PDFlib PLOP: PDF Linearization, Optimization, Protection Page inserted by evaluation version www.pdflib.com – [email protected]
Posttranscriptional Regulation of Myelin Protein Gene Expression" ANTHONY T. CAMPAGNONI: JOSEPH M. VERDI, A. NEIL VERITY, SHASHI Ah4UR-UMARJEE, AND SUJATHA BYRAVAN Mental Retardation Research Center U.C.L.A.Centerfor the Health Sciences Los Angeles, California 90024 Regulation of gene expression may occur at a number of points between the transcription of a gene and the appearance of a functional protein in its appropriate location within the cell. In recent years we have begun to learn a great deal about the factors that regulate the activity and transcription of genes and the intracellular targeting of proteins after synthesis. However, less is known about the factors that affect the stability and translation of mRNAs, the role that specific mRNAs may play in the intracellular translocation of macromolecules, and the importance of mRNA structure in influencing these processes. Certainly, the importance of the structure of the 5'-untranslated region in the general regulation of translation' as well as cis-acting elements within this region that regulate the translation of specific mRNAs, such as that encoding ferritin,' is being recognized. Evidence also indicates that the 3'-untranslated region of mRNAs may play a role in regulating the translation and stability of mRNAs and, possibly, their translocation within the ce1l.j The expression of the genes encoding the major myelin proteins - myelin basic proteins (MBP), proteolipid proteins (PLPs), myelin-associated glycoprotein (MAG), and 2',3'-cyclic nucleotide-3'-phosphodiesterase(CNP) - have been the subject of intense investigation over the last several years. (See ref. 4 for a review.) Expression of all of these genes is developmentally regulated and probably intimately associated with the differentiation of the oligodendrocyte. All of these genes undergo alternative splicing, producing at least two mRNAs, and the proportions of the alternatively spliced mRNAs and/or the protein isoforms produced change significantly with development, in most cases. Recent investigations suggest that the expression of one of these genes, the MBP gene, may be regulated at a number of different levels including (1) promoter utilization, (2) splicing, (3) mRNA translocation, (4)translation, and (5) posttranslational events. This paper presents further evidence for the posttranscriptional regulation of MBP gene expression, with most examples coming from work done on the expression of the MBP gene.
'This work was supported by NIH grants NS 23022 and NS 23322 and a generous donation from the Ciba-Geigy Corporation, Summit, New Jersey. bAddress for correspondence: Dr. A. T. Campagnoni, Mental Retardation Research Center, U.C.L.A. Center for the Health Sciences, 760 Westwood Plaza, Los Angeles, CA 90024. 178
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
179
MBP mRNA TRANSLOCATION It has been known for about 10 years that the MBPs were synthesized on "free" polyribosome~~~~ and the myelin PLP was synthesized on ribosomes bound to the endoplasmic reticulum.6 Other experiments, both in vivo and in tissue slices, indicated that although MBPs were incorporated very rapidly (within minutes) into myelin, newly synthesized PLP was incorporated into myelin much more slowly ( 30 minute^).^ Furthermore, the incorporation of PLP, but not the MBPs, into the membrane could be interrupted by treatment with monensin, suggesting that assembly of PLP into myelin involved transport through the Golgi apparatusP The available biochemical evidence supported the notion that the two classes of proteins were assembled into myelin by separate routes. Subcellular fractionation experiments demonstrated a 20-fold enrichment of the MBP mRNAs in myelin fractions compared to PLP mRNAs6 This observation suggested that ribosomes containing MBP mRNAs might be located within the cytoplasmic processes connecting the oligodendrocyte somas with the myelin sheaths and, perhaps, the cytoplasmic channels infiltrating the sheaths. If MBP synthesis occurred in these processes and/or channels, then it could explain why incorporation of these proteins into the myelin membrane occurred faster than did incorporation of proteins synthesized in the cell bodies which would require posttranslational transport out to the sites of myelin assembly. Proof of such a model could be obtained by in situ hybridization histochemistry (ISH) through the localization of MBP and PLP mRNAs, but it had to await the isolation of appropriate cDNA probes. Recently, several groups have used ISH to localize myelin-specific mRNAs in vivo9-" and in vitro.'2-'6In vivo, the pattern of distribution of the MBP mRNAs is quite different from that of the PLP mRNAs. With radioactive probes, the distribution of silver grains generated with the PLP probes tended to be clustered over cell bodies, whereas those generated with the MBP probes were diffusely distributed over myelinated regions in the brain.'.'' This difference in gross distribution of these two mRNAs was consistent with the MBP mRNA translocation hypothesis. Further evidence was provided by developmental studies. Verity and Campagnoni" examined the distribution of MBP and PLP mRNAs in the medulla oblongata of the mouse brain at daily intervals beginning at birth. They observed that before the appearance of histologically stained myelin in this region (0-2 days), silver grains representing MBP mRNAs and PLP mRNAs were clustered over cell bodies. Between 2 and 3 days postpartum the distribution of MBP mRNAs changed dramatically, from a clustered to a diffuse distribution, and this change in distribution coincided with the appearance of histologically recognizable myelin. Additional proof that MBP mRNAs are translocated from the cell bodies of oligodendrocytes to their processes has come from studies performed in primary brain cell culture^.'^^'^ Although localization of silver grains to oligodendrocyte processes has been a bit more apparent in vitro than in vivo, the most convincing localizations have been obtained in vitro with nonradioactive probes labeled with dig~xigenin'~ or biotin,16 which are detected immunochemically. In these studies, MBP mRNAs could be clearly localized within oligodendrocyte processes and the translocation of these mRNAs visualized with time in c u l t ~ r e . 'In ~ contrast, PLP mRNAs, MAG mRNAs, and greater than 99% of CNP mRNAs remained associated with the oligodendrocyte cell bodies." Thus, substantial biochemical and cell biologic evidence has now accumulated indicating that MBP mRNAs are translocated within
-
180
ANNALS NEW YORK ACADEMY OF SCIENCES
oligodendrocytes from the cell somas to their processes and, presumably, the cytoplasmic channels infiltrating the myelin sheaths. The protein products of the other major myelin protein genes appear to be targeted to myelin by other mechanisms subsequent to translation. The intracellular compartmentalization of mRNAs has been described for a number of mRNAs including those that encode mit~chondrial,’~ dendritic,18 and cytoskeletal proteins.’’ The mechanisms by which this intracellular translocation of mRNAs occurs is not yet known, but they could involve the participation of specific carrier proteins binding to cytoskeletal elements in a fashion similar to the transport of messenger ribonucleoproteins from the nucleus to the cytoplasm.” It is also unclear whether the MBP mRNAs or the nascent MBP polypeptide chains play a role in the translocation process. In this regard, MBP has been shown to interact with cytoskeletal proteins:’ and there is evidence that “free” polyribosomes can associate with cytoskeletal elements through their mRNAs.” MBP mRNA AND PLP mRNA TRANSLOCATION IN JZMPY AND QUAKING MICE
Quaking is an autosomal, recessive mutation that maps to mouse chromosome 17 and does not appear to be a mutation in one of the major myelin protein genes. The mutation is characterized by pronounced hypomyelination of the central nervous ~ystem.2~ Work from this laboratory has shown that expression of the MBP gene is altered in the quaking brain up to about 18-21 days, after which the levels of MBP mRNA and the synthesis of MBP appear to be close to n0rma1.2~In marked contrast, the steady-state levels of MBP, as measured by radioimmunologic techniques, are only 5 2 5 % of normal in the central nervous system. Inasmuch as unbound MBP is readily susceptible to protease attack, these results are consistent with an earlier hypothesis that quaking mice are defective in their ability to incorporate newly synthesized MBPs into m y e h a In view of evidence indicating that MBP mRNAs are translocated within the oligodendrocyte, it was of interest to determine if this inability of the MBPs to be incorporated into quaking myelin was due to a defect in this translocation phenomenon or to some other posttranslational event. ISH examination of the quaking brain indicated that MBP mRNAs could be detected at approximately normal levels in the 20-day-oldquaking brain (FIG.1)and that the distribution of mRNAwas as diffuse as that in the controls. These results suggest that MBP mRNA translocation is unimpaired in the quaking brain. A corollary to these results is that yet another regulatory point must exist in the assembly of the MBPs into the myelin sheath. It has been suggested that as the MBPs are located on the cytoplasmic face of the unit bilayer comprising the myelin sheath, the assembly of MBPs into the membrane occurs by diffusion shortly after their synthesis in the oligodendrocyte processes.’ Our data suggest that synthesis of MBPs close to the growing myelin sheath is not sufficient to assure proper assembly of the proteins into the membrane. Some additional event, such as a posttranslational modification of the protein, must occur for proper assembly of the MBPs into the membrane. Alternatively, the presence of a receptor for MBP in the membrane might be required before incorporation of the protein.26 The jimpy mutation appears to be a point mutation in the splice acceptor site of phenotype is somewhat complex, affecting the the fifth exon of the PLP gene.27s28The oligodendrocyte and the expression of several of the myelin protein genes, probably through an alteration in the differentiation/maturation/death of the oligodendro-
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
181
cyte.2%32 The levels of MBP mRNA, even though severely reduced in this mutant, have generally been higher than the steady-state levels of polypeptide in jimpy brain^.",'^ This finding suggests that the small amount of protein made in the mutant brain is not incorporated effectively into the membrane. ISH have confirmed other biochemical data34that MBP mRNA levels are reduced in jimpy brains. However, in the mutant MBP mRNA-specific silver grains appear more clustered on oligodendrocyte cell bodies and less diffuse than in either normal or quaking brains (FIG.1). This pattern of labeling was most prominent in the caudate putamen and corpus callosum, but the pattern was similar in all regions of the brain and persisted until the animals died at 3-4 weeks. These data suggest impairment in the translocation of MBP mRNAs in jimpy brain. It is unlikely that impairment of MBP mRNA translocation is due to a reduction or truncation of oligodendrocyte cellular processes because two groups have independently reported that the numbers and general appearance of processes with antigalactocerebroside staining appear normal in the m~tant.~’.~’ It is noteworthy that jirnpy oligodendrocytes contain numerous lipid inclusions and have a substantial reduction in m i c r o t ~ b u l e s .If~ ~MBP mRNA translocation involves the cytoskeletal network, both of these factors might contribute to impaired translocation of MBP mRNA in the mutant.
