Critical role of extracellularly secreted neuronal pentraxin 1 in ischemic neuronal death

Abstract

Background: Developing brain is highly susceptible to hypoxic-ischemic injury leading to severe neurological disabilities in surviving infants and children. Previously we reported induction of neuronal pentraxin 1 (NP1) in hypoxic-ischemic injury in neonatal brain and NP1 co-localization with the excitatory AMPA receptors GluR1 at the synaptic sites. However, how NP1 contributes to hypoxic-ischemic neuronal injury is not completely understood.

Results: Here we report that extracellular secretion of NP1 is required for ischemic neuronal death. Primary cortical neurons at days in vitro (DIV) 12 were subjected to oxygen glucose deprivation (OGD), an in vitro model of ischemic stroke, for different time periods (2-8 h). Oxygen glucose deprivation showed characteristic morphological changes of dying cells, OGD time-dependent induction of NP1 (2-4-fold) and increased neuronal death. In contrast, the NP1-KO cortical neurons were healthy and showed no sign of dying cells under similar conditions. NP1gene silencing by NP1-specific small interfering RNA (NP1-siRNA) protected cortical neurons from OGD-induced death. Conditioned media (CM) collected from OGD exposed WT cortical cultures caused neurotoxicity when added to a subset of DIV 12 normoxia control WT cortical cultures. In contrast, CM from OGD-exposed NP1-KO cultures did not induce cell toxicity in control WT cultures, suggesting a role for extracellular NP1 in neuronal death. However, NP1-KO neurons, which showed normal neuronal morphology and protection against OGD, sustained enhanced death following incubation with CM from WT OGD-exposed cultures. Western blot analysis of OGD exposed WT CM showed temporal increase of NP1 protein levels in the CM. Most strikingly, in contrast to NP1-KO CM, incubation of normal cortical cultures with CM from OGD exposed NP2-KO cultures showed neurotoxicity similar to that observed with CM from OGD exposed WT neuronal cultures. Western immunoblotting further confirmed the increased presence of NP1 protein in OGD-exposed NP2-KO CM. Live immunofluorescence analysis show intense cell surface clustering of NP1 with AMPA GluR1 receptors.

Conclusions: Collectively, our results demonstrate that extracellular release of NP1 promote hypoxic-ischemic neuronal death possibly via surface clustering with GluR1 at synaptic sites and that NP1, not its family member NP2, is involved in the neuronal death mechanisms.

Figure 1

The OGD time-dependent progression of cytotoxicity in primary cortical neurons. Primary cortical cultures at DIV 12 were submitted to OGD conditions for indicated time periods as described in the Methods. Light microscopic images show morphological evidence of degenerated neurons and disintegration of processes (yellow arrow) of WT cortical neurons (upper panels) and healthy cell bodies with intact processes of NP1-KO neurons (lower panels) observed following OGD exposure as compared to respective normoxia control neurons. Scale bar, 100 μm.

Figure 2

The NP1 induction is associated with OGD-induced cortical neuronal death. A) Total cellular RNA was extracted and NP1 mRNA expression levels were analyzed by RT-qPCR. Data show relative quantification of Nptx1 expression at different time periods of OGD exposure. Fold induction is the ratio of NP1 to internal control HPRT, which remained stable throughout the OGD period (mean ± SEM, n = 6; *p < 0.05, **p < 0.01). B) Total cellular protein was analyzed by SDS-PAGE and immunoblotted for NP1 protein using NP1-specific antibody that detected NP1-immunoreactive single band of molecular mass 47 kDa. The β-actin serves as loading control. Quantitative densitometry values normalized to β-actin (NP1/β-actin ratio, n = 6) are also shown. Representative bands are shown. C & D) OGD exposure of WT cortical neurons caused cell death, while NP1-KO neurons were protected against OGD. Quantification of cell death as indicated by LDH release showed OGD time dependent increase of LDH release at 2, 4, 6, and 8 h of OGD exposures of WT cortical neurons. LDH release remained at the control level or non-significant increase in OGD-exposed NP1-KO neurons. Data are expressed as % LDH release normalized to normoxia control (mean ± SEM, n = 8; *p < 0.05, ***p < 0.001). We found ~ 50% cytotoxicity occurred at 6 h of OGD.

