Promoter paper

Isolation and characterization of the promoter region of the ratkidney-type glutaminase gene

Lynn Taylor, Xiangdong Liu, William Newsome, Richard A. Shapiro,Maithreyan Srinivasan, Norman P. Curthoys *

Department of Biochemistry and Molecular Biology, Colorado State University, Fort Collins, CO 80523-1870, USA

Received 27 November 2000; received in revised form 19 January 2001; accepted 22 January 2001

Abstract

A VEMBL3 rat genomic library was screened to clone a phage that contained the promoter region of the kidney-type mitochondrialglutaminase gene. The resulting VGA1 phage contained 13.7 kb of genomic DNA that was mapped by Southern blotting and restrictionanalysis. The 2.22 kb and 0.83 kb SacI fragments of VGA1 were sequenced and the transcription initiation site was identified by RNasemapping. The reported sequence contains 2287 bp of the promoter, the entire exon 1 (542 bp), and 223 bp of the initial intron of theglutaminase gene. The initial exon contains 141 bp of 5P-nontranslated sequence and 401 bp of coding sequence that encodes the 72-aminoacid mitochondrial targeting presequence and 61 amino acids from the N-terminus of the mature 66 kDa glutaminase subunit. Varioussegments of the GA promoter were cloned into a chloramphenicol acetyltransferase (CAT) expression vector. The resulting GA-CATconstructs were transfected into LLC-PK1-F� kidney cells to assess the promoter function of the isolated genomic DNA. The GA3402CATconstruct produced a 10-fold greater CAT activity than the promoter-less pCAT vector. Analysis of various deletion constructs indicatedthat elements located between 3402 and 363 bp must act in synergy with more proximal elements to create a functional promoter. Theinitial 402 bp segment lacks a TATA sequence but is GC-rich and contains two CCAAT boxes and two Sp1 sites. ß 2001 Elsevier ScienceB.V. All rights reserved.

Keywords: Glutaminase; Promoter; Kidney gene; Glutamine catabolism; Constitutive expression

Glutamine is an important metabolic fuel that is con-stitutively metabolized in liver [1], intestinal epithelium [2],lymphocytes [3], brain [4] and various transformed cells[5]. In contrast, renal catabolism of glutamine is activatedonly during metabolic acidosis [6]. However, in each ofthese tissues, the primary pathway of glutamine catabo-lism is initiated by a mitochondrial glutaminase (GA).Mammals contain at least two distinct genes that encodea mitochondrial GA [7]. The liver-type GA is generatedfrom a single 2.4 kb mRNA that is expressed in adulthepatocytes, brain, pancreas, and breast cancer cells [8].In liver, expression of this GA isoform is regulated similarto the enzymes of the urea cycle and of hepatic gluconeo-genesis [9]. The complete liver GA cDNA and 1 kb of itspromoter region have been cloned and sequenced [10]. Thepromoter segment of the liver GA gene lacks a TATA boxbut contains putative response elements for mediating the

e¡ects of cAMP and glucocorticoids. cDNAs that encodethe kidney-type GA have been cloned from kidney [11],brain [12] and intestine [13]. Various renal GA mRNAsthat range from 3.4 to 6.0 kb in length are produced byuse of alternative polyadenylation sites [11,12] or by alter-native splicing [13]. The alternatively spliced mRNA en-codes a protein that contains a distinct C-terminal se-quence and thus constitutes a variant of the kidney-typeGA. Increased expression of the GA gene in kidney duringchronic metabolic acidosis [14] is mediated at the level ofmRNA stabilization [15,16]. However, expression of therenal GA gene in other cells is likely to be regulated atthe level of transcription. For example, the level of GAactivity and GA mRNA expressed in intestinal cells isregulated by the availability of glutamine [17] or by epi-dermal growth factor [18]. GA activity is also increased inneutrophils that are incubated with adrenaline or cAMP[19]. To facilitate the future characterization of these re-sponses, a rat genomic DNA that contains the promoterregion of the kidney-type GA gene was isolated and char-acterized.