REGULATION OF MYELIN PROTEIN mRNA TRANSLATION BY STEROIDS
Steroids are well known to control gene expression at the transcriptional level; however, a number of studies indicate that steroids also may regulate gene expression posttran~criptionally.~’-~~ It has been suggested that steroids can alter the stability of the mRNAs4” and/or influence the rates of translation of responsive ~RNAS.~~,~~ None of the myelin protein genes appears to be transcriptionally regulated by steroids, but two recent studies suggest that steroids might act at the posttranscriptional level. Kumar et aL4*observed that hydrocortisone can stimulate the expression of MBP and PLP in primary oligodendrocyte cultures, without any increase in the transcription rates of these genes. We have found that the translation of MBP and PLP mRNAs is stimulated by hydrocortisone and other steroids in cell-free sy~tems.4~ Our most recent efforts have been aimed toward determining (1) if steroids could exert a direct effect on the translation of mRNAs, (2) if sequences within the MBP mRNA were involved in this phenomenon, and (3) which protein synthetic step is modulated by the steroid. Toward this end we developed a cell-free translation system that could be programmed with synthetic mRNAs to determine the effects of steroids on the translation of specific synthetic mRNAs. The use of this system permitted us (1) to assess the effect of steroids on mRNA translation independent of effects on transcription, (2) to assess directly the effect of steroids on the rates of translation of specific mRNAs, and (3) to determine possible regulatory sequences in the 5’untranslated region by programming the system with structurally modified mRNAs. Using this system we found that several classes of steroids could stimulate the translation of MBP mRNAs in this system, with glucocorticoids being more effective than sex steroids. These same steroids served to inhibit the translation of the CNP mRNA in an equal, but opposite direction to the effects observed on the MBP mRNA. No effect was observed on the translation of either message with the mineralocorticoid, aldosterone, or the sterols cholesterol and desmosterol. Transla-
182
ANNALS NEW YORK ACADEMY OF SCIENCES
FLGURE 1. In situ hybridization histochemistry of MBP mRNAs within the caudate putamen of 15-day-old normal (+/+) mice and the dysmyelinating mutants quaking (qk),jimpy up), and shiverer (sh). In normal and quaking mice, silver grains representing MBP mRNAs are diffusely distributed over myelinated tracts indicative of normal translocation of MBP mRNAs. In contrast, in the jimpy mutant, MBP mRNA translocation is impaired as seen by increased clustering of silver grains over oligodendrocyte cell bodies (arrows). As expected, no detectable MBP transcripts are observed in shiverer mice.
m w
c
184
ANNALS NEW YORK ACADEMY OF SCIENCES
tion of the PLP mRNA was also stimulated by hydrocortisone, but the translation of MAG mRNA was not affected. Kinetic analysis of the data indicated that in this system the steroids were affecting the rates of translation of the responsive mRNAs. Furthermore, experiments suggested that steroids were acting on translation through a novel mechanism that did not involve the “classic” glucocorticoid receptor: (1) RU 38486, an antagonist of the “classic” glucocorticoid receptor, did not inhibit the stimulation of MBP mRNA translation by glucocorticoids. (2) The order of effectiveness among the glucocorticoids was different from that of the glucocorticoid receptor-mediated effect on transcription. (3) The dose response curves were more complex than observed with the classic glucocorticoid receptor, reaching maxima in the physiologically relevant range of 10-R-10-9M steroid. (4) The nucleotide sequence of the element in the MBP mRNA that is responsive to hydrocortisone is unlike that of the glucocorticoid-responsive element in genes that are transcriptionally regulated by this steroid. We have found that the translation of all alternatively spliced forms of the MBP mRNAs transcribed from the major promoter of the gene are stimulated by hydroc~rtisone.~~ Transcription from the major promotor of the MBP gene begins 48 nt upstream of the initiator codon. However, the MBP gene also possesses a second, minor upstream transcription start site. Transcription from this secondary promoter produces MBP mRNAs with a longer 5’-untranslated region that consists of the 48 nt leader plus an additional upstream sequence of at least 350 nt.44The translation of these minor MBP mRNAs with a longer 5‘-untranslated region is not stimulated by hydrocorti~one.~~ These results suggested that the 5‘-untranslated region might be important in the steroid modulation of MBP translation. This notion was supported by studies in which the translation of an MBP mRNA, which contained little of the 3’-untranslated region but a normal 5‘-untranslated region, was stimulated by hydrocortisone. This finding prompted us to investigate further the role of the 5‘-untranslated region in this p h e n ~ m e n o nA . ~series ~ of cDNA clones encoding the 14 kD and 18.5 kD MBP isoforms, but differing in the lengths of their 5’-untranslated regions were generated using the polymerase chain reaction and subcloned into a transcription vector. Site-directed changes were also introduced into selected nucleotides within the 5‘-untranslated region of the MBP mRNAs using this technique. The transcribed RNAs were examined in the cell-free translation system in the presence and absence of hydrocortisone, permitting us to localize a nine-base region within the 5’untranslated region of MBP mRNA between -29 and -37 that appears necessary for the stimulation of translation by hydrocortisone. Several lines of evidence indicate the importance of this region: (1) Removal of this region of the message eliminated the translational response to steroid. (2) Site-directed mutagenesis of several bases within this region reduced or eliminated the response. (3) A synthetic oligonucleotide of the same sequence could entirely compete out the response at a 10:1 molar ratio of oligonucleotide:message, but no such competition was observed with an oligonucleotide containing only two base changes. (4) Transferral of the sequence to a nonresponsive message rendered the translation of that mRNA responsive to hydrocortisone. The steroid modulatory element in the 5’-untranslated region of the MBP mRNA is included within the sequence AGAAGACCC. However, this sequence alone probably does not account for the entire translational response to steroid because translation of the minor MBP mRNA with a longer 5‘-untranslated region was not influenced by steroids, even though it contains the nine nucleotide element. Examination of the most stable secondary structure of the 5’-untranslated region of
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
185
