their tumorigenic potential was observed. The therapeutic potential of hFWE gene knockdown was tested in combination with standard-of-care chemotherapy against the tumorigenic potential of these cancer cell lines. DU-145 and HCT-116 cells had significantly higher tumorigenic potential and generated larger tumours than CCL-218 and OVCAR-8 cells. hFWE knockdown significantly reduced tumour volumes. Control xenografts and those treated with scrambled shRNA generated much larger tumours than xenografts treated with anti-hFWE shRNA or chemotherapy (fluorouracil (5-FU) for colon cancer, or cisplatin and docetaxel for prostate and ovarian cancer). Treatment with a combination of anti-hFWE shRNA and chemotherapy further reduced, and in some cases completely eliminated, tumorigenesis. Control experiments demonstrated the efficiency of hFWE shRNA in all tumours. Metastatic probability was reduced markedly upon hFWE knockdown and even further by combined treatment with chemotherapy. Thus, it is concluded that human hFWE proteins can have a powerful effect on tumorigenicity and propose that therapies targeting these proteins have the potential to impair cancer growth and metastasis (Yekelchyk, 2021).

This study has shown that the human Flower protein code mediates competitive interactions that depend on cell contact and caspase-dependent apoptosis but not on calcium transport. Homogeneous expression of any of the four Flower proteins does not trigger cell elimination or affect cell fitness in terms of proliferation, survival, or clonogenicity. Cell elimination requires different levels of Win and Lose isoforms in neighbouring cells, demonstrating that relative levels of these proteins are more important than absolute levels for cell selection. In addition, it was found that human tumours use cell fitness markers to gain a competitive advantage over neighbouring cells. This work represents a conceptual advance in understanding the role of cell fitness comparisons in the progression of human cancer. It describes a concrete molecular mechanism that regulates oncogenic growth and metastasis via the Flower fitness comparison system. This appears to be a striking general feature of solid tumours and an unrecognized prerequisite for tumour growth. Notably, it was also shown that inhibition of cell competition can be combined with standard-of-care chemotherapy to further prevent tumour growth and metastasis of colon, ovarian and prostate cancers (Yekelchyk, 2021).

Cytotoxic granule endocytosis depends on the Flower protein

Cytotoxic T lymphocytes (CTLs) kill target cells by the regulated release of cytotoxic substances from granules at the immunological synapse. To kill multiple target cells, CTLs use endocytosis of membrane components of cytotoxic granules. This study examined the potential calcium dependence of endocytosis in mouse CTLs on Flower, which mediates the calcium dependence of synaptic vesicle endocytosis. In Drosophila melanogaster Flower is predominantly localized on intracellular vesicles that move to the synapse on target cell contact. Endocytosis is entirely blocked at an early stage in Flower-deficient CTLs and is rescued to wild-type level by reintroducing Flower or by raising extracellular calcium. A Flower mutant lacking binding sites for the endocytic adaptor AP-2 proteins fails to rescue endocytosis, indicating that Flower interacts with proteins of the endocytic machinery to mediate granule endocytosis. Thus, these data identify Flower as a key protein mediating granule endocytosis (Chang, 2017).

Search PubMed for articles about Drosophila Flower

Baxter, S.L., Allard, D.E., Crowl, C. and Sherwood, N.T. (2014). Cold temperature improves mobility and survival in Drosophila models of autosomal-dominant hereditary spastic paraplegia (AD-HSP). Dis Model Mech 7: 1005-1012. PubMed ID: 24906373

Chang, H. F., et al. (2017). Cytotoxic granule endocytosis depends on the Flower protein. J Cell Biol [Epub ahead of print]. PubMed ID: 29288152

Diaz, B. and Moreno, E. (2005). The competitive nature of cells. Exp Cell Res 306: 317-322. PubMed ID: 15925586

Fernandez-Hernandez, I., Rhiner, C. and Moreno, E. (2013). Adult neurogenesis in Drosophila. Cell Rep 3: 1857-1865. PubMed ID: 23791523

Levayer, R., Hauert, B. and Moreno, E. (2015). Cell mixing induced by myc is required for competitive tissue invasion and destruction. Nature 524: 476-480. PubMed ID: 26287461

Merino, M. M., Rhiner, C., Portela, M. and Moreno, E. (2013). 'Fitness fingerprints' mediate physiological culling of unwanted neurons in Drosophila. Curr Biol 23: 1300-1309. PubMed ID: 23810538

Merino, M. M., Rhiner, C., Lopez-Gay, J. M., Buechel, D., Hauert, B. and Moreno, E. (2015). Elimination of unfit cells maintains tissue health and prolongs lifespan. Cell 160: 461-476. PubMed ID: 25601460

Moreno, E., Fernandez-Marrero, Y., Meyer, P. and Rhiner, C. (2015). Brain regeneration in Drosophila involves comparison of neuronal fitness. Curr Biol [Epub ahead of print]. PubMed ID: 25754635

Petrova, E., Lopez-Gay, J. M., Rhiner, C. and Moreno, E. (2012). Flower-deficient mice have reduced susceptibility to skin papilloma formation. Dis Model Mech 5: 553-561. PubMed ID: 22362363

Portela, M., et al. (2010). Drosophila SPARC is a self-protective signal expressed by loser cells during cell competition. Dev. Cell 19(4): 562-73. PubMed ID: 20951347

Rhiner, C., Lopez-Gay, J. M., Soldini, D., Casas-Tinto, S., Martin, F. A., Lombardia, L. and Moreno, E. (2010). Flower forms an extracellular code that reveals the fitness of a cell to its neighbors in Drosophila. Dev Cell 18: 985-998. PubMed ID: 20627080

Yao, C. K., Lin, Y. Q., Ly, C. V., Ohyama, T., Haueter, C. M., Moiseenkova-Bell, V. Y., Wensel, T. G. and Bellen, H. J. (2009). A synaptic vesicle-associated Ca2+ channel promotes endocytosis and couples exocytosis to endocytosis. Cell 138: 947-960. PubMed ID: 19737521

Yao, C. K., Liu, Y. T., Lee, I. C., Wang, Y. T. and Wu, P. Y. (2017). A Ca2+ channel differentially regulates Clathrin-mediated and activity-dependent bulk endocytosis. PLoS Biol 15(4): e2000931. PubMed ID: 28414717

Yekelchyk, M., Madan, E., Wilhelm, J., Short, K. R., Palma, A. M., Liao, L., Camacho, D., Nkadori, E., Winters, M. T., Rice, E. S., Rolim, I., Cruz-Duarte, R., Pelham, C. J., Nagane, M., Gupta, K., Chaudhary, S., Braun, T., Pillappa, R., Parker, M. S., Menter, T., Matter, M., Haslbauer, J. D., Tolnay, M., Galior, K. D., Matkwoskyj, K. A., McGregor, S. M., Muller, L. K., Rakha, E. A., Lopez-Beltran, A., Drapkin, R., Ackermann, M., Fisher, P. B., Grossman, S. R., Godwin, A. K., Kulasinghe, A., Martinez, I., Marsh, C. B., Tang, B., Wicha, M. S., Won, K. J., Tzankov, A., Moreno, E. and Gogna, R. (2021). Flower lose, a cell fitness marker, predicts COVID-19 prognosis. EMBO Mol Med: e13714. PubMed ID: 34661368

date revised: 23 June 2023

Home page: The Interactive Fly © 2011 Thomas Brody, Ph.D.