Skip to main content

Isoform-specific functions of NAD-synthesising enzyme NMNAT in compartmentalised neuronal death

Final Report Summary - NEURONAD (Isoform-specific functions of NAD-synthesising enzyme NMNAT in compartmentalised neuronal death)

The project aim was the investigation of the role of NAD-synthesising enzyme NMNAT in the maintenance and sensitivity to insult of neurons. In particular, we tested the hypothesis that nuclear NMNAT1 and cytosolic NMNAT2 isozymes control neuron maintenance and survival in the cell body and in the axon, respectively, with the purpose of identifying new targets for intervention in pathological conditions of the nervous system. Preserving the health of neurons represents a primary medical target in our ageing society, specially in the European Community, where the increased prevalence of psychiatric and neurodegenerative disorders profoundly affects quality of life and threatens to overwhelm the healthcare systems.

There is little information about the contribution of NMNAT1 to neuron maintenance, primarily because specific inhibitors are not available and complete knock out (KO) of the relative gene in mouse is lethal. We proposed therefore to generate a neuron specific Nmnat1 conditional KO, investigating whether NMNAT1 is required for the integrity of the neuron.

Nmnat1 conditional KO was obtained by utilizing the Cre/loxP system, both in vitro, by microinjecting a Cre recombinase-expressing vector in primary neurons of Nmnat1flox/flox mice, and in vivo, by crossing Nmnat1flox/flox with SLICK-V mice, which express Cre recombinase in a subset of YFP-labeled neurons.

The in vitro results showed the deleterious effect of NMNAT1 depletion in dissociated superior cervical ganglia (SCG) neurons, with cell body degeneration occurring faster and to a greater extent than the control cells, represented by SCG microinjected with the empty vector (Fig. 1).

For the in vivo study SLICK-V mice have been mated with the Nmnat1flox/flox mice, to obtain the SLICK x Nmnat1flox/+ heterozygous mice. Consequently, The first littermates have been crossed to obtain the SLICK x Nmnat1flox/flox homozygous mice. 3- to 5-month old SLICK x Nmnat1flox/flox mice have been treated with tamoxifen to activate Cre recombinase. Analysis of the brain of these mice at different time points, even at 5 months after NMNAT1 inactivation, however, has not shown any morphological changes in neurons (Fig. 2). Based on the recent discovery that NMNAT1 mutations are associated with vision defect in the retina, we also analysed the morphology of labelled neurons in the retina. Also in this case, we could not detect any morphological change in fluorescent neurons with respect to mice treated with the vehicle. We cannot exclude a long-term stability of residual NMNAT1 after gene disruption, or the inability to find significant differences due to the limited number of fluorescent neurons.

The second main objective of this project was the demonstration that NMNAT2 promotes axon survival by scavenging its substrate nicotinamide mononucleotide (NMN). The importance of NMNAT2 in the degeneration of injured axons has been previously demonstrated, but the mechanism involved is still unknown. Our preliminary data led us to propose that the NMNAT substrate NMN is the trigger of axon degeneration. We demonstrated that pharmacological inhibition of NMN synthesis by using 2 different drugs (FK-866 and CHS-828) significantly delays injury induced axon degeneration. This effect is surmounted by NMN at various concentrations (Fig. 3).

In order to further demonstrate that the protective property of NMNAT2 is depending on its ability to scavenge NMN, we utilized the bacterial protein NMN Deamidase, which, converting NMN to its deamidated form, provides an NMN scavenging system independent of NMNAT2.

Escherichia coli (wild type and two enzimatically inactive mutants) and Shewanella oneidensis NMN Deamidase fused to the marker protein eGFP were expressed in mouse SCG neurons. While SCG neurites microinjected with the empty vector completely degenerated 8 h after axotomy, those expressing E. coli or S. oneidensis NMN Deamidase were strongly protected 48-72 h after cut, with some axons still intact 100 h after cut (Fig. 4). The expression of two E. coli enzyme mutants provided a limited neuroprotection for 24 and 8 h in line with their negligible but still detectable activity.
To extend these observations in vivo, we generated a transgenic mouse expressing E. coli NMN Deamidase. We confirmed the expression of the bacterial protein by measuring the relative enzymatic activity in the brain. By using different approaches we demonstrated that the degeneration of sciatic nerve after cut (Wallerian degeneration) in this mouse is strongly delayed, even more than 10 fold with respect to that of the wild type mouse. Indeed, neurofilament heavy chain western blots of injured sciatic and tibial nerves (Fig. 5), light microscopy images of injured sciatic nerve sections (Fig. 6) and fluorescence images of labelled axons in sciatic nerves explants of NMN Demidase mice crossed to mice expressing the fluorescent protein YFP (Fig. 7) clearly showed the preservation of the sciatic nerve of the transgenic mouse several days after lesion. Most importantly, in collaboration with the fellow’s previous lab, we demonstrated that NMN accumulation in sciatic nerve after cut is inhibited in the transgenic mouse. We also confirmed the neuroprotective properties of NMN Deamidase in cultured primary neurons of NMN Deamidase pups: in this case neurites expressing the bacterial protein not only were protected against degeneration induced by physical injury (Fig. 8), but also against degeneration induced by the neurotoxin vincristine or by NGF deprivation (Fig. 9). Finally, in addition to the structural preservation of the sciatic nerve, we also verified its functional preservation, demonstrating its prolonged capacity of transmission of an evoked electric potential three days after cut, in collaboration with Lucy Donaldson group at University of Nottingham.

Wallerian degeneration shows morphological and functional features in common with the axon pathology typical of several neurodegenerative diseases. Therefore, the NMN Deamidase tg mouse that we have generated, which is highly resistant to injury-induced neurodegeneration, will be a useful tool to test the hypothesis that an increase in NMN causes axon loss in models of neurodegenerative diseases. This will be of great interest for the pharmacological research and will have a strong socio-economic impact on the civil society of the European Community.