this unresponsive message indicates that the element is involved in a base paired stem, whereas in the “normal” MBP message six of the nine nucleotides (AGAAGA) are found in an accessible loop structure. This group of six nucleotides, but not the entire nine nucleotide region, is found twice in the 5’-untranslated region of the PLP mRNA, the translation of which is also stimulated by hydrocortisone. The position of the glucocorticoid modulatory element in the 5‘-untranslated region of the MBP mRNA, the many reports of alterations in the peptide chain initiation step of protein synthesis by steroids, and the fact that the most common form of translational regulation occurs during peptide chain initiation& suggested that this is the most likely step that is being affected. To examine this further we measured directly the formation of 80s initiation complexes by programming reticulocyte lysates with MBP mRNAs containing “normal” and altered 5’-untranslated regions in the presence of sparsomycin (to prevent the peptide elongation step). With mRNAs, such as that encoding globin, which are nonresponsive to translational stimulation by steroids, the formation of the 80s initiation complex could easily be seen on glycerol gradients. When the reticulocyte lysates were programmed with normal MBP mRNAs, the appearance of a complex of 100-110s could be seen in addition to the 80s initiation complex. The formation of the 100-110s complex was dependent upon the presence of the nine nucleotide steroid modulatory element in the MBP mRNA. It was abolished if the sequence was deleted from the message or if site-directed changes were introduced into the sequence. Formation of the 100-110s complex could also be prevented if an oligomer of the nine nucleotide sequence was added to the incubation medium. The 100-110s complex appears to be composed of an 80s initiation complex and an additional 40s ribosomal subunit bound to the 5‘-untranslated region of the message. Because elongation is inhibited by sparsomycin, the complex, which represents the beginning of an additional initiation round, is “frozen” in place with further initiation rounds being prevented. These data are consistent with a model in which the steroid serves to promote an increase in the initiation rate of protein synthesis in those messages of appropriate structure, such as the MBP mRNA. We propose that the steroids bind to some normal component of the protein synthetic apparatus (a protein factor) and that this steroid-protein complex interacts with the steroid-responsive element in such a way as to facilitate more frequent rounds of chain initiation by enhancing the interaction of the 40s ribosomal subunit with the 5’ end of the message. It is possible that the effects of steroids on message stability and rates of translation might be related mechanistically. For example, a steroid-induced increase in polypeptide chain initiation would be expected to result in increased ribosomal density per message, thereby leading to increased message stability. For messages whose translation was inhibited by steroids, ribosome density would be expected to decrease in the presence of steroid, thereby leading to decreased message stability. There is evidence that transcripts that are not undergoing active translation or have fewer ribosomes per message are good substrates for degradati~n.~’ The magnitude of the effects of steroids on the posttranscriptional modulation of myelin protein gene expression observed by us and others is only two- to threefold. Nonetheless, the importance of this form of regulation should not be underestimated. During most active myelination, the oligodendrocyte produces substantial amounts of myelin estimated at more than three times the weight of the cell body each day.4RDuring this period, expression of the myelin protein genes is extremely active, and the myelin proteins produced, when incorporated into myelin, are quite stable, with low turnover rates.49Over this same developmental period (10-40 days in
186
ANNALS NEW YORK ACADEMY OF SCIENCES
the mouse) the myelin membrane undergoes major compositional changes, with the proportion of MBP and PLP increasing significantly relative to CNP. Consequently, a two- to threefold increase in the synthesis of the MBP and PLP polypeptides with an equivalent decrease in the amount of CNP could produce a profound shift in the composition of the membrane during this developmental period. Thus, even though the steroid effects on translation of the myelin protein mRNAs are relatively small, they could result in significant changes in the proportions of proteins produced, and, consequently, the composition of the myelin membrane.
SUMMARY
Regulation of myelin protein gene expression occurs at many different levels including transcription, mRNA translocation, translation, and posttranslational modification of myelin proteins prior to their assembly into the membrane. Translocation of myelin basic protein (MBP) mRNAs into oligodendrocyte processes was observed in vivo and in primary cultures, but no such translocation was observed for the mRNAs encoding the proteolipid protein (PLP) or myelin-associated glycoprotein. More than 99% of the mRNAs encoding 2'3'-cyclic nucleotide phosphodiesterase (CNP) remained associated with cell bodies. In the jimpy mutant, MBP mRNA translocation appeared to be impaired, but translocation occurred normally in quaking brains in vivo. We have found that steroids, such as glucocorticoids, stimulate the translation of MBP and PLP mRNAs in cell-free systems and inhibit the translation of CNP mRNA. This pattern of regulation is consistent with compositional changes noted in myelin during development. We have localized a nine nucleotide segment within the 5'-untranslated region of the MBP mRNA that is involved in the action of steroids on translation of this mRNA. We have also determined that the protein synthetic step modulated by the steroids is chain initiation, enhancing the rate at which new ribosomal subunits bind to the MBP mRNAs.
REFERENCES
1. KOZAK,M. 1989. Circumstances and mechanisms of inhibition of translation by secondary structure in eucaryotic mRNAs. Mol. Cell. Biol. 9 5134-5142. R. D. & J. B. HARFORD.1989. Cis-trans models for post-transcriptional gene 2. KLAUSNER, regulation. Science 246: 870-872. 3. JACKSON, 1990. Do the poly(A) tail and 3' untranslated region R. J. & N. STANDART. control mRNA translation? Cell 6 2 15-24. A. T. & W. B. MACKLIN.1988. Cellular and molecular aspects of myelin 4. CAMPAGNONI, protein gene expression. Mol. Neurobiol. 2 41-89. A. T., G. D. CAREY& Y.-T. Yu. 1980. In vitro synthesis of the myelin basic 5. CAMPAGNONI, proteins: Subcellular site of synthesis. J. Neurochem. 3 4 677-686. D. R., G. KREIBICH, A. B. FREY& D. SABATINI. 1982. Synthesis and incorpora6. COLMAN, tion of myelin polypeptides into CNS myelin. J. Cell. Biol. 9 5 598-608. J. A. & P. MORELL.1978. Proteins of myelin and their metabolism. Neuro7. BENJAMINS, chem. Res. 3: 137-174. L. E. & J. A. BENJAMINS. 1983. Effects of monensin on posttranslational 8. TOWNSEND, processing of myelin proteins. J. Neurochem. 4 0 1333-1339. 9. TRAPP,B. D., T. MOENCH,M. PULLEY,E. BARBOSA, G. TENNEKOON & J. GRIFFIN.1987. Spatial segregation of mRNA encoding myelin-specific proteins. Proc. Natl. Acad. Sci. USA 8 4 7773-7777.
