Shi, L., Twary, S.N., Yoshioka, H., Gregerson, R.G., Miller, S.S. ...

The Plant Cell, Vol. 9, 1339-1356, August 1997 O 1997 American Society of Plant Physiologists

Nitrogen Assimilation in Alfalfa: lsolation and Characterization of an Asparagine Synthetase Gene Showing Enhanced Expression in Root Nodules and Dark-Adapted Leaves

Lifang Shi,asb Scott N. Twary,c Hirofumi Yoshiokaybld Rober t G. Gregerson,e Susan S. Miller,b Deborah A. Samacyayf J. Stephen Gantt,g Pat J. Unkeferyc and Carroll P. Vancea.b*l

a U.S. Department of Agriculture, Agricultura1 Research Service, Plant Science Research Unit, 41 1 Borlaug Hall, 1991 Upper Buford Circle, University of Minnesota, St. Paul, Minnesota 55108

St. Paul, Minnesota 551 08 C Los Alamos National Laboratory, CST-18, Mail Stop C922, Los Alamos, New Mexico 87545

Plant Pathology Laboratory, Nagoya University, Chikusa, Nagoya 464-01, Japan e Department of Biology, Lyon College, Batesville, Arkansas 72503 ‘Department of Plant Pathology, University of Minnesota, St. Paul, Minnesota 551 08

Department of Plant Biology, University of Minnesota, St. Paul, Minnesota 551 08

Department of Agronomy and Plant Genetics, University of Minnesota, 41 1 Borlaug Hall, 1991 Upper Buford Circle,

Asparagine, the primary assimilation product from N2 fixation in temperate legumes and the predominant nitrogen transport product in many plant species, is synthesized via asparagine synthetase (AS; EC 6.3.5.4). Here, we report the isolation and characterization of a cDNA and a gene encoding the nodule-enhanced form of AS from alfalfa. The AS gene is comprised of 13 exons separated by 12 introns. The 5’ flanking region of the AS gene confers nodule-enhanced reporter gene activity in transformed alfalfa. This region also confers enhanced reporter gene activity in dark-treated leaves. These results indicate that the 5’ upstream region of the AS gene contains elements that affect expression in root nodules and leaves. Both AS mRNA and enzyme activity increased -10- to 20-fold during the development of effective nodules. lneffective nodules have strikingly reduced amounts of AS transcript. Alfalfa leaves have quite low levels of AS mRNA and protein; however, exposure to darkness resulted in a considerable increase in both. In situ hybridization with effective nodules and P-glucuronidase staining of nodules from transgenic plants showed that AS is expressed in both infected and uninfected cells of the nodule symbiotic zone and in the nodule parenchyma. RNA gel blot analysis and in situ hybridization results are consistent with the hypothesis that initial AS expression in nodules is independent of nitrogenase activity.

INTRODUCTION

Asn is the major nitrogen transport product in many plant species, particularly in temperate legumes such as alfalfa, pea, trefoil, and clovers (Sieciechowicz et al., 1988; Lea et al., 1990; Vance, 1990; Waterhouse et al., 1996). Asn may comprise as much as 80% of the xylem sap free amino acid pool in these species (Sieciechowicz et al., 1988; Ta et al., 1988). The usual source of Asn in temperate legumes is from effective root nodules (Vance et al., 1994). Ta et al. (1988) calculated, from in vivo feeding of I5N2 and I4CO2, that the rate of Asn synthesis in alfalfa is 150 nmol min-l g fresh weight nodule-I. The xylem sap of ineffectively nodulated plants, on the other hand, contains little or no Asn (Maxwell et al., 1984; Rosendahl et al., 1990). Moreover, when effective

To whom correspondence should be addressed. E-mail vance004 Qmaroon.tc.umn.edu; fax 61 2-649-5058.

nodules are removed from alfalfa, clover, and bird’s-foot trefoil, Asn concentrations in xylem sap decrease significantly (Maxwell et al., 1984; Kim et al., 1993). Higher amounts of Asn are also formed in plant tissues that are stressed and/or carbon limited. The classic example of Asn accumulation in response to stress is documented in asparagus spears in which concentrations increase five- to 1 O-fold within 24 hr of harvest (King et al., 1990). Pea leaves darkened for 72 hr have an -15-:old increase in free Asn compared with leaves exposed to the normal photoperiod (Joy et al., 1983). Similarly, sugar-starved maize root tips (Brouquisse et al., 1992) and sycamore suspension cells (Genix et al., 1990) have markedly higher amounts of free Asn compared with control tissues.

The synthesis of Asn is catalyzed by asparagine synthetase (AS; EC 6.3.5.4) (Lea et al., 1990; McGrath and Coruzzi, 1991). The enzyme catalyzes the Gln- and/or

1340 The Plant Cell

ammonia-dependent amidation of aspartate with the simul- taneous hydrolysis of ATP to AMP and PPi. Two forms of AS are found in Escherichia coli One, dependent on ammonia, is encoded by the asnA gene (Nakamura et al., 1981); the other, dependent on Gln, is encoded by the asnB gene (Scofield et al., 1990). Plant AS enzyme activity is known to be quite unstable and has only been partially purified from pea (Joy et al., 1983), lupin (Lea and Fowden, 1975; Rognes, 1975), maize (Oaks and Ross, 1984), and soybean (Huber and Streeter, 1985). However, recently, alfalfa root nodule AS was purified to near homogeneity, and antibodies were raised against the protein (Twary et al., 1994; S.N. Twary, P.J. Unkefer, and C.P. Vance, unpublished data). Although all plant AS enzymes studied to date appear to be Gln dependent, many catalyze Asn synthesis in the presence of high ammonia concentrations in vitro. lncreased Gln-dependent AS activity has been correlated with the accumulation of Asn in soybean (Huber and Streeter, 1984) and alfalfa root nodules (Ta et al., 1988), sugar-starved maize tips (Brouquisse et al., 1992), and the radicle of germinating white lupin (Lea and Fowden, 1975).

Significant progress has been made in understanding molecular aspects of AS through the isolation and characterization of AS cDNAs from pea (Tsai and Coruzzi, 1990), asparagus (Davies and King, 1993), Arabidopsis (Lam et al., 1994), broccoli (Downs et al., 1995), trefoil (Waterhouse et al., 1996), and maize (Chevalier et al., 1996). With the excep- tion of the deduced amino acid sequence from maize, these cDNAs are >80% similar. These cDNA clones suggest that all plant AS proteins are composed of 579 to 591 amino acids, yielding a molecular mass of -65 kD, and contain a Gln binding site. The two pea AS cDNAs, AS1 and AS2 (Tsai and Coruzzi, 1990), show enhanced expression in root nodules, with AS1 transcripts being more highly expressed than those of AS2. The pea AS1 cDNA hybridizes with a highly enhanced 2.2-kb mRNA from alfalfa nodules (Vance et al., 1994). 60th pea AS1 and Arabidopsis AS (Tsai and Coruzzi, 1990; Lam et al., 1994) transcripts are repressed by light, apparently by a phytochrome-mediated signaling process. When placed in the dark, AS transcripts in pea and Arabi- dopsis increase substantially. When dark-adapted Arabi- dopsis is then treated with sucrose, AS mRNA is repressed. Thus, dark induction/light repression of AS gene expression in these species may be related to leaf carbon status. Simi- larly, in maize and asparagus, synthesis of AS mRNAs is enhanced by the depletion of sugars (Davies and King, 1993; Chevalier et al., 1996; Davies et al., 1996). Moreover, AS expression in maize and Arabidopsis is responsive to nitrogen addition in the form of amino acids. Thus, AS expression appears to be regulated at least in part by a lack of sugars su- perimposed upon the nitrogen status of the tissues in question (Lam et al., 1995; Davies et al., 1996).

Because AS catalyzes the synthesis of Asn, which is the primary product of nitrogen assimilation in alfalfa, efforts to improve N, fixation and nitrogen assimilation will require a

thorough understanding of the genes encoding this enzyme. Other than a brief mention of the AS gene from asparagus (Moyle:et al., 1996), plant AS genes have not been characterized. Because our principal goal is to understand genetic factors controlling N2 fixation and nitrogen metabolism, we thought it essential to characterize an alfalfa AS gene. Here, we report the isolation of an AS cDNA by screening an expression library with AS antibodies and characterize an AS gene that is highly expressed in root nodules and is repressed in leaves by light. We also show that although AS expression is highly correlated with effective nodule development, enhanced expression does not have an absolute re- quirement for effective nodules. In addition, we investigated the cellular localization of AS mRNA in effective and ineffective nodules by using in situ hybridization. Lastly, we demon- strate that the 5' region, upstream from the translation start site of this gene'directs high reporter gene activity in root nodules and affects dark-induced expression of AS in leaves.

R ESU LTS

lsolation and Characterization of an AScDNA

A Xgt22 cDNA library, constructed from poly(A)+ RNA extracted from 20-day-old alfalfa root nodules (Gregerson et al., 1993), was screened with antibodies raised' against purified nodule AS (Twary et al., 1994). Six recombinant AS antigen- producing bacteriophages were purified and subcloned into pBluescript KS-. The cDNAs ranged in length from 1.6 to 2.2 kb. Partia1 sequencing of all six cDNAs showed that they were produced from a single species of mRNA. The largest cDNA insert, designated AS13, contains a 2184-bp insert with a single long open reading frame. The predicted AS coding region (1 758 bp) is preceded by 183 bp of 5' untranslated sequence and is followed by 243 bp of 3 ' untranslated sequence before the polyadenylation site is reached. An in- frame stop codon is found 18 nucleotides upstream of the ATG starting at nucleotide 184, the first potential initiation site in this cDNA. The complete nucleotide sequence of this insert has GenBank accession number U89923.

The alfalfa AS13 open reading frame encodes a 5 8 6 amino acid protein with a predicted molecular mass of 66,500 D and a PI of 6.01. A search of protein databases with the deduced amino acid sequence, by using the Genet- ics Computer Group BLAST program, revealed that the protein encoded by AS1 3 shares significant sequence identity with all plant AS cDNAs currently listed with GenBank. The percentage of similarity to other plant AS proteins, as com- puted by the Genetics Computer Group FASTA program, varies from 75.6 to 95.1 %, with alfalfa AS1 3 encoding a protein most similar to pea AS1. In addition, alfalfa AS shares 45.6 to 54.3% similarity with the Gln-dependent AsnB protein

Asparagine Synthetase Gene Structure 1341

of E. coli, the two yeast AS proteins, and animal AS proteins. In contrast, the alfalfa AS13 protein was only 22.7% similar to the ammonia-dependent AsnA protein of E. coli, suggesting that the alfalfa AS protein is more closely related to the Gln-dependent AS of E. coli than is the ammonia- dependent form.

To examine similarities and differences in individual amino acid positions, in Figure 1 we aligned the deduced amino acid sequence of AS13 with six other AS proteins represent- ing the most similar plant AS (pea ASí) to the most similar prokaryotic AS (E. coli AsnB). The N-terminal portion of alfalfa AS is quite conserved with respect to other plant AS proteins, whereas the last 30 to 40 amino acid residues at the C terminus show considerable variation, as previously pointed out by Davies and King (1993) and Lam et al. (1994). In the human AS protein, Cys-2 has been shown to be es- sentia1 for Gln-dependent activity (Van Heeke and Schuster, 1989). This Cys residue is conserved in alfalfa AS and all other plant AS proteins. The Gln-amide transfer function of purf-type glutamine amidotransferases is thought to reside in the catalytic triad Cys-2-Asp-29-His-102 (Mei and Zalkin, 1989). These residues align to positions 2, 34, and 104, respectively, in alfalfa and other plant AS proteins (Lam et al., 1994). Conservation of these residues in alfalfa AS further suggests that alfalfa AS is a Gln-dependent enzyme.

