| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
FJ
EXPRESS SUMMARY ARTICLE The Full-length version of this article is also available, published online January 27, 2006 as doi:10.1096/fj.05-4338fje. |
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Neuroscience, Centre Médical Universitaire, Geneva, Switzerland
1Correspondence: Department of Neuroscience CMU, 1 rue Michel Servet, 1211 Genève 4, Switzerland. E-mail: alex.baertschi{at}medecine.unige.ch
SPECIFIC AIMS
Neuropeptides and hormones play vital roles in pain, sleep, reproduction, growth, and metabolism. While the secretory pathway has been extensively investigated, potential intracellular targets of secretory proteins in endocrine cells are little known. The purposes of this study were to 1) Identify intracellular sites (in cardiocytes, HL-1, and hypothalamic cells) expressing preproneuropeptide-Y (preproNPY), preprogrowth hormone (preproGH) and preproatrial natriuretic factor (preproANP) constructs; 2) determine whether the cell environment or mutations causing secretion defects could modulate intracellular targeting, even when signal peptide coding sequences were intact; and 3) estimate the probability of intracellular targeting for other prohormones.
PRINCIPAL FINDINGS
1. Intracellular expression of prohormone-derived peptides: identification of N-terminal truncated peptides in endocrine cells
Within atrial myocytes, a plasmid (N111-E) encoding human wild-type preproNPY[1-38]-EGFP was expressed either in secretory vesicles, or in a complex meshwork, or in both (Fig. 1
A, B; online Fig. 1; films 1A-z,1B-z). A mitochondrial marker specifically increased the fluorescence of the complex meshwork, thus identifying it as mitochondria (films 1C1, 1C2). Constructs without the EGFP tag (N111, N011, N111-CPON) were also expressed in mitochondria (Fig. 1C, D
). Mitochondrial expression of N111 was detected by immunofluorescence with midportion but not N-terminal sensitive anti-NPY antiserum. As a control, mutations of the first initiation site (construct N011-E) resulted in a shorter fusion protein that was no longer recognized by an N-terminal sensitive anti-NPY antibody (online Fig. 2, center). N011-E was expressed in mitochondria of hypothalamic slices and atrial myocytes, while mutation of the second initiation site (construct N101-E) resulted in vesicular expression (online Fig. 2; films 2A-t, 2C-t, 2D-t, 2E-t, 2F-z). Similarly, expression of preproANP[1-127]-EGFP and preproGH[1-22]-EGFP resulted in expression of green fluorescence in nucleus and cytoplasm in addition to trans-GOLGI network (TGN) and secretory vesicles, and yielded truncated proteins in HL-1 extracts (not shown).
|
2. Subcellular expression of secretory peptides: modulation by culture media and mutations of N-terminal proANP
Omission of zinc in culture media reduced expression of proNPY[1–38]-EGFP derived peptide in mitochondria (Fig. 2
A), and enhanced expression of proGH[1–22]-EGFP derived peptide in cytoplasm (Fig. 2B
). While vesicle and TGN fluorescence were linearly related, there was a high mitochondrial expression of NPY[1-38]-EGFP and high cytoplasmic expression of GH[1–22]-EGFP-derived peptides when TGN fluorescence was reduced (Fig. 2C, D
). Deletion of the C-terminal of proANP in the preproANP[1–127]-EGFP construct did not change the high expression of green fluorescence in vesicles and TGN (online Figs. 4, 5A; film 4B-t). Deletion of the 132 nucleotides coding for the N-terminal strongly reduced the expression of green fluorescence in vesicles and TGN (online Fig. 5B). Retaining only 36 nucleotides coding for preproANP[1–11], and adding nucleotides coding for flag-EGFP eliminated all vesicular expression, and caused high expression in cytoplasm when TGN fluorescence was reduced (online Fig. 5C). Without exception, the complex subcellular expression profiles were explained, using P-SORT software, by a 3' shift in translation initiation, resulting in new targeting sequences directing the N-terminal truncated prohormones to mitochondria, nucleus, and cytoplasm.
|
3. Predictions of intracellular targeting of other prohormones and proneuropeptides
A total of 18 prohormone and proneuropeptide sequences (Fig. 3
A) were analyzed by P-SORT. For translation starting at the first AUG (Fig. 3B
), the highest probability was found for the secretory pathway, with lower probabilities for cytoplasm, nucleus, and mitochondria. For translation starting at the second AUG (Fig. 3A, C
), the highest probabilities were found for cytoplasm, nucleus, and mitochondria, with lower probability for the secretory pathway (online Fig. 7). For only 22% of all sequences, the second AUG site initiated translation of a signal sequence directing the protein into the secretory pathway.
