NELF is a nuclear protein involved in hypothalamic GnRH neuronal migration

Abstract

Nasal embryonic LHRH factor (NELF) has been hypothesized to participate in the migration of GnRH and olfactory neurons into the forebrain, a prerequisite for normal hypothalamic-pituitary-gonadal function in puberty and reproduction. However, the biological functions of NELF, which has no homology to any human protein, remain largely elusive. Although mRNA expression did not differ, NELF protein expression was greater in migratory than postmigratory GnRH neurons. Pituitary Nelf mRNA expression was also observed and increased 3-fold after exogenous GnRH administration. Contrary to a previous report, NELF displayed predominant nuclear localization in GnRH neurons, confirmed by mutagenesis of a putative nuclear localization signal resulting in impaired nuclear expression. NELF knockdown impaired GnRH neuronal migration of NLT cells in vitro. These findings and the identification of two putative zinc fingers suggest that NELF could be a transcription factor. Collectively, our findings implicate NELF as a nuclear protein involved in the developmental function of the reproductive axis.

Figure 1

A) Northern blot analysis with a ³²P labeled human NELF cDNA probe. Although not well see in the figure, low expression levels were observed in the small intestine, skeletal muscle, and peripheral WBCs. B) QRT-PCR of mouse Nelf mRNA expression in each of the three neuronal GnRH cell lines. Triplicate experiments (n=9 for each cell line) showed that all three GnRH cell lines express Nelf, but levels were not significantly different among cell lines. C) QRT-PCR of mouse Nelf mRNA expression following exogenous GnRH agonist treatment in LβT2 pituitary cells (n=3/group). Nelf expression increased three-fold at three hours, while Chd7 expression did not change after GnRH agonist treatment. Nelf expression was greater at both 3 and 12 hr vs. control (P<0.001). Although there was a statistically significant rise in Chd7 expression at 3 hr, this effect is only ~ 40% increase over baseline, questioning any biological significance.

Figure 2

A) Protein expression by western blot analysis with our new anti-NELF antibody is greater in migratory NLT and GN11 cells than postmigratory GT1-7 cells. B) Following cellular fractionation, NELF expression is greater in the nucleus than the cytoplasm in all three GnRH cell lines as represented by a ~63 kDa band. A smaller ~57kDa nonspecific band was seen in the nuclear fraction, which could represent a NELF splice variant. Note absence of the 63 kDa band in COS-7 cells. The lower figure clearly shows greater nuclear expression in all three GnRH cell lines. Beta-actin was used as a loading control. C) Following sucrose density gradient separation, NELF expression was seen in the nucleus. The most intense expression of NELF was seen in the final nuclear pellet P2. Abbreviations are as follows: H = histone; S1 = supernatant 1; P = pellet 1; 1.6 M and 1.9 M sucrose layers; P2 = pellet 2, which comprises the nuclear pellet. A 100 kDa nonspecific band is seen in some fractions.

Figure 3

A) Immunofluorescence with confocal microscopy (63X) demonstrates predominant NELF expression in the nucleus in all three cell lines, as shown by DAPI and anti-NELF antibody staining. As a negative control, peptide used to generate the antibody was preincubated with the antibody, and no NELF staining was seen (not shown). B) Western blot analysis of the exogenous NELF/GFP fusion product in NLT cells is shown. The expected 93kDa band is shown using both anti-NELF and anti-GFP antibodies. Beta-actin is the loading control. C) Nuclear/cytoplasmic (N/C) ratios for wild type and 3A NELF mutant in NLT cells are shown. In the inset, immunofluorescence demonstrates nuclear GFP staining in the wild type, but diffuse staining in the 3A mutant, consistent with its reduced nuclear/cytoplasmic expression.

Figure 4

Boyden chamber NLT cell migration results are shown. On the left, NELF knockdown (KD) results in a three-fold reduction in migration compared with the control. On the right, when fluorescence-activated cell sorting (FACS) was done to enrich for transfected cells, GnRH neuron migration was reduced more than 12-fold.

Figure 5

Amino acid alignment of human NELF (NP_056352) and rat Jacob (AJ293697) using CLUSTAL 2.0.11 is shown. The locations of the myristoylation site (brackets above amino acids 1-7) and bipartite NLS of NELF (bold line above amino acids 243-260 of NELF) corresponding to the identical sequence in Jacob protein shaded in aqua). Included in the NLS is the sequence RRKR (mutated to AAKA—the 3A mutant) and adjacent to NLS is the RK sequence (mutated to AA—the 2A mutant), both shaded in magenta. In addition, the peptide sequence used to generate the anti-NELF antibody by Kramer et al [Kramer and Wray 2000] is shown (underline beginning at NELF amino acid 360) and the C-terminal location for our new anti-NELF antibody (boxed sequence at the very C-terminus), which was used for all figures in the current study. Shown, shaded in yellow, are potential sites of phosphorylation of the Jacob protein [Dieterich et al. 2008].

Figure 6

Schematic presentation of putative zinc finger domains of NELF are shown. Two zinc finger domains with HH-CC and CH-HC consensus are shown consecutively. These zinc finger domains were represented in the Pfam program of PF08080 (http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?PF08080) and PF09337 (http://www.sanger.ac.uk/cgi-bin/Pfam/getacc?PF09337). Red colored circles represent cysteine or histidine residues binding directly with zinc ions. The alignment of NELF proteins containing two putative zinc finger domains from five different species of human, mouse, rat, chicken and zebrafish. The fully conserved putative zinc finger domains of HH-CC and CH-HC are boxed in red.