A phylogenetic analysis of the 16 deduced AS amino acid sequences in GenBank, with deletion of 85 divergent C-terminal amino acids, was used to compare the relationship of the protein encoded by alfalfa AS13 to other AS proteins. As shown in Figure 2, four well-separated branches of AS proteins (animal, plant, yeast, and bacteria) can be identified in the phylogenetic tree. With inclusion of E. coli AsnB as an outgroup, the separation of clusters is well supported by the results of 1 O00 bootstrap replicas. Monocotyledonous and dicotyledonous AS proteins form two different subbranches within the plant grouping. Legume and nonlegume proteins can be traced to two separate clusters on the dicotyledonous subbranch. Moreover, the two nodule-enhanced AS proteins, pea AS1 and alfalfa pAS13, originate from a single clade, suggesting they have a single phylogenetic origin. It should be noted that inclusion of the 85 divergent C-termi- na1 amino acids in the analysis had no effect on the phylogenetic relationships shown in Figure 2 (data not shown).

To ascertain whether there was more than one AS mRNA species in alfalfa, >30 individual AS cDNA clones were isolated from alfalfa root and nodule cDNA libraries. At least 1 O separate cDNAs from each library were sequenced at the 5' and 3' ends. Other than a 75-nucleotide variation in position of the polyadenylation site, these partia1 sequences were 99% identical to AS13 (data not shown). Although occa- sional nucleotide changes were found, none resulted in a change in deduced amino acids. These results suggest that a single AS mRNA species is present in both roots and nodules, and any nucleotide change probably reflects allelic variation in this tetraploid outcrossing species.

Organ-SpecWic Expression of Alfalfa AS

The organ-specific expression pattern of AS transcripts in alfalfa mRNA isolated from roots, nodules, cotyledons, stems, and leaves of effective Saranac plants grown in an 18-hr-light and 6-hr-dark cycle was characterized by RNA gel blot analysis, as shown in Figure 3A. The blot was also hybridized with either the 3' or 5' untranslated region or the coding region of pAS13. lndependent of the probe used and the tissue type examined, an mRNA of -2.2 kb hybridized with 32P-labeled AS13 cDNA. The abundance of AS mRNA was significantly greater in effective root nodules than in any other tissue. Im- age analysis, both radioanalytic and autoradiographic, showed that nodules consistently contained 20- to 40-fold more AS mRNA per microgram of poly(A)+ RNA than did roots, stems, leaves, and cotyledons.

As demonstrated in Figure 36, AS protein levels in soluble protein extracted from the various tissues used for RNA blots and evaluated by antibody staining of protein blots reflect AS mRNA content. Effective nodules (Figure 36, lane N) contain 1 O- to 20-fold more AS protein than any other tissue tested. AS protein was detected in other tissues, including ineffective nodules, but the amount of total soluble protein loaded into wells had to be increased by 10-fold to obtain consistent positive staining on blots. The molecular mass of the AS protein, as determined on protein blots, was 65,000 D, which is similar to the calculated mass of the protein deduced from the cDNA sequence.

Previous studies have shown that AS enzyme activity in pea leaves (Joy et al., 1983) and mRNA in leaves of pea (Tsai and Coruzzi, 1990), Nicofiana spp (Tsai and Coruzzi, 1991), and Arabidopsis (Lam et al., 1994) are repressed by light. In contrast, AS expression in asparagus (Davies and King, 1993) and maize roots (Chevalier et al., 1996) appears to be unaffected by light. To examine whether light represses AS gene expression in alfalfa leaves, total RNA and soluble protein from leaves of effective Saranac plants grown under various light regimes were evaluated by RNA and protein gel blotting. As shown in Figure 4, AS mRNA was very low in leaves of plants grown with continuous light for 2 or 4 days (Figure 4A, lanes 1 and 4). When transferred to continuous dark conditions, AS mRNA increased significantly within 24 hr (Figure 4A, lane 2) and remained high after 72 hr of darkness (Figure 4A, lane 3). When the dark-treated plants were then exposed to light, AS mRNA dropped precipitously within 6 hr of light exposure (Figure 4A, lane 5) and was vir- tually absent after 24 hr of continuous light (Figure 4A, lane 6). Light did not appear to affect AS mRNA accumulation in roots (data not shown).

The repression of AS mRNA accumulation in response to light was also reflected in part by the amount of the AS protein. Leaves of effective Saranac plants grown in either 2 or 4 days of continuous light had little AS protein (Figure 46, lanes 1 and 4). However, after 24 hr of darkness, the AS protein was apparent (Figure 48, lane 2), with a further increase

1342 The Plant Cell

* * MS-AS MCGILAVLGCSDDSQAKRVRILELS~~RGPD---WSGLHQHGDNYLAHQRLAIVDPASGDQPLFNEDKSII-V 75 PS-AS1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 At-ASN1 .................... 75

Hum-AS . . . . W.LF.SD.CLSVQCLSAMK1A---- . . . . . AFRFENVNGYTNCCFGFH . . . LF.M..IRVKKYPYLWLCY. 75 SC-ASN1 . . . . F.AFR-HE.VHRYKPKA.Q..K.IR.....---...NAI~STIFV.E.....GVE..A..ITSS.GEYM-~.. 74 Ec-AsnB . . S.FG.FDIKT.AVEL.KKA.....LMR. . . . IYASDNAI . . . E..S..."A.A...Y.QQ.THV-LA.. 75

---

Zm-AS . . . . . . . . . V V E V .

* Ms-AS G E I Y N H E D W ( Q L - P N H K F R T Q C D C D V I A H L Y E N G I G - E N F V D M L I X ; I F S F V L L 152 PS-AS1 . . . . . . . ........................................................................ 152

At-ASN1 ....... E...R.-K......GS..E..........-VD..... . . . . . . . . . . . . . . . M............. Zm-AS . . . . . . . E.KAK.-KT.E.Q.GS..E..........-.E.......M..........K...A......ICP..M

HUm-AS . . . . . . ~ Q Q H F - - EFCIQ .KV. GE1.L . . . DKG.1 .QTIC....V.A......ANKKVF~..TY..RP.FKAETrE. 152 SC-ASN1 1 5 1 EC-AsnB . . . . . . Q A . . A C Y G D R Y Q . Q . G S . . E . . L A . . Q . K . - P E . L . D . Q . M . A . A . Y . S E I G . . H L . I I P . . M . Y D E H 153

MS-AS G - - S V W ~ S E L K G L N D E C M F E V F P P G H L Y S S K D R E F R R W Y N P P F 229 PS-AS1 .............................................. 229 At-ASN1 S..M.....D.....T.....F....LGG.KQ.........S.-.S...E..AI.R...~............ 228 Zm-AS FS . . M.A.S.D . . R.IT . . . . . . . . TGGL.........S.T.-.S...NA.F..EM............... 228

. . . . . . IQ . . EECAD-YE.G.LS .. EP.IPM .LKHD- IDAP K Y . . . M.AWT.Y .AKQDRIVA...P..I.T..M.RSSA

Hum-AS .FLA.-- C . . A . . . ~ ~ S A T P . L K V E P F L P G H Y . V L D L K P N G K V A S V E M V K . E Sc-ASN1 SPKT. Y F . . . . . C.T.D .~ITA.....V.D..TDKIT.YFT.D.LD.KR..S..I.YMAI.HSL....R EC-AsnB Q L W . . . M.A.VPV.RTIKE..A.SYLW.Q.G.I.SY.HRD..D~~~.KNE..Q.L.DS.K

MS-AS G V L L S G G L D S S L V A A V - - - - - - - - - - - - - - - T - - - - A R Y L F 285 __-- -_-- PS-AS1 . . . . _ _ _ . . . . . . _ . . . . . . _ . . . . . . . . . . . P...............K

At-ASN1 _ _ . . . . _ _ _ . . . . . . _ _ . . _ . . . . . . . . PQ . . . . . . . . E.S . . . . . . K SI--------------- _ _ _ _ H _ _ _ - Zm-AS . . . . . . . . . . . . . . S.---------------A----S.H.NE.----.VDR...N...T..I....S.....A.....Y 284 Hum-AS E I E ~ R I . F I ~ L L S G G ~ S S L V . A . L L K Q L ~ . V Q Y P . ~ . A I . ~ S . . . L . A . K . . . H 305 SC-ASN1 . . . . . . . . . . . . I.SIARRETAKATNDVEPS .YDSK..H...IDDm.~A~S....AI..PN....Q.A.K..K. 309

I1S.I--------------- .KKYA..RVEDQ----ERSEA.WPQ....A...P.S.....AQ...NH 290

MS-AS LG

At-ASN1 . . . .

Ms-AS Ps-ASl At-ASN1 Zm-AS Hum-AS Sc-ASN1 Ec-AsnB

MS-AS PS-AS1 At-ASN1 Zm-AS Hum-AS Sc-ASN1 Ec-As&

E F H Q E T C R K I K A L H R Y I X ! L R A N K S T Y A W G L E A R V P F L D K D - - - - D E E ... ..... . . . . . . . . . . . . . . . . . F.....D....... . . . ..... ........NT.. SL . . . S....P........ .. LE . . . . . . . . . . L........A.S...V.........S..S........WN...RDL......VM.....----.D. KAEE.SE.LLRE.YLF.V . . . DRT.A.H ... L......HR.FSYYLSLP..MRIP.NG---...HL..ET.EDSNL--- ... T.SVQRV.N..LA.........M............RE.LQ~.N...NE....PK......Y.......TTGEPDA .L.E..V..LL...M...A....AMS...V.........K.LD...R.N.QD..CGN--.KM..H...EC.E-------

RMMLNASHIFPFNTPLTKEIFERFFPQNSARL . . . ................................... K..S..G....H...N......................

. . . N....L.SFPEQQ...E..N..AQM..Y...VN............L...D...E --- I..E..W.P..A....ITS.I(NS.FKILQEYVEHQ.D.A..A..AQK......K...G....QV...HY.GRADW. K....EE..W.........---...S....L..T.EAVIS.E.FASPKSDI.T....FW..LK.DAL...KTVAD -A...ASVAW.........---...S...TL.EV..QQ.S.QQL~.RFR..Y...TS....L..E...EL..LP..AE

--__ 438 438 437 437 457 466 438

514 514 513 513 533 542 513

Ms-AS TVPGGASVA---CSTEKAIEWDASWSSNLDPSGRAAFGVHNSA~QVN~~~P~IIP~-ISNLGVAIQS-- 586 PS-AS1 . . . . . P...---..............N.........L...V....H.I.P.T.G.........IG-V.P......T-- 586 At-ASN1 . . . . . . T..---...A..V.......N.M.......I...L...CGK--N..L.IP.L.A.DN.PMMMGQ..V...-- 584 Zm-AS . . . W.P.1 PA . . . . VEQ.KASN.....FISSHDSA.TDH--TA.SRRWPTAAAR.ANG~KD.WPIAV 586 H ~ - A S S~~KWINATDPSARTLTHYK,AVKA-------------------------------------------------- 561 Sc-ASN1 ,,~WIpK,WJG,-pSGRyAQI~~~E------------------------------------------------ 572 Ec-AsnB 554

Figure 1. Amino Acid Sequence Alignment of Alfalfa AS with AS from Other Species.

The alfalfa AS (Ms-AS) sequence was aligned with AS sequences from Pisum sativum (Ps-ASI; GenBank accession number X52179; Tsai and Coruzzi, 1990), Arabidopsis tbaliana (At-ASN1 ; GenBank accession number L29083; Lam et al., 1994), Zea mays (Zm-AS; GenBank accession number X82849; Chevalier et al., 1996), humans (Hum-AS; GenBank accession number M27396; Andrulis et al., 1987), yeast (Sc-ASNI; Gen- Bank accession number U40829), and E. coli (Ec-AsnB; GenBank accession number J05554; Scofield et al., 1990). ldentical residues are indicated by dots. Dashes indicate gaps introduced to maximize sequence similarity. Numbers to the right of each sequence indicate the position of residues on the diagram. Asterisks indicate the residues of the putative purf-type glutamine binding triad: boldface C, Cys-2; boldface D, Asp- 34; and boldface H, His-104.