CAMPAGNONI et nl.: REGULATION OF GENE EXPRESSION
187
A. N. &A. T. CAMPAGNONI. 1988. Regional expression of myelin protein genes in 10. VERITY, the developing mouse brain: In situ hybridization studies. J. Neurosci. Res. 21: 238-248. D., V. FRIEDRICH, 11. JORDAN, JR. & M. DUBOIS-DALCQ. 1989. In situ hybridization analysis of myelin gene transcripts in developing mouse spinal cord. J. Neurosci. 9 248-257. 1989. Developmental profile and differential 12. SHIOTA,C., M. MIURA& K. MIKOSHIBA. localization of mRNAs of myelin proteins (MBP and PLP) in oligodendrocytes in the brain and in culture. Dev. Brain Res. 4 5 83-94. & R. A. LAZZARINI.1985. The timely 13. ZELLER,N. K., T. N. BEHAR,M. E. DUBOIS-DALCQ expression of myelin basic protein gene in cultured rat brain oligodendrocytes is independent of continuous neuronal influences. J. Neurosci. 5: 2955-2962. E., G. HERMANSON, R. COLE& J. DE VELLIS.1988. Developmental expression of 14. HOLMES, glial-specific mRNAs in primary cultures of rat brain visualized by in situ hybridization. J. Neurosci. Res. 1 9 389-396. S. G., L. HALL & A. T. CAMPAGNONI. 1990. Spatial distribution of 15. AMUR-UMARJEE, mRNAs for myelin proteins in primary cultures of mouse brain. Dev. Neurosci. 12: 262-272. M. S. & R. P. SKOFF.1991. Double labeling in situ hybridization analysis of 16. GHANDOUR, mRNAs for carbonic anhydrase I1 and myelin basic protein: Expression in developing cultured glial cells. Glia 4: 1-10, 1987. A mammalian mitochondria1 RNA processing 17. CHANG,D. D. & D. A. CLAYTON. activity contains nucleus-encoded RNA. Science 235: 1178-1 184. 18. DAVIS,L., G. A. BANKER& 0. STEWARD. 1987. Selective dendritic transport of RNA in hippocampal neurons in culture. Nature 330 477-479. 19. LAWRENCE, J. B. & R. H. SINGER.1986. Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell 4 5 407-415. H. D., M. BACHMAN, B. DIEHL-SIEFERT 20. SCHRODER, & W. E. MULLER.1987. Transport of mRNA from nucleus to cytoplasm. Prog. Nucleic Acid Res. Mol. Biol. 43: 89-142. B. & Z. DOBROWOLSKI. 1984. Ca’+-calmodulin-dependent regulation of F-actin21. BARYLKO, myelin basic protein interaction. Eur. J. Cell Biol. 3 5 327-335. & S. PENMAN.1977. A cytoskeletal structure with 22. LENK,R., L. RANSOM,Y. KAUFMAN associated polyribosomes obtained from HeLa cells. Cell 1 0 67-78. 23. HOGAN,E. L. & S. GREENFIELD. 1984. Animal models of genetic disorders of myelin. In Myelin. P. Morell, ed.: 489-534. Plenum Press. New York. 1987. Developmental expression of the 24. SORG,B. J. A,, M. M. SMITH& A. T. CAMPAGNONI. myelin proteolipid protein and basic protein mRNAs in normal and dysmyelinating mutant mice. J. Neurochem. 4 9 1146-1154. S., S. BROSTOFF & E. HOGAN. 1977. Evidence for defective incorporation of 25. GREENFIELD, proteins in myelin of the quaking mutant mouse. Brain Res. 120: 507-515. 26. BARBARESE, E., C. BARRY,C. H. CHOU,D. J. GOLDTEIN,G. A. NAKOS,R. HYDEDERUYSCHER, K. SCHELD & J. H. CARSON.1988. Expression and localization of myelin basic protein in oligodendrocytes and transfected fibroblasts. J. Neurochem. 51: 17371745. K. D. KING& K. KAMPF.1987. An AG + G G 27. MACKLIN,W. B., M. V. GARDINIER, transition at a splice site in the myelin proteolipid protein gene in jimpy mice results in the removal of an exon. FEBS Lett. 223: 417-421. 28. NAVE,K.-A., F. E. BLOOM& R. J. MILNER.1987. A single nucleotide difference in the gene for myelin proteolipid protein defines the jimpy mutation in mouse. J. Neurochem. 4 9 1873-1877. 29. KNAPP,P. E., R. P. SKOFF& D. W. REDSTONE. 1986. Oligodendroglial cell death in jimpy mice: An explanation for the myelin deficit. J. Neurosci. 6 2813-2822. 30. SKOFF,R. P. 1982. Increased proliferation of oligodendrocytes in the hypomyelinated mouse mutant-jimpy. Brain Res. 248 19-31. 31. KNAPP,P. E. & R. P. SKOFF.1987. A defect in the cell cycle of neuroglia in the myelin deficient jimpy mouse. Dev. Brain Res. 3 5 301-306. 1990. Gene expression in the jimpy 32. VERITY,A. N., M. S. LEVINE& A. T. CAMPAGNONI. mutant: Evidence for fewer oligodendrocytes expressing myelin protein genes and impaired translocation of myelin basic protein mRNA. Dev. Neurosci. 12: 359-372.
188
ANNALS NEW YORK ACADEMY OF SCIENCES
J. H., N. N. HERSCHKOWITZ & P. E. BRAWN. 1975. Synthesis and degradation of 33. CARSON, myelin basic protein in normal and jimpy mouse brain. Trans. Am. SOC.Neurochem. 6 207. 1987. Developmental expression of the 34. SORG,B. A., M. M. SMITH& A. T. CAMPAGNONI, myelin proteolipid protein and basic protein mRNAs in normal and dysmyelinating mutant mice. J. Neurochem. 49: 1146-1154. M. S. & R. P. SKOFF.1988. Expression of galactocerebroside in developing 35. GHANDOUR, normal and jimpy oligodendrocytes in situ. J. Neurocytol. 17: 485498. & A. BISCHOFF.1974. Morphological and biochemical 36. MEIER,C., N. HERSCHKOWITZ observations in the jimpy spinal cord. Acta Neuropathol. 27: 349-362. P., D. M. ROBINS& R. T. SCHIMKE.1978. Regulation of translation of 37. PENNEQUIN, ovalbumin messenger RNA by estrogens and progesterone in oviduct of withdrawn chicks. Eur. J. Biochem. 9 0 51-33. O., A. E. THOMPSON, JR. & R. P. PERRY.1987. Glucocorticoids selectively 38. MEYUHAS, inhibit translation of ribosomal protein mRNAs in P1798 lymphosarcoma cells. Mol. Cell. Biol. 7: 2691-2699. P. CHAMBON, R. L. LINDSEY,M. PONGLIKITMONGKOL, M. 39. SACEDA,M., M. E. LIPPMAN, PUENTE& M. B. MARTIN.1988. Regulation of the estrogen receptor in MCF-7 cells by estradiol. Mol. Endocrinol. 2: 1157-1162. 40. SHAPIRO,D. J., J. E. BLUME& D. A. NIELSEN.1985. Regulation of messenger RNA stability in eukaryotic cells. Bioessays 6 221-226. 41. LABATE,M. E., S. M. WHELLY& K. L. BARKER.1986. Ribosome-associated estradiol binding components in the uterus and their relationship to the translational capacity of uterine ribosomes. Endocrinology 119 140-151. 42. KUMAR,S., R. COLE, F. CHIAPPELLI & J. DE VELLIS.1989. Differential regulation of oligodendrocyte markers by glucocorticoids: Post-transcriptional regulation of both proteolipid protein and myelin basic protein and transcriptional regulation of glycerol phosphate dehydrogenase. Proc. Natl. Acad. Sci. USA 8 6 6807-6811. 1989. Translational regulation of myelin 43. VERDI,J. M., K. KAMPF& A. T. CAMPAGNONI. protein synthesis by steroids. J. Neurochem. 52: 321-324. K., S. L. NEWMAN, 44. KITAMUKA, C. W. CAMPAGNONI, J. M. VERDI,T. MOHANDAS, V. W. HANDLEY & A. T. CAMPAGNONI. 1990. Expression of a novel transcript of the myelin basic protein gene. J. Neurochem. 54: 2032-2041. 1990. Translational regulation by steroids: Identifica45. VERDI,J. M. & A. T. CAMPAGNONI. tion of a steroid modulatory element in the 5’ untranslated region of the myelin basic protein messenger RNA. J. Biol. Chem. 265: 20314-20320. 46. PAIN, V. M. 1986. Initiation of protein synthesis in mammalian cells. Biochem. J. 235: 625-631. 47. KELLY,R., D. R. SHAW& H. L. ENNIS.1987. Role of protein synthesis in decay and accumulation of mRNA during spore germination in the cellular slime mold Diciyostelium discoideum. Mol. Cell. Biol. 7: 799-805. 48. MORELL,P. & A. D. TOEWS.1984. In vivo metabolism of oligodendroglial lipids. In Oligodendroglia. W. T. Norton, ed.: 47-86. Plenum Press. New York. 49. BENJAMINS, J. A. 1984. Protein metabolism of oligodendroglial cells in vivo. In Oligodendroglia. W. T. Norton, ed.: 87-124. Plenum Press. New York.
