- Email: [email protected]
The Influence of mRNA Stability on Glucose Transporter (GLUT1) Gene Expression
Biochemical and Biophysical Research Communications 263, 265–269 (1999) Article ID bbrc.1999.1328, available online at http://www.idealibrary.com on
BREAKTHROUGHS AND VIEWS The Influence of mRNA Stability on Glucose Transporter (GLUT1) Gene Expression Chen Qi and Phillip H. Pekala 1 Department of Biochemistry, School of Medicine, East Carolina University, Greenville, North Carolina 27858-4354
Received August 10, 1999
One mechanism for modification of glucose transport activity occurs through regulation of the cellular content of transporter protein by alteration of transcript stability. Regulated mRNA decay has been shown to play an important role in control of posttranscriptional gene expression. Implicated, as a pivotal element in this regulation is the 3*-untranslated region (UTR) of the message. Recent work from several labs has focused on sequence motifs within the 3*-UTR of glucose transporter (GLUT1) mRNA that serve as destabilizing or stabilizing elements and recognition of these elements by specific proteins. In this review, we address several critical studies each of which has identified elements in the GLUT1 3*-UTR that are involved in the control of transcript stability and demonstrated that these sequence motifs are recognized by specific binding proteins. © 1999 Academic Press Key Words: mRNA stability; GLUT1; 3*-untranslated region; RNA binding proteins.
The facilitated diffusion of glucose across the plasma membrane is the rate-limiting step for subsequent glucose metabolism and energy production within the cell (1, 2). A family of tissue-specific integral membrane proteins known as the glucose transporters (GLUTs) (3) catalyzes this process. Transporter activity can be regulated by hormones, growth factors and metabolites through the redistribution of transporter proteins from intracellular vesicular storage to the plasma membrane (translocation), modulation of transporter intrinsic activity and stimulation of new protein synthesis through increased transcription and/or alteration in transcript stability (4). Our work, in large part, has focused on control of the transcription of the glucose transporter genes. During the course of these studies it 1
To whom correspondence should be addressed. Fax: (252) 8163383. E-mail: [email protected].
became readily apparent that most agents affecting transporter transcription also modulated transporter mRNA stability. With this consideration, our review will focus on regulated mRNA stability as a vehicle for control of transporter gene expression, with particular attention to GLUT1, the ubiquitously expressed, homeostatic glucose transporter. GLUT1 EXPRESSION IN HEALTH AND DISEASE The GLUT1 protein is expressed at high levels in all fetal tissues, while in the adult it is expressed ubiquitously but predominantly at high levels within several biologically important tissues in cell types to include, fibroblasts, endothelial cells and erythrocytes (2). Major localization also includes the endothelial cells at blood-tissue barriers such as placenta and the bloodbrain barrier (5–7). GLUT1 is also the primary transporter in the choroid plexus, ependyma and glia (8). Seidner et al. recently provided the first identification of a molecular defect underlying dysfunctional bloodbrain barrier glucose transport resulting in disease (9). Mutations in one allele of the GLUT1 gene leading to truncation of the GLUT1 protein result in decreased availability of glucose to the brain. While these individuals have one functional copy of the GLUT1 gene, the phenotype is severe, characterized by infantile seizures, delayed development and acquired microcephaly (9). The metabolic importance of the GLUT1 transporter is further supported by the failure to obtain mice homozygous for an inactivated GLUT1 gene (10). Altered expression of GLUT1 has been correlated with several disease states. Increased GLUT1 expression has been associated strongly with neoplastic progression in the colon (11), as well as both sporadic and inherited clear cell renal carcinoma associated with von Hippel-Lindau disease (12). In contrast, patients with Alzheimer’s disease exhibited a reduced GLUT1 protein content at the blood brain barrier (13).
265
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Regulation of GLUT1 mRNA Stability Cell type/tissue Mouse fibroblasts BALB/c 3T3 3T3-L1
NIH 3T3-L1 Mouse adipocytes 3T3 F442A
3T3-L1 Rat muscle cells L6 myocytes
L6 myotubes Brain and neural cells Bovine brain endothelial cells Glioma Rat neuronal and astroglia cells ECL cells C6 Glioma Muscle tissue Rat hindlimb Red skeletal muscle White skeletal muscle
Agent
PDGF TNF-a cAMP Phorbol ester Okadaic acid Vanadate Insulin Piogtilazone Insulin 1 Pioglitazone Arachidonic acid
Fold increase in half-life
Reference
4 3 6 6 5.6 2–3
(24) (4, 31, 32)
2.5 1.6 .10 .2
(33) (27)
(29)
Insulin/IGF-1 Glucose deprivation Insulin 1 Glucose deprivation Tunicamycin 2-deoxyglucose Glucosamine Oxidative stress
2 2 4 2 2 3
(17, 34)
Glucose deprivation VEGF Hypoxia Cerebrolysin Actinomycin D
2 4.8 1.2 1.7 2.3
(16) (36) (37) (38) (39)
Perfusion Perfusion
3.5 2.5
(40)
Regulation of mRNA stability plays a major role in control of GLUT1 gene expression (4, 14, 15). Several laboratories have reported alteration of the stability of the GLUT1 transcript by conditions to include glucose deprivation (16 –18), experimental diabetes (19), hemangioblastoma (20 –22), inhibition of oxidative phosphorylation (23) as well as cytokine, hormone and metabolite stimulation (24, 17, 14, 25–30). As demonstrated in Table 1, a variety of unrelated hormones, pharmaceuticals and physiological situations have been demonstrated to alter the stability of the GLUT1 transcript in various cell lines and tissues. In terms of structure, the GLUT1 transcript is 2544 nucleotides in length with a 39-UTR of 884 nucleotides. The 39-UTR is approximately 56% A 1 U overall but regions of 150 to 200 bases exist where the A 1 U content is nearly 70%. In this context, the 39-UTR would be considered A 1 U rich. GENERAL ASPECTS OF mRNA TURNOVER In order that the reader may appreciate the studies that have addressed GLUT1 mRNA stability detailed in the following sections, we will first briefly examine a few general principles of RNA turnover.