100 Hum-AS (M27396, Human)

Ham-AS (M27838, Hamster)

—— Zm-AS (X82849, Maize)

—Os-AS (U55873, Rice)

I——PS-AS2 (X52180, Pea)

l_Lj-AS1 (X89409, Lotus)

PS-AS1 (X52179, Pea)

Ms-AS (U89923, Alfalfa)

Lj-AS2(X8941 O.Lotus)

Gm-AS (U55874, Soybean)

Bo-AS (X84448, Broccoli)

AI-ASN1 (L29083, Arabidopsis)

——Ao-AS (X67958, Asparagus)

Sc-ASW (U40829, Yeast)

SC-ASN2 (Z72909, Yeast)

Ec-AsnB (J05554, E. coli) 50 aa changes

Figure 2. Unrooted Phylogenetic Tree of 16 AS Proteins.The 16 amino acid sequences were aligned using the PILEUP pro-gram. The phylogram of the aligned amino acid sequences was con-structed after removal of 85 divergent C-terminal amino acids.Parsimony analysis was performed using the heuristic search optionof phylogenetic analysis program (PAUP) using parsimony version3.0. For each AS protein, the organism and GenBank accessionnumber are listed in parentheses. The value on each branch indi-cates the percentage of 1000 bootstrap replicas supporting thebranch. The scale bar indicates the number of amino acid substitu-tions required to obtain branch lengths, aa, amino acids.

occurring after 3 days of darkness. In contrast to AS mRNA,the AS protein was still prominent after dark-treated leaveswere exposed to light (Figure 4B, lanes 5 and 6). At each as-say period when AS protein was detectable in leaves, ASenzyme activity could be detected in cell-free extracts (datanot shown).

AS mRNA Accumulation and Enzyme Activity in Nodulesof Effective and Ineffective Alfalfa

AS mRNA abundance in total RNA isolated from developingnodules of effective Saranac and plant gene-controlled inef-fective Saranac (/r^Sar) was determined by RNA gel blotanalysis, as shown in Figure 5A. During the course of thisexperiment, nitrogen fixation, as measured by acetylene re-duction, was first detected on day 9 after planting and inoc-ulation, with maximum nitrogenase activity occurring on day12 and remaining constant through day 33 (Gantt et al.,1992). AS mRNA in effective nodules increased in abun-dance from 9 to 33 days after planting (Figure 5A). By com-

parison, RNA extracted from in-^Sar nodules at the sametime points had strikingly reduced amounts of AS mRNA(Figure 5A). Ethidium bromide staining and labeling with32P-polyuridine indicated that each lane had equal amountsof total RNA (data not shown). The amount of the radioactiveAS13 probe that hybridized with the RNA blot shown in Fig-ure 5A was quantitated, and the results are presented in Fig-ure 5B. Relative to a basal level of expression on days 5 and7, which was very similar for the two genotypes, radioactivecounts in AS mRNA of effective Saranac at succeeding timepoints increased 2-, 8-, 23-, 34-, and 50-fold at days 8, 9,12, 19, and 33, respectively. For in^Sar nodule mRNA, theincreases were 1-, 2-, 4-, 8-, and 10-fold at the same timepoints given above. Thus, the maximum accumulation of ASmRNA in in^Sar nodules was approximately equivalent tothat seen in day 9 effective Saranac nodules.

Although the overall pattern of AS mRNA accumulation inin^Sar nodules was similar to that in effective Saranac, therewere two notable differences: (1) from day 9, the amount ofradioactive pAS13 cDNA probe that hybridized with /r^SarRNA was one-fourth to one-fifth of that bound by effectiveSaranac; and (2) the increase in AS mRNA abundance thatoccurred after nitrogenase was detected and coincident withmaximum nitrogenase activity in effective Saranac was notobserved in in-^Sar nodules.

To assess whether AS enzyme activity reflected AS mRNAabundance, in vitro AS activity was measured at the same

R N C S L

- 2.2 kb

B • —-——. -65kD

Figure 3. Analysis of AS mRNA and Protein in Various AlfalfaTissues.(A) Poly(A)+ RNA (2 u.g) was isolated from roots (lane R), effectivenodules (lane N), cotyledons (lane C), stems (lane S), and leaves(lane L) of alfalfa cultivar Saranac, electrophoresed in a 1.5% formal-dehyde-agarose gel, transferred to a ZetaProbe membrane, andprobed with a 32P-labeled AS13 probe corresponding to aminoacids 172 to 364 of the translated region of the cDNA. The size ofthe AS transcript (2.2 kb) is indicated at right.(B) Total protein extracted from effective Saranac roots, nodules,cotyledons, stems, and leaves was separated by SDS-PAGE, trans-ferred to a nitrocellulose membrane, and probed with AS antisera(Twary et al., 1994). The protein loaded into each well was 5 u.g fornodules and 50 ^g for all other organs. The molecular mass (65 kD)of the AS polypeptide is indicated at right.

1344 The Plant Cell

578 9 12 19 33

-65kD

Figure 4. Light Represses the Expression of AS mRNA in AlfalfaLeaves.

(A) RNA gel blot analysis of AS transcripts in alfalfa leaves grown un-der various light treatments. Total RNA (15 pig) was electrophoresedin a formaldehyde-agarose gel, transferred to a ZetaProbe mem-brane, and probed with a 32P-labeled AS13 probe corresponding toamino acids 172 to 364 of the translated region of the cDNA. Treat-ment days or hours refer to continuous exposure. Lane 1 containsRNA from leaves grown under 2 days of light; lane 2, 2 days of lightand then 1 day of dark; lane 3, 2 days of light and then 3 days ofdark; lane 4, 4 days of light; lane 5, 2 days of light, 3 days of dark,and then 6 hr of light; and lane 6, 2 days of light, 3 days of dark, andthen 1 day of light. The size of the AS transcript (2.2 kb) is indicatedat right.(B) Total protein extracted from leaves of effective Saranac plantsseparated by SDS-PAGE, transferred to a nitrocellulose membrane,and probed with AS antisera. Each lane contains 100 p.g of protein.Treatment designations are the same as those indicated in (A). Themolecular mass of the AS polypeptide (65 kD) is indicated at right.

time points that nodule mRNA was isolated. As shown inFigure 5C, AS activity in effective Saranac nodules was notdetectable until day 9, even though small white outgrowthsof nodules begin to emerge by day 8. AS activity in the ef-fective Saranac nodules increased through day 12 andthereafter remained at a rather high constant level. By com-parison, AS activity in /n,Sar nodules at all developmentalstages was low to undetectable, except for a small increaseoccurring on day 9. The highest AS activity in /V^Sar noduleswas only ~10% of that found in day 19 effective nodules.

To ascertain whether the reduced amount of AS mRNA innon-N2-fixing nodules is unique to the /H, Sar genotype, ASmRNA levels in nodules of effective Saranac inoculated witheither effective Rhizobium meliloti 102F51 or with one of fourmutant R. meliloti strains that produce ineffective noduleson Saranac plants were evaluated by RNA gel blot hybrid-ization, as shown in Figure 6. Equal amounts of RNA wereloaded in each lane and verified by ethidium bromide stain-ing and hybridization with 32P-polyuridine. In the effectivesymbiosis (Figure 6A, 102F51, first three lanes), AS mRNAaccumulation was similar to that shown in Figure 5. The hy-bridization pattern for AS mRNA from all plants inoculatedwith ineffective R. meliloti strains was similar to that of effec-tive nodules through day 9. However, on day 12, the AStranscripts in nodules of all plants inoculated with ineffectiveR. meliloti strains were much lower than were those in effec-tive nodules. Figure 6B shows the radioactivity detected ineach AS hybridizing band in Figure 6A. The amount of

Sar

imSar

B15000

0 10 20 30 40

Days after Inoculation

Days after Inoculation

Figure 5. AS Enzyme Activity and mRNA Accumulation during De-velopment of Effective and Ineffective Saranac (/n,Sar) AlfalfaNodules.

(A) RNA gel blot analysis of AS mRNA accumulation in effective andineffective Saranac nodules. Total RNA (10 p.g) isolated from roots(day 5) or nodules (days 7, 8, 9, 12, 19, and 33) was applied to eachlane, electrophoresed on a 1.5% formaldehyde-agarose gel, trans-ferred to a ZetaProbe membrane, and hybridized with 32P-labeledAS13.(B) Radioactivity in the 2.2-kb AS band on each sampling date asdetermined by AMBIS radioanalytic image analysis (Scanalytics, Bil-lerica, MA).(C) AS enzyme activity throughout the development of effective(open square) and ineffective (open circle) Saranac nodules. Onehundred milligrams of roots (day 5) or nodules (days 7, 9,12,19, and33) were extracted, and AS activity was measured as described byEglietal. (1989).Sar, effective Saranac; inlSar, ineffective Saranac.


radioactivity detected in AS transcripts from ineffective nod-ules was never >25% of that found in day 12 effectivenodules.

Localization of AS Transcripts in Effective andIneffective Alfalfa Nodules

To ascertain the cellular localization of AS transcripts, weused either a conserved coding region or a 3' untranslatedregion as probes for in situ hybridization with 12-day-oldeffective Saranac and in-^Sar nodules. As shown in Figure 7,four zones are observed along the nodule axis in longitudi-nal sections of 12-day-old effective Saranac nodules (Vasseet al., 1990; Franssen et al., 1992; Hirsch, 1992): (1) the api-cal meristem zone (zone I); (2) the invasion and prefixingzone (zone II); (3) the amyloplast-rich interzone (*, zones II toIII); and (4) the nitrogen-fixing zone (zone III) (Figures 7A and7C). The senescent zone (zone IV) is not seen in these ex-amples but is apparent in 12-day-old /n,Sar nodules (Fig-ures 7G and 7H). The nodule parenchyma (inner cortex;Van de Wiel et al., 1990), nodule vascular bundles, and the

R 9 12 R 9 12 R 9 12 R 9 12 R 9 12

102F51 T202 G456 1491 F642

B

Q-O

5000

4000 -

5 3000 -

2000 -

1000 -

102F51 T202 G456 1491 F642

Figure 6. RNA Gel Blot Analysis of AS mRNA Levels in Effective Sa-ranac Nodules Induced by Effective or Ineffective R. meliloti Strains.(A) Total RNA (10 .̂g), isolated from roots (R; day 5 or 6) and nod-ules (days 9 and 12) of Saranac plants inoculated with either effec-tive R. meliloti 102F51 or ineffective strains T202, G456, 1491, orF642, as indicated, was electrophoresed in a 1.5% formaldehyde-agarose gel, transferred to a ZetaProbe membrane, and probed with32P-labeled AS13.(B) Radioactivity in the 2.2-kb AS band on each sampling date wasdetermined by AMBIS radioanalytic image analysis.

outer cortex are seen at the periphery of the central nitro-gen-fixing tissue of the nodule.

After hybridization of the antisense probe containing theconserved coding sequence of AS13 to the longitudinal sec-tion, dark-field microscopy showed that AS mRNA waslocalized predominantly in the cells of the late symbiotic, ni-trogen-fixing zone III and the nodule parenchyma of matureeffective 12-day-old Saranac nodules (Figures 7A and 7B).The control sense probe showed little or no signal (Figures7C and 7D). No AS transcripts were detected in the mer-istem zone I, invasion zone II, nodule vascular bundles, orouter cortex. At a higher magnification, as shown in Figures7E and 7F, AS transcripts were identified in the infected anduninfected cells as well as in the nodule parenchyma. AS ex-pression in the symbiotic region began in the first cell layeradjacent to the interzone II to III (Figures 7E and 7F). Celllayer-specific AS expression was also noted in the nodule pa-renchyma, where silver grains were restricted to the peripheryof zone III (Figures 7B and 7F). In contrast to AS mRNA accu-mulation in effective Saranac nodules, hybridization with/n-iSar nodules revealed that the number of silver grains wasquite reduced, with little or no signal detected in the noduleparenchyma and senescent zone IV (Figures 7G and 7H). ASmRNA expression was limited to a band of cells encompass-ing the youngest portion of zone III in ineffective nodules.

To compare the pattern of AS gene expression obtainedwith a conserved probe capable of hybridizing with all ASmRNAs with the pattern of a potentially gene-specific probe,we conducted in situ hybridization using a probe derivedfrom the 3' untranslated portion of the AS13 cDNA (Figures71 and 7J). Dark-field microscopy of longitudinal sections of12-day-old effective nodules revealed a pattern of expressionnearly identical to that observed with the conserved probe.