0167-4781 / 01 / $ ^ see front matter ß 2001 Elsevier Science B.V. All rights reserved.PII: S 0 1 6 7 - 4 7 8 1 ( 0 1 ) 0 0 1 8 3 - X

* Corresponding author. Fax: +1-970-491-0494;E-mail : ncurth@lamar.colostate.edu

BBAEXP 91526 12-3-01

Biochimica et Biophysica Acta 1518 (2001) 132^136

www.bba-direct.com

A V library containing 2.3U106 independent clones ofSprague-Dawley rat genomic DNA inserted into the Bam-HI site of EMBL-3 (Clontech) was screened with a 299 bpEcoRI-PstI fragment from the 5P-end of the rat kidneyGA cDNA [11]. The isolated phage (VGA1) tested positivein secondary and tertiary screens. It contains a 13.7 kbinsert that was mapped by Southern blotting and restric-tion analysis (Fig. 1A). The 299 bp probe hybridized spe-ci¢cally to the 0.83 kb SacI fragment. Thus, the 0.83 kbfragment and the preceding 2.22 kb SacI fragment were

subcloned into pBluescriptII SK(3) and sequenced in bothdirections by dideoxynucleotide sequencing [20] usingprimers designed from known sequence or by cloning con-venient restriction fragments into pBluescriptII SK(3) andsequencing with T7 and T3 primers. An RNase protectionassay was used to identify the transcription initiation siteof the kidney GA gene. A SacI-NarI fragment was sub-cloned into pBluescriptII SK(3) to yield pGASac/Nar.This plasmid was used to transcribe an antisense RNAthat contained 196 nucleotides of complementary se-quence. RNase mapping with this RNA yielded a pro-tected fragment of 133 nucleotides (Fig. 1B). This ¢ndingwas used to determine the transcription initiation site andto establish the length of the 5P-nontranslated region ofthe GA mRNA.

The nucleotide sequence of the 2.22 kb and 0.83 kb SacIfragments of VGA1 is illustrated in Fig. 2. The combinedfragments contain 2287 bp of promoter sequence, the en-tire sequence of exon 1 (542 bp) and 223 bp of intron 1.The initial exon contains 141 bp of 5P-nontranslated se-quence and the sequence that encodes the initial 133 ami-no acids of the rat kidney GA. The sequence of exon 1from bp 84 to 542 is identical to that reported for theinitial 459 bp of pGA117 [11]. This identity con¢rmsthat the isolated genomic DNA is derived from the genethat encodes the rat kidney-type GA. The proximal pro-moter region of the renal GA gene lacks an identi¢ableTATA box. However, analysis of the promoter seg-ment using Omega software (Oxford Molecular Group)identi¢ed two CAATT boxes and putative binding sitesfor Sp1, NF-1 and serum response transcription factors(Fig. 2).

Various GA-chloramphenicol acetyltransferase (CAT)constructs were synthesized and used to establish thatthe isolated genomic sequence can function as a promoter(Fig. 3). The GA3402CAT construct, when transfected intoLLC-PK1-F� cells [21], produced a 10-fold greater level ofchloramphenicol acetyltransferase activity [22] than thatobserved with the promoter-less pCAT construct. Addi-tion of longer segments of genomic sequence reduced theexpression of CAT activity in the kidney cells. Each of thethree constructs that extended upstream from 31552 bp to32287 bp produced approximately half of the activityderived from GA3402CAT. In contrast, the GA363CATconstruct yielded less activity than the pCAT construct.This suggests that promoter elements located between bp363 and 3402 are essential for basal expression of the GAgene in kidney cells. However, deletion of the segmentfrom bp 363 to +110 also results in loss of promoteractivity. Thus, the proteins that bind to the upstream ele-ments are not su¤cient to initiate transcription but theymust function in synergy with proteins that bind to moreproximal elements. The initial 402 bp segment is GC-richand contains two CCAAT boxes and two Sp1 sites whichare characteristic of a TATA-less promoter. All of theseelements bind factors that may facilitate the basal tran-

