Cloning genomic INGAP: a Reg-related family member with distinct transcriptional regulation sites

Biochimica et Biophysica Acta 1638 (2003) 83 – 89 www.bba-direct.com

Short sequence-paper

Cloning genomic INGAP: a Reg-related family member with distinct $ transcriptional regulation sites David A. Taylor-Fishwick a,b,*, Sharon Rittman a, Hidayah Kendall a, Lipika Roy a, Wenjing Shi a, Yong Cao a, Gary L. Pittenger a,c, Aaron I. Vinik a,c a

Department of Medicine, The Leonard Strelitz Diabetes Institutes, Eastern Virginia Medical School, 855 W. Brambleton Avenue, Norfolk, VA 23510, USA b Department of Microbiology and Molecular Cell Biology, Eastern Virginia Medical School, 855 W. Brambleton Avenue, Norfolk, VA 23510, USA c Department of Anatomy and Neurobiology, Eastern Virginia Medical School, 855 W. Brambleton Avenue, Norfolk, VA 23510, USA Received 12 December 2002; received in revised form 17 March 2003; accepted 21 March 2003

Abstract The protein product of hamster islet neogenesis-associated protein (INGAP) cDNA induces new pancreatic islet development. Manipulation of this process provides a new therapeutic strategy for the treatment of diabetes. As regulators of INGAP gene expression are unknown over 6 kb of hamster genomic INGAP has been cloned. Sequence analysis has identified a 3 kb 5-prime region with core promoter elements that is rich in transcription factor binding sites and six exons for the coding region. Analysis of promoter activity reveals stimulusresponsive DNA elements which have been identified though deletion analysis. Comparison of transcription factor binding sites in INGAP to the related gene RegIIIy exposes potential sites for differential gene regulation. D 2003 Elsevier Science B.V. All rights reserved. Keywords: Diabetes; Islet; INGAP; Neogenesis; Pancreas; Gene expression; PMA

1. Introduction While distinct in etiology, the two major forms of diabetes mellitus ultimately result from a failure of the insulin-producing pancreatic beta cell mass [1]. In type one diabetes, the beta cells undergo autoimmune destruction [2], whereas type two diabetes is associated with severe insulin resistance [3], although diabetes does not ensue until there is a failure or loss of the beta cell mass [4]. A clear rationale to address the biology of diabetes is replacement of the destroyed or depleted beta cell mass with new fully functional islets [5]. In recent years, numerous islet re-population approaches have been targeted towards this challenge. The feasibility of islet transplantation has been proven in humans [6]. However, the practicality of this and other exogenous islet-replacement approaches as a therapy for diabetes is currently restricted both by the $ The sequence of genomic INGAP has been registered with GenBank; Accession #AY184211. * Corresponding author. Tel.: +1-757-446-5912; fax: +1-757-4465868. E-mail address: [email protected] (D.A. Taylor-Fishwick).

limited availability of suitable donor pancreases and by the need to induce long-term immunological tolerance [7,8]. In a distinct approach, our group has focused on identifying proteins or factors that ‘‘re-educate’’ the pancreas to produce the desired new functional beta cell mass endogenously. One such protein has been termed INGAP or islet neogenesis-associated protein [9]. Expression of INGAP is regulated in a pancreas-specific manner [9]. INGAP or a bioactive peptide derived from INGAP has been shown in animals to both initiate new islet cell growth and reverse chemical-induced diabetes [10 – 12]. Since INGAP is an endogenous protein that stimulates the production of ‘‘self’’ beta cells, therapy mediated through INGAP is expected to bypass the immunological challenges encountered in other re-population strategies. While the utility of INGAP or derived peptide in human diabetes is as yet unknown, the phenomenon of islet cell regeneration is open to exploitation. INGAP is a factor that is detected in the earliest stages of pancreas organogenesis [13] and its expression in the adult pancreas appears to initiate a primordial cascade resulting in new islet cell development from a precursor cell population residing in the pancreatic duct [10,12,14]. Knowledge of which pathways regulate

0925-4439/03/$ - see front matter D 2003 Elsevier Science B.V. All rights reserved. doi:10.1016/S0925-4439(03)00042-5