Isolation and Characterization of an AS Gene

An alfalfa genomic library (Gregerson et al., 1994) wasscreened by hybridization with the AS13 cDNA. Four hybrid-izing clones were obtained and digested with a variety of re-striction enzymes. DNA blot analysis of the restricted DNAindicated that one clone with an insert of ~8 kb, comprisedof two EcoRI fragments, contained the entire region homolo-gous to the pAS13 cDNA. This 8-kb insert, designatedAS18, was subcloned into pBluescript KS- and sequencedfrom 2748 bp upstream of the translation start codon to 734bp beyond the polyadenylation site. The complete nucle-otide sequence of AS18 (7962 bp) has GenBank accessionnumber L40327. A comparison of nucleotide sequence ofAS18 to that of the AS13 cDNA revealed differences at 23positions (98.8% identity). However, no nucleotide differ-ence resulted in an amino acid change.

By aligning the sequence of AS13 with that of AS18, wedetermined the exon-intron organization of the alfalfa ASgene. As shown in Figure 8, the alfalfa AS gene is composed

1346 The Plant Cell

Figure 7. Localization of AS Transcripts in Effective and Ineffective Saranac Nodules by in Situ Hybridization.

(A) Bright-field microscopy of a day 12 effective Saranac nodule. The nodule was sectioned longitudinally and hybridized with an AS 35S-labeledantisense RNA probe corresponding to the conserved region of AS13. Zones identified are as follows: I, meristem; II, infection; *, transition II toIII; and III, nitrogen-fixing zone, esz, early symbiotic zone; iz, cell invasion zone; Isz, late symbiotic zone; m, meristem; vb, vascular bundle. Bar =130 urn.(B) Dark-field microscopy of the tissue shown in (A). AS transcripts are indicated by bright dots (silver grains) and are found throughout nitrogen-fixing zone III and the nodule parenchyma. Day 12 nodules have no senescent zone.(C) Bright-field microscopy of a longitudinal serial section from the same nodule shown in (A) hybridized with AS 35S-labeled sense RNA probe.(D) Dark-field microscopy of the tissue shown in (C). No significant signal is shown with control sections.(E) Bright-field microscopy of detail magnification of the boxed region in (A) showing transition from infection zone II to nitrogen-fixing zone IIIand the nodule cortex. The arrow denotes the interzone II to III cell layer, in, infected cells; nc, nodule cortex; np, nodule parenchyma; un, unin-fected cells. Bar = 25 j^m.(F) Dark-field microscopy of the tissue shown in (E) showing that the high expression of AS transcripts is initiated in the first cell layer after tran-sition zone II to III. AS transcripts are evident in both infected and uninfected parenchyma cells. The arrow denotes the cell layer correspondingto interzone II to III.(G) Bright-field microscopy of a longitudinal section of a day 12 /n,Sar nodule hybridized with the 35S-labeled antisense AS probe. By day 12,in^Sar nodules begin to senesce at the base, sez, senescent zone. Bar = 100 p.m.(H) Dark-field microscopy of the tissue in (G) showing a reduced amount of signal for AS transcripts in the symbiotic zone, with little or no AStranscripts in the senescent zone.(I) Bright-field microscopy of a longitudinal section of a day 12 effective Saranac nodule hybridized with a 35S-labeled antisense RNA probe fromthe 3' untranslated region of AS13. Bar = 100 ^m.(J) Dark-field microscopy of the tissue in (I) showing the distribution of AS transcripts throughout nitrogen-fixing zone III and the nodule paren-chyma. The AS transcript and abundance are very similar to that seen in (B), with an AS13 conserved region used as a probe.

Hindlll

ATG

Xmnl Hindlll Bglll2 3 4 5 6 7


TAA

EcoRI

8 9 1011 12 13

Figure 8. Diagrammatic Representation of the Structure of the Alfalfa Nodule AS Gene.

Exons are represented by the numbered boxed regions; introns and 5' and 3' nontranscribed regions are represented by lines. The open (white)regions of exons 1 and 13 correspond to the untranslated sequences.

of 13 exons interrupted by 12 introns, and all but one ofthese predicted introns have the 5'-GT ... AG-3' consensusintron boundary sequences (Shapiro and Senapathy, 1988).The border sequences of intron 1 are 5'-GC . . . AG-3'. Thelengths of the introns vary from 88 to 1012 bp, and thelengths of the exons vary from 80 to 475 bp. The overallstructure of the alfalfa AS is similar to that reported for aspar-agus (Moyle et al., 1996). Plant and human AS genes have anequal number of exons (Zhang et al., 1989), whereas thehamster AS has only 12 (Andrulis et al., 1987). In addition, thesplicing position of intron 10 is conserved between plant andhuman AS. The three catalytic triad-forming residues in thepurF-type glutamine binding domain, Cys-2, Asp-34, andHis-104, are located in exons 1, 2, and 3 of alfalfa AS18 andexons 2 and 3 of the human AS gene, respectively.

To determine the genomic organization and gene copynumber of AS in alfalfa, gel blot analysis of genomic DNAwas performed. As demonstrated in Figure 9, equimolaramounts of P. meliloti genomic DNA cut with EcoRI gave nohybridization signal (lane 1), whereas hybridizing fragmentswere seen with alfalfa genomic DNA cut with Xbal (lane 2),EcoRI (lane 3), or EcoRV (lane 4). Multiple hybridizing frag-ments were seen in each lane, suggesting either a smallgene family or allelic variation for a single gene in this tetra-ploid species. Similar results were obtained under low-strin-gency hybridization conditions (data not shown).

Figure 10 shows the sequence of 2000 nucleotides in the 5'flanking region upstream of the AS18 putative translationalstart codon. Sandal et al. (1987) identified two sequencesconserved in legume nodule-enhanced genes: CTCTT andAAAGAT, each of which was found in one or more copieswithin 200 bp of the transcription initiation site in the soybeanN24, Lb, and N23 genes. These two sequences were alsofound to be conserved in other Lb genes from soybean nodu-lin 23 (Stougaard et al., 1990) and Sesbania rostratum leghe-moglobin (Welters et al., 1993), and in the soybean nodulingenes 20, 22, and 44 (Sandal et al., 1987). In the flanking re-gion of alfalfa AS18, the CTCTT motif is found four times inthe 5' untranslated region in the direct orientation, whereasthe AAAGAT motif is found once at 1210 nucleotides up-stream of the translation start codon (Figure 10). Interest-ingly, 14 perfect direct repeats with lengths varying from 7 to21 bp were found near each other upstream of the transla-tional start codon (Figure 10). Among them, one 13-bp repeat

(AAGGTGACAAAAA) is separated by 20 nucleotides, whereasanother 9-bp repeat (AAATAAGGT) is located partially withinthe gap and slightly overlapping the 13-bp direct repeats. An-other direct repeat (ATTAAATAAAt/cAt/aCATTT), found 575nucleotides upstream of the translation start codon, is AT rich(Figure 10). The significance of these repeats in the regulationof AS gene expression has not been evaluated.

Expression of an AS Promoter-p-Glucuronidase FusionConstruct in Root Nodules of Transgenic Alfalfa

To evaluate the role of the 5' flanking region of AS18 in con-ferring enhanced expression to root nodules, a transcrip-tional fusion was made with a 2.7-kb fragment starting 15nucleotides upstream of the putative translation start codonof the AS18 gene and the p-glucuronidase (GUS) gene. The

1 2 3 4

12.2-

6.0-

5.0-

4.0-

3.0-

Figure 9. DNA Gel Blot Analysis of AS Gene Sequences in the Al-falfa Genome.

Eighteen micrograms of alfalfa genomic DNA was digested withXbal (lane 2), EcoRI (lane 3), or EcoRV (lane 4), electrophoresedthrough an 0.8% agarose gel, denatured, transferred to a nylonmembrane, and probed with 32P-labeled pAS13 at high stringency.Genomic DNA (0.01 (j.g) from R. meliloti was digested with EcoRI(lane 1) and used as a negative control. Numbers at left indicate thepositions of the molecular length markers in kilobases.

1348 The Plant Cell

GACCATATATTTGGCTAGTTTATTTCCCAGATGTAATTMT~CGTTTTAGGCGMTCGMCGCATMTCTCACTTGATCACTGCA~AAAAGCAT~ a CCAACCATAAATMCCTGTTTACAATCGGCMGTTTATCACGCCTMGGACGGACMGTTCTAATTTATG~TAGTGGTCACGCATGTGT~ TGTGGTTATACAGACCATATATTTGGCTAGTTATCT~CCCAGGAGT~TAATATCGT~TAGGCAAAATATATATATGCGMGATTTCCCC~TTCC A A C G A A G A C A A C C A C C G C A A C G A C C G C C G T C C A C C A T T A A C T C ACACACTMTTTCTTCGGATTCGATCTTCAAAAATCTTC~TGATGGTTCTA~~GCAGATTTGGTATCTTATATGCTAATGAATATGAA~TTTGACTAGAGTA

AAATCACTGAACAATTAACTTCCAGMTCTTTTGGTTTGTATTCCTCCTMTCCATTT~AGAAATTTGTG~CCACAATGTAAAACC~GAGACAC

TGTGGTTGATGAAATTACTCGAACGTCAGTCAGTCACTCGTGTTTGTTCACCTTTGCCGTCTTTGT~TGTCAATTCATCGGCTTTG~CACTCTCCT~GCC

b’ b C

C C

GTCTTTGTTTT~TCAGTGACAGTTATGGAGTTCCGATTGACATTTMGGAGGT~rTTTCCGGTAGGGTGGCTGGATCCGCGGCG~GCTC~GAC C d d-

A C G G T A C T T ~ C C T T G G T C G A T C A G T C A T T C A A T C C T T

GATGTTGAAAATCTTGTTGCTGGTTATCTGTTTGTTTGTGTTTT~GGACAGAGGAGACMGGAGAGAGGCCAGAATGAAATGAAG~GAAGCATCGCT~C ****** e m

e m __k GTCGTCGCTGGAAATAGAAGAGAAATAmffiACGTTATCGTCACTAAAAACGGAAGA~ GACGGCGTCGOAAqCGGAATAGGAAGAGGCGGGGMCG

ACGTCGTCACCGAAAACATCGCCGOAAACAGAAGMGCGTGGTGGGCGGCGACGCTGG~G~GGATCTGTACAAATCATAGGAGAAAGCC~T~G~TGGGTG -7 g h i h @

J

ATAAAGTAAGGGTGAAGGAGAGAGGAGAAATGTGTGT~TCC~CTT-~~CTTGTTGCATTAATTTGTGT~TGTGTAGGGGTGGATGCCA~TGGGG j m k‘ k

GTGTACACCAATCACTACTCATTATTATTAGCTGACGTAG~C~GA~GGTATCMCTAGCAAA~C~CAGTGTGGATAAACCATTCTCACA C T T C T A T T A T T T G G G A G G M G C A G C A A A C C G C T C M G C A T ~ C T A G M T C ~ T ~ T G ~ T ~ ~ ~ ~ ~ A ~ C ~ T ~ T ~ A C G T ~ - ~ - ~ ~ ~ ~ ~

c A T _ T T ~ G ~ G n C A T G M C A A T ~ T T C C A G T ~ T M G G G C T T G C T A T G T A ~ T G A G A C A C T C T T T ~ C A A T A T ~ T A A A T A

ATTATTCTTTAAAAATATCAATTATTTTGATTAATAATCACATTAAATATAGGAGAAACATTATCAGTTTATCACATATGATTCTTAAGCAC~CTTAACATTTT

C C T A A A C A A T T A A A A C T A A A A T G M G ~ G G T G A C ~ T A A ~ T T C ~ T C A ~ T A A G G T ~ C ~ T A T C C C A T A G C ~ C ~ C G T C A ~ C ~ C G