(35)
Messenger RNA decay plays an important role in control of posttranscriptional gene expression (41– 43). Numerous studies, predominantly in the yeast Saccharomyces cerevisiae, detailing the mRNA decay characteristics for a variety of transcripts have supported the existence of several pathways for the degradation of eukaryotic mRNAs (43– 45). A decay pathway in which turnover of eukaryotic mRNA is initiated by shortening of the poly (A) tail, followed by decapping and subsequent 59 to 39 exonucleolytic degradation of the message appears to predominate (reviewed in ref. 46). Additionally a 39 to 59 pathway has been described where decay of the poly (A) tail continues after the initial phase of deadenylation followed by degradation of the 39-untranslated region (39-UTR) (47, 48). Recent evidence (49) suggests that the 39 to 59 pathway exists and may predominate in mammalian cells. While not meant to represent an exclusive list of potential pathways of mRNA decay, the 39-UTR (including the poly (A) tail) is implicated as being a pivotal element for the regulation of mRNA decay in both models (50). RNA-binding proteins that recognize specific consensus sequences (particularly in the 39-UTR) are also thought to play a role in the degradation process. The identification and characterization of these proteins is
266
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
important in defining the mechanisms involved not only in control of mRNA turnover but in posttranscriptional regulation in general. RNA-protein interactions have been shown to influence many processes including: translation (51–53), RNA stability (21, 43, 52, 54 –57), mRNA transport and localization (58 – 60), splicing (61– 63) and polyadenylation (64). We will now turn our attention to several instances where protein binding has influenced the stability of the GLUT1 transcript. GLUT1 EXPRESSION AND TUMOR NECROSIS FACTOR-a-INDUCED RNA BINDING PROTEINS Treatment of 3T3-L1 fibroblasts with the cytokine tumor necrosis factor-a (TNF) produced a 3 to 4-fold increase in the half-life of the GLUT1 mRNA. The increased GLUT1 mRNA content was demonstrated to result in the synthesis and plasma membrane localization of new GLUT1 protein which lead to increased glucose transport activity (8- to 10-fold over basal) (14). Coincident with the stabilization of the message was the up-regulation of RNA binding proteins of molecular weights 37 and 40 kDa with apparent binding specificity for the GLUT1 39-UTR (32). Further studies demonstrated that: 1. the GLUT1 39-UTR was a destabilizing UTR; 2. the destabilizing element mapped to a portion of the GLUT1 39-UTR located between bases 2242 and 2347; and 3. this region mediated the stabilizing response to TNF. This region is predominantly GC-rich ('60%) and was not homologous to either a GC-rich stability element reported for the TGF-b1 gene (65) or to a TNF-responsive stability element identified within the surfactant-B protein mRNA (67). Deletion of this region confirmed that it mediated both the rapid decay of the GLUT1 message as well as GLUT1 mRNA accumulation (stabilization) in response to TNF (15). Using RNA gel-mobility shift assays, the two TNFinducible RNA-binding proteins were identified as ligands for the region between nucleotides 2240 and 2347 of the 39-UTR. The RNA-binding activity of these proteins was observed to increase coordinately with the stabilization of the GLUT1 message following TNF stimulation. When the 2240 through 2347 region of the 39-UTR was deleted, coincident with the loss of protein binding, the message was stabilized and the response to TNF was lost. These observations demonstrate that the instability element, the TNF-response element and protein binding are all localized to the 106-nucleotide motif. While these elements may be separable within the motif, the data are consistent with congruence between or among them. GLUT1 EXPRESSION AND BRAIN-DERIVED RNA BINDING PROTEINS Pardridge and colleagues have examined GLUT1 expression in human brain tumors and have described
two cis-elements in the 39-UTR of the GLUT1 mRNA that bind protein and regulate expression in an apparent tumor specific fashion (20, 21, 39, 67). The first element is A 1 U rich and localized between nucleotides 1885 and 1906 that binds a protein of 44 kDa and suggested to be an instability element (20, 21, 39, 67). The second element resides between nucleotides 21862203 and binds a protein of 80 kDa. The 44 kDa complex was selectively down regulated in hemangioblastoma (a brain tumor that overexpresses GLUT1) as compared to glioblastoma multiform. While the 80 kDa protein was selectively expressed in hemangioblastoma (20, 21, 39, 67). The enhancer role of the region between nucleotides 2100 to 2300 was confirmed by subcloning the region into a luciferase reporter construct which was then transfected into C6 rat glioma cells. Luciferase expression increased fivefold relative to controls and a 228% increase in the message half-life was observed (21, 39, 67). Deletion of nucleotides 21812190, the putative binding site for the 80 kDa protein, eliminated the enhancement of the luciferase reporter (21, 39, 67). These data support the hypothesis that the 2181-2190-nucleotide region is responsible for increased GLUT1 expression via enhanced GLUT1 mRNA stabilization. Thus, the two counter-regulatory elements and the tumor specific protein expression are responsible for controlling the level of GLUT1 expression. GLUT1 EXPRESSION AND THE RNA BINDING PROTEIN—Hel-N1 Hel-N1 is an mRNA binding protein (68, 69) and is a mammalian homologue of the Drosophila gene ELAV, which is required for differentiation and maintenance of neurons (70, 71). However, the mechanism by which it functions remains unknown. The strongest homology between Hel-N1 and ELAV resides in three RNA recognition motifs (RRMs) (71). Levine et al. (68) have shown using in vitro studies, that Hel-N1 binds preferentially to the 39UTRs of mRNAs which have short stretches of uridylate residues such as those encoding c-myc, c-fos, GM-CSF and Id. Gao et al. (69) defined brain mRNA binding targets of Hel-N1 in vitro, using a combinatorial library representing naturally derived 39UTRs. This approach allowed the identification of potential binding ligands, which included certain tumor markers and growth regulatory proteins. Levine et al. (68) demonstrated that in in vitro assays, Hel-N1 bound to the 39UTR of c-myc while Jain et al. (72) documented binding to an A 1 U rich site in the 39UTR (nucleotides 1920 through 1950) of the GLUT1 message in vitro. Studies with 3T3-L1 adipocytes ectopically expressing Hel-N1 demonstrated an increased (8- to 10-fold) content of GLUT1 protein. The data supported a model where the increase in GLUT1 protein content resulted from increased synthesis due