84

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89

INGAP expression and the proteins that switch INGAP on or indeed keep INGAP turned off is vital to our ability to exploit this cascade. The identification of the INGAP genomic sequence including the INGAP promoter reported here will ultimately lead to answers to these important questions. 1.1. Genomic cloning and sequencing of the hamster INGAP promoter region Genomic DNA was isolated from Syrian Golden hamster liver and used to construct four restriction libraries which were prepared using a universal genome walking kit [15] (Clontech, Palo Alto, CA). The library construction and genome walks were carried out following the manufacturers instructions. Starting with antisense gene-specific primers (GSPs), INGEN1_3 and INGEN2_3 (located 79 –105 and 48 –74, respectively, in INGAP cDNA), a series of sequential 5-prime genome walks were performed to build up the described sequence (Fig. 1B). The sense primers were the adapter primers (AP-1 and AP-2) supplied in the kit

(Clontech). Additional antisense primers for the subsequent genome walks were designed GSPs (Fig. 1C and D). The buffer conditions for each genome walk were optimized by performing the PCR reaction with GC-MELT (Clontech) at 0, 0.5, 1 or 1.5 mM to allow amplification of GC-rich DNA regions of the genomic DNA. For each genome walk, the primary PCR was carried out using the AP-1/GSP-1 primer pair (initial 94 jC, 1 min; 7 cycles 94 jC, 2 s, 68 jC, 3 min; 32 cycles 94 jC, 2 s, 63 jC, 3 min; 1 cycle 63 jC, 4 min). The secondary PCR was carried out using the AP-2/GSP-2 primer pair (initial 94 jC, 1 min; 5 cycles 94 jC, 2 s, 68 jC, 3 min; 20 cycles 94 jC, 2 s, 63 jC, 3 min; 1 cycle 63 jC, 4 min). PCR amplified bands which were unique to reactions including both sense and antisense primers were isolated from agarose gels and column purified (Qiagen, Valencia, CA); being either sequenced directly or subcloned into pGEM-T (Promega, Madison, WI) before sequencing. DNA sequencing was performed by DSSF, Iowa State University. Six genome walks were performed each on four different restriction libraries. Genome walk fragments always overlapped and a 3-prime genome walk into INGAP

Fig. 1. Cloning of genomic hamster INGAP has identified five introns within the coding sequence (A). The genomic cloning strategy, illustrated in B, shows the positions of the resolved genomic fragments from genome walks (circles to arrowhead) and genomic PCR (diamond to diamond). The positions of the TATA box and poly A signal are labeled as are the restriction sites (Top) used to make the promoter – reporter constructs. The location in the gene of the primers used (C) and their sequences (D) are also shown.

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89

with sense GSPs INGEN1_2 and INGEN 2_2, respectively confirmed the sequence was from the 5-prime of the INGAP gene (Fig. 1B). Fragments obtained from the genome walks were aligned and pieced together at overlapping regions using sequence analysis software Omiga 2.0 (Oxford Molecular, Burlington, MA). The resolved 5-prime region of INGAP was analyzed for conserved promoter motifs and transcription factor binding sites (www.Genomatix.com). The results of these searches identified both a minimal promoter and several binding sites for known transcription factors. A TATA-box, CCAAT-like box and GC-rich motif were identified in a topology consistent with mammalian promoters. The transcriptional start site was identified to be at position 3123 of the INGAP genomic sequence, a threebase modification which has been previously reported [9]. Additionally, predicted binding sites for transcription factors such as AP-1, NFnB, NFAT, STAT, MyoD and PDX are clustered throughout the 5-prime sequence. These data provide an invaluable point from which to determine the transcription factors and promoter elements that are crucial for the regulation of INGAP gene expression. 1.2. Genomic cloning of hamster INGAP coding region To resolve the genomic sequence for the INGAP coding region, nested genomic PCR was performed on hamster liver genomic DNA. Sense/antisense primers INGAP 1_1L/R and INGAP 2_1L/R (Fig. 1C and D) were reacted in primary (1_1L/R) and secondary (2_1L/R) PCRs using reaction conditions described above for the genome walk. Two distinct PCR products were obtained, the primary PCR product was approximately 2.5 kbp, whereas the secondary PCR product was approximately 1.5 kbp suggesting the nested primers spanned an exon/ intron boundary. Sequencing of the primary and secondary PCR products confirmed this observation and alignment with INGAP cDNA identified the location of the exon/ intron boundaries. Resolution of the INGAP coding region was achieved by a 3-prime genomic walk using the sense GSPs INGEN13_5 and INGEN14_5 with adapter primers (AP-1/AP-2), respectively. Over 6 kb of the hamster INGAP gene has been resolved including 3 kb of the 5-prime regulatory region. Comparison of INGAP genomic DNA with INGAP cDNA revealed six exons generated by five intron insertions. The five introns depicted in Fig. 1A are located at positions 3150 –3426, 3508 – 4442, 4562 – 4735, 4874 – 5459, and 5587 – 5843 respectively in the INGAP genomic sequence. All splicing sites conform to the GT/AG rule for exon/intron boundaries. 1.3. INGAP promoter- -galactosidase reporter deletion constructs To construct sequential deletions for the upstream region of INGAP, the cloned fragments from the genome walks