Fig. 1. Restriction and RNase mapping of VGA1 DNA. (A) The VGA1DNA was treated with SacI (S), BamHI (B) and XbaI (X), restrictionenzymes that release the inserted rat genomic DNA without cutting theEMBL-3 VDNA. The individual SacI fragments were sized (lengths inkb) and oriented by further restriction analysis. The 5P-end of the cod-ing region was identi¢ed by Southern analysis using the EcoRI-PstIfragment of pGA117 cDNA as a probe. The restriction map is drawnto scale. (B) Total RNA was isolated from the kidneys of acidotic ratsand poly(A)�RNA was puri¢ed by chromatography on oligo(dT)-cellu-lose. Antisense RNA was transcribed from pGASac/Nar in the presenceof 50 WCi of [K-32P]CTP. Approx. 106 cpm of the labeled antisenseRNA was incubated with 10 Wg of poly(A)�RNA for 10 min at 85³C.The samples were hybridized overnight at 53³C and then treated with0.6 Wg of RNase T1 and 12 Wg of RNase A for 1 h at 30³C. Subse-quently, the samples were digested with 100 Wg of proteinase K for 30min at 37³C and then extracted with phenol/CHCl3. After addition of20 Wg of tRNA, the protected antisense RNA fragments were concen-trated by ethanol precipitation and sized on a polyacrylamide gel usingspeci¢c RNA fragments as standards. The length of the protected frag-ment (133 nucleotides) was used as illustrated to de¢ne the sites for ini-tiation of transcription and translation.

BBAEXP 91526 12-3-01

L. Taylor et al. / Biochimica et Biophysica Acta 1518 (2001) 132^136 133

scription of the kidney GA gene. The only potential reg-ulatory element identi¢ed in the reported sequence is anupstream serum response element at bp 31258. Consistentwith these observations, the activity of the various GA-

CAT constructs was not a¡ected by treatment with phor-bol esters or cotransfection with an expression vector thatencodes the catalytic subunit of protein kinase A. How-ever, the serum response element may contribute to the

BBAEXP 91526 12-3-01

L. Taylor et al. / Biochimica et Biophysica Acta 1518 (2001) 132^136134

high level of kidney-type glutaminase that is expressed inrapidly growing and transformed cells [2,3,5].

The proximal promoter of the rat liver-type GA genelacks a functional TATA box but it contains putativeHNF-1, HNF-5 and C/EBP binding sites that may medi-ate the basal expression of this gene in liver [10]. None ofthese sequences are found in the initial 2.3 kb segment ofthe promoter of the kidney-type GA gene. Previous studieshave established that glucocorticoids induce expression ofthe kidney-type GA gene in the small intestine [23] and inlymphocytes [24]. However, a consensus GRE sequencewas not present in the reported promoter sequence.Thus, the necessary GRE may be located further up-stream, in an alternative promoter, or as part of a complexregulatory unit that di¡ers from the consensus sequence.The available pGA-CAT constructs could be used to testthe latter possibility. Experiments with various CAT re-porter constructs indicate that the initial 402 bp of theidenti¢ed promoter functions to initiate a modest, but sig-ni¢cant level of transcription in LLC-PK1-F� kidney cells.However, the pGA3402CAT construct produced only one-seventh of the activity obtained with pRSV-CAT that con-tains the strong promoter from the long terminal repeat ofthe Rous sarcoma virus (data not shown).

