Inhibition of porcupine by Wnt-C59 triggers the loss of neurites in embryonic cultures of hippocampal neurons
Neurons construct a characteristic polarized dendrite axis, which is physiologically important for receiving and processing axon synaptic signals. In low-density cultures, polarization and active axon growth occur between DIV 2-6; therefore, this time is ideal to study the role of endogenous Wnt signaling in dendritic tree formation and axonal processes. For this purpose, we blocked Wnt signaling with the acylase inhibitor Wnt-C59 which is specific for the acylase PORCN. Rat hippocampal neurons DIV 4, were treated with different Wnt-C59 concentrations (0, 0.1, 1, 10, and 100 nM) for 36 h, and stained for microtubule-associated protein 1B (MAP1B) which is abundantly expressed in actively spreading neurons (29, 30). Neurites were visualized with phalloidin which labels F-actin and nuclei were stained with ATF-2. Our results showed that PORCN inhibition significantly decreased both the number of neurites/neurons (Fig. 1A; see representative micrographs, **p<0.01) and the length of the neurites (Fig. 1B, **p<0.01). Interestingly, a large amount of terminal filopodia/lamellipodia was observed with Wnt-C59 treatment (Fig. 1A, see white arrow).
Inhibition of porcupine by Wnt-C59 results in loss of contact with other cells in embryonic cultures of hippocampal neurons
Another phenomenon that was verified in our experiments is the loss of contacts with other types of cells present in primary cultures. This phenomenon was evaluated by counting the number of neuronal processes in contact with another type of nonneuronal cell. The number of contacts with another cell type decreases by half when the neuronal cultures are treated with the Wnt-C59 inhibitor (Fig. 2A, see representative micrographs and lower graph, **p<0.01). The treatment of hippocampal neuron cultures with the inhibitor also results in a significant decrease in b-catenin, a key protein in the Wnt signaling pathway, with an apparent increase in the inactive form of GSK-3b kinase (phosphorylated in serine 9). The Wnt3a ligand also decreased after Wnt-C59 treatment (Fig. 2B, see representative blot and quantification of different proteins, *p<0.05). See Supplementary Fig. 1, for further studies on the changes in b-catenin and Wnt3a in L cells.
Inhibition of porcupine by Wnt-C59 treatment does not change axonal polarity in embryonic hippocampal neurons
The polarized sorting and trafficking of newly synthesized proteins to the somato-dendritic and axonal domains of neurons occurs by selective incorporation into distinct populations of vesicular transport carriers. Voltage-gated Na+ and K+ channels are crucial for the efficient initiation and saltatory propagation of action potentials along myelinated axons of vertebrates, are clustered at the axon initial segment (AIS) and nodes of Ranvier. The clustering of these channels is mediated by ankyrin-G (AnkG) (31). Previous studies concluded that AIS, an actin-based filter, selectively prevents the passage of somato-dendritic vesicles into the axon (32). Conventional kinesin-1 is a major anterograde motor that operates in axons and consists of a heavy-chain (KIF5A, KIF5B, or KIF5C) dimer and two light chains (KLC) binding to the C termini of the dimer (33). One of the questions that arises from our experiments is whether the inhibition of PORCN, with the consequent lack of ligands and activation of the Wnt pathway, might alter neuronal polarity in some way. For this purpose, embryonic neurons were treated with the inhibitor Wnt-C59 and two proteins of the AIS were studied, AnkG and KIF5 (32, 34). Our results showed that both proteins were correctly located in the neuronal axon of the neuron (Fig 3). Embryonic neurons were treated with different concentrations of Wnt-C59 (0.1, 1, 10 nM) and stained for AnkG and Piccolo, a protein present in the Piccolo-Bassoon transport vesicle (PTV), which are dense particles that form the active zone of the neurons (15). Our experiments show that both proteins are correctly located, together with actin activity in the same place (Fig. 3A, see representative micrograph, white arrow shows AIS). Embryonic hippocampal neurons transfected with KIF5 and treated with 10 nM Wnt-C59, showed a similar localization to control cells together with specific proteins of the neuronal axon (Fig. 3B, see representative micrographs). We conclude that inhibition of porcupine with Wnt-C59 did not affect the neuronal polarity of the axon.
Different exogenous Wnt ligands recover the morphology of embryonic hippocampal neurons in culture
To determine the specific role of specific Wnt ligands, embryonic young primary hippocampal neurons were treated with Wnt-C59 to shut down Wnt signaling activity. We evaluated the individual activity of Wnt ligands with canonical activity, such as Wnt 3a and 7a, and a known non canonical Wnt5a ligand.
A. Treatment with exogenous canonical Wnt3a ligand on neuronal morphology
Embryonic hippocampal neurons (DIV 4), were treated with 10 nM Wnt-C59 from DIV 2 to DIV 4. These cultures were treated with different concentrations of the Wnt3a ligand for 24 h (50, 100, 200 ng/ml) and stained with MAP-1B (red), ATF-2 a nuclear transcription factor (blue) and phalloidin to observe actin (green), which can be seen at the tips of neuronal processes and to a lesser extent in neurons treated with Wnt-C59 (see white arrows). Embryonic neurons treated with Wnt3a showed a recovery of the length of neurites similar to the control (Fig. 4A, see representative micrographs). Another parameter that was recovered was the dendritic arbor complexity, concomitant with the recovery of secondary and tertiary projections (Fig. 4B, see graphs).
B. Treatment with exogenous noncanonical Wnt5a ligand on neuronal morphology
Sister embryonic hippocampal neurons (DIV 4), were treated with 10 nM Wnt-C59 from DIV 2 to DIV 4. These cultures were treated with different concentrations of Wnt5a for 24 h (50, 100, 200 ng/ml) and stained with MAP1B (red), ATF (blue) and phalloidin to visualize actin (green), which can be observed at the tips of neuronal processes and to a lesser extent in neurons treated with Wnt-C59 (Fig. 5; representative micrographs). Our results showed that embryonic neurons treated with different concentrations of Wnt5a showed a recovery of the length of neurites to a longer extension than control neurons, by 40-50% of the control depending on Wnt5a concentration (Fig. 5A, see representative micrographs) F-actin activity was marked with phalloidin (see white arrow). The complexity of the dendritic arbor was also recovered, together with the secondary and tertiary projections (Fig. 5B, see graph).
C. Effect of treatment with exogenous canonical Wnt7a ligand on neuronal morphology
Embryonic hippocampal neurons (DIV 4), were treated with 10 nM Wnt-C59 from DIV 2 to DIV 4. These cultures were treated with different concentrations of the Wnt7a ligand for 24 h (50, 100, 200 ng/ml) and stained with MAP-1B (red), ATF-2 a nuclear transcription factor (blue) and phalloidin to visualize actin (green), which can be observed at the tips of neuronal processes (see white arrow). The embryonic neurons treated with Wnt7a ligand showed a recovery of the length of neurites to a much longer extension than the control, but at a much lower concentration of Wnt7a than of Wnt5a (Fig. 6A, see representative micrographs; B, left graph). According to our results, 24-hour treatment of embryonic neurons restores dendritic tree complexity, to an extent similar to that observed with the control (Fig. 6B, see right graph).