85

were serially ligated together using enzymes unique to the entire INGAP 5-prime region. Initially fragment 2_3 was PCR amplified to introduce the additional restriction sites, HindIII (3-prime) and Xba (5-prime). Using these added restriction sites, this fragment was subcloned into the cloning plasmid, pKs (Invitrogen, Carlsbad, CA). For promoter assembly, fragments 2_3, 8_3, 14_3, 16_3 and 19_3 were serially ligated together at the restriction sites SphI, BbsI, PstI and AccB7I, respectively (Fig. 1B). All ligations incorporated directional cloning; the 5-prime restriction site in all cases was XmaI, a site originating from the AP-2 primer and therefore common to all fragments, yet unique in the INGAP 5-prime sequence. For digestions with AccB7I the dcm methylase-deficient E. coli INV110 (Invitrogen) were first transformed with the cloning vectors, prior to fragment isolation, digestion and ligation. All fragments, once constructed, were digested with XmaI (5-prime) and BglII (3-prime) and subcloned into pBgal-basic vector (Clontech), a promoterless h-galactosidase expression vector. The BglII site is unique in the INGAP promoter and is located 3-prime to the identified TATA box (Fig. 1B). The plasmid, p2_3T, was generated by a StuI digestion of p2_3 followed by blunt-end ligation. The first ATG, start codon, downstream of the TATA-box in the reporter gene constructs is that for h-galactosidase. Reporter genes were sequenced in both directions. The resulting promoter deletion reporter constructs are illustrated in Fig. 2A. 1.4. Transient transfection assays 293T cells, a non-INGAP expressing cell line grown in DMEM medium supplemented with 10% fetal bovine serum (Invitrogen), were transfected with the INGAP promoter – reporter plasmids. Per transfection, 400,000 cells in wells of a six-well tissue culture plate were transfected with 700 ng of a plasmid containing the INGAP promoter-h-galactosidase reporter construct and 700 ng of pAP4 [16] (a secretory alkaline phosphatase expression plasmid used to control for transfection efficiency). Transfection was performed by incubating the cells with a DNA and Lipofectamine (Invitrogen) mixture for 5 h at 37 jC as per manufacturers instruction. For induction studies, transfected cells were stimulated the day following transfection with PMA and incubated for 2 days prior to analysis. To calculate transfection efficiency, 10 Al of culture media, which had been heat-inactivated at 70 jC for 10 min to destroy endogenous alkaline phosphatase activity, was assayed in duplicate with pNPP substrate buffer (Sigma, St Louis, MO) to determine secretory alkaline phosphatase activity. Cells were harvested, pelleted and washed with PBS, before h-galactosidase activity was assayed using a luminescent-based h-galactosidase detection kit (Clontech), following the manufacturers’ instructions. For data analysis, transfection efficiencies were normalized using the alkaline phosphatase activity with the normalized h-galactosidase activity being expressed as fold increase above control transfections with pBASIC (the

86

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89

Fig. 2. Identification of INGAP promoter activity. A diagram of the serially deleted INGAP promoter fragments (black line) linked to the h-galactosidase reporter gene (hatched line) is shown (A) adjacent to the fold-induction in reporter gene activity mediated by the INGAP promoter fragments (B). Basal intrinsic activity for the INGAP promoter fragments in unstimulated, control, cells (black bars) is compared to the corresponding activity in cells stimulated with 50 ng/ml PMA (hatched bars). PMA induction for p14_3 is significantly higher ( P < 0.05) than that for p2_3 (t-test) data shows means and S.E. (n = 8). Electrophoretic mobility shift analysis of a fluorescine-labeled probe from INGAP fragment 2_3T is shown in (C). A specific shifted band (arrow) is detected (lanes 2 and 5) following incubation of the fluo-probe with nuclear extracts from control (lane 2 and 3) or PMA stimulated (lane 5 and 6) cells. No band is detected in samples containing fluo-probe in the absence nuclear extract (lanes 1 and 4) or samples containing fluo-probe, nuclear extract and a 25-fold excess of unlabelled probe (lanes 3 and 6). The response of promoter fragments to increasing doses of PMA is shown in (D).

