There are four major findings from this study: 1) age-dependent abnormal climbing in dfmr1 mutant flies can be genetically rescued, 2) excessive grooming is identified as a new behavioral defect in dfmr1 mutant flies, 3) excessive grooming can be suppressed by reserpine, and 4) dVMAT mRNA and protein levels are increased in the absence of dFMRP.
In a previous study we revealed abnormal climbing activity in dfmr1 mutant flies that progresses with age. Our results in this study confirm this finding, and additionally show that introducing a wild-type dfmr1 transgene into the dfmr1 mutant background restores normal climbing behavior. Further, a frameshift mutation in the open reading frame of the transgene abolishes the rescue of climbing behavior. These results demonstrate that the abnormal climbing in dfmr1 mutant flies is directly caused by the loss of dFMRP.
In this study we have also identified excessive grooming as an important and novel behavioral defect in the fly model of FXS. Our results show that dfmr1 mutant flies groom significantly more than control flies, and that mutant flies also have significantly longer grooming bouts. Further, this excessive grooming intensifies with age in dfmr1 mutant flies, whereas control flies show essentially no change in grooming activity over time. A wild-type copy of the dfmr1 gene can rescue the excessive grooming defect in dfmr1 mutant flies. It is worth noting that FS mutants show more dramatic climbing and grooming defects compared to dfmr1 mutants. We do not know the exact underlying cause, but it may be due to the presence of either mRNA or a truncated peptide produced from the FS rescue fragment, having gain of function effects.
In video recordings, control flies mostly walk around the observation chamber and groom occasionally, but rarely stand motionless. In contrast, the mutant fly spends more time grooming. It is possible that grooming is a default activity that occurs whenever a fly is not walking. If this is the case, excessive grooming in dfmr1
mutant flies could result indirectly from problems in walking. This explanation would be consistent with our observation that dfmr1
mutant flies exhibit postural problems and uncoordinated movement. However, the dfmr1
flies are capable of climbing after a brief period of mechanical disturbance (i.e., knocking them down in the graduated cylinder), albeit at a slower speed compared to control flies. Further, reserpine suppresses grooming in dfmr1
mutant flies without improving walking (data not shown). These observations suggest that grooming is not simply a default behavior in the absence of walking and that dfmr1
mutations specifically cause excessive grooming. Notably, Fmr1
KO mice have also been reported to exhibit excessive grooming when presented with social stimuli 
. A study of self-injurious behavior in FXS patients reported a prevalence of harmful rubs and scratches 
. Hence, heightened repetitive activity such as grooming is a common behavioral defect in FXS.
Although reducing mGluR signaling has been shown to rescue learning and memory defects in both mouse and fly FXS models, we find that the mGluR antagonist MPEP enhances excessive grooming in dfmr1
mutant flies. This is not completely surprising, as the absence of dFMRP likely alters numerous signaling pathways and developmental processes of the nervous system. MPEP also fails to rescue abnormal sleep 
and circadian rhythm 
mutant flies, which may impact locomotor activity like grooming. It is worth noting that dfmr1
mutant flies did not groom more when treated with LiCl, suggesting that mGluR antagonists and LiCl may have different neuronal targets. An interesting question that arises from these results is whether an mGluR agonist might suppress grooming in dfmr1
mutant flies. Previous results have shown that glutamate at concentrations as low as 5 µM is toxic to dfmr1
mutant flies and significantly affects various behaviors in the fly 
. This makes it difficult to assess the potential benefit of mGluR agonists on grooming.
Previous work shows that dopamine plays a role in FXS in both mice 
and Drosophila 
, and that biogenic monoamines stimulate fly grooming 
. In our studies, blocking dVMAT with reserpine suppresses excessive grooming in dfmr1
mutant flies, but only significantly at 50 µM. Control flies groom significantly less when treated with just 10 µM. These results indicate that dfmr1
mutant flies are less sensitive to reserpine's effect on grooming. However, we cannot exclude the possibility that reserpine has additional targets and therefore generally sedates the fly. Both suppression of dVMAT as well as a non-specific target could slow down most motor activities including grooming. Alternatively, it is possible that basal monoamine activity is required for grooming, and therefore shutting down monoamine signaling may block the behavior.
In our study, we find elevated levels of dVMAT
transcript and protein in dfmr1
mutant flies. Although these increases are not statistically significant in some instances, they are consistent in both mutant lines. However, it is not clear from our results how the loss of dFMRP leads to increased dVMAT
expression. The transcription of dVMAT
may be directly increased. Alternatively, degradation of dVMAT
mRNA may decrease in the absence of dFMRP, a distinct possibility as FMRP has been previously indicated to regulate mRNA stability 
. How dFMRP regulates dVMAT protein levels is also unclear. Elevated dVMAT protein levels may occur exclusively because of increased transcript levels, but could also result from increased translation or reduced degradation of the protein. Nonetheless, our observations are in agreement with the known function of FMRP as a regulator of transcription and translation 
Many factors may contribute to the excessive grooming in dfmr1
mutant flies, and our data do not resolve whether upregulation of dVMAT directly influences this behavior. Overexpression of dVMAT stimulates grooming in flies 
, and dopamine levels are increased in dfmr1
mutant brains 
. Monoamines could directly or indirectly modulate multiple downstream signaling pathways involved in grooming. The hyposensitivity to reserpine seems to suggest that a greater number of dVMATs are present on mutant synaptic vesicles, as a higher concentration of the drug is required to reduce grooming. We note that overexpression of dVMAT in serotonergic and dopaminergic neurons leads to hypersensitivity to reserpine on grooming 
. One likely explanation of these differences is that dfmr1
mutations affect not only monoamine cells but also other cells such as neurons in the mushroom bodies and neurons postsynaptic to monoamine cells. We are also are aware that dopamine signaling is reduced in the forebrain of Fmr1
KO mice 
. Thus, while plausible given the effect of reserpine, we cannot establish a clear causal relationship between excessive grooming and dVMAT expression levels.
Understanding how FMRP functions in development and aging will be crucial for effective treatment of FXS 
. Studies in mouse and Drosophila
indicate that FMRP is temporally regulated and that treatment requires proper timing 
. Our results add to the growing evidence of the importance of FMRP in age-related processes, and also demonstrate that hyperactivity and repetitive behavior increase with age in the Drosophila
model. Interestingly, the severity of autistic behavior and anxiety has been found to increase with age in studies of FXS patients 
. Our results indicating that reserpine is effective in adult dfmr1
mutant flies could help develop or improve treatment, as they suggest that hyperactive and repetitive behavior in older patients is potentially reversible.
Although we believe excessive grooming in dfmr1 mutant flies is a model of an impulsive and repetitive behavior, animal models can never completely recapitulate human disorders. The mechanisms underlying repetitive behaviors in FXS patients are likely much more complex. Nonetheless, we demonstrate a correlation between monoamine signaling and the excessive grooming phenotype in dfmr1 mutant flies and that VMAT is a protein that merits further study in FXS. Importantly, our study provides potentially useful information for improving the pharmaceutical treatment of FXS symptoms in human patients.