Posttranscriptional Regulation of Myelin Protein Gene Expression" ANTHONY T. CAMPAGNONI: JOSEPH M. VERDI, A. NEIL VERITY, SHASHI Ah4UR-UMARJEE, AND SUJATHA BYRAVAN Mental Retardation Research Center U.C.L.A.Centerfor the Health Sciences Los Angeles, California 90024 Regulation of gene expression may occur at a number of points between the transcription of a gene and the appearance of a functional protein in its appropriate location within the cell. In recent years we have begun to learn a great deal about the factors that regulate the activity and transcription of genes and the intracellular targeting of proteins after synthesis. However, less is known about the factors that affect the stability and translation of mRNAs, the role that specific mRNAs may play in the intracellular translocation of macromolecules, and the importance of mRNA structure in influencing these processes. Certainly, the importance of the structure of the 5'-untranslated region in the general regulation of translation' as well as cis-acting elements within this region that regulate the translation of specific mRNAs, such as that encoding ferritin,' is being recognized. Evidence also indicates that the 3'-untranslated region of mRNAs may play a role in regulating the translation and stability of mRNAs and, possibly, their translocation within the ce1l.j The expression of the genes encoding the major myelin proteins - myelin basic proteins (MBP), proteolipid proteins (PLPs), myelin-associated glycoprotein (MAG), and 2',3'-cyclic nucleotide-3'-phosphodiesterase(CNP) - have been the subject of intense investigation over the last several years. (See ref. 4 for a review.) Expression of all of these genes is developmentally regulated and probably intimately associated with the differentiation of the oligodendrocyte. All of these genes undergo alternative splicing, producing at least two mRNAs, and the proportions of the alternatively spliced mRNAs and/or the protein isoforms produced change significantly with development, in most cases. Recent investigations suggest that the expression of one of these genes, the MBP gene, may be regulated at a number of different levels including (1) promoter utilization, (2) splicing, (3) mRNA translocation, (4)translation, and (5) posttranslational events. This paper presents further evidence for the posttranscriptional regulation of MBP gene expression, with most examples coming from work done on the expression of the MBP gene.
'This work was supported by NIH grants NS 23022 and NS 23322 and a generous donation from the Ciba-Geigy Corporation, Summit, New Jersey. bAddress for correspondence: Dr. A. T. Campagnoni, Mental Retardation Research Center, U.C.L.A. Center for the Health Sciences, 760 Westwood Plaza, Los Angeles, CA 90024. 178
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
179
MBP mRNA TRANSLOCATION It has been known for about 10 years that the MBPs were synthesized on "free" polyribosome~~~~ and the myelin PLP was synthesized on ribosomes bound to the endoplasmic reticulum.6 Other experiments, both in vivo and in tissue slices, indicated that although MBPs were incorporated very rapidly (within minutes) into myelin, newly synthesized PLP was incorporated into myelin much more slowly ( 30 minute^).^ Furthermore, the incorporation of PLP, but not the MBPs, into the membrane could be interrupted by treatment with monensin, suggesting that assembly of PLP into myelin involved transport through the Golgi apparatusP The available biochemical evidence supported the notion that the two classes of proteins were assembled into myelin by separate routes. Subcellular fractionation experiments demonstrated a 20-fold enrichment of the MBP mRNAs in myelin fractions compared to PLP mRNAs6 This observation suggested that ribosomes containing MBP mRNAs might be located within the cytoplasmic processes connecting the oligodendrocyte somas with the myelin sheaths and, perhaps, the cytoplasmic channels infiltrating the sheaths. If MBP synthesis occurred in these processes and/or channels, then it could explain why incorporation of these proteins into the myelin membrane occurred faster than did incorporation of proteins synthesized in the cell bodies which would require posttranslational transport out to the sites of myelin assembly. Proof of such a model could be obtained by in situ hybridization histochemistry (ISH) through the localization of MBP and PLP mRNAs, but it had to await the isolation of appropriate cDNA probes. Recently, several groups have used ISH to localize myelin-specific mRNAs in vivo9-" and in vitro.'2-'6In vivo, the pattern of distribution of the MBP mRNAs is quite different from that of the PLP mRNAs. With radioactive probes, the distribution of silver grains generated with the PLP probes tended to be clustered over cell bodies, whereas those generated with the MBP probes were diffusely distributed over myelinated regions in the brain.'.'' This difference in gross distribution of these two mRNAs was consistent with the MBP mRNA translocation hypothesis. Further evidence was provided by developmental studies. Verity and Campagnoni" examined the distribution of MBP and PLP mRNAs in the medulla oblongata of the mouse brain at daily intervals beginning at birth. They observed that before the appearance of histologically stained myelin in this region (0-2 days), silver grains representing MBP mRNAs and PLP mRNAs were clustered over cell bodies. Between 2 and 3 days postpartum the distribution of MBP mRNAs changed dramatically, from a clustered to a diffuse distribution, and this change in distribution coincided with the appearance of histologically recognizable myelin. Additional proof that MBP mRNAs are translocated from the cell bodies of oligodendrocytes to their processes has come from studies performed in primary brain cell culture^.'^^'^ Although localization of silver grains to oligodendrocyte processes has been a bit more apparent in vitro than in vivo, the most convincing localizations have been obtained in vitro with nonradioactive probes labeled with dig~xigenin'~ or biotin,16 which are detected immunochemically. In these studies, MBP mRNAs could be clearly localized within oligodendrocyte processes and the translocation of these mRNAs visualized with time in c u l t ~ r e . 'In ~ contrast, PLP mRNAs, MAG mRNAs, and greater than 99% of CNP mRNAs remained associated with the oligodendrocyte cell bodies." Thus, substantial biochemical and cell biologic evidence has now accumulated indicating that MBP mRNAs are translocated within
-
180
ANNALS NEW YORK ACADEMY OF SCIENCES
oligodendrocytes from the cell somas to their processes and, presumably, the cytoplasmic channels infiltrating the myelin sheaths. The protein products of the other major myelin protein genes appear to be targeted to myelin by other mechanisms subsequent to translation. The intracellular compartmentalization of mRNAs has been described for a number of mRNAs including those that encode mit~chondrial,’~ dendritic,18 and cytoskeletal proteins.’’ The mechanisms by which this intracellular translocation of mRNAs occurs is not yet known, but they could involve the participation of specific carrier proteins binding to cytoskeletal elements in a fashion similar to the transport of messenger ribonucleoproteins from the nucleus to the cytoplasm.” It is also unclear whether the MBP mRNAs or the nascent MBP polypeptide chains play a role in the translocation process. In this regard, MBP has been shown to interact with cytoskeletal proteins:’ and there is evidence that “free” polyribosomes can associate with cytoskeletal elements through their mRNAs.” MBP mRNA AND PLP mRNA TRANSLOCATION IN JZMPY AND QUAKING MICE
Quaking is an autosomal, recessive mutation that maps to mouse chromosome 17 and does not appear to be a mutation in one of the major myelin protein genes. The mutation is characterized by pronounced hypomyelination of the central nervous ~ystem.2~ Work from this laboratory has shown that expression of the MBP gene is altered in the quaking brain up to about 18-21 days, after which the levels of MBP mRNA and the synthesis of MBP appear to be close to n0rma1.2~In marked contrast, the steady-state levels of MBP, as measured by radioimmunologic techniques, are only 5 2 5 % of normal in the central nervous system. Inasmuch as unbound MBP is readily susceptible to protease attack, these results are consistent with an earlier hypothesis that quaking mice are defective in their ability to incorporate newly synthesized MBPs into m y e h a In view of evidence indicating that MBP mRNAs are translocated within the oligodendrocyte, it was of interest to determine if this inability of the MBPs to be incorporated into quaking myelin was due to a defect in this translocation phenomenon or to some other posttranslational event. ISH examination of the quaking brain indicated that MBP mRNAs could be detected at approximately normal levels in the 20-day-oldquaking brain (FIG.1)and that the distribution of mRNAwas as diffuse as that in the controls. These results suggest that MBP mRNA translocation is unimpaired in the quaking brain. A corollary to these results is that yet another regulatory point must exist in the assembly of the MBPs into the myelin sheath. It has been suggested that as the MBPs are located on the cytoplasmic face of the unit bilayer comprising the myelin sheath, the assembly of MBPs into the membrane occurs by diffusion shortly after their synthesis in the oligodendrocyte processes.’ Our data suggest that synthesis of MBPs close to the growing myelin sheath is not sufficient to assure proper assembly of the proteins into the membrane. Some additional event, such as a posttranslational modification of the protein, must occur for proper assembly of the MBPs into the membrane. Alternatively, the presence of a receptor for MBP in the membrane might be required before incorporation of the protein.26 The jimpy mutation appears to be a point mutation in the splice acceptor site of phenotype is somewhat complex, affecting the the fifth exon of the PLP gene.27s28The oligodendrocyte and the expression of several of the myelin protein genes, probably through an alteration in the differentiation/maturation/death of the oligodendro-
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
181
cyte.2%32 The levels of MBP mRNA, even though severely reduced in this mutant, have generally been higher than the steady-state levels of polypeptide in jimpy brain^.",'^ This finding suggests that the small amount of protein made in the mutant brain is not incorporated effectively into the membrane. ISH have confirmed other biochemical data34that MBP mRNA levels are reduced in jimpy brains. However, in the mutant MBP mRNA-specific silver grains appear more clustered on oligodendrocyte cell bodies and less diffuse than in either normal or quaking brains (FIG.1). This pattern of labeling was most prominent in the caudate putamen and corpus callosum, but the pattern was similar in all regions of the brain and persisted until the animals died at 3-4 weeks. These data suggest impairment in the translocation of MBP mRNAs in jimpy brain. It is unlikely that impairment of MBP mRNA translocation is due to a reduction or truncation of oligodendrocyte cellular processes because two groups have independently reported that the numbers and general appearance of processes with antigalactocerebroside staining appear normal in the m~tant.~’.~’ It is noteworthy that jirnpy oligodendrocytes contain numerous lipid inclusions and have a substantial reduction in m i c r o t ~ b u l e s .If~ ~MBP mRNA translocation involves the cytoskeletal network, both of these factors might contribute to impaired translocation of MBP mRNA in the mutant.