promoterless h-galactosidase reporter plasmid). These cells had undergone equivalent stimulation. The data in Fig. 2B shows that the INGAP promoter in 293T cells has low constitutive activity. This basal activity can be elevated by the addition of PMA, a phorbol ester which enhances transcription factor activity by stimulating intracellular signaling cascades, principally PKC and Raf [17]. The response to PMA is bimodal being detected in all INGAP promoter deletions tested and peaking in the p14_3 INGAP promoter fragment that contains the region 1769– 3134 of the INGAP genomic sequence. These data both confirm promoter activity associated with the INGAP 5-prime sequence and reveal that the promoter is subject to at least two PMA-sensitive DNA elements, one located proximal to the minimal promoter and one located within 1769 – 2168 of the INGAP genomic sequence. Extension of the 5-prime region indicates that a possible negative regulatory DNA element neutralizes the activity of the PMA-responsive region located 1769– 2168 of the INGAP genomic sequence. The response of the INGAP promoter to PMA occurs at a dose of PMA that is physiologically relevant [17]. Moreover, the promoter response to PMA is dosedependent plateauing near 100 ng/ml with an EC50 of approximately 17 ng/ml. Increasing the dose of PMA up to 300 ng/ml did not overcome the inhibitory effects of the DNA elements located in the 5-prime of p19_3 (Fig. 2C), region 1 – 1400 in the INGAP genomic sequence. Many mammalian promoters require occupation of several transcriptionally active DNA elements for optimal gene expression in vivo/in vitro. For example, the loss of one

transcriptionally active site in the interleukin-2 promoter severely reduces the level of IL-2 gene transcription [18]. The role of a negative DNA element(s) located within 1– 1400 of the INGAP genomic sequence could be critical to regulating the tissue-specific expression of INGAP. Previous reports have shown INGAP expression can be induced in the exocrine pancreas in vivo [9]. It is therefore anticipated that other pancreatic exocrine-specific transcription factors are likely to be involved in regulating INGAP expression. These speculations will be strengthened through transfections of the INGAP promoter – reporter constructs into exocrine cells or cell lines which endogenously express INGAP. To date, transfection studies with the amphocrine rat pancreatic cell line AR42J have not demonstrated increased constitutive activity. Since this is a tumor-derived cell line, it is possible that the cells may either be inappropriate candidates for INGAP expression or in the wrong transition state; these cells have the capacity to differentiate in a stimulus-dependent manner into neuroendocrine-like or exocrine-like cells [19]. Supporting this contention, INGAP mRNA was not detected in these cells (data not shown). In support of the observations for a physiologically functional DNA enhancer element located proximal to the transcriptional start site of INGAP, fluorescent oligonucleotides from 2_3T, region 2931– 3134 of the INGAP genomic sequence, were evaluated in an electrophoretic mobility shift assay [20] when incubated with nuclear extracts from 293T cells stimulated with and without PMA. A specific shifted band, as determined by its displacement in the presence of a 25-fold excess of unlabeled probe, was observed in both nuclear extracts. Currently, we are mapping the actual

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89 87

Fig. 3. Comparison of the regulatory regions of INGAP and RegIIIy. Sequence alignment of the 5-prime regions of the two genes are shown in A, where areas of sequence identity are shaded. Regions for DNA insertions in RegIIIy (open line), INGAP (closed line) and the position of the TATA-box (checkered line) are marked. AP-1 binding sites in INGAP are outlined. The exocrine enhancer consensus-motifs present in RegIIIy (hashed lines) and the equivalent region in INGAP is boxed for both genes. The locations of predicted transcription factor binding sites for RegIIIy and INGAP are compared in B. Sites on the sense strand (up triangles) and antisense strand (down triangles) were predicted using default vertebrate settings for MatInspector Professional (www.Genomatix.com). The positions of predicted binding sites for the transcription factors AP-1 (A), NFnB (R), NFAT (N), STAT (S), MyoD (D) and PDX (P) are shown, for reference, the CAAT-like box (C) and TATA-box (T) are also marked.