|
CONCLUSIONS AND SIGNIFICANCE
Prohormones normally are exported via secretory vesicles. In atrial myocytes and hypothalamic cells, N-terminal truncated prohormones were also expressed in cytoplasm, nucleus, and mitochondria. This conclusion is based on 8 constructs encoding prohormone-EGFP fusion proteins; 6 of these were expressed not only in TGN and secretory vesicles, but also in cytoplasm, nucleus, and mitochondria. Removal of the EGFP tag in 3 additional constructs did not prevent targeting of N-terminal truncated NPY to mitochondria (Fig. 1C, D
). Intracellular destinations of prohormones were not due to overexpression, since overexpression in TGN resulted in low or no expression in cytoplasm or mitochondria (Fig. 2
; online Fig. 5); nor were they due to cell damage, since they were observed in healthy looking contracting myocytes. Evidence for N-terminal truncation stems from online Fig. 2 (center) showing a lower molecular weight band that was no longer recognized by the N-terminal sensitive anti-NPY antibody, and by immunofluorescence studies in online Fig. 1E, F showing staining of endogenous NPY in mitochondria by midportion but not N-terminal sensitive antibody.
Intracellular targeting occurred even in presence of intact signal peptide coding sequences and was modulated by extracellular zinc. P-SORT accurately predicted the intracellular expression of tested constructs when translation started at downstream initiation sites. Skipping of the first AUG in mRNA encoding human preproNPY by-passes the sequence coding for signal-peptide (MLGNKRLGLSG...), and initiates translation of MARYYSALRHY..., correctly predicted to target mitochondria. Similar reasoning applies to the preproGH[1-22] encoding mRNA, where the first AUG was present but skipped, resulting in translation of MLRAHRL..., a peptide targeted to cytoplasm as predicted. In case of proANP[1-127]-EGFP, the Kozak sequence context was weaker for the 2nd initiation site, consistent with the weak cytoplasmic expression. Of particular interest, zinc deficiency resulted in much lower expression of NPY in mitochondria and higher expression of GH-derived peptide in cytoplasm (Fig. 2)
. Zinc is known to cause aggregation of proGH. Since expression of proANP[1-127]-EGFP was enhanced in secretory vesicles and TGN, changes in expression were not due to nonspecific effects of zinc deficiency on translation, degradation, sorting or energy production. Thus cell environment could change the destination of prohormones from extracellular to intracellular signaling. Results with proANP[1–11]-flag-EGFP fusion protein (online Fig. 5) and a previous study on GH suggest that defects in the secretory pathway may shift translation to the second AUG initiation site. The exact mechanism, perhaps signaling of a transport defect to the mRNA scanning complex, remains to be established.
Approximately 80% of all prohormones are predicted to be potentially expressed in cytoplasm, nucleus, or mitochondria. Some of these are known to be naturally expressed there. This third conclusion is based on P-SORT predictions for 18 prohormone and proneuropeptide sequences (Fig. 3A
). Since predictions of P-SORT were correct for all 11 constructs that were tested, they are likely correct for non-tested sequences as well. The model of Fig. 3
could explain reports of cytoplasmic or mitochondrial expression of secretory proteins. For example, endogenous cytoplasmic expression of ANP has been reported for equine atrial myocytes, and endogenous expression of NPY in mitochondria for human umbilical vein endothelial cells (HUVEC) (online Fig. 1E, F).
The role of intracellular hormones remains to be clearly established. Peptides targeted to mitochondria, nucleus and cytoplasm could affect cellular metabolism, gene expression and cytoplasmic enzymes. Mitochondrial NPY depolarized mitochondria by an estimated 14 mV, suggesting an inhibitory action on cellular energy metabolism (online Fig. 6). So far, mitochondrial NPY receptors have not yet been described, but entry of positively charged NPY could by itself lead to mitochondrial depolarization. N-terminal truncated 43 kDa T3 receptors also target mitochondria and act as transcription factors for mitochondrial activation by T3. A truncated PTH related hormone is directed to the nucleus and activates the proliferation of arterial smooth muscle. Few other mammalian examples exist, and no common rule emerges so far. Yet, first described for viruses and plants, N-terminal truncation emerges as an exciting mechanism for mammalian secretory gene expression.
FOOTNOTES
To read the full text of this article, go to http://www.fasebj.org/cgi/doi/10.1096/fj.05-4338fje;
This article has been cited by other articles:
|
|
 |
|
  P. Philip-Couderc, N. I. Tavares, A. Roatti, R. Lerch, C. Montessuit, and A. J. Baertschi Forkhead Transcription Factors Coordinate Expression of Myocardial KATP Channel Subunits and Energy Metabolism Circ. Res., February 1, 2008; 102(2): e20 - e35. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