TAGCAGTGGTTCACCGCCACCAACCGGTTCCGGTCACCCGGTCACCAAA~CCCAAAACC~CGACACCG~TTCGACCCACCTTACGCATATAAATAGATGGTG

tactaaactctcgtatctcaccattctacggttggttattcaca~tcaccttcttttcgtgtctt~ttctttttatcctctctcacatttta

ccaatcctttttttctttctCtcttcaaacaaacaaaacccttatgaaactatt~taatcatacatg

1 1

n n m m o - o’ -

-

* * * * * *****

* * * * * *i***

Figure 10. Sequence of the Alfalfa AS18 5’ Flanking Region

-1900 -1800

-1700 - 1 6 0 0 -1500 - 1 4 0 0

- 1 3 0 0

-1200

-1100

-1000

- 9 0 0

- 8 0 0

-700

-600

- 5 0 0 - 4 0 0

- 3 0 0

-200

-100 ’

-1

The 2164-nucleotide flanking region of the AS78 gene upstream from the putative translation start codon is shown. Uppercase letters correspond to the 2000-bp region of the AS18 putative promoter, whereas lowercase letters correspond to the 164-bp 5’ untranslated region of the cDNA. The first nucleotide upstream of the pASl3 cDNA is labeled -1. Thirteen perfect direct repeats are in boldface and underlined by solid arrows, with corresponding letters given beneath. A fourteenth 9-bp perfect repeat, n, lying within repeat m, is overlined with a solid arrow. An- other AT-rich repeat is indicated by dashed arrows. Two conserved sequence motifs found in other nodulins (TCTCC and AAGAT) are indicated by asterisks.

chimeric promoter-GUS construct was transformed into alfalfa via Agrobacterium. Nodules from four independent, fully transformed and regenerated plants were assayed his- tochemically for GUS, and GUS enzyme activity was measured in flowers, stems, roots, leaves, and nodules. Stained nodules were also sectioned and evaluated by dark-field microscopy for cellular localization of GUS activity. Figures 11A and 11 B show the typical expression observed in root nodules of transformed, regenerated alfalfa. GUS activity was detected throughout zone 111, the symbiotic zone, and in the nodule parenchyma. Within the symbiotic zone, GUS activity appeared in both infected and uninfected cells. Activity in the senescent zone, zone IV, was reduced and sometimes not detectable (data not shown).

Dark-field microscopy of nodule thin sections (Figure 11 B) was consistent with in situ hybridization, showing that GUS expression driven by the AS promoter was evident in both

infected and uninfected cells as well as parenchyma. Little to no GUS staining was evident in the meristem zone I, invasion zone 11, and prefixing interzone II to III (Figures 11A and 11 B). The blue X-gluc precipitate observed in the interior of nodules (Figure 11A) by using bright-field illumination appears red under dark-field conditions (Figure 11 B). In other organs of alfalfa, GUS activity was limited to weak staining in vascular bundles (data not shown). As shown in Table 1, enzyme assays of GUS activity, as measured with the substrate 4-methylumbelliferyl p-o-glucuronide, further confirmed that the AS promoter directed high GUS activity to root nodules. The average GUS activity from four independent transformants showed that nodules had 4.5, 11.7-, 18.7-, and 33.6-fold more activity than did roots, flowers, stems, and leaves, respectively.

To ascertain whether this 5’ flanking region of AS78 is involved in conferring enhanced AS expression to dark-


I II III

Figure 11. GUS Activity Staining in Nodules of Transgenic Alfalfa.

The plants were transformed with Agrobacterium containing theGUS reporter gene driven by the 2.7-kb 5' flanking region of theAS18 gene starting 15 nucleotides upstream of the putative transla-tion start codon.(A) Bright-field microscopy of a nodule stained for GUS activity. Af-ter staining for 12 to 16 hr, nodules were cut in half, and the interiorface was photographed. Activity is indicated by the blue color. NoteGUS activity throughout nitrogen-fixing zone III. Little or no activityis evident in meristem zone I, infection zone II. or the nodule cortex.(B) Dark-field microscopy of a nodule stained for GUS activity. Afterstaining for 12 to 16 hr, nodules were cut into serial sections (4 ^mthick) and evaluated at a magnification of x10. GUS activity, whichappears red in dark-field microscopy, was detected in infected anduninfected cells of nitrogen-fixing zone III and is evident in noduleparenchyma cells. Little or no activity is found in meristem zone I, in-fection zone II, or the cortex. Arrows identify infected (in) and unin-fected (un) cells.esz, early symbiotic zone; in, infected cells; iz, infection zone; Isz,late symbiotic zone; m, merislem; nc, nodule cortex; np, nodule pa-renchyma; un, uninfected cells; vb, vascular bundle. Zonation is ac-cording to Vasse et al. (1990): zone I, meristem; zone II, infection; *,interzone II to III; zone III, nitrogen fixing.

adapted leaves, we made cuttings from four independenttransformants and evaluated GUS activity in leaves after 3days of darkness. As shown in Table 2, GUS activity in-creased in the leaves of all transformants. The increaseranged from 1.3-fold in plants derived from AS17 to nearlysixfold in AS2-derived plants. It should be noted that all ofthe cuttings used in this experiment had intense blue stain-ing in nodules both at the beginning and end of the darktreatment (data not shown).

DISCUSSION

Biosynthesis of Asn, the primary product of N2 fixation in tem-perate legumes (Vance, 1990) and the predominant nitrogentransport product in many plant species (Sieciechowicz et al.,1988; Lea et al., 1990), is catalyzed by AS (Lam et al., 1995).The AS enzyme has been recalcitrant to study because of itslow activity in cell-free preparations and inherent in vitro in-stability (Streeter, 1977; Stulen and Oaks, 1977; Lam et al.,1995). However, in recent years, molecular studies of severalcDNAs for this enzyme have provided new insight into regu-lation of this critical step in nitrogen assimilation. In thisstudy, we extend the understanding of AS by producing an-tibodies to the enzyme protein and by characterizing an al-falfa gene encoding the nodule-enhanced form of this enzyme.Moreover, we document that the 5' upstream flanking re-gion of AS controls expression in root nodules and in dark-exposed leaves.

Although multiple AS cDNAs, presumably derived fromdistinct genes, have been identified in Arabidopsis (Lam et al.,1995), pea (Tsai and Coruzzi, 1990), and trefoil (Waterhouseet al., 1996), we found little evidence to support a secondAS gene in alfalfa. Whereas there are 23 nucleotide differ-ences between the transcribed region of the AS18 gene andpAS13 cDNA, the deduced protein encoded by the AS18gene is identical to that deduced from the AS cDNA. Thesedata, coupled with the fact that 20 independent AS cDNAs(10 from nodules and 10 from roots) were 99% identical withno changes in deduced amino acids, suggest that the AS18gene and that represented by AS13 correspond to differentalleles of a single AS gene. Allelic variation as reported forother alfalfa genes (Gregerson et al., 1994; Vance et al.,1995) would be expected in this outcrossing tetraploidspecies.

The deduced amino acid sequence of alfalfa AS is similarto those reported previously for other plant AS proteins (Fig-ures 1 and 2), suggesting that the enzyme is a Gin-depen-dent form of AS. This interpretation is consistent with 15N-and 14C-labeling studies (Fujihara and Yamaguchi, 1980;Snapp and Vance, 1986; Ta et al., 1988) and enzyme kineticdata (Lea and Fowden, 1975; Rognes, 1975; Stulen et al.,1979; Ta et al., 1988). The N-terminal portion of alfalfa AScontains the invariant Cys-2 residue, which has been shownthrough site-directed mutagenesis of the human AS gene tobe essential for Gin-dependent AS activity (Van Heeke andSchuster, 1984). In addition, the Cys-2-Asp-29-His-101 cat-alytic triad postulated to be critical for Gin binding in purF-type Gin amidotransferases (Mei and Zalklin, 1989) is foundat positions 2, 104, and 34, respectively, in alfalfa AS. How-ever, the lack of His-101 residues in yeast Asn1 and Asn2 aswell as the Gin-dependent E. coli AsnB raises questions aboutwhether this member of the catalytic triad is absolutely re-quired for Gin-dependent activity.

Even though Gin appears to be the preferred substrate forAS, most studies indicate that the enzyme from plants and

1350 The Plant Cell

Table 1. Organ-Specific GUS Activity in Transgenic Alfalfa Containing the GUS Gene Driven by the AS Promoter

GUS Activity (pmol min-l mg protein-l)

Plantsa

Organ TAS 1 TAS2 ' TAS4 TASl 7 Average 35s-GUSb ControlC

Leaf 75 40 152 350 195 26,380 4 Stem 111 70 366 690 350 34,350 9 Flower 44 250 1,378 N Dd 557 9,490 12 Root 294 2,630 1,603 230 1,411 8,150 16 Nodule 5,668 9,900 7,129 3,370 6,566 6,920 52

alndependent transformants TAS1, TAS2, TAS4, and TASl7. b35S-GUS, transgenic alfalfa containing the GUS reporter gene driven by cauliflower mosaic virus 35s promoter. =The control is regenerated nontransgenic alfalfa. ND, not determined.

animals may also utilize NH4+, albeit less efficiently, as a nitrogen donor (Stulen et al., 1979; Pfeiffer et al., 1987; Sieciechowicz et al., 1988). Alfalfa nodule AS can use NH4+ as a substrate, but it has 20-fold less affinity for NH4+ than for Gln (S.N. Twary and P.J. Unkefer, unpublished data). Moreover, the free NH4+ concentration is so low in nodules as to be almost undetectable (Meeks et al., 1978; Streeter, 1989). In a strategy to determine whether NH4+-dependent AS activity could be enhanced in plants, Brears et al. (1993) created a site-specific mutation in the pea AS1 cDNA that specifically deleted the three amino acids required for Gln binding. When this construct driven by the cauliflower mosaic virus 35s promoter was introduced into the heterologous host tobacco, two independent lines of transgenic tobacco showing overexpression of the mutated Gln-dependent AS continued to show high levels of Asn. Although the authors interpreted this result to mean that the transgenic lines could assimilate NH4+ directly into Asn, neither enzyme activity nor substrate specificity were measured for the modi- fied AS expressed in tobacco.

The availability of an AS cDNA clone provided the means to characterize mRNA levels on RNA gel blots and mRNA localization through in situ hybridization. In plants grown under a normal 16-hr-light and 8-hr-dark photoperiod, AS mRNA was most abundant in mature (day 19) effective nodules, with much lower amounts present in roots, stems, cotyledons, and leaves (Figure 3A). Within effective nodules, AS mRNA abundance increased coincident with the increase in nitrogenase activity (Figure 5A). Compared with previous studies in our laboratory showing that mRNA expression for nodule-enhanced forms of aspartate aminotransferase (AAT-2), glutamate synthase (NADH-GOGAT), glutamine synthetase (GS), and phosphoenolpyruvate carboxylase (PEPC) increases before detectable nitrogenase activity (Gantt et al., 1992; Pathirana et al., 1992; Gregerson et al., 1993; Vance et al., 1994), the increase in AS mRNA is dekyed by 24 hr. This delay could reflect the position of AS in the biosynthetic pathway for Asn. Synthesis of Asn first requires initial assimilation of NH4+ by the GWGOGAT cycle into Gln; then Asp is

derived from oxaloacetate via PEPC and AAT, with AS cata- lyzing the final step of Asp to Asn (Lea et al., 1990; Vance et al., 1994; Lam et al., 1995). Alternatively, AS expression may respond to regulatory signals that are formed later than those affecting GS, GOGAT, AAT-2, and PEPC. Although reduced in amount compared with effective nodules, the synthesis of AS mRNA in both plant genecontrolled and bacterially controlled ineffective nodules (Figures 5 and 6) raises questions regarding the signal(s) involved in nodule-enhanced AS expression.