267
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
to translational (acceleration of the formation of the translation initiation complexes, detected through the use of polysome profiles) and posttranslational (GLUT1 mRNA stability) mechanisms. In consideration of specific function, we hypothesized that Hel-N1 may function by preventing the binding of destabilizing factor(s), perhaps by competing with endogenous factors that normally regulate mRNA translation or stability. This then would give rise to the secondary effects of altered gene expression (52). These studies have provoked speculation on the existence of a Hel-N1 homologue in the adipocytes functioning to control translation and stability of specific mRNAs. GLUT1 AS A LIGAND FOR DEHYDROGENASES There are data to suggest that several catalytically active proteins such as the NAD 1-dependent dehydrogenases possess a secondary activity as RNA-binding proteins (73, 74). Based on these observations, in vitro binding of several dehydrogenases to the GLUT1 39UTR was examined. Studies were performed with commercially obtained dehydrogenases, which were selected for their relevance to glucose metabolism (75). These included: glucose 6-phosphate (G6PDH), lactate (LDH), isocitrate, glyceraldehyde 3-phosphate (G3PDH) and glutamate dehydrogenases. Of this group G6PDH, LDH, and G3PDH bound with high affinity to the full length GLUT1 39UTR. Interestingly, G3PDH bound specifically to an A 1 U rich region localized between positions 1820 and 1950 of the 39UTR. It would appear that the dehydrogenases prefer to bind to A 1 U rich regions of the message (75). While with respect to the data presented above, the binding of dehydrogenases to the GLUT1 39-UTR is somewhat enigmatic, we would argue that these observations prompt consideration of these enzymes as potential regulators of mRNA stability and expand the functional role of dehydrogenases with previously defined roles in glucose metabolism. SUMMARY AND PERSPECTIVES While they are not the exclusive determinants of mRNA stability, cis-elements in the 39UTR play a significant role in dictating the half-life of the message. In a manner similar to promoter-transcription factor interaction, binding of a protein to these elements appears to control the turnover of the message and thus gene expression. From the studies described above, the GLUT1 39UTR contains at least four distinct ciselements, which are, recognized by a minimum of four proteins or protein complexes. The identification of these proteins and the further characterization of their cell biology will shed light on their function in the turnover of mRNA.
ACKNOWLEDGMENTS The authors acknowledge the support of NIH Grant DK55769 and thank Dr. Joseph G. Cory, Ashlie Pruett, Natalie Lombardi, and Ryan O’Neal for their thoughtful comments on the manuscript.
REFERENCES 1. Baldwin, S. (1993) Biochim. Biophys. Acta 1154, 17–19. 2. Olsen, A. L., and Pessin, J. E. (1996) Annu. Rev. Nutrition 16, 235–256. 3. Bell, G. I., Burant, C. F., Takeda, J., and Gould, G. W. (1993) J. Biol. Chem. 268, 19161–19164. 4. McGowan, K. M., Long, S. D., and Pekala, P. H. (1995) Pharmac. Ther. 66, 465–505. 5. Takata, K., and Hirano, O. (1997) Micro. Resh. Tech. 38, 145– 152. 6. Maher, F., Vannucci, S., and Simpson, I. (1994) FASEB J. 8, 1003–1011. 7. McCall, A. L., Millington, W. R., and Wurtman, R. J. (1982) Proc. Natl., Acad. Sci. USA 79, 5406 –5410. 8. Vannucci, S. J., Clark, R. R., Koehler-Stec, E., Li, K., Smith, C. B., Davies, P., Maher, F., and Simpson, I. A. (1998) Dev. Neurosci. 20, 369 –379. 9. Seidner, G., Alvarez, M. G., Yeh, H. I., O’Driscoll, K. R., Klepper, J., Stump, T. S., Wang, D., Spinner, N. B., Birnbaum, M. J., and DeVivo, D. C. (1998) Nature Genet. 18, 188 –191. 10. Longo, N., and Elsas, L. J. (1998) Adv. Ped. 45, 293–313. 11. Haber, R. S., Rathan, A., Weiser, K. R., Pritsker, A., Itzkowitz, S. H., Bodian, C., Slater, G., Weiss, A., and Burstein, D. E. (1998) Cancer 83, 34 – 40. 12. Schoenfeld, A., Davidowitz, E. J. and Burk, R. D. (1998) Proc. Natl. Acad. Sci. USA 95, 8817– 8822. 13. Mooradian, A. D., Chung, H. C., and Shah, G. N. (1997) Neurobiol. Aging 18, 469 – 474. 14. Cornelius, P., Marlowe, M., Lee, M. D., and Pekala, P. H. (1990) J. Biol. Chem. 265, 20506 –20516. 15. McGowan, K. M., Police, S., Winslow, J., and Pekala, P. H. (1997) J. Biol. Chem. 272, 1331–1337. 16. Boado, R., and Pardridge, W. M. (1993) J. Neurochemistry 60, 2290 –2296. 17. Maher, R., and Harrison, L. C. (1990) Biochem. Biophys. Res. Commun. 171, 210 –215. 18. Kumagai, A. K., Kang, Y. S., Boado, R. J., and Pardridge, W. M. (1995) Diabetes 44, 1399 –1404. 19. Pardridge, W. M., Triguero, D., and Farrell, C. R. (1990) Diabetes 39, 1040 –1044. 20. Tsukamoto, H., Boado, R. J., and Pardridge, W. M. (1996) J. Clin. Invest. 97, 2828 –2832. 21. Tsukamoto, H., Boado, R. J., and Pardridge, W. M. (1997) J. Neurochem. 68, 2587–2592. 22. Cornford, E. M., Hayman, S., Black, K. L., Cornford, M. E., Vinters, H. V., and Pardridge, W. M. (1995) J. Neuropathol. Exp. Neurol. 54, 842– 841. 23. Shetty, M., Ismail-Beigi, N., Loeb, J. W., and Ismail-Beigi, F. (1993) Am. J. Physiol. 265, C1224-C1229. 24. Rollins, B. J., Morrison, E. D., Usher, P., and Flier, J. S. (1988) J. Biol. Chem. 263, 16523–16526. 25. Stephens, J., and Pekala, P. H. (1991) J. Biol. Chem. 266, 21834 –21845. 26. Stephens, J., and Pekala, P. H. (1992) J. Biol. Chem. 267, 13580 –13584.