REGULATION OF MYELIN PROTEIN mRNA TRANSLATION BY STEROIDS
Steroids are well known to control gene expression at the transcriptional level; however, a number of studies indicate that steroids also may regulate gene expression posttran~criptionally.~’-~~ It has been suggested that steroids can alter the stability of the mRNAs4” and/or influence the rates of translation of responsive ~RNAS.~~,~~ None of the myelin protein genes appears to be transcriptionally regulated by steroids, but two recent studies suggest that steroids might act at the posttranscriptional level. Kumar et aL4*observed that hydrocortisone can stimulate the expression of MBP and PLP in primary oligodendrocyte cultures, without any increase in the transcription rates of these genes. We have found that the translation of MBP and PLP mRNAs is stimulated by hydrocortisone and other steroids in cell-free sy~tems.4~ Our most recent efforts have been aimed toward determining (1) if steroids could exert a direct effect on the translation of mRNAs, (2) if sequences within the MBP mRNA were involved in this phenomenon, and (3) which protein synthetic step is modulated by the steroid. Toward this end we developed a cell-free translation system that could be programmed with synthetic mRNAs to determine the effects of steroids on the translation of specific synthetic mRNAs. The use of this system permitted us (1) to assess the effect of steroids on mRNA translation independent of effects on transcription, (2) to assess directly the effect of steroids on the rates of translation of specific mRNAs, and (3) to determine possible regulatory sequences in the 5’untranslated region by programming the system with structurally modified mRNAs. Using this system we found that several classes of steroids could stimulate the translation of MBP mRNAs in this system, with glucocorticoids being more effective than sex steroids. These same steroids served to inhibit the translation of the CNP mRNA in an equal, but opposite direction to the effects observed on the MBP mRNA. No effect was observed on the translation of either message with the mineralocorticoid, aldosterone, or the sterols cholesterol and desmosterol. Transla-
182
ANNALS NEW YORK ACADEMY OF SCIENCES
FLGURE 1. In situ hybridization histochemistry of MBP mRNAs within the caudate putamen of 15-day-old normal (+/+) mice and the dysmyelinating mutants quaking (qk),jimpy up), and shiverer (sh). In normal and quaking mice, silver grains representing MBP mRNAs are diffusely distributed over myelinated tracts indicative of normal translocation of MBP mRNAs. In contrast, in the jimpy mutant, MBP mRNA translocation is impaired as seen by increased clustering of silver grains over oligodendrocyte cell bodies (arrows). As expected, no detectable MBP transcripts are observed in shiverer mice.
m w
c
184
ANNALS NEW YORK ACADEMY OF SCIENCES
tion of the PLP mRNA was also stimulated by hydrocortisone, but the translation of MAG mRNA was not affected. Kinetic analysis of the data indicated that in this system the steroids were affecting the rates of translation of the responsive mRNAs. Furthermore, experiments suggested that steroids were acting on translation through a novel mechanism that did not involve the “classic” glucocorticoid receptor: (1) RU 38486, an antagonist of the “classic” glucocorticoid receptor, did not inhibit the stimulation of MBP mRNA translation by glucocorticoids. (2) The order of effectiveness among the glucocorticoids was different from that of the glucocorticoid receptor-mediated effect on transcription. (3) The dose response curves were more complex than observed with the classic glucocorticoid receptor, reaching maxima in the physiologically relevant range of 10-R-10-9M steroid. (4) The nucleotide sequence of the element in the MBP mRNA that is responsive to hydrocortisone is unlike that of the glucocorticoid-responsive element in genes that are transcriptionally regulated by this steroid. We have found that the translation of all alternatively spliced forms of the MBP mRNAs transcribed from the major promoter of the gene are stimulated by hydroc~rtisone.~~ Transcription from the major promotor of the MBP gene begins 48 nt upstream of the initiator codon. However, the MBP gene also possesses a second, minor upstream transcription start site. Transcription from this secondary promoter produces MBP mRNAs with a longer 5’-untranslated region that consists of the 48 nt leader plus an additional upstream sequence of at least 350 nt.44The translation of these minor MBP mRNAs with a longer 5‘-untranslated region is not stimulated by hydrocorti~one.~~ These results suggested that the 5‘-untranslated region might be important in the steroid modulation of MBP translation. This notion was supported by studies in which the translation of an MBP mRNA, which contained little of the 3’-untranslated region but a normal 5‘-untranslated region, was stimulated by hydrocortisone. This finding prompted us to investigate further the role of the 5‘-untranslated region in this p h e n ~ m e n o nA . ~series ~ of cDNA clones encoding the 14 kD and 18.5 kD MBP isoforms, but differing in the lengths of their 5’-untranslated regions were generated using the polymerase chain reaction and subcloned into a transcription vector. Site-directed changes were also introduced into selected nucleotides within the 5‘-untranslated region of the MBP mRNAs using this technique. The transcribed RNAs were examined in the cell-free translation system in the presence and absence of hydrocortisone, permitting us to localize a nine-base region within the 5’untranslated region of MBP mRNA between -29 and -37 that appears necessary for the stimulation of translation by hydrocortisone. Several lines of evidence indicate the importance of this region: (1) Removal of this region of the message eliminated the translational response to steroid. (2) Site-directed mutagenesis of several bases within this region reduced or eliminated the response. (3) A synthetic oligonucleotide of the same sequence could entirely compete out the response at a 10:1 molar ratio of oligonucleotide:message, but no such competition was observed with an oligonucleotide containing only two base changes. (4) Transferral of the sequence to a nonresponsive message rendered the translation of that mRNA responsive to hydrocortisone. The steroid modulatory element in the 5’-untranslated region of the MBP mRNA is included within the sequence AGAAGACCC. However, this sequence alone probably does not account for the entire translational response to steroid because translation of the minor MBP mRNA with a longer 5‘-untranslated region was not influenced by steroids, even though it contains the nine nucleotide element. Examination of the most stable secondary structure of the 5’-untranslated region of
CAMPAGNONI et al.: REGULATION OF GENE EXPRESSION
185