Developmentally, all of the ineffective nodules evaluated by RNA blot analysis in Figures 5 and 6 attain a relatively similar stage, which equates to a day 7 or 8 effective nodule (Hirsch et al., 1983; Virts et al., 1988; Egli et al., 1989; Yarosh et al., 1989; Driscoll and Finan, 1993). Bacteria are released from infection threads, infected and uninfected cells can be found in the nodule interior, and small amounts of Lb can be detected. Although we do not know where the

Table 2. Effect of Dark Adaptation on GUS Activity in Leaves from Transgenic Alfalfa Containing the GUS Gene Driven by the AS Promoter

Transgenic Planta Cuttings Analyzedb Average Fold lncrease ~~

TAS2 12 5.8 rt 2.8 TAS4 4 4.1 rt 1.0 TAS 1 5 1.7 rt 0.6 TASl 7 14 1.3 2 0.4

a Four independent transformants were analyzed by preparing cuttings and analyzing GUS activity in leaves immediately before transferring to total darkness. The same cuttings were assayed again after 3 days in total darkness, and the fold increase in activity was determined. bCuttings from each plant were grown for 6 weeks in vermiculite inoculated with R. meliloti 102F51. GUS staining in nodules was verified for each cutting. Cuttings were then transferred to total darkness for 3 days, and leaves were assayed for 4-methyllumbelliferyl 8-o-glucuronide activity.


lesion occurs in in,Sar, the mutations in the bacterial strains affect the following: in strain T202, oxygen regulation is affected; in G456, malic enzyme is affected; in 1491, nitrogenase H is affected; and in F642, dicarboxylic acid uptake is affected. Thus, the events giving rise to ineffectiveness occur concurrent with bacteroid differentiation. In each example of ineffective nodules, AS mRNA accumulated to approximately the same level as that found in day 9 effective nodules, suggesting that the initial signal for AS induction is related to events late in nodule development, perhaps asso- ciated with bacteroid release and development. This factor must either increase in the effective symbiosis, resulting in even further enhanced expression of AS, or remain relatively constant or decrease in ineffective nodules. Alternatively, the continued increase in AS mRNA in effective nodules occurring after day 9 may involve a second separate signal as- sociated with N,-fixing nodules and/or N metabolites, which is absent in the ineffective nodules.

In situ hybridization data taken from seria1 sections of >20 effective Saranac and 20 in,Sar nodules further extend our understanding of AS expression by showing the cellular distribution of AS mRNAs (Figure 7). In situ hybridization with longitudinal sections of indeterminate alfalfa nodules allows for direct assessment of mRNA not only in various cell types but also at different stages of development (de Billy et al., 1991). AS hybridization was initially detected in the first cell layer adjacent to the interzone II to 111 (Figure 7); thus, the expression pattern for AS is quite similar to that described for the late nodulins pGS100 (Temple et al., 1995) and Lb (de Billy et al., 1991). However, unlike GSl O0 and Lb mRNAs that were detected only within infected cells, AS expression in effective alfalfa nodules was fairly uniform in both infected and uninfected cells of the nitrogen-fixing zone and the parenchyma. The detection of AS in uninfected cell types strengthens our interpretation that the initial signal for AS expression in nodules is unrelated to the presence of func- tional nitrogenase. Furthermore, the presence of AS transcripts in cell types that do not contain bacteroids has important implications for nitrogen assimilation.

For AS to function catalytically, there must be a source of Asp and Gln or possibly NH4+ in the uninfected cell types. Gln could occur in these cells through synthesis via GS or diffusion from the infected cells. Whether GS occurs in uninfected cells of nodules is not yet resolved. Two alfalfa GS genes appear to be expressed only in infected cells (Temple et al., 1995). However, the expression patterns of other alfalfa GS genes remain to be evaluated. Experiments using promoter-GUS chimeric genes transformed into heterologous hosts have shown that the pea GS3A gene and the common bean gln-y gene are active in infected cells of alfalfa and Lotus nodules, respectively. By contrast, the common bean gln-p and the soybean GS20 genes are active in parenchyma and uninfected cells of Lotus (Forde et al., 1989; Mia0 et al., 1991).

lmmunogold localization studies have shown that common bean nodule-enhanced GSnl is localized to both infected

and uninfected cells of bean nodules (Datta et al., 1991). Likewise, recent immunogold localization studies from our laboratory have shown that nodule-enhanced AAT-2 and PEPC proteins, both of which are important for synthesis of Asp, are present in infected and uninfected cells (Robinson et al., 1994, 1996). Inclusively, these observations suggest that uninfected cells and perhaps even nodule parenchyma have a full complement of enzymes required for assimilation of nitrogen into Gln and Asn.

lndirect evidence indicates that Gln may move across cell types in nodules. Amino acid analysis of nodules and xylem sap of effective and ineffective alfalfa, Lotus, and pea (Maxwell et al., 1984; Rosendahl et al., 1990) show that effective plants synthesize i4C-labeled Gln in nodules, and some of this Gln can move into xylem sap. In comparison, little Gln can be detected in nodules and xylem sap of ineffective plants. Because most synthesis of Gln is thought to occur in the infected cells (Forde et al., 1989; Streeter, 1991), the presence of Gln in xylem sap reflects movement from the symbiotic zone (zone 111) to vascular bundles embedded in the nodule parenchyma.

Promoter-GUS chimeric reporter gene results support the idea that AS78 expression is related to root nodulation. GUS staining indicated that the AS promoter is active mainly in nodules of alfalfa. Within nodules, GUS staining patterns were very similar to mRNA expression patterns obtained by in situ hybridization. The 5' upstream region of AS18 directed high GUS activity to infected and uninfected cells of the symbiotic zone and the nodule parenchyma with no activity in the outer cortex, meristem, and invasion zone. In vitro assays of GUS activity in various organs of transgenic alfalfa confirmed the GUS staining patterns observed visu- ally (Table 1). Analysis of four independent transformants showed GUS activity to be considerably higher in nodules than any other tissue. Thus, the pattern of GUS expression regulated by the putative promoter region of AS78 is consistent with the pattern of AS transcripts measured by RNA blots and in situ hybridization. These data suggest that the expression of AS is controlled at least in part at the transcriptional level in root nodules, and this control is mediated through signals interacting with elements in the 5' upstream region of the AS18 gene.

Root nodules are the predominant site for AS expression in alfalfa plants grown in a normal photoperiod; however, when plants are placed in 24 hr of continuous darkness, AS mRNA and protein increase substantially in leaves (Figures 4A and 4B). As dark-treated leaves are exposed to light, within 6 hr the amount of AS mRNA rapidly decreases. Iden- tical results were obtained whether the blots were hybridized with an AS13 3' untranslated-specific probe, a 5 ' untranslated-specific probe, or a coding region probe. These results suggest that the AS gene, which is repressed by light in alfalfa leaves, is the same as that which is enhanced in root nodules. Our observations are consistent with and further confirm those reported for pea AS1 (Tsai and Coruzzi, 1990, 1991) and Arabidopsis AS (Lam et al., 1994). These

1352 The Plant Cell

studies show that AS expression is repressed by light. In efforts to extend our understanding of light control of AS expression in leaves, cuttings of transgenic alfalfa plants containing the AS promoter-GUS chimeric reporter gene were exposed to darkness, and GUS activity was measured in leaves (Table 2). The results show that the 2.7-kb fragment of the 5’ flanking region of AS78 was active in dark- adapted leaves, further confirming that the putative promoter of AS78 controls expression in both nodules and leaves. These observations suggest that there are cis-acting elements in the 5‘ upstream region of AS78 that respond to light or dark signals from leaves. Whether the element(s) controlling expression in leaves is the same as that controlling expression in nodules awaits AS78 deletion analysis. However, there is precedence for the same or similar elements of nitrogen-assimilating genes controlling expression in leaves and nodules. A 132-bp element in the pea GS3A gene has been shown to function in alfalfa leaves and nodules and was qualitatively equivalent to the 1800-bp “full- length” promoter (Brears et al., 1991).

lmprovement of nitrogen metabolism has been a long- term goal in agricultura1 research (Vance and Graham, 1995). lsolation and characterization of the genes involved in nitrogen assimilation provide tools for new strategies to improve nitrogen metabolism through direct gene manipula- tion. We have previously isolated the alfalfa genes encoding nodule-enhanced AAT-2 (Gregerson et al., 1994), PEPC (Pathirana et al., 1992), NADH-GOGAT (Vance et al., 1995), and GS (C.P. Vance, R.G. Gregerson, and J.S. Gantt, unpublished data). With the isolation of a gene encoding AS, we are now able to initiate experiments to modify nitrogen metabolism in alfalfa. Currently, transgenic plants containing con- structs designed either to overexpress or to underexpress these enzymes are being generated. It will be important and informative to assess growth and metabolism in these transformants. Direct genetic modification of alfalfa with transgene technology may result in enhanced nitrogen assimilation, which could have significant benefits for agriculture by reduc- ing the need for and use of industrial nitrogen fertilizer.

METHODS

Plant Material and Bacterial Strains

Seeds of wild-type effective alfalfa (Medicago sativa) cultivar Saranac and its mutant “ineffective” Saranac (in,Sar), which contains a single recessive mutation that results in early senescing ineffective nodules (Peterson and Barnes, 1981), were obtained from D.K. Barnes (U.S. Department of Agriculture-Agricultura1 Research Service, St. Paul, MN). The fact that alfalfa is an outcrossing tetraploid species pre- cludes the formation of isogenic lines; however, S O % of the in,Sar genotype is from the Saranac background.

In most experiments, plants were grown in greenhouse sand benches inoculated with effective Rhizobium meliloti 102F51, as described previously (Egli et al., 1989). In the experiments examining

asparagine synthetase (AS) expression in bacterially conditioned ineffective nodules, effective Saranac seeds were planted in pots of sterile sand and inoculated with one of four ineffective R. meliloti strains, T202, G456, 1491, and F642, whose relevant genotypes were summarized previously (Gregerson et al., 1993; Vance et al., 1995). In all cases, the sand was inoculated on the day when seeds were planted, designated as day O. For experiments analyzing developmental expression of AS, roots (day 5) or nodules (days 8, 9, 12, 19, and 33) were collected by hand, placed on ice, weighed, and used immediately for either protein or RNA extraction. In dark induction experiments, sterile Saranac seeds were planted in flats of unin- oculated sterile sand (40 x 20 X 1 O cm) in a growth chamber with an 18-hr-light and 6-hr-dark cycle at 21°C for 9 days and then continuous light for 2 days to minimize circadium rhythm effects. Plants were then placed in black boxes in a dark growth chamber for either 1 or 3 days. For light treatments, 3-day-old dark-adapted plants were removed from dark boxes and exposed to continuous light for either 6 or 24 hr. Before initiation of the dark treatments, plants were watered with sterile half-strength Hoagland’s solution with 1 O mM KNO, every other day.

Various alfalfa organs, including uninfected roots, nodules, stems, leaves, and cotyledons, were hand collected onto ice and used immediately for either RNA isolation or protein extraction, except that the leaves and roots of the dark-treated plants were harvested under green light and immediately frozen in liquid NP.

RNA lsolation and Gel Blot Analysis

Total RNA was isolated from freshly collected tissues, such as roots, nodules, stems, leaves, and cotyledons, as described by Gregerson et al. (1993). Poly(A)+ RNA was obtained by a single pass over an oligo(dT) column (Gibco BRL).

RNA gel blotting was performed essentially as described by Gantt et al. (1 992). Either 1 O to 15 pg of total RNA or 2 pg of poly(A)+ RNA was separated on denaturing agarose gels, transferred to ZetaProbe (Bio-Rad) membranes, and hybridized with a 32P-labeled AS cDNA fragment, as described previously (Pathirana et al., 1992). Blots were independently repeated two to three times. Labeling of all RNA blots with 3zP-labeled polyuridine showed that RNA loading in each lane was comparable.

lsolation of AS cDNA and Genomic DNA Clones

An oligo(dT)-primed alfalfa nodule cDNA library (Gregerson et al., 1993) constructed in the vector hgt22 by using a cDNA synthesis kit (Life Technologies, Grand Island, NY) was screened with primaty antisera prepared against the alfalfa AS protein (Twary et al., 1994) by using horseradish peroxidase-conjugated goat anti-rabbit antibody as the secondary stain. From a screening of -5 X 1 O4 plaque-forming units of the cDNA library, six recombinant antigen-producing bacteriophages were isolated and purified, the sizes of the cDNA inserts were determined, and the inserts were subcloned into pBluescript KS- (Stratagene) and partially sequenced. The nucleotide sequence of the largest cDNA, designated AS13 (-2.2 kb), was determined from nested deletion fragments (Henikoff, 1987) by the dideoxy ter- mination method using Sequenase (Amersham). Both DNA strands were sequenced, and the overlapping regions were determined using the Genetics Computer Group (Madison, WI) gel program. A total of 5 x 105 recombinant bacteriophages from an alfalfa genomic Ii- brary was screened with the AS13 cDNA, essentially as described by


Gregerson et al. (1994). Four clones were obtained and partially characterized. The largest of them, MS78, containing both the 5' and 3' end sequences of the AS13 cDNA, was sequenced as described above.

nodules were separated by electrophoresis in a 10% SDS-polyacryl- amide gel and electrophoretically trgnsferred to nitrocellulose. The resulting filters were probed with AS antisera (Twary et al., 1994), as described previously (Miller et al., 1987). Protein blots were independently repeated two or three times.