268
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
27. Sandouk, T., Reda, D., and Hoffman, C. (1993) Endocrinology 133, 352–359. 28. White, M., DeVente, J. D., Robbins, P. J., Canupp, D. M., Mayo, M. W., Steelman, L. S., and McCubrey, J. A. (1994) Oncol. Rep. 1, 17–26. 29. Tebbey, P. W., McGowan, K., Stephens, J., Buttke, T., and Pekala, P. H. (1994) J. Biol. Chem. 269, 639 – 644. 30. Long, S. D., and Pekala, P. H. (1996) Biochem. Biophys Res. Commun. 220, 949 –953. 31. Cornelius, R., Marlowe, M., Call, K., and Pekala, P. H. (1991) J. Cell. Physiol. 146, 298 –308. 32. Stephens, J. M., Carter, B. Z., Pekala, P. H., and Malter, J. S. (1992) J. Biol. Chem. 267, 8336 – 8341. 33. Mountjoy, K. J., and Flier, J. S., (1990) Endocrinology 127, 2025–2034. 34. Maher, F., and Harrison, L. C. (1991) Biochim. Biophys. Acta 1089, 27–32. 35. Kozlovsky, N., Rudich, A., Potashnik, R., Ebina, Y., Murakami, T., and Bashan, N. (1997) J. Biol. Chem. 272, 33367–33372. 36. Stein, I., Neeman, M., Shweiki, D., Itin, A., and Keshet, E. (1995) Mol. Cell. Biol. 15. 5363–5368. 37. Bruckner, B. A., Ammini, C. V., Otal, M. P., Raizaka, M. K., and Stacpoole, P. W. (1999) Metabolism, 48, 422– 431. 38. Boado, R. J. (1996) Neurosci. Lett. 220, 53–56. 39. Boado, R. J., and Pardridge, W. M. (1998) Mol. Brain Res. 59, 109 –113. 40. Neufer, P. D., Devente, J. E., Tapscott, E. B., and Dohm, G. L. (1995) Am. J. Physiol. 268, E866-E872. 41. Peltz, S. W., Brewer, G., Bernstein, P., Hart, P. A., and Ross, J. (1991) Crit. Rev. Euk. Gene Express. 1, 99 –126. 42. Belasco, J., and Brawerman, G. (1993) in Control of mRNA Stability (Belasco, J. and Brawerman, G., Eds.), pp. 149 –159. Academic Press, San Diego. 43. Ross, J. (1996) Trends Genet. 12, 171–175. 44. Sachs, A. B. (1993) Cell 74, 413– 421. 45. Decker, C. J., and Parker, R. (1994) TIBS 19, 336 –340. 46. Jacobson, A., and Peltz, S. W. (1996) Annu. Rev. Biochem. 65, 693–739. 47. Mitchell, P., Petfalski, E., Shevhenko, A., Mann, M., and Tollervery, D. (1997) Cell 91, 457– 466. 48. Jacob Anderson, J. S., and Parker, R. (19880 EMBO J. 17, 1497–1506. 49. Brewer, G. (1999) J. Biol. Chem. 274, 16174 –16179. 50. Beelman, C. A. and Parker, R. (1995) Cell 81, 179 –183. 51. Klausner, R. D., Rouault, T., and Harford, J. B. (1993) Cell 72, 19 –28.
52. Jain, R. G., Andrews, L. G., McGowan, K. M., Pekala, P. H., and Keene, J. D. (1997) Mol. Cell Bio. 17, 954 –962. 53. Antic, D., Lu, N., and Keene, J. D. (1999) Genes Dev. 13, 449 – 461. 54. Laroia, G., Cuesta, R., Brewer, G., and Schneider, R. J. (1999) Science 284, 499 –502. 55. DeMaria, C. T., and Brewer, G. (1996) J. Biol. Chem. 271, 12179 –12184. 56. Sirenko, O. I., Lofquist, A. K., Demaria, C. T., Morris, J. S., Brewer, G., and Haskill, J. S. (1997) Mol. Cell Biol. 17, 3898 – 3906. 57. Savant-Bhonsale, S., and Cleveland, D. W. (1992) Genes Dev. 6, 1927–1939. 58. Keene, J. D. (1999) Proc. Natl. Acad. Sci. USA 96, 5–7. 59. Fan, X. C., and Steitz, J. A. (1998) Proc. Natl. Acad. Sci. USA 95, 15293–15298. 60. Schwartz, S. P., Aisenthal, L., Elisha, Z., Oberman, F., and Yiseraeli, J. K. (1991) Proc. Natl. Acad. Sci. USA 89, 11895– 11899. 61. Minvielle-Sebastia, L., Beyer, K., Krecic, A. M., Hector, R. E., Swanson, M. S., and Keller, W. (1998) EMBO J. 17, 7454 –7468. 62. Wilusz, J., Whenk, T., Takagaki, Y., and Manley, (1990) Mol. Cell Biol. 10, 1244 –1248. 63. Weiss, E. A., Gilmartin, G., and Nevins, J. R. (1991) EMBO J. 10, 215–219. 64. Wahle, E., and Keller, W. Annu. Rev. Biochem. 61, 419 – 440. 65. Scotto, L., and Assoian, R. K. (1993) Mol. Cell. Biol. 13, 3588 –3597. 66. Pryhuber, G. S., Church, S. L., Kroft, T., Panchal, A., and Whitset, J. A. (1994) Am. J. Physiol. 267, L16 –L24. 67. Dwyer, K. J., Boado, R. J., and Pardridge, W. M. (1996) J. Neurochem. 66, 449 – 458. 68. Levine, T. D., Gao, F. B., King, P. H., Andrews, L. G., and Keene, J. D. (1993) Mol. Cell Biol. 13, 3494 –3504. 69. Gao, F., Carson, C., Levine, T., and Keene, J. (1994) Proc. Natl. Acad. Sci. USA 91, 11207–11211. 70. Yao, K. M., Samson, M. L., Reeves, R., and White, K. (1993) J. Neurobiology 24 723–739. 71. King, P. H., Levine, T. D., Fremeau, R. T., Jr., and Keene, J. D. (1994) J. Neurosci. 14, 1943–1952. 72. Jain, R. G., Andrews, L. G., McGowan, K. M., Gao, F., Keene, J. D., and Pekala, P. H.(1996) Nucleic Acids Sympos. Ser. 33, 209 –211. 73. Nagy, E., and Rigby, W. F. (1995) J. Biol. Chem. 270, 2755–2763. 74. Hentze, M. W. (1994) TIBS 19, 101–103. 75. McGowan, K. M., and Pekala, P. H. (1996) Biochem. Biophys. Res. Commun. 221, 42– 45.