Figure 3

Knockdown of NP1 by siRNA targeted against NP1 mRNA significantly protected against OGD-induced neuronal death. We have achieved >90% reduction in NP1 protein compared to scramble siRNA (shown in inset). LDH cytotoxicity and MTT cell viability (A-B) showed significantly less cell death in NP1-siRNA transfected WT cells vs. non-transfected cells following exposure to OGD (6 h). Data are expressed as percent LDH release normalized to normoxia control (mean ± SEM, n = 8; **p < 0.01 compared to control normoxia; +p < 0.01compared to OGD scramble siRNA group) C) TUNEL staining of cortical neurons transfected with either control scramble or NP1-siRNA revealed significant increase in the number of TUNEL (+ve) cells in scramble SsiRNA transfected cultures, which was significantly reduced in NP1-siRNA transfected neurons. Data are expressed as percent TUNEL/DAPI cells (mean ± SEM, n = 8; **p < 0.01 compared to control normoxia; +p < 0.01compared to OGD scramble siRNA group).

Figure 4

OGD-conditioned culture media caused increased neuronal death in cortical neuronal cultures. Primary cortical cultures at DIV 12 were exposed to OGD and conditioned media (CM) were collected, centrifuged and concentrated as described in the methods. A) Western blot analysis of CM from OGD-exposed WT primary cortical cultures showed OGD time-dependent increase of NP1 protein levels. B) This CM of OGD exposed WT cultures (WT-OGD CM) was added to a subset of control cortical cultures for 24 h (Control + WT-OGD CM). In addition, another subgroup of cells was also exposed to 6 h of OGD (WT-OGD CM + OGD). Data shown are mean ± SEM (n = 8 in each group) and repeated two times, **p < 0.01 vs. normoxia controls; +p < 0.01 vs. OGD only and Control + WT-OGD CM group.

Figure 5

Absence of NP1 protein in NP1-KO OGD conditioned medium does not cause neuronal death in WT normoxia cortical cultures. Morphological evidence of degenerated cell bodies and processes reveal that WT OGD CM induced neuronal degeneration in control WT cultures (A- A1) whereas, OGD conditioned NP1-KO CM did not cause cell death in WT cultures (B- B1). In contrast, NP1-KO cortical cultures, which are protected against OGD, showed neuronal death when treated with WT-OGD CM (C- C1). Representative light microscopic images are shown, n = 6, Scale, 100 μm.

Figure 6

NP2-KO OGD conditioned medium caused neurotoxicity and death in WT normoxia cortical cultures. Quantification of cell death by LDH release assay revealed significantly higher percentage of cell death when WT-OGD CM was added to WT control cultures (A). B) The NP2-KO OGD CM caused similar extent of cytotoxicity when added to the WT control cultures as compared to control CM (Con-CM). Similarly, the NP2-KO OGD CM in combination with OGD further enhanced cell death. Data shown are mean ± SEM (n = 8 in each group) and repeated two times, **p < 0.01 vs. normoxia controls. C) Western blot analysis of OGD exposed NP2-KO CM showed OGD time-dependent increased levels of NP1 protein present in the NP2-KO OGD CM, suggesting NP1, but not the NP1, contributes to neuronal death. Representative bands are shown.

Figure 7

Extracellular release of NP1 protein following OGD enhances interaction between NP1 and GluR1 at the synaptic sites. Live immunostaining of DIV 12 primary cortical neurons with NP1 and GluR1 antibodies following OGD (4 h). Immunofluorescence microscopy and merged images show increased number of NP1 (red) and GluR1 (green) co-localized clusters (yellow) at cortical dendrites and axons. Representative images are shown, n = 6, Scale, 100 μm.