Plasmid Construction and Alfalfa Transformation In Situ Hybridization

A 2.7-kb flanking region of AS78 starting from 15 nucleotides upstream of the putative translational start codon was excised by digestion with restriction endonucleases (Hindlll and Xmnl) and in- serted into the Hindlll and Smal sites in plant transformation vector pBI101.2 (Jefferson 1987), producing a transcriptional fusion construct pBICAS-GUS.

pBICAS-GUS was introduced into Agrobacterium strain LBA4404 by electroporation. Sequence analysis of pBICAS-GUS demonstrated that the insert was correctly cloned. Transgenic M. sativa cv Regen SY plants were obtained via Agrobacterium-mediated transformation, essentially by the method of Austin et al. (1996). After root- ing, plants were inoculated with R. meliloti 102F51. Histochemical analysis of p-glucuronidase (GUS) activity in nodules, roots, leaves, or flowers was performed according to the protocol of Stomp (1992). Stained nodules were fixed, dehydrated, and embedded in London Resin White (Polysciences Inc., Warrington, PA), as described by Robinson et al. (1996). Sections (4 pm) were cut using an MT-2000 UI- tramicrotome (Research and Manufacturing Co., Tucson, Az), mounted on slides, and observed using dark-field microscopy. The blue color seen with X-gluc staining appears red in dark-field microscopy.

Fluorometric GUS assays were performed essentially as described by Jefferson (1987). Fresh or frozen nodules, stems, flowers, leaves, and roots (-25 mg) of transgenic alfalfa were ground in lysis buffer. Protein concentrations were determined by the Bio-Rad method. Approxi- mately 5 to 10 pg of protein was incubated with 1 mM 4-methylumbelliferyl p-D-glucuronide in 100 pL of lysis buffer at 37°C for 10, 20, and 30 min. After reactions were stopped with 2 mL of 0.2 M sodium carbonate, fluorescence of the 4-methylumbelliferone product was determined in a minifluorometer (model TKO-I 00; Hoefer Scientific, San Francisco, CA). Fhorescence of a solution of 0.1 mM 4-methylumbelliferone in 0.2 M sodium carbonate was used for calibration.

Plant DNA Extraction and DNA Gel Analysis

Genomic DNA was isolated from 12-day-old alfalfa cultivar Saranac seedlings as described by Shure et al. (1983). For DNA gel analysis, approximately equimolar amounts of genomic DNA from alfalfa (18 pg) and R. meliloti (0.01 pg) were digested with a variety of restriction enzymes (EcoRI, EcoRV, or Xbal), electrophoresed through an 0.8% agarose gel, and transferred to an Immobilon-N membrane (Milli- pore, Bedford, MA), as recommended by the manufacturer. The membrane was hybridized with a 32P-labeled AS13 cDNA, as described previously (Gregerson et al., I 993). Hybridization and washing were performed at 65°C for high stringency and 50°C for low stringency, respectively.

In situ hybridization was performed essentially as described by Van de Wiel et al. (1990). At least 20 day 12 Saranac and in,Sar nodules were fixed with 4% paraformaldehyde and 0.25% glutaraldehyde in 50 mM sodium phosphate buffer, pH 7.2, at room temperature. The tissues were then rinsed twice in the same buffer and twice in water and dehydrated in a graded ethanol series. After the absolute ethanol was replaced with xylene, tissues were embedded in paraffin. The embedded tissues were sectioned (7 pm thick) and affixed to poly-~- lysine-coated slides. Hybridization was performed at 46°C overnight, and the final washing condition was 0.1 x SSC (1 x SSC is 0.15 M NaCI, 0.015 M sodium citrate) and 1 mM DTT at 50°C for 20 min. Af- ter 1 to 4 weeks of exposure at 4"C, the emulsion was developed, and the sections were then stained with 0.05% toluidine blue O, dehydrated, and mounted with Permount (Fisher Scientific Co., Fair Lawn, NJ). Sections were viewed and photographed with a Labophot mi- croscope (Nikon Inc., Garden City, NY) equipped with dark- and bright-field optics.

An Xmnl fragment encoding amino acids 172 to 364 was excised from AS13, ligated with EcoRV-digested pBluescript KS-, and sequenced to determine its cloning orientation. For antisense RNA probe preparation, the plasmid was cut with Hindlll and transcribed with T7 RNA polymerase. For sense RNA probe preparation, the plasmid was cut with EcoRl and transcribed with T3 RNA polymerase. To prepare a probe derived from the 3' untranslated region of AS13, a 260-bp fragment (from nucleotides 1901 to 2160) was amplified by polymerase chain reaction, subcloned into pBluescript KS-, and sequenced to determine its orientation. For antisense RNA probe preparation, the plasmid was cut with Xbal and transcribed with T7 RNA polymerase; for sense RNA probe preparation, the plasmid was cut with BamHl and transcribed with T3 RNA polymerase. All probes were radioactively labeled with 35S-UTP (1000 Ci mmol-I; Amersham) and degraded to -1 50 nucleotides before hybridization.

Phylogenetic Analysis

AS amino acid sequences were retrieved from GenBank and aligned using the PILEUP program from the Genetics Computer Group soft- ware package. Parsimony analysis was performed for AS polypeptides with and without the deletion of the divergent C-terminal 80 amino acids, respectively. The phylogenetic trees were constructed with the heuristic search option of the phylogenetic analysis program (PAUP) using parimony, version 3.0 (D.L. Swofford, lllinois Natural History Survey, Champaign, IL), using Escherichia coli AsnB as an outgroup.

Protein Extraction and Protein lmmunoblotting ACKNOWLEDGMENTS

Total protein was extracted from different organs of Saranac and in,Sar, as described previously (Gantt et al., 1992). For protein immu- noblotting analysis, 5 pg of protein from Saranac nodules and from 50 to 100 pg of protein from other Saranac tissues and in,Sar

We thank Don Helinski, Ann Hirsch, Brian Driscoll, and Turlough Finan for generously sharing ineffective Rhizobium strains. We also thank Les Szabo for assistance in performing the phylogenetic

1354 The Plant Cell

analysis of AS proteins. This work is supported in part by U.S. De- partment of Agriculture (USDA) Grant No. NRI 94-37305-0575 and National Science Foundation Grant No. IBN-9206890. This article is a joint contribution from the Plant Science Research Unit, USDA, Ag- ricultural Research Service, and the Minnesota Agricultura1 Experiment Station (Paper No. 971 13007, Scientific Journal Series). Mention of a trademark, proprietary product, or vendor does not constitute a guarantee or warranty to the product by USDA and does not imply its approval to the exclusion of other products or vendors that might also be suitable.

Received March 18, 1997; accepted May 29, 1997.

REFERENCES

Andrulis, I.L., Chen, J., and Ray, P.N. (1987). lsolation of human cDNAs for asparagine synthetase and expression in Jensen rat sarcoma cells. MOI. Cell. Biol. 7, 2435-2443.

Austin, S., Bingham, E.T., Mathews, DE, Shahan, M.N., Will, J., and Burgess, R.R. (1996). Production and field performance of transgenic alfalfa (Medicago sativa L.) expressing alpha-amylase and manganese-dependent lignin peroxidase. Euphytica 85,381393,

Brears, T., Walker, E.L., and Coruzzi, G. (1991). A promoter sequence involved in cell-specific expression of the pea glutamine synthetase GS3A gene in organs of transgenic tobacco and alfalfa. Plant J. 1,235-244.

Brears, T., Liu, C., Knight, T.J., and Coruui, G.M. (1993). Ectopic overexpression of asparagine synthetase in transgenic tobacco. Plant Physiol. 103, 1285-1290.

Brouquisse, R., James, F., Pradet, A., and Raymond, P. (1992). Asparagine metabolism and nitrogen distribution during protein degradation in sugar-starved maize root tips. Planta 188,384-395.

Chevalier, C., Bourgeois, E., Just, D., and Raymond, P. (1996). Metabolic regulation of asparagine synthetase gene expression in maize (Zea mays L.) root tips. Plant J. 9, 1-1 1.

Datta, D.B., Cai, X., Wong, P.P., and Triplett, E.W. (1991). Immu- nocytochemical localization of glutamine synthetase in organs of fhaseolus vulgaris L. Plant Physiol. 96,507-512.

Davies, K.M., and King, G.A. (1 993). lsolation and characterization of a cDNA for harvest-induced asparagine synthetase from Aspar- agus officinalis L. Plant Physiol. 102, 1337-1340.

Davies, K.M., Seelye, J.F., Irving, D.E., Borst, W.M., Hurst, P.L., and King, G.A. (1996). Sugar regulation of harvest related genes in Asparagus. Plant Physiol. 11 1, 877-883.

de Billy, F., Barker, D.G., Gallusci, P., and Truchet, G. (1991). Leghemoglobin gene transcription is triggered in a single cell layer in the indeterminate nitrogen-fixing root nodule of alfalfa. Plant J.

Downs, C.G., Pogson, B.J., Davies, K.M., and Amira, E.C. (1995). An asparagine synthetase cDNA clone (GenBank X84448) from broccoli (PGR95-016). Plant Physiol. 108, 1342.

Driscoll, B.T., and Finan, T.M. (1 993). NAD+-dependent malic enzyme of Rhizobium meliloti is required for symbiotic nitrogen fixation with Medicago sativa. MOI. Microbiol. 7, 865-875.

1,27-35.

Egli, M.A., Griffith, S.M., Miller, S.S., Anderson, M.A., and Vance, C.P. (1 989). Nitrogen assimilatoty enzyme activities and enzyme protein during development and senescence of effective and plant gene-controlled ineffective alfalfa nodules. Plant Physiol. 91,

Forde, B.G., Day, H.M., Turton, J.F., Shen, W.-i., Cullimore, J.V., and Oliver, J.E. (1 989). Two glutamine synthetase genes from fhaseolus vulgaris L. display contrasting developmental and spa- tia1 patterns of expression in transgenic Lotus corniculatus plants. Plant Cell I, 391-401.

Franssen, H.J., Vijn, I., Yang, W.-C., and Bisseling, T. (1992). Developmental aspects of the Rhizobium-legume symbiosis. Plant MOI. Biol. 19, 89-107.

Fujihara, S., and Yamaguchi, M. (1980). Asparagine formation in soybean nodules. Plant Physiol. 66, 139-141.

Gantt, S.J., Larson, R.J., Farnham, M.W., Pathirana, S.M., Miller, S.S., and Vance, C.P. (1 992). Aspartate aminotransferase in effective and ineffective alfalfa nodules. Cloning of a cDNA and determination of enzyme activity, protein, and mRNA levels. Plant Physiol. 98, 868-878.

Genix, P., Bligny, R., Martin, J.-B., and Douce, R. (1990). Transient accumulation of asparagine in sycamore cells after a long period of sucrose starvation. Plant Physiol. 94, 717-722.

Gregerson, R.G., Miller, S.S., Twary, S.N., Gantt, J.S., and Vance, C.P. (1 993). Molecular characterization of NADH-dependent glutamate synthase from alfalfa nodules. Plant Cell 5, 215-226.

Gregerson, R.G., Miller, S.S., Petrowski, M., Gantt, J.S., and Vance, C.P. (1 994). Genomic structure, expression and evolution of the aspartate aminotransferase genes of alfalfa. Plant MOI. Biol. 27,387-399.

Henikoff, S. (1 987). Unidirectional digestion with exonuclease III in DNA sequence analysis. Methods Enzymol. 155, 156-1 65.

Hirsch, A.M. (1 992). Developmental biology of legume nodulation. Tansley review No. 40. New Phytol. 122, 21 1-237.