88

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89

binding sites located in both the 2931– 3134 region and the 1769 – 2168 regions of the INGAP genomic sequence to reveal the origins of the PMA response. In summary, we report here the genomic sequence of hamster INGAP including initial analysis of the 5-prime regulatory region. These data confirm the identification of the INGAP promoter through the presence of transcriptionally active, stimulus-responsive DNA elements proximal to the transcriptional start of INGAP. Resolution of the INGAP gene has identified the coding region of INGAP to consist of six exons spanning approximately 3 kb of genomic sequence (Fig. 1A). This organization is analogous to the related gene family of Reg proteins. Reg-related proteins are a group of homologous calcium dependent (C-type) lectins [21] that have been classified into four sub-types based upon structural similarities [22,27]. INGAP is a C-type lectin whose structural features are consistent with a type III Reg-related protein. The genomic organization and protein structure of INGAP has the greatest homology to mouse RegIIIy (72.2% similarity in amino acid sequence). The biological function of RegIIIy has not been defined and while expressed in the pancreas, the expression of RegIIIy in other organs is ambiguous. Abe et al. [23] report the expression of RegIIIy to be restricted to the pancreas but a parallel report of a mouse protein termed INGAPrP whose cDNA is identical to RegIIIy, describe a broader tissue expression of RegIIIy which includes pancreas, stomach, duodenum and skeletal muscle [24]. This broad tissue expression is distinct to that reported for INGAP which is restricted to the pancreas and pancreas/duodenal border [9]. Furthermore, while the actual tissue expression profile of mouse RegIIIy remains to be resolved, both reports describe the constitutive expression of RegIIIy in the pancreas. This is distinct to the expression reported for INGAP in the hamster pancreas which was not detected in the normal (unwrapped) pancreas [9]. It is difficult to conclude whether INGAP and RegIIIy are orthologs. Conceivably, the disparity in expression may reflect different genes or species differences in the regulation of the same gene. A detailed analysis of the promoter for INGAP and the consequent regulation of INGAP expression is expected to resolve how expression of the two genes are differentially regulated. There is a 60% sequence similarity between the reported 5prime region of RegIIIy [23] and the equivalent region of the INGAP promoter (Fig. 3A); however, analysis of the two promoter regions (MatInspector professional and Alibaba2.1; www.Genomatix.com) reveals a disparate distribution of predicted transcription factor binding sites. Of the predicted DNA binding sites for RegIIIy and INGAP (109 and 139 respectively) only 12 transcription factor binding sites were identified to have an analogous location in the two promoters and several of the binding sites are unique to one of the two promoters (Fig. 3B). Sequence alignment suggests several regions of DNA insertions between INGAP and RegIIIy, which in INGAP, correspond to consensus

recognition sequences for additional transcription factor binding sites. Moreover, the region of the RegIIIy promoter that contains the dual motif pancreatic exocrine enhancer [25,26] is not present in the INGAP promoter. The equivalent region in the INGAP promoter only has one of these motifs (Fig. 3A). Based upon the distinct patterns for predicted transcription factor binding sites, it would appear feasible that the regulation of INGAP and RegIIIy is distinct, further analysis will reveal the significance of the differences noted here. Regulation of INGAP presents a new approach to stimulating the endogenous production of islets as a therapeutic rational for treating diabetes. Identification of the INGAP promoter reported here is an important achievement towards defining the cascade of factors that regulate islet neogenesis.

Acknowledgements This research was sponsored by grants from the Diabetes Institutes Foundation, Cosmopolitan International and GMP? Companies Inc. We are grateful to Dr. John Flanagan for providing pAP4 and acknowledge the assistance of Dr. Manas Ray in primer design.