this unresponsive message indicates that the element is involved in a base paired stem, whereas in the “normal” MBP message six of the nine nucleotides (AGAAGA) are found in an accessible loop structure. This group of six nucleotides, but not the entire nine nucleotide region, is found twice in the 5’-untranslated region of the PLP mRNA, the translation of which is also stimulated by hydrocortisone. The position of the glucocorticoid modulatory element in the 5‘-untranslated region of the MBP mRNA, the many reports of alterations in the peptide chain initiation step of protein synthesis by steroids, and the fact that the most common form of translational regulation occurs during peptide chain initiation& suggested that this is the most likely step that is being affected. To examine this further we measured directly the formation of 80s initiation complexes by programming reticulocyte lysates with MBP mRNAs containing “normal” and altered 5’-untranslated regions in the presence of sparsomycin (to prevent the peptide elongation step). With mRNAs, such as that encoding globin, which are nonresponsive to translational stimulation by steroids, the formation of the 80s initiation complex could easily be seen on glycerol gradients. When the reticulocyte lysates were programmed with normal MBP mRNAs, the appearance of a complex of 100-110s could be seen in addition to the 80s initiation complex. The formation of the 100-110s complex was dependent upon the presence of the nine nucleotide steroid modulatory element in the MBP mRNA. It was abolished if the sequence was deleted from the message or if site-directed changes were introduced into the sequence. Formation of the 100-110s complex could also be prevented if an oligomer of the nine nucleotide sequence was added to the incubation medium. The 100-110s complex appears to be composed of an 80s initiation complex and an additional 40s ribosomal subunit bound to the 5‘-untranslated region of the message. Because elongation is inhibited by sparsomycin, the complex, which represents the beginning of an additional initiation round, is “frozen” in place with further initiation rounds being prevented. These data are consistent with a model in which the steroid serves to promote an increase in the initiation rate of protein synthesis in those messages of appropriate structure, such as the MBP mRNA. We propose that the steroids bind to some normal component of the protein synthetic apparatus (a protein factor) and that this steroid-protein complex interacts with the steroid-responsive element in such a way as to facilitate more frequent rounds of chain initiation by enhancing the interaction of the 40s ribosomal subunit with the 5’ end of the message. It is possible that the effects of steroids on message stability and rates of translation might be related mechanistically. For example, a steroid-induced increase in polypeptide chain initiation would be expected to result in increased ribosomal density per message, thereby leading to increased message stability. For messages whose translation was inhibited by steroids, ribosome density would be expected to decrease in the presence of steroid, thereby leading to decreased message stability. There is evidence that transcripts that are not undergoing active translation or have fewer ribosomes per message are good substrates for degradati~n.~’ The magnitude of the effects of steroids on the posttranscriptional modulation of myelin protein gene expression observed by us and others is only two- to threefold. Nonetheless, the importance of this form of regulation should not be underestimated. During most active myelination, the oligodendrocyte produces substantial amounts of myelin estimated at more than three times the weight of the cell body each day.4RDuring this period, expression of the myelin protein genes is extremely active, and the myelin proteins produced, when incorporated into myelin, are quite stable, with low turnover rates.49Over this same developmental period (10-40 days in
186
ANNALS NEW YORK ACADEMY OF SCIENCES
the mouse) the myelin membrane undergoes major compositional changes, with the proportion of MBP and PLP increasing significantly relative to CNP. Consequently, a two- to threefold increase in the synthesis of the MBP and PLP polypeptides with an equivalent decrease in the amount of CNP could produce a profound shift in the composition of the membrane during this developmental period. Thus, even though the steroid effects on translation of the myelin protein mRNAs are relatively small, they could result in significant changes in the proportions of proteins produced, and, consequently, the composition of the myelin membrane.
SUMMARY
Regulation of myelin protein gene expression occurs at many different levels including transcription, mRNA translocation, translation, and posttranslational modification of myelin proteins prior to their assembly into the membrane. Translocation of myelin basic protein (MBP) mRNAs into oligodendrocyte processes was observed in vivo and in primary cultures, but no such translocation was observed for the mRNAs encoding the proteolipid protein (PLP) or myelin-associated glycoprotein. More than 99% of the mRNAs encoding 2'3'-cyclic nucleotide phosphodiesterase (CNP) remained associated with cell bodies. In the jimpy mutant, MBP mRNA translocation appeared to be impaired, but translocation occurred normally in quaking brains in vivo. We have found that steroids, such as glucocorticoids, stimulate the translation of MBP and PLP mRNAs in cell-free systems and inhibit the translation of CNP mRNA. This pattern of regulation is consistent with compositional changes noted in myelin during development. We have localized a nine nucleotide segment within the 5'-untranslated region of the MBP mRNA that is involved in the action of steroids on translation of this mRNA. We have also determined that the protein synthetic step modulated by the steroids is chain initiation, enhancing the rate at which new ribosomal subunits bind to the MBP mRNAs.
REFERENCES
1. KOZAK,M. 1989. Circumstances and mechanisms of inhibition of translation by secondary structure in eucaryotic mRNAs. Mol. Cell. Biol. 9 5134-5142. R. D. & J. B. HARFORD.1989. Cis-trans models for post-transcriptional gene 2. KLAUSNER, regulation. Science 246: 870-872. 3. JACKSON, 1990. Do the poly(A) tail and 3' untranslated region R. J. & N. STANDART. control mRNA translation? Cell 6 2 15-24. A. T. & W. B. MACKLIN.1988. Cellular and molecular aspects of myelin 4. CAMPAGNONI, protein gene expression. Mol. Neurobiol. 2 41-89. A. T., G. D. CAREY& Y.-T. Yu. 1980. In vitro synthesis of the myelin basic 5. CAMPAGNONI, proteins: Subcellular site of synthesis. J. Neurochem. 3 4 677-686. D. R., G. KREIBICH, A. B. FREY& D. SABATINI. 1982. Synthesis and incorpora6. COLMAN, tion of myelin polypeptides into CNS myelin. J. Cell. Biol. 9 5 598-608. J. A. & P. MORELL.1978. Proteins of myelin and their metabolism. Neuro7. BENJAMINS, chem. Res. 3: 137-174. L. E. & J. A. BENJAMINS. 1983. Effects of monensin on posttranslational 8. TOWNSEND, processing of myelin proteins. J. Neurochem. 4 0 1333-1339. 9. TRAPP,B. D., T. MOENCH,M. PULLEY,E. BARBOSA, G. TENNEKOON & J. GRIFFIN.1987. Spatial segregation of mRNA encoding myelin-specific proteins. Proc. Natl. Acad. Sci. USA 8 4 7773-7777.