Hirsch, A.M., Bang, M., and Ausubel, F.M. (1983). Ultrastructural analysis of ineffective alfalfa nodules formed by nif::Tn5 mutants of Rhizobium meliloti. J. Bacteriol. 155,362-380.

Huber, T.A., and Streeter, J.G. (1 984). Asparagine biosynthesis in soybean nodules. Plant Physiol. 74, 605-61 O.

Huber, T.A., and Streeter, J.G. (1 985). Purification and properties of asparagine synthetase from soybean root nodules. Plant Sci.

Jefferson, R.A. (1987). Assaying chimeric genes in plants: The GUS gene fusion system. Plant MOI. Biol. Rep. 5, 387-405.

Joy, K.W., Ireland, R.J., and Lea, P.J. (1983). Asparagine synthesis in pea leaves and the occurrence of an asparagine synthetase inhibitor. Plant Physiol. 73, 165-168.

Kim, T.H., Bigot, J., Ourry, A., and Boucand, J. (1993). Amino acid content in xylem sap of regrowing alfalfa (Medicago sativa L.): Relations with N uptake, N2 fixation and N remobilization. Plant Soil 149, 167-174.

King, G.A., Woollard, D.C., Irving, D.E., and Borst, W.M. (1990). Physiological changes in asparagus spear tips after harvest. Physiol. Plant. 80,393-400.

Lam, H.-M., Peng, S.S.-Y., and Coruzzi, G.M. (1994). Metabolic regulation of the gene encoding glutamine-dependent asparagine synthetase in Arabidopsis thaliana. Plant Physiol. 106, 1347-1 357.

898-904.

42, 9-1 7.


Lam, H.-M., Coschigano, K., Schultz, C., Melo-Oliveria, R., Tjaden, G., Oliveria, I., Ngai, N., Hsieh, M.-H., and Coruzzi, G. (1995). Use of Arabidopsis mutants and genes to study amide amino acid biosynthesis. Plant Cell 7, 887-898.

Lea, P.J., and Fowden, L. (1975). The purification and properties of glutamine-dependent asparagine synthetase isolated from Lupi- nus albus. Proc. R. SOC. Lond. Ser. B 192, 13-26.

Lea, P.J., Robinson, S.A., and Stewart, G.R. (1990). The enzymol- ogy and metabolism of glutamine, glutamate, and asparagine. In The Biochemistry of Plants, Vol. 16, lntermediary Nitrogen Metab- olism, B.J. Miflin and P.J. Lea, eds (San Diego, CA: Academic Press), pp. 121-159.

Maxwell, C.A., Vance, C.P., Heichel, G.H., and Stade, S. (1984). C02 fixation in alfalfa and birdsfoot trefoil root nodules and parti- tioning of 14C to the plant. Crop Sci. 24,257-264.

McGrath, R.B., and Coruzzi, G.M. (1991). A gene network controlling glutamine and asparagine biosynthesis in plants. Plant J. 1,

Meeks, J.C., Wolk, C.P., Schilling, N., Shaffer, P.W., Avissar, Y., and Chien, W.4. (1978). lnitial organic products of fixation of [13N]dinitrogen by root nodules of soybean (Glycine max). Plant Physiol. 61, 980-983.

Mei, B., and Zalkin, H. (1 989). A cysteinehistidine-aspartate catalytic triad is involved in glutamine amide transfer function in purf-type glutamine amidotransferases. J. Biol. Chem. 264, 16613-16619.

Miao, G.-H., Hirel, B., Marsolier, M.C., Ridge, R.W., and Verma, D.P.S. (1 991). Ammonia-regulated expression of a soybean gene encoding cytosolic glutamine synthetase in transgenic Lotus corniculatus. Plant Cell 3, 11-22,

Miller, S.S., Boylan, K.L.M., and Vance, C.P. (1987). Alfalfa root nodule carbon dioxide fixation. 111. lmmunological studies of nodule phosphoenolpyruvate carboxylase. Plant Physiol. 84, 501-508.

Moyle, R.L., Davies, K.M., King, G.A., and Farnden, K.J.F. (1996). Nucleotide sequence of the asparagine synthetase gene (accession no. X99552) from Asparagus officinalis L. (PGR 96-096). Plant Physiol. 112, 1397.

Nakamura, M., Yamada, M., Hisota, Y., Sugimoto, K., Oka, A., and Takanami, M. (1981). Nucleotide sequence of the asnA gene coding for asparagine synthetase of E. coli K-12. Nucleic Acids Res. 9,46694676.

Oaks, A., and Ross, D.W. (1984). Asparagine synthetase in Zea mays. Can. J. Bot. 62, 68-73.

Pathirana, S.M., Vance, C.P., Miller, S.S., and Gantt, J.S. (1992). Alfalfa root nodule phosphoenolpyruvate carboxylase: Character- ization of the cDNA and expression in effective and ineffective nodules. Plant MOI. Biol. 20, 437-450.

Peterson, M.A., and Barnes, D.K. (1 981). lnheritance of ineffective nodulation and nonnodulation in alfalfa. Crop Sci. 21, 611-616.

Pfeiffer, N.E., Mehlhaff, P.M., Wylie, D.E., and Schuster, S.M. (1987). Topographical separation of the catalytic sites of asparagine synthetase with monoclonal antibodies. J. Biol. Chem. 262,

Robinson, D.L., Kahn, M.L., and Vance, C.P. (1994). Cellular localization of nodule enhanced aspartate aminotransferase in Medi- cago sativa L. Planta 92,202-210.

275-280.

11 565-1 1570.

Robinson, D.L., Pathirana, S.M., Gantt, J.S., and Vance, C.P. (1 996). lmmunogold localization of nodule-enhanced phosphoenolpyruvate carboxylase in alfalfa. Plant Cell Environ. 19,602-608.

Rognes, S.E. (1975). Glutamine-dependent asparagine synthetase from Lupinus luteus. Phytochemistry 14,1975-1982.

Rosendahl, L., Vance, C.P., and Pederson, W.B. (1990). Products of dark C02 fixation in pea root nodules support bacteroid metabolism. Plant Physiol. 93, 12-19.

Sandal, N.N., Bojsen, K., and Marcker, K.A. (1987). A small family of nodule specific genes from soybean. Nucleic Acids Res. 15,

Scofield, M.A., Lewis, W.S., and Schuster, S.M. (1990). Nucle- otide sequence of Escherichia coli asnB and deduced amino acid sequence of asparagine synthetase B. J. Biol. Chem. 265,

Shapiro, M.B., and Senapathy, P. (1988). RNA splice junctions of different classes of eukaryotes: Sequence statistics and func- tional implication in gene expression. Nucleic Acids Res. 15,

Shure, M., Wessler, S., and Fedoroff, N. (1983). Molecular identifi- cation and isolation of the Waxy locus in maize. Cell 35, 235-242.

Sieciechowicz, K.A., Joy, K.W., and Ireland, R.J. (1988). The metabolism of asparagine in plants. Phytochemistry 27, 663-671.

Snapp, S.S., and Vance, C.P. (1 986). Asparagine biosynthesis in alfalfa (Medicago sativa L.) root nodules. Plant Physiol. 82,390-395.

Stomp, A.M. (1 992). Histochemical localization of p-glucuronidase. In GUS Protocols: Using the GUS Gene as a Reporter of Gene Expression, S.R. Gallagher, ed (San Diego, CA: Academic Press),

Stougaard, J., Joergenson, J.E., Christensen, T., Kuhle, A., and Marcker, K.A. (1 990). lnterdependence of specificity of cis-acting regulatory elements in soybean leghemoglobin lbc3 and N23 gene promoters. MOI. Gen. Genet. 220, 353-360.

Streeter, J.G. (1 989). Estimation of ammonium concentration in the cytosol of soybean nodules. Plant Physiol. 90, 779-782.

Streeter, J.G. (1 991). Transport and metabolism of carbon and nitrogen in legume nodules. Adv. Bot. Res. 18, 129-187.

Streeter, J.M. (1 977). Asparaginase and asparagine transaminase in soybean leaves and root nodules. Plant Physiol. 60, 235-239.

Stulen, I., and Oaks, A. (1977). Asparagine synthetase in corn roots. Plant Physiol. 60, 680-683.

Stulen, T., Israelstam, G.F., and Oaks, A. (1979). Enzymes of asparagine synthesis in maize roots. Planta 146, 237-241.

Ta, T.-C., Macdowall, F.D.H., and Faris, M.A. (1988). Metabolism of nitrogen fixed by nodules of alfalfa (Medicago sativa L.). II. Asparagine synthesis. Biochem. Cell. Biol. 66, 1349-1 354.

Temple, S.J., Heard, J., Ganter, G., Dunn, K., and Sengupta- Gopalan, C. (1 995). Characterization of a nodule-enhanced glutamine synthetase from alfalfa: Nucleotide sequence, in situ localization, and transcript analysis. MOI. Plant-Microbe Interact.

Tsai, F.-Y., and Corurzi, G.M. (1990). Dark-induced and organ- specific expression of two asparagine synthetase genes in Pisum sativum. EMBO J. 9,323-332.

1507-1 51 9.

12895-1 2902.

71 55-71 74.

pp. 103-1 13.

8,218-227.

1356 The Plant Cell

Tsai, F.-Y., and Coruzzi, G.M. (1 991). Light represses transcription of asparagine synthetase genes in photosynthetic and nonphoto- synthetic organs of plants. MOI. Cell. Biol. 11, 4966-4972.

Twary, S.N., Gregerson, R.G., Unkefer, P.J., and Vance, C.P. (1 994). ldentification and characterization of a cDNA for alfalfa nodule-enhanced asparagine synthetase. Plant Physiol. 105 (suppl.), 152 (abstr.).

Vance, C.P. (1 990). Symbiotic nitrogen fixation: Recent genetic advances. In The Biochemistry of Plants, Vol. 16, lntermediary Nitrogen Metabolism, B.J. Miflin and P.J. Lea, eds (San Diego, CA: Academic Press), pp. 43-88.

Vance, C.P., and Graham, P.H. (1995). Nitrogen fixation in agriculture. In Nitrogen Fixation: Fundamentals and Applications, I.A. Tikhonovich, N.A. Provorov, V.I. Romanov, and W.E. Newton, eds (Dordrecht, The Netherlands: Kluwer Academic Publishers), pp.

Vance, C.P., Gregerson, R.G., Robinson, D.L., Miller, S.S., and Gantt, J.S. (1994). Primary assimilation of nitrogen in alfalfa nodules: Molecular features of the enzymes involved. Plant Sci. 101, 51-64.

Vance, C.P., Miller, S.S., Gregerson, R.G., Samac, DA., Robinson, D.L., and Gantt, J.S. (1 995). Alfalfa NADH-dependent glutamate synthase: Structure of the gene and importance in symbiotic N, fixation. Plant J. 5,215-226.

Van de Wiel, C., Scheres, B., Franssen, H., Van Lierop, M.J., Van Lammeren, A., Van Kammen, A., and Bisseling, T. (1990). The early nodulin transcript ENOD2 is located in the nodule paren-

77-87.

chyma (inner cortex) of pea and soybean root nodules. EMBO J. 9, 1-7.

Van Heeke, G., and Schuster, S.M. (1989). The N-terminal cysteine residue of human asparagine synthetase is essential for glutamine-dependent activity. J. Biol. Chem. 264, 19475-1 9478.

Vasse, J., De Billy, F., Camut, S., and Truchet, G. (1 990). Correla- tion between ultrastructural differentiation of bacteroids and nitrogen fixation in alfalfa nodules. J. Bacteriol. 172, 4295-4306.

Virts, E.L., Stanfield, S.W., and Helinski, D.R. (1988). Common regulatory elements control symbiotic and microaerobic induction of nifA in Rhizobium meliloti. Proc. Natl. Acad. Sci. USA85,3062-3065.

Waterhouse, R.N., Smyth, A.J., Massonneau, A., Prosser, I.M., and Clarkson, D.T. (1 996). Molecular cloning and characterization of asparagine synthetase from Lotus japonicus: Dynamics of asparagine synthesis in N-sufficient conditions. Plant MOI. Biol.

Welters, P., Metz, B