References [1] [2] [3] [4] [5] [6]

[7] [8]

[9] [10] [11]

[12] [13] [14] [15] [16] [17] [18]

[19]

G.I. Bell, K.S. Polonsky, Nature 414 (2001) 788 – 791. D. Mathis, L. Vence, C. Benoist, Nature 414 (2001) 792 – 798. R.A. DeFronzo, E. Ferrannini, Diabetes Care 14 (1991) 173 – 194. U.K.P.D.S. Group, Lancet 352 (1998) 837 – 853. A. Vinik, R. Rafaeloff, G. Pittenger, L. Rosenberg, W. Duguid, Horm. Metab. Res. 29 (1997) 278 – 293. A.M. Shapiro, J.R. Lakey, E.A. Ryan, G.S. Korbutt, E. Toth, G.L. Warnock, N.M. Kneteman, R.V. Rajotte, N. Engl. J. Med. 343 (2000) 230 – 238. D.W. Gray, Ann. N.Y. Acad. Sci. 944 (2001) 226 – 239. D. Pipeleers, B. Keymeulen, L. Chatenoud, C. Hendrieckx, Z. Ling, C. Mathieu, B. Roep, D. Ysebaert, Ann. N.Y. Acad. Sci. 958 (2002) 69 – 76. R. Rafaeloff, G.L. Pittenger, S.W. Barlow, X.F. Qin, B. Yan, L. Rosenberg, W.P. Duguid, A.I. Vinik, J. Clin. Invest. 99 (1997) 2100 – 2109. L. Rosenberg, W.P. Duguid, R.A. Brown, A.I. Vinik, Diabetes 37 (1988) 334 – 341. G. Gold, C. Broderick, M. Caragna, G. Williams, G.L. Pittenger, R. Rafaeloff, A. Reifel-Miller, T. Borts, J. Hale, L. Churgay, B. Gerlitz B. Grinnell, A. Vinik, Diabetes 47 (1998) A245. L. Rosenberg, R. Wang, J. Li, G.L. Pittenger, W.P. Duguid, A. Vinik, Diabetes 49 (2000) A256. R. Rafaeloff, E. Schmitt, H. Edlund, G. Gold, A. Vinik, Diabetes 47 (1998) A259. L. Rosenberg, Microsc. Res. Tech. 43 (1998) 337 – 346. P.D. Siebert, A. Chenchik, D.E. Kellogg, K.A. Lukyanov, S.A. Lukyanov, Nucleic Acids Res. 23 (1995) 1087 – 1088. J.G. Flanagan, H.J. Cheng, Methods Enzymol. 327 (2000) 198 – 210. D.A. Taylor-Fishwick, J.N. Siegel, Eur. J. Immunol. 25 (1995) 3215 – 3221. D.A. Taylor-Fishwick, C.A. June, J.N. Siegel, in: M.M. Harnett, K.P. Rigley (Eds.), Lymphocyte Signalling; Mechanisms, Subversion and Manipulation, Wiley, London, 1997, pp. 31 – 50. H. Mashima, H. Ohnishi, K. Wakabayashi, T. Mine, J. Miyagawa, T.

D.A. Taylor-Fishwick et al. / Biochimica et Biophysica Acta 1638 (2003) 83–89

[20]

[21] [22] [23]

Hanafusa, M. Seno, H. Yamada, I. Kojima, J. Clin. Invest. 97 (1996) 1647 – 1654. M. Laniel, A. Beliveau, S. Guerin, in: T. Moss (Ed.), 2nd ed., DNA – Protein Interactions: Principles and Protocols, vol. 148, Humana Press, Tatowa, NJ, 2001, pp. 13 – 30. K. Drickamer, Curr. Opin. Struct. Biol. 3 (1993) 393 – 400. H. Okamoto, J. Hepatobiliary Pancr. Surg. 6 (1999) 254 – 262. M. Abe, K. Nata, T. Akiyama, N.J. Shervani, S. Kobayashi, T. Tomioka-Kumagai, S. Ito, S. Takasawa, H. Okamoto, Gene 246 (2000) 111 – 122.

89

[24] K. Sasahara, T. Yamaoka, M. Moritani, K. Yoshimoto, Y. Kuroda, M. Itakura, Biochim. Biophys. Acta 1500 (2000) 142 – 146. [25] S.L. Weinrich, A. Meister, W.J. Rutter, Mol. Cell. Biol. 11 (1991) 4985 – 4997. [26] A. Meister, S.L. Weinrich, C. Nelson, W.J. Rutter, J. Biol. Chem. 264 (1989) 20744 – 20751. [27] J.C. Hartupee, H. Zhang, M.F. Bonaldo, M.B. Soares, B.K. Dieckgraefe, Biochim. Biophys. Acta 1518 (2001) 287 – 293.