CAMPAGNONI et nl.: REGULATION OF GENE EXPRESSION
187
A. N. &A. T. CAMPAGNONI. 1988. Regional expression of myelin protein genes in 10. VERITY, the developing mouse brain: In situ hybridization studies. J. Neurosci. Res. 21: 238-248. D., V. FRIEDRICH, 11. JORDAN, JR. & M. DUBOIS-DALCQ. 1989. In situ hybridization analysis of myelin gene transcripts in developing mouse spinal cord. J. Neurosci. 9 248-257. 1989. Developmental profile and differential 12. SHIOTA,C., M. MIURA& K. MIKOSHIBA. localization of mRNAs of myelin proteins (MBP and PLP) in oligodendrocytes in the brain and in culture. Dev. Brain Res. 4 5 83-94. & R. A. LAZZARINI.1985. The timely 13. ZELLER,N. K., T. N. BEHAR,M. E. DUBOIS-DALCQ expression of myelin basic protein gene in cultured rat brain oligodendrocytes is independent of continuous neuronal influences. J. Neurosci. 5: 2955-2962. E., G. HERMANSON, R. COLE& J. DE VELLIS.1988. Developmental expression of 14. HOLMES, glial-specific mRNAs in primary cultures of rat brain visualized by in situ hybridization. J. Neurosci. Res. 1 9 389-396. S. G., L. HALL & A. T. CAMPAGNONI. 1990. Spatial distribution of 15. AMUR-UMARJEE, mRNAs for myelin proteins in primary cultures of mouse brain. Dev. Neurosci. 12: 262-272. M. S. & R. P. SKOFF.1991. Double labeling in situ hybridization analysis of 16. GHANDOUR, mRNAs for carbonic anhydrase I1 and myelin basic protein: Expression in developing cultured glial cells. Glia 4: 1-10, 1987. A mammalian mitochondria1 RNA processing 17. CHANG,D. D. & D. A. CLAYTON. activity contains nucleus-encoded RNA. Science 235: 1178-1 184. 18. DAVIS,L., G. A. BANKER& 0. STEWARD. 1987. Selective dendritic transport of RNA in hippocampal neurons in culture. Nature 330 477-479. 19. LAWRENCE, J. B. & R. H. SINGER.1986. Intracellular localization of messenger RNAs for cytoskeletal proteins. Cell 4 5 407-415. H. D., M. BACHMAN, B. DIEHL-SIEFERT 20. SCHRODER, & W. E. MULLER.1987. Transport of mRNA from nucleus to cytoplasm. Prog. Nucleic Acid Res. Mol. Biol. 43: 89-142. B. & Z. DOBROWOLSKI. 1984. Ca’+-calmodulin-dependent regulation of F-actin21. BARYLKO, myelin basic protein interaction. Eur. J. Cell Biol. 3 5 327-335. & S. PENMAN.1977. A cytoskeletal structure with 22. LENK,R., L. RANSOM,Y. KAUFMAN associated polyribosomes obtained from HeLa cells. Cell 1 0 67-78. 23. HOGAN,E. L. & S. GREENFIELD. 1984. Animal models of genetic disorders of myelin. In Myelin. P. Morell, ed.: 489-534. Plenum Press. New York. 1987. Developmental expression of the 24. SORG,B. J. A,, M. M. SMITH& A. T. CAMPAGNONI. myelin proteolipid protein and basic protein mRNAs in normal and dysmyelinating mutant mice. J. Neurochem. 4 9 1146-1154. S., S. BROSTOFF & E. HOGAN. 1977. Evidence for defective incorporation of 25. GREENFIELD, proteins in myelin of the quaking mutant mouse. Brain Res. 120: 507-515. 26. BARBARESE, E., C. BARRY,C. H. CHOU,D. J. GOLDTEIN,G. A. NAKOS,R. HYDEDERUYSCHER, K. SCHELD & J. H. CARSON.1988. Expression and localization of myelin basic protein in oligodendrocytes and transfected fibroblasts. J. Neurochem. 51: 17371745. K. D. KING& K. KAMPF.1987. An AG + G G 27. MACKLIN,W. B., M. V. GARDINIER, transition at a splice site in the myelin proteolipid protein gene in jimpy mice results in the removal of an exon. FEBS Lett. 223: 417-421. 28. NAVE,K.-A., F. E. BLOOM& R. J. MILNER.1987. A single nucleotide difference in the gene for myelin proteolipid protein defines the jimpy mutation in mouse. J. Neurochem. 4 9 1873-1877. 29. KNAPP,P. E., R. P. SKOFF& D. W. REDSTONE. 1986. Oligodendroglial cell death in jimpy mice: An explanation for the myelin deficit. J. Neurosci. 6 2813-2822. 30. SKOFF,R. P. 1982. Increased proliferation of oligodendrocytes in the hypomyelinated mouse mutant-jimpy. Brain Res. 248 19-31. 31. KNAPP,P. E. & R. P. SKOFF.1987. A defect in the cell cycle of neuroglia in the myelin deficient jimpy mouse. Dev. Brain Res. 3 5 301-306. 1990. Gene expression in the jimpy 32. VERITY,A. N., M. S. LEVINE& A. T. CAMPAGNONI. mutant: Evidence for fewer oligodendrocytes expressing myelin protein genes and impaired translocation of myelin basic protein mRNA. Dev. Neurosci. 12: 359-372.
188
ANNALS NEW YORK ACADEMY OF SCIENCES
J. H., N. N. HERSCHKOWITZ & P. E. BRAWN. 1975. Synthesis and degradation of 33. CARSON, myelin basic protein in normal and jimpy mouse brain. Trans. Am. SOC.Neurochem. 6 207. 1987. Developmental expression of the 34. SORG,B. A., M. M. SMITH& A. T. CAMPAGNONI, myelin proteolipid protein and basic protein mRNAs in normal and dysmyelinating mutant mice. J. Neurochem. 49: 1146-1154. M. S. & R. P. SKOFF.1988. Expression of galactocerebroside in developing 35. GHANDOUR, normal and jimpy oligodendrocytes in situ. J. Neurocytol. 17: 485498. & A. BISCHOFF.1974. Morphological and biochemical 36. MEIER,C., N. HERSCHKOWITZ observations in the jimpy spinal cord. Acta Neuropathol. 27: 349-362. P., D. M. ROBINS& R. T. SCHIMKE.1978. Regulation of translation of 37. PENNEQUIN, ovalbumin messenger RNA by estrogens and progesterone in oviduct of withdrawn chicks. Eur. J. Biochem. 9 0 51-33. O., A. E. THOMPSON, JR. & R. P. PERRY.1987. Glucocorticoids selectively 38. MEYUHAS, inhibit translation of ribosomal protein mRNAs in P1798 lymphosarcoma cells. Mol. Cell. Biol. 7: 2691-2699. P. CHAMBON, R. L. LINDSEY,M. PONGLIKITMONGKOL, M. 39. SACEDA,M., M. E. LIPPMAN, PUENTE& M. B. MARTIN.1988. Regulation of the estrogen receptor in MCF-7 cells by estradiol. Mol. Endocrinol. 2: 1157-1162. 40. SHAPIRO,D. J., J. E. BLUME& D. A. NIELSEN.1985. Regulation of messenger RNA stability in eukaryotic cells. Bioessays 6 221-226. 41. LABATE,M. E., S. M. WHELLY& K. L. BARKER.1986. Ribosome-associated estradiol binding components in the uterus and their relationship to the translational capacity of uterine ribosomes. Endocrinology 119 140-151. 42. KUMAR,S., R. COLE, F. CHIAPPELLI & J. DE VELLIS.1989. Differential regulation of oligodendrocyte markers by glucocorticoids: Post-transcriptional regulation of both proteolipid protein and myelin basic protein and transcriptional regulation of glycerol phosphate dehydrogenase. Proc. Natl. Acad. Sci. USA 8 6 6807-6811. 1989. Translational regulation of myelin 43. VERDI,J. M., K. KAMPF& A. T. CAMPAGNONI. protein synthesis by steroids. J. Neurochem. 52: 321-324. K., S. L. NEWMAN, 44. KITAMUKA, C. W. CAMPAGNONI, J. M. VERDI,T. MOHANDAS, V. W. HANDLEY & A. T. CAMPAGNONI. 1990. Expression of a novel transcript of the myelin basic protein gene. J. Neurochem. 54: 2032-2041. 1990. Translational regulation by steroids: Identifica45. VERDI,J. M. & A. T. CAMPAGNONI. tion of a steroid modulatory element in the 5’ untranslated region of the myelin basic protein messenger RNA. J. Biol. Chem. 265: 20314-20320. 46. PAIN, V. M. 1986. Initiation of protein synthesis in mammalian cells. Biochem. J. 235: 625-631. 47. KELLY,R., D. R. SHAW& H. L. ENNIS.1987. Role of protein synthesis in decay and accumulation of mRNA during spore germination in the cellular slime mold Diciyostelium discoideum. Mol. Cell. Biol. 7: 799-805. 48. MORELL,P. & A. D. TOEWS.1984. In vivo metabolism of oligodendroglial lipids. In Oligodendroglia. W. T. Norton, ed.: 47-86. Plenum Press. New York. 49. BENJAMINS, J. A. 1984. Protein metabolism of oligodendroglial cells in vivo. In Oligodendroglia. W. T. Norton, ed.: 87-124. Plenum Press. New York.