The initial precursor of the kidney-type mitochondrialGA contains 674 amino acids [11] of which 133 residuesare encoded by exon 1. In contrast, the correspondingprecursor for the liver-type GA contains only 536 aminoacids that are 86% identical to residues 140^674 of thekidney-type GA [10]. Thus, the initial exon of the kidneyGA gene encodes nearly all of the residues that are uniqueto the kidney-type GA. This suggests that the kidney GAgene may have evolved by duplication of the liver GAgene followed by addition of the sequence contained inexon 1. The complete coding segment of the kidney GAcDNA has been cloned into a baculovirus and expressedin Sf9 cells (L. Taylor and N.P. Curthoys, manuscript inpreparation). The recombinant protein can be puri¢ed bya modi¢cation of the procedure developed to purify therat renal GA [25]. When stored at 4³C, the recombinantGA undergoes degradation to produce a 60 kDa peptidethat retains full catalytic activity. Automated Edman se-quencing indicated that the N-terminus of the truncatedGA begins with the glycine residue that corresponds to

amino acid 132 of the primary sequence. This result sug-gests that the residues encoded by the initial exon may belargely unfolded and are not essential for catalytic activity.However, this segment may impart unique properties tothe kidney-type GA. Thus, it will be interesting to deter-mine if this sequence accounts for the ability of kidney-type GA to associate with the inner surface of the mito-chondrial membrane [26,27] or to potentially interact withthe mitochondrial glutamine transporter [28] or subse-quent enzymes of glutamine metabolism.

The nucleotide sequence results reported are availablethrough the GenBank Nucleotide Sequence Databaseunder accession No. AF302091. This work was supported

Fig. 3. Functional analysis of the GA promoter. The various promoter-CAT constructs that were used to test the function of the isolated ge-nomic DNA are drawn to scale. (B) LLC-PK1-F� cells were grown in a50/50 mixture of Dulbecco's modi¢ed Eagle's and Ham's F12 mediumcontaining 5 mM glucose and 10% fetal bovine serum in a 5% CO2/95%air atmosphere at 37³C. The cells were split and replated to 30% con£u-ence. After 2 days in culture, the cells were transfected with 10 Wg of aGA-CAT construct and 5 Wg of pRSVLgal. After 16 h the excess precip-itate was removed and the cells were cultured for 32 h in fresh medium.The cells were then homogenized and assayed for L-galactosidase activ-ity. Samples of homogenate (50^100 Wl) containing equivalent units ofL-galactosidase activity were used to measure CAT activity. The acety-lated products and the unreacted substrate were separated by thin layerchromatography and the percent conversion was quantitated using aPhosphorimager (Molecular Dynamics). The presented data are mean þS.E.M. with an n of 3^7 separate samples.

Fig. 2. Nucleotide sequence of the 2.22 kb and 0.83 kb SacI fragmentsof VGA1. The nucleotide sequence is numbered from the transcriptioninitiation site that is designated as +1. The ATG that encodes the initialmethionine of the mitochondrial glutaminase precursor is located at po-sition +142. The amino acid sequence translated from the subsequentcoding region is illustrated with one-letter abbreviations. The initial se-quence of intron 1 is doubly underlined. Indicated in bold italic are thepositions of the SacI and NarI restriction sites that were used to gener-ate the antisense RNA for RNase mapping. Potential CCAAT boxesare shaded and putative binding sites that are identical to the consensuselements for known transcription factors are underlined and labeled.6

BBAEXP 91526 12-3-01

L. Taylor et al. / Biochimica et Biophysica Acta 1518 (2001) 132^136 135

in part by grant DK 37124 from the National Institutes ofHealth.

References

[1] D. Haussinger, Adv. Enzymol. Relat. Areas Mol. Biol. 72 (1998) 43^86.

[2] R. McCauley, S.E. Kong, K. Heel, J.C. Hall, Int. J. Biochem. CellBiol. 31 (1999) 405^413.