269
BREAKTHROUGHS AND VIEWS The Influence of mRNA Stability on Glucose Transporter (GLUT1) Gene Expression Chen Qi and Phillip H. Pekala 1 Department of Biochemistry, School of Medicine, East Carolina University, Greenville, North Carolina 27858-4354
Received August 10, 1999
One mechanism for modification of glucose transport activity occurs through regulation of the cellular content of transporter protein by alteration of transcript stability. Regulated mRNA decay has been shown to play an important role in control of posttranscriptional gene expression. Implicated, as a pivotal element in this regulation is the 3*-untranslated region (UTR) of the message. Recent work from several labs has focused on sequence motifs within the 3*-UTR of glucose transporter (GLUT1) mRNA that serve as destabilizing or stabilizing elements and recognition of these elements by specific proteins. In this review, we address several critical studies each of which has identified elements in the GLUT1 3*-UTR that are involved in the control of transcript stability and demonstrated that these sequence motifs are recognized by specific binding proteins. © 1999 Academic Press Key Words: mRNA stability; GLUT1; 3*-untranslated region; RNA binding proteins.
The facilitated diffusion of glucose across the plasma membrane is the rate-limiting step for subsequent glucose metabolism and energy production within the cell (1, 2). A family of tissue-specific integral membrane proteins known as the glucose transporters (GLUTs) (3) catalyzes this process. Transporter activity can be regulated by hormones, growth factors and metabolites through the redistribution of transporter proteins from intracellular vesicular storage to the plasma membrane (translocation), modulation of transporter intrinsic activity and stimulation of new protein synthesis through increased transcription and/or alteration in transcript stability (4). Our work, in large part, has focused on control of the transcription of the glucose transporter genes. During the course of these studies it 1
To whom correspondence should be addressed. Fax: (252) 8163383. E-mail: [email protected].
became readily apparent that most agents affecting transporter transcription also modulated transporter mRNA stability. With this consideration, our review will focus on regulated mRNA stability as a vehicle for control of transporter gene expression, with particular attention to GLUT1, the ubiquitously expressed, homeostatic glucose transporter. GLUT1 EXPRESSION IN HEALTH AND DISEASE The GLUT1 protein is expressed at high levels in all fetal tissues, while in the adult it is expressed ubiquitously but predominantly at high levels within several biologically important tissues in cell types to include, fibroblasts, endothelial cells and erythrocytes (2). Major localization also includes the endothelial cells at blood-tissue barriers such as placenta and the bloodbrain barrier (5–7). GLUT1 is also the primary transporter in the choroid plexus, ependyma and glia (8). Seidner et al. recently provided the first identification of a molecular defect underlying dysfunctional bloodbrain barrier glucose transport resulting in disease (9). Mutations in one allele of the GLUT1 gene leading to truncation of the GLUT1 protein result in decreased availability of glucose to the brain. While these individuals have one functional copy of the GLUT1 gene, the phenotype is severe, characterized by infantile seizures, delayed development and acquired microcephaly (9). The metabolic importance of the GLUT1 transporter is further supported by the failure to obtain mice homozygous for an inactivated GLUT1 gene (10). Altered expression of GLUT1 has been correlated with several disease states. Increased GLUT1 expression has been associated strongly with neoplastic progression in the colon (11), as well as both sporadic and inherited clear cell renal carcinoma associated with von Hippel-Lindau disease (12). In contrast, patients with Alzheimer’s disease exhibited a reduced GLUT1 protein content at the blood brain barrier (13).
265
0006-291X/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS TABLE 1
Regulation of GLUT1 mRNA Stability Cell type/tissue Mouse fibroblasts BALB/c 3T3 3T3-L1
NIH 3T3-L1 Mouse adipocytes 3T3 F442A
3T3-L1 Rat muscle cells L6 myocytes
L6 myotubes Brain and neural cells Bovine brain endothelial cells Glioma Rat neuronal and astroglia cells ECL cells C6 Glioma Muscle tissue Rat hindlimb Red skeletal muscle White skeletal muscle
Agent
PDGF TNF-a cAMP Phorbol ester Okadaic acid Vanadate Insulin Piogtilazone Insulin 1 Pioglitazone Arachidonic acid
Fold increase in half-life
Reference
4 3 6 6 5.6 2–3
(24) (4, 31, 32)
2.5 1.6 .10 .2
(33) (27)
(29)
Insulin/IGF-1 Glucose deprivation Insulin 1 Glucose deprivation Tunicamycin 2-deoxyglucose Glucosamine Oxidative stress
2 2 4 2 2 3
(17, 34)
Glucose deprivation VEGF Hypoxia Cerebrolysin Actinomycin D
2 4.8 1.2 1.7 2.3
(16) (36) (37) (38) (39)
Perfusion Perfusion
3.5 2.5
(40)
Regulation of mRNA stability plays a major role in control of GLUT1 gene expression (4, 14, 15). Several laboratories have reported alteration of the stability of the GLUT1 transcript by conditions to include glucose deprivation (16 –18), experimental diabetes (19), hemangioblastoma (20 –22), inhibition of oxidative phosphorylation (23) as well as cytokine, hormone and metabolite stimulation (24, 17, 14, 25–30). As demonstrated in Table 1, a variety of unrelated hormones, pharmaceuticals and physiological situations have been demonstrated to alter the stability of the GLUT1 transcript in various cell lines and tissues. In terms of structure, the GLUT1 transcript is 2544 nucleotides in length with a 39-UTR of 884 nucleotides. The 39-UTR is approximately 56% A 1 U overall but regions of 150 to 200 bases exist where the A 1 U content is nearly 70%. In this context, the 39-UTR would be considered A 1 U rich. GENERAL ASPECTS OF mRNA TURNOVER In order that the reader may appreciate the studies that have addressed GLUT1 mRNA stability detailed in the following sections, we will first briefly examine a few general principles of RNA turnover.
