I. NF-kB signaling

1. Identification of three novel components of the NF-kB signaling pathway (TNAP, NIBP, TALL-1)

 1) TNAP.

Using yeast two-hybrid system with NIK (NFkB-inducing kinase) as bait to screen cDNA libraries from brain, spleen, leukocyte and HEK293T cells, several known proteins such as TRAF3, PAX6, CDC23 and proteasome subunit PSMA3, and two novel proteins, TNAP and NIBP; the latter were characterized in detail and shown to interact with NIK.  TNAP (TRAFs and NIK-Associated Protein) specifically inhibited TNF-a and IL-1b-induced NF-κB activation by interacting with NIK and TRAF2/3, and suppressing NIK kinase activity. Thus, TNAP regulated both classical (IkBa phosphorylation and degradation) and non-classical (p100 processing to p52) pathways of NF-kB activation. TNAP also suppressed TNFa-induced and NIK-mediated Ser⁵³⁶ phosphorylation of p65.

        2)  NIBP.  

NIBP (NIK and IKKb Binding Protein) was mainly expressed in brain, muscle, heart and kidney, and moderately expressed in immune tissues such as spleen, thymus and peripheral blood leukocytes, where NF-kB was known to modulate immune function. NIBP physically interacted with NIK, IKKb, but not IKKa or IKKg.  NIBP over-expression potentiated TNFa and IL-1b-induced NF-kB activation through increased phosphorylation of the IKK complex and its downstream IkBa and p65 substrates. Knockdown of NIBP expression by lentiviral vector-mediated small interfering RNA reduced TNFa-induced NF-kB activation, prevented NGF-induced neuronal differentiation and decreased the expression of NFkB-dependent gene, Bcl-xL, in PC12 cells. These data demonstrated that NIBP, by interacting with NIK and IKKb, was a novel enhancer of  cytokine-induced NF-kB signaling.

3)  TALL-1.

 Using amino acid homology analysis, another novel member of TNF ligand family was identified and designated TALL-1 (for TNF- and ApoL-related Leukocyte expressed Ligand 1).  Simultaneously, TALL-1 was discovered by others and designated BAFF, BLyS, and zTNF4. TALL-1 was a potent modulator of B-cell proliferation via its receptors BCMA (B cell maturation antigen) and TACI (transmembrane activator and CAML-interactor), and was expressed by monocytes/macrophages and dendritic cells.

2.  Characterization of NFkB signaling pathways

As shown above, TNAP suppressed and NIBP enhanced cytokine-induced NF-kB activation. Our earlier studies demonstrated that receptors for TRAIL (TNF-related apoptosis-inducing ligand) induced apoptosis, and NF-kB and JNK activation through distinct signaling pathways. The apoptosis-inducing adaptors FADD (Fas-associated via Death Domain), Casper and Caspase-8 potently activated NF-kB, whereas “activated” Caspase-8 blocked NF-kB activation by inactivating NIK. These data were corroborated by several other groups.

 Neurons and their neighboring cells employ the NF-κB pathway for distinctive functions, ranging from development to neuronal plasticity and coordination of cellular responses to injury.  As part of our studies,  IκBa-dominant mutant transgenic mice were generated which stably expressed mutant IκBa under the control of an astrocyte-specific promoter, GFAP (glial fibrillary acidic Protein) or a neuron-specific promoter, synapsin.  Selective inactivation of astroglial NF-kB in transgenic mice led to marked improvement in function 8 weeks after contusive spinal cord injury.  The mice showed reduced expression of pro-inflammatory chemokines and cytokines, such as CXCL10, CCL2, and TGF-b2.  Inactivation of astroglial NF-kB in transgenic mice led to a significant deficit in learning and memory.

3.  RGS7 expression in inflammatory macrophage and microglia cells of CNS

Previous studies demonstrated that TNF-a prevented proteasome-dependent degradation of RGS7 in neurons via p38 MAP kinase-mediated mechanism.  Our studies confirmed the expression of RGS7 in spinal neurons and found that RGS7 was inducibly expressed in microglia and macrophages after spinal cord injury.