[3] E.A. Newsholme, P.C. Calder, Nutrition 13 (1997) 728^730.[4] E. Kvamme, in: D. Haussinger, H. Sies (Eds.), Glutamine Metabo-

lism in Mammalian Tissue, Springer-Verlag, Berlin, 1984, pp. 33^48.[5] A. Colquhoun, E.A. Newsholme, Biochem. Mol. Biol. Int. 41 (1997)

583^596.[6] R.L. Tannen, in: Handbook of Physiology, Renal Physiology, vol. I,

Am. Physiol. Soc., Bethesda, MD, 1992, sect. 8, Ch. 23, pp. 1017^1059.

[7] N.P. Curthoys, M. Watford, Annu. Rev. Nutr. 15 (1995) 133^159.[8] P.M. Gomez-Fabre, J.C. Aledo, A. Del Castillo-Olivares, F.J. Alon-

so, I. Nunez De Castro, J.A. Campos, J. Marquez, Biochem. J. 345(2000) 365^375.

[9] J.D. McGivan, K. Boon, F.A. Doyle, Biochem. J. 274 (1991) 103^108.

[10] M. Chung-Bok, N. Vincent, U. Jhala, M. Watford, Biochem. J. 324(1997) 193^200.

[11] R.A. Shapiro, L. Farrell, M. Srinivasan, N.P. Curthoys, J. Biol.Chem. 266 (1991) 18792^18796.

[12] T. Holcomb, L. Taylor, J. Trohkimoinen, N.P. Curthoys, Mol. BrainRes. 76 (2000) 56^63.

[13] K.M. Elgadi, R.A. Mequid, M. Qian, W.W. Souba, S.F. Abcouwer,Physiol. Genomics 1 (1999) 51^62.

[14] N.P. Curthoys, O.H. Lowry, J. Biol. Chem. 248 (1973) 162^168.[15] O.F. Laterza, W.R. Hansen, L. Taylor, N.P. Curthoys, J. Biol.

Chem. 272 (1997) 22481^22488.[16] O.F. Laterza, N.P. Curthoys, Am. J. Physiol. Renal Physiol. 278

(2000) F970^F977.[17] S.E. Kong, J.C. Hall, D. Cooper, R.D. McCauley, Biochim. Biophys.

Acta 1475 (2000) 67^75.[18] G.X. Zhang, J.H. Lai, T.W. Jia, W.Z. Wang, J.Y. Wang, Nutrition

13 (1997) 652^655.[19] C. Garcia, T.C. Pithon-Curi, M. de Lourdes-Firmano, M. Pires de

Melo, P. Newsholme, R. Curi, Clin. Sci. 96 (1999) 549^555.[20] F. Sanger, S. Nicklen, A.R. Coulsen, Proc. Natl. Acad. Sci. USA 74

(1977) 5463^5467.[21] G. Gstraunthaler, J.S. Handler, Am. J. Physiol. Cell Physiol. 252

(1987) C232^C238.[22] C.M. Gorman, L.F. Mo¡at, B.H. Howard, Mol. Cell. Biol. 2 (1982)

1044^1051.[23] P. Sarantos, A. Abouhamze, W.W. Souba, Surgery 112 (1992) 278^

283.[24] P. Sarantos, A. Abouhamze, E.M. Copeland, W.W. Souba, J. Surg.

Res. 57 (1994) 227^231.[25] N.P. Curthoys, T. Kuhlenschmidt, S.S. Godfrey, Arch. Biochem.

Biophys. 174 (1976) 82^89.[26] N.P. Curthoys, R. Weiss, J. Biol. Chem. 249 (1974) 3261^3266.[27] B. Roberg, I.A. Torgner, J. Laake, Y. Takumi, O.P. Ottersen, E.

Kvamme, Am. J. Physiol. Cell Physiol. 279 (2000) C648^C657.[28] C. Indiveri, G. Abruzzo, I. Stipani, F. Palmieri, Biochem. J. 333

(1998) 285^290.

BBAEXP 91526 12-3-01

L. Taylor et al. / Biochimica et Biophysica Acta 1518 (2001) 132^136136