(35)
Messenger RNA decay plays an important role in control of posttranscriptional gene expression (41– 43). Numerous studies, predominantly in the yeast Saccharomyces cerevisiae, detailing the mRNA decay characteristics for a variety of transcripts have supported the existence of several pathways for the degradation of eukaryotic mRNAs (43– 45). A decay pathway in which turnover of eukaryotic mRNA is initiated by shortening of the poly (A) tail, followed by decapping and subsequent 59 to 39 exonucleolytic degradation of the message appears to predominate (reviewed in ref. 46). Additionally a 39 to 59 pathway has been described where decay of the poly (A) tail continues after the initial phase of deadenylation followed by degradation of the 39-untranslated region (39-UTR) (47, 48). Recent evidence (49) suggests that the 39 to 59 pathway exists and may predominate in mammalian cells. While not meant to represent an exclusive list of potential pathways of mRNA decay, the 39-UTR (including the poly (A) tail) is implicated as being a pivotal element for the regulation of mRNA decay in both models (50). RNA-binding proteins that recognize specific consensus sequences (particularly in the 39-UTR) are also thought to play a role in the degradation process. The identification and characterization of these proteins is
266
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
important in defining the mechanisms involved not only in control of mRNA turnover but in posttranscriptional regulation in general. RNA-protein interactions have been shown to influence many processes including: translation (51–53), RNA stability (21, 43, 52, 54 –57), mRNA transport and localization (58 – 60), splicing (61– 63) and polyadenylation (64). We will now turn our attention to several instances where protein binding has influenced the stability of the GLUT1 transcript. GLUT1 EXPRESSION AND TUMOR NECROSIS FACTOR-a-INDUCED RNA BINDING PROTEINS Treatment of 3T3-L1 fibroblasts with the cytokine tumor necrosis factor-a (TNF) produced a 3 to 4-fold increase in the half-life of the GLUT1 mRNA. The increased GLUT1 mRNA content was demonstrated to result in the synthesis and plasma membrane localization of new GLUT1 protein which lead to increased glucose transport activity (8- to 10-fold over basal) (14). Coincident with the stabilization of the message was the up-regulation of RNA binding proteins of molecular weights 37 and 40 kDa with apparent binding specificity for the GLUT1 39-UTR (32). Further studies demonstrated that: 1. the GLUT1 39-UTR was a destabilizing UTR; 2. the destabilizing element mapped to a portion of the GLUT1 39-UTR located between bases 2242 and 2347; and 3. this region mediated the stabilizing response to TNF. This region is predominantly GC-rich ('60%) and was not homologous to either a GC-rich stability element reported for the TGF-b1 gene (65) or to a TNF-responsive stability element identified within the surfactant-B protein mRNA (67). Deletion of this region confirmed that it mediated both the rapid decay of the GLUT1 message as well as GLUT1 mRNA accumulation (stabilization) in response to TNF (15). Using RNA gel-mobility shift assays, the two TNFinducible RNA-binding proteins were identified as ligands for the region between nucleotides 2240 and 2347 of the 39-UTR. The RNA-binding activity of these proteins was observed to increase coordinately with the stabilization of the GLUT1 message following TNF stimulation. When the 2240 through 2347 region of the 39-UTR was deleted, coincident with the loss of protein binding, the message was stabilized and the response to TNF was lost. These observations demonstrate that the instability element, the TNF-response element and protein binding are all localized to the 106-nucleotide motif. While these elements may be separable within the motif, the data are consistent with congruence between or among them. GLUT1 EXPRESSION AND BRAIN-DERIVED RNA BINDING PROTEINS Pardridge and colleagues have examined GLUT1 expression in human brain tumors and have described
two cis-elements in the 39-UTR of the GLUT1 mRNA that bind protein and regulate expression in an apparent tumor specific fashion (20, 21, 39, 67). The first element is A 1 U rich and localized between nucleotides 1885 and 1906 that binds a protein of 44 kDa and suggested to be an instability element (20, 21, 39, 67). The second element resides between nucleotides 21862203 and binds a protein of 80 kDa. The 44 kDa complex was selectively down regulated in hemangioblastoma (a brain tumor that overexpresses GLUT1) as compared to glioblastoma multiform. While the 80 kDa protein was selectively expressed in hemangioblastoma (20, 21, 39, 67). The enhancer role of the region between nucleotides 2100 to 2300 was confirmed by subcloning the region into a luciferase reporter construct which was then transfected into C6 rat glioma cells. Luciferase expression increased fivefold relative to controls and a 228% increase in the message half-life was observed (21, 39, 67). Deletion of nucleotides 21812190, the putative binding site for the 80 kDa protein, eliminated the enhancement of the luciferase reporter (21, 39, 67). These data support the hypothesis that the 2181-2190-nucleotide region is responsible for increased GLUT1 expression via enhanced GLUT1 mRNA stabilization. Thus, the two counter-regulatory elements and the tumor specific protein expression are responsible for controlling the level of GLUT1 expression. GLUT1 EXPRESSION AND THE RNA BINDING PROTEIN—Hel-N1 Hel-N1 is an mRNA binding protein (68, 69) and is a mammalian homologue of the Drosophila gene ELAV, which is required for differentiation and maintenance of neurons (70, 71). However, the mechanism by which it functions remains unknown. The strongest homology between Hel-N1 and ELAV resides in three RNA recognition motifs (RRMs) (71). Levine et al. (68) have shown using in vitro studies, that Hel-N1 binds preferentially to the 39UTRs of mRNAs which have short stretches of uridylate residues such as those encoding c-myc, c-fos, GM-CSF and Id. Gao et al. (69) defined brain mRNA binding targets of Hel-N1 in vitro, using a combinatorial library representing naturally derived 39UTRs. This approach allowed the identification of potential binding ligands, which included certain tumor markers and growth regulatory proteins. Levine et al. (68) demonstrated that in in vitro assays, Hel-N1 bound to the 39UTR of c-myc while Jain et al. (72) documented binding to an A 1 U rich site in the 39UTR (nucleotides 1920 through 1950) of the GLUT1 message in vitro. Studies with 3T3-L1 adipocytes ectopically expressing Hel-N1 demonstrated an increased (8- to 10-fold) content of GLUT1 protein. The data supported a model where the increase in GLUT1 protein content resulted from increased synthesis due
267
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
to translational (acceleration of the formation of the translation initiation complexes, detected through the use of polysome profiles) and posttranslational (GLUT1 mRNA stability) mechanisms. In consideration of specific function, we hypothesized that Hel-N1 may function by preventing the binding of destabilizing factor(s), perhaps by competing with endogenous factors that normally regulate mRNA translation or stability. This then would give rise to the secondary effects of altered gene expression (52). These studies have provoked speculation on the existence of a Hel-N1 homologue in the adipocytes functioning to control translation and stability of specific mRNAs. GLUT1 AS A LIGAND FOR DEHYDROGENASES There are data to suggest that several catalytically active proteins such as the NAD 1-dependent dehydrogenases possess a secondary activity as RNA-binding proteins (73, 74). Based on these observations, in vitro binding of several dehydrogenases to the GLUT1 39UTR was examined. Studies were performed with commercially obtained dehydrogenases, which were selected for their relevance to glucose metabolism (75). These included: glucose 6-phosphate (G6PDH), lactate (LDH), isocitrate, glyceraldehyde 3-phosphate (G3PDH) and glutamate dehydrogenases. Of this group G6PDH, LDH, and G3PDH bound with high affinity to the full length GLUT1 39UTR. Interestingly, G3PDH bound specifically to an A 1 U rich region localized between positions 1820 and 1950 of the 39UTR. It would appear that the dehydrogenases prefer to bind to A 1 U rich regions of the message (75). While with respect to the data presented above, the binding of dehydrogenases to the GLUT1 39-UTR is somewhat enigmatic, we would argue that these observations prompt consideration of these enzymes as potential regulators of mRNA stability and expand the functional role of dehydrogenases with previously defined roles in glucose metabolism. SUMMARY AND PERSPECTIVES While they are not the exclusive determinants of mRNA stability, cis-elements in the 39UTR play a significant role in dictating the half-life of the message. In a manner similar to promoter-transcription factor interaction, binding of a protein to these elements appears to control the turnover of the message and thus gene expression. From the studies described above, the GLUT1 39UTR contains at least four distinct ciselements, which are, recognized by a minimum of four proteins or protein complexes. The identification of these proteins and the further characterization of their cell biology will shed light on their function in the turnover of mRNA.
