Neuronal heparanase 1 and its involvement in homeostatic plasticity

Abstract

 Heparanase1 (HPA1) is the mammalian β-D-endoglucuronidase that acts to cleave heparan sulfate (HS). The pathological roles of HPA1 have been extensively studied in malignancy progression and immune disorders. However, the physiological functions of HPA1 remain largely elusive. Knowledge of biochemical properties of HPA1 was mainly derived from studies with non-neuronal mammalian cells. Earlier lab members found (1) HPA1 expression in neurons and astrocytes generally in the central nervous system; (2) specifically, treatment of hippocampal slice (rat) with recombinant proHPA1 led to dose-dependent decrease in both the synaptic strength and the level of induced LTP in the circuit. If HPA1 is delivered to the synapse, does network activity determine the rate of delivery and release into the perisynaptic environment? We therefore hypothesized that HPA1 transportation to synaptic terminals is sensitive to prevailing activities of neuronal circuits for roles in homeostatic plasticity. We aimed first to document subcellular localization of HPA1 in neurons. Purified hippocampal neurons were maintained in culture for the study. Immunocytochemistry revealed perinuclear localization of HPA1 in hippocampal neurons as has been observed in non-neuronal mammalian cells. HPA1-immunoreactivity was also found along both axons and dendrites. Double immunofluorescence revealed colocalization of HPA1 with distinct presynaptic markers but rarely with post-synapses at dendritic spine or shaft. Western blot analysis of synaptosomal fractions revealed enrichment of the 50 KDa subunits of mature HPA1.Gel mobility shift assay demonstrated HS-cleaving enzymatic activity in the synaptosomal fractions, in support of the localization of mature HPA1 at presynaptic terminals. With the hippocampal neurons in culture (18 DIV) and transfection with the HPA1-mCherry construct for live-cell imaging and kymograph recording, the trafficking of HPA1 along both dendrites and axons was monitored. The mCherry-marked dots were found to be transported in-step with EGFP-Lamp2 marked vesicles, contrasting diffuse mCherry signals demonstrated by transfectants with a HPA1-mCherry construct devoid of the signal sequence. Dendritic and axonal HPA1 differed in trafficking dynamics: dendritic HPA1-mCherry were frequently found to be stationary or in a tug-of-war mode of movement, contrasting the directional trafficking in axons. HPA1 trafficking along both dendrites and axons was then studied in cultures where spontaneous network activity was either enhanced by treatment with bicuculline or suppressed by treatment with tetrodotoxin as compared with the vehicle control. The dendritic and axonal trafficking of HPA1 were found to be regulated by neuronal activity. Moreover, network enhancement or suppression for 8 hours or 48 hours led to distinct alterations of HPA1 trafficking. For dendritic trafficking, 8-hour treatments brought about similar extents of deceleration of mCherry-marked dots in the tug-of-war mode of movement irrespective of the drug applied. The reductions in velocity were similar in extent for neurons subjected to either activity manipulation. Extending activity suppression to 48 hours resulted in further deceleration of HPA1 trafficking whereas after activity enhancement for 48 hours, HPA1 trafficking resumed speed, more so in the anterograde than the retrograde direction. For axonal trafficking, 8-hour treatment with both drugs brought about anterograde acceleration of HPA1 trafficking whereas retrograde acceleration of HPA1 trafficking was observed only in the case of network activity suppression. Prolonging the suppression of network activity to 48 hours resulted in a different scenario of axonal HPA1 trafficking, however, brought about return of both anterograde and retrograde trafficking of HPA1 to that exhibited by neurons exposed to the vehicle control. By contrast, network activity enhancement for 48 hours led to marked acceleration of HPA1 trafficking in both directions. To determine the roles of HPA1 in synaptic plasticity, shRNA-mediated knockdown of HPA1 was conducted and presynaptic calcium influx was monitored by the fluorescence emitted by the calcium sensor GCaMP6s fused to the C-terminus of the presynaptic marker, synaptophysin. Baseline presynaptic calcium influx was not affected by HPA1 knockdown. After network activity enhancement for 48 hours, HPA1 knockdown with two shRNA constructs bearing different knock-down efficiency led to opposite changes of presynaptic calcium influx. In addition, the 48-hour network activity suppression induced-homeostatic enhancement of presynaptic calcium influx was reversed by HPA1 knock-down. Our study revealed that trafficking of HPA1 along dendrites and axons are distinctly regulated by prolonged perturbations of the network activity. Moreover, anterograde and retrograde trafficking are modulated in different manners during homeostatic plasticity for both dendritic and axonal HPA1. Our study also highlighted the essential roles of HPA1 in homeostatic upregulation of presynaptic calcium influx after ongoing suppression of the neuronal activity.