Faculty Advisor

Dr. David S. Adams

Faculty Advisor

Dr. Ronald D. Cheetham

Faculty Advisor

Dr. Daniel G. Gibson


Ependymin (EPN) is a neurotrophic factor (NTF) that functions in goldfish long-term memory formation and optic nerve elongation (Shashoua, 1976; Shashoua, 1977; Shashoua, 1985). Goldfish EPN, or CMX-8933 (a short goldfish EPN mimetic studied by our lab), surprisingly have several demonstrated effects on mammalian cells, including neuroregenerative effects in a rat stroke model (Shashoua et al, 2003), and the activation of therapeutic superoxide dismutase (SOD) (Parikh, 2003) and transcription factor AP-1 (Adams et al, 2003) in mouse neuroblastoma cells or rat primary cortical neuronal cultures. Among its various functions, AP-1 can function as a master switch in long-term memory consolidation (Sanyal et al, 2002), so it may be a key event in EPN's mechanism of action. AP-1 activation is also a characteristic associated with other full-sized neurotrophic factors, including nerve growth factor and brain-derived nerve growth factor. This thesis was divided into three parts. The purpose of part I was to determine whether our previously observed upregulation of SOD by CMX-8933 is dependent upon (or merely concurrent with) AP-1 activation. Four independent SOD immunoblot experiments demonstrated that pre-treatment of rat primary cortical cultures with trifluoromethyl pyrimidine carboxylate (TFPC), a specific inhibitor of AP-1, significantly (p = 0.0004) decreased cellular levels of SOD by 67% at its IC50 concentration of 1 ìM, and completely inhibited the upregulation at 10 and 100 ìM concentrations. Thus, the CMX-8933-induced upregulation of SOD appears to depend (directly or indirectly) on AP-1 activation. Part II of this thesis included the use of bioinformatics to re-verify exciting recent observations that EPN-like proteins exist in mammals, termed mammalian-ependymin-related proteins or MERPs (Apostolopoulos et al, 2001). If our analyses were convincing, human EPN mimetics would then be designed and tested for AP-1 activation. Computer alignments and hydropathy plots performed with EPN amino acid sequences deduced from gene entries in GenBank verified the existence of mammalian homologs containing highly conserved domains with fish EPN's, suggesting the possibilities of similar protein conformation and function. Two human EPN mimetics were designed, hEPN-1 (8 aa long, corresponding to the same region as CMX-8933) and hEPN-2 (14 aa long, containing CMX-8933 and 6 upstream aa). Several mimetic doses were tested on mouse Neuro-2a cultures for nuclear translocation of c-Jun and c-Fos proteins (comprising the AP-1 particle upregulated by fish CMX-8933). Seven independent c-Jun immunoblot experiments, and five c-Fos experiments, demonstrated a strong (as high as 25-fold) dose-dependent increase in the nuclear titers of the AP-1 proteins. Both peptides had statistically equivalent effects. Thus, human EPN appears to exist, and two mimetics derived from its sequence appear to be biologically active against mouse neuroblastoma cells. Since hEPN-1 and -2 have only a few residues in common with CMX-8933, we hypothesize that the mimetic shape rather than sequence may be important for biological activity. In part III of this thesis, the biological effects of hEPN-1 and hEPN-2 on mouse Neuro-2a cells were studied further using RT-PCR to analyze potential increases in specific mRNAs. mRNAs related to growth, energy production, and protein translation were tested since previous data in our lab (Kaska, 2003) indicated mRNAs for translational elongation factor-2 (EF-2), and ribosomal proteins L19 and S12 were upregulated in rat primary cortical cultures by fish mimetic CMX-8933 (Kaska, 2003). Treatment of Neuro-2a cells with 1.0 ìg/ml hEPN-1 (the highest dose tested for the AP-1 translocation experiments) for 24 hrs appeared to increase (N = 1) mRNAs for ATP Synthase-C, ribosomal protein L19, and translational EF-2, relative to the levels of housekeeper polyubiquitin. Thus hEPN-1 may be involved in processes related to growth. Altogether, the data from this thesis extends our knowledge of fish EPN mimetic CMX-8933 (showing that its induction of SOD requires AP-1), demonstrates that human EPN may exist (bioinformatics), shows that two human EPN mimetics are biologically active (induce AP-1 translocation), and shows that one mimetic hEPN-1 may activate several mRNAs related to growth in mouse Neuro-2a cells.


Worcester Polytechnic Institute

Degree Name



Biology & Biotechnology

Project Type


Date Accepted





AP-1, Ependymin, Neurotrophic factor, Neurotropin, Neurotrophic functions, Transcription factors, Ependymin