ACKNOWLEDGMENTS The authors acknowledge the support of NIH Grant DK55769 and thank Dr. Joseph G. Cory, Ashlie Pruett, Natalie Lombardi, and Ryan O’Neal for their thoughtful comments on the manuscript.
REFERENCES 1. Baldwin, S. (1993) Biochim. Biophys. Acta 1154, 17–19. 2. Olsen, A. L., and Pessin, J. E. (1996) Annu. Rev. Nutrition 16, 235–256. 3. Bell, G. I., Burant, C. F., Takeda, J., and Gould, G. W. (1993) J. Biol. Chem. 268, 19161–19164. 4. McGowan, K. M., Long, S. D., and Pekala, P. H. (1995) Pharmac. Ther. 66, 465–505. 5. Takata, K., and Hirano, O. (1997) Micro. Resh. Tech. 38, 145– 152. 6. Maher, F., Vannucci, S., and Simpson, I. (1994) FASEB J. 8, 1003–1011. 7. McCall, A. L., Millington, W. R., and Wurtman, R. J. (1982) Proc. Natl., Acad. Sci. USA 79, 5406 –5410. 8. Vannucci, S. J., Clark, R. R., Koehler-Stec, E., Li, K., Smith, C. B., Davies, P., Maher, F., and Simpson, I. A. (1998) Dev. Neurosci. 20, 369 –379. 9. Seidner, G., Alvarez, M. G., Yeh, H. I., O’Driscoll, K. R., Klepper, J., Stump, T. S., Wang, D., Spinner, N. B., Birnbaum, M. J., and DeVivo, D. C. (1998) Nature Genet. 18, 188 –191. 10. Longo, N., and Elsas, L. J. (1998) Adv. Ped. 45, 293–313. 11. Haber, R. S., Rathan, A., Weiser, K. R., Pritsker, A., Itzkowitz, S. H., Bodian, C., Slater, G., Weiss, A., and Burstein, D. E. (1998) Cancer 83, 34 – 40. 12. Schoenfeld, A., Davidowitz, E. J. and Burk, R. D. (1998) Proc. Natl. Acad. Sci. USA 95, 8817– 8822. 13. Mooradian, A. D., Chung, H. C., and Shah, G. N. (1997) Neurobiol. Aging 18, 469 – 474. 14. Cornelius, P., Marlowe, M., Lee, M. D., and Pekala, P. H. (1990) J. Biol. Chem. 265, 20506 –20516. 15. McGowan, K. M., Police, S., Winslow, J., and Pekala, P. H. (1997) J. Biol. Chem. 272, 1331–1337. 16. Boado, R., and Pardridge, W. M. (1993) J. Neurochemistry 60, 2290 –2296. 17. Maher, R., and Harrison, L. C. (1990) Biochem. Biophys. Res. Commun. 171, 210 –215. 18. Kumagai, A. K., Kang, Y. S., Boado, R. J., and Pardridge, W. M. (1995) Diabetes 44, 1399 –1404. 19. Pardridge, W. M., Triguero, D., and Farrell, C. R. (1990) Diabetes 39, 1040 –1044. 20. Tsukamoto, H., Boado, R. J., and Pardridge, W. M. (1996) J. Clin. Invest. 97, 2828 –2832. 21. Tsukamoto, H., Boado, R. J., and Pardridge, W. M. (1997) J. Neurochem. 68, 2587–2592. 22. Cornford, E. M., Hayman, S., Black, K. L., Cornford, M. E., Vinters, H. V., and Pardridge, W. M. (1995) J. Neuropathol. Exp. Neurol. 54, 842– 841. 23. Shetty, M., Ismail-Beigi, N., Loeb, J. W., and Ismail-Beigi, F. (1993) Am. J. Physiol. 265, C1224-C1229. 24. Rollins, B. J., Morrison, E. D., Usher, P., and Flier, J. S. (1988) J. Biol. Chem. 263, 16523–16526. 25. Stephens, J., and Pekala, P. H. (1991) J. Biol. Chem. 266, 21834 –21845. 26. Stephens, J., and Pekala, P. H. (1992) J. Biol. Chem. 267, 13580 –13584.