II. Secondary spinal cord injury and neural growth inhibition

Around twenty years ago, the pathophysiologic mechanism of the secondary spinal cord injury was a hot spot in the field of central nervous system injury. Using evoked potentials to evaluate spinal cord function and biomicrosphere technique to measure spinal cord blood flow, my studies demonstrated that ischemia in the white matter is closely correlated with spinal cord dysfunction after balloon compressive injury in dogs. My further works together with my colleagues corroborated the important role of excitotoxicity and lipoxidation in secondary spinal cord injury. These studies were awarded a Second Prize of Military Science and Technology Achievement in 1998.

During my PhD study, I continued the research on the excitotoxic mechanism of spinal cord injury using pharmacological animal model, focusing on the role of NMDA-Ca2+-NOS/NO pathway. NO was a "molecule of the year 1992 in science". My research demonstrated for the first time that neurotoxic dose of dynorphin (an endogenous opioid peptide) induces high expression of both neuronal and inducible nitric oxide synthases (nNOS and iNOS) in the spinal cord of rats. Selective inhibition of either nNOS or iNOS is neuroprotective while non-selective NOS inhibition aggravates dynorphin-induced spinal cord injury. NMDA receptor functional activity is significantly elevated in the ventral spinal cord of rats with dynorphin spinal neurotoxicity. In cultured spinal cord neurons, high concentration of dynorphin produces persistent calcium overload, which is antagonized by pretreatment with both NMDA receptor antagonist and kappa opioid receptor antagonist. These data were granted a Second Prize of Beijing Science and Technology Achievement in 2000.

Excitatory amino acids transporters (EAAT) are essential to prevent excitotoxicity and to terminate glutamatergic neurotransmission. During my postdoctoral training in Miami Project to Cure Paralysis, I observed, unexpectedly, that EAAT4 immunoreactivity is highly enriched in the spinal cord. Further studies demonstrated that EAAT4 is expressed in the astrocytes of spinal cord at both protein and mRNA levels. This astrocytic localization of EAAT4 may reveal some new function of EAAT4 in the spinal cord.

To develop a better understanding of the mechanisms responsible for the functions of Nogo, an important myelin-derived nerve growth inhibitor after spinal cord injury, I performed a yeast two-hybrid screen of human brain cDNA library using Nogo-66 as bait.  A novel mitochondrial protein designated NIMP (for Nogo-Interacting Mitochondrial Protein) is highly conserved and ubiquitously expressed in neurons and astrocytes. Other two-mitochondrial proteins UQCRC1/2 also interact with Nogo, indicating that Nogo may affect mitochondrial functions.

III. Receptor mechanism for S1P signaling

Sphingosine-1-phosphate (S1P) regulates diverse biological processes through five receptor types, S1P_1-5. S1P induces an initial Ca²⁺)-dependent contraction followed by a sustained Ca²⁺-independent, RhoA-mediated contraction in rabbit gastric smooth muscle cells. The cells coexpress S1P₁ and S1P₂ receptors, but the signaling pathways initiated by each receptor type and the involvement of one or both receptors in contraction are not known. Lentiviral vector-mediated siRNA silencing of S1P₁ receptors abolished S1P-stimulated activation of Ga_i3 and partially inhibited activation of Ga_i1, whereas silencing of S1P₂ receptors abolished activation of Ga_q, Ga₁₃, and Ga_i2 and partially inhibited activation of Ga_i1. Silencing of S1P₂ but not S1P₁ receptors suppressed S1P-stimulated PLC-b and Rho kinase activities, implying that both signaling pathways were mediated by S1P₂ receptors. The results obtained by receptor silencing were corroborated by receptor inactivation. The selective S1P₁ receptor agonist SEW2871 did not stimulate PLC-b or Rho kinase activity or induce initial and sustained contraction; when this agonist was used to protect S1P₁ receptors so as to enable chemical inactivation of S1P₂ receptors, S1P did not elicit contraction, confirming that initial and sustained contraction was mediated by S1P₂ receptors. Thus S1P₁ and S1P₂ receptors are coupled to distinct complements of G proteins. Only S1P₂ receptors activate PLC-b and Rho kinase and mediate initial and sustained contraction.