268
Vol. 263, No. 2, 1999
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
27. Sandouk, T., Reda, D., and Hoffman, C. (1993) Endocrinology 133, 352–359. 28. White, M., DeVente, J. D., Robbins, P. J., Canupp, D. M., Mayo, M. W., Steelman, L. S., and McCubrey, J. A. (1994) Oncol. Rep. 1, 17–26. 29. Tebbey, P. W., McGowan, K., Stephens, J., Buttke, T., and Pekala, P. H. (1994) J. Biol. Chem. 269, 639 – 644. 30. Long, S. D., and Pekala, P. H. (1996) Biochem. Biophys Res. Commun. 220, 949 –953. 31. Cornelius, R., Marlowe, M., Call, K., and Pekala, P. H. (1991) J. Cell. Physiol. 146, 298 –308. 32. Stephens, J. M., Carter, B. Z., Pekala, P. H., and Malter, J. S. (1992) J. Biol. Chem. 267, 8336 – 8341. 33. Mountjoy, K. J., and Flier, J. S., (1990) Endocrinology 127, 2025–2034. 34. Maher, F., and Harrison, L. C. (1991) Biochim. Biophys. Acta 1089, 27–32. 35. Kozlovsky, N., Rudich, A., Potashnik, R., Ebina, Y., Murakami, T., and Bashan, N. (1997) J. Biol. Chem. 272, 33367–33372. 36. Stein, I., Neeman, M., Shweiki, D., Itin, A., and Keshet, E. (1995) Mol. Cell. Biol. 15. 5363–5368. 37. Bruckner, B. A., Ammini, C. V., Otal, M. P., Raizaka, M. K., and Stacpoole, P. W. (1999) Metabolism, 48, 422– 431. 38. Boado, R. J. (1996) Neurosci. Lett. 220, 53–56. 39. Boado, R. J., and Pardridge, W. M. (1998) Mol. Brain Res. 59, 109 –113. 40. Neufer, P. D., Devente, J. E., Tapscott, E. B., and Dohm, G. L. (1995) Am. J. Physiol. 268, E866-E872. 41. Peltz, S. W., Brewer, G., Bernstein, P., Hart, P. A., and Ross, J. (1991) Crit. Rev. Euk. Gene Express. 1, 99 –126. 42. Belasco, J., and Brawerman, G. (1993) in Control of mRNA Stability (Belasco, J. and Brawerman, G., Eds.), pp. 149 –159. Academic Press, San Diego. 43. Ross, J. (1996) Trends Genet. 12, 171–175. 44. Sachs, A. B. (1993) Cell 74, 413– 421. 45. Decker, C. J., and Parker, R. (1994) TIBS 19, 336 –340. 46. Jacobson, A., and Peltz, S. W. (1996) Annu. Rev. Biochem. 65, 693–739. 47. Mitchell, P., Petfalski, E., Shevhenko, A., Mann, M., and Tollervery, D. (1997) Cell 91, 457– 466. 48. Jacob Anderson, J. S., and Parker, R. (19880 EMBO J. 17, 1497–1506. 49. Brewer, G. (1999) J. Biol. Chem. 274, 16174 –16179. 50. Beelman, C. A. and Parker, R. (1995) Cell 81, 179 –183. 51. Klausner, R. D., Rouault, T., and Harford, J. B. (1993) Cell 72, 19 –28.
52. Jain, R. G., Andrews, L. G., McGowan, K. M., Pekala, P. H., and Keene, J. D. (1997) Mol. Cell Bio. 17, 954 –962. 53. Antic, D., Lu, N., and Keene, J. D. (1999) Genes Dev. 13, 449 – 461. 54. Laroia, G., Cuesta, R., Brewer, G., and Schneider, R. J. (1999) Science 284, 499 –502. 55. DeMaria, C. T., and Brewer, G. (1996) J. Biol. Chem. 271, 12179 –12184. 56. Sirenko, O. I., Lofquist, A. K., Demaria, C. T., Morris, J. S., Brewer, G., and Haskill, J. S. (1997) Mol. Cell Biol. 17, 3898 – 3906. 57. Savant-Bhonsale, S., and Cleveland, D. W. (1992) Genes Dev. 6, 1927–1939. 58. Keene, J. D. (1999) Proc. Natl. Acad. Sci. USA 96, 5–7. 59. Fan, X. C., and Steitz, J. A. (1998) Proc. Natl. Acad. Sci. USA 95, 15293–15298. 60. Schwartz, S. P., Aisenthal, L., Elisha, Z., Oberman, F., and Yiseraeli, J. K. (1991) Proc. Natl. Acad. Sci. USA 89, 11895– 11899. 61. Minvielle-Sebastia, L., Beyer, K., Krecic, A. M., Hector, R. E., Swanson, M. S., and Keller, W. (1998) EMBO J. 17, 7454 –7468. 62. Wilusz, J., Whenk, T., Takagaki, Y., and Manley, (1990) Mol. Cell Biol. 10, 1244 –1248. 63. Weiss, E. A., Gilmartin, G., and Nevins, J. R. (1991) EMBO J. 10, 215–219. 64. Wahle, E., and Keller, W. Annu. Rev. Biochem. 61, 419 – 440. 65. Scotto, L., and Assoian, R. K. (1993) Mol. Cell. Biol. 13, 3588 –3597. 66. Pryhuber, G. S., Church, S. L., Kroft, T., Panchal, A., and Whitset, J. A. (1994) Am. J. Physiol. 267, L16 –L24. 67. Dwyer, K. J., Boado, R. J., and Pardridge, W. M. (1996) J. Neurochem. 66, 449 – 458. 68. Levine, T. D., Gao, F. B., King, P. H., Andrews, L. G., and Keene, J. D. (1993) Mol. Cell Biol. 13, 3494 –3504. 69. Gao, F., Carson, C., Levine, T., and Keene, J. (1994) Proc. Natl. Acad. Sci. USA 91, 11207–11211. 70. Yao, K. M., Samson, M. L., Reeves, R., and White, K. (1993) J. Neurobiology 24 723–739. 71. King, P. H., Levine, T. D., Fremeau, R. T., Jr., and Keene, J. D. (1994) J. Neurosci. 14, 1943–1952. 72. Jain, R. G., Andrews, L. G., McGowan, K. M., Gao, F., Keene, J. D., and Pekala, P. H.(1996) Nucleic Acids Sympos. Ser. 33, 209 –211. 73. Nagy, E., and Rigby, W. F. (1995) J. Biol. Chem. 270, 2755–2763. 74. Hentze, M. W. (1994) TIBS 19, 101–103. 75. McGowan, K. M., and Pekala, P. H. (1996) Biochem. Biophys. Res. Commun. 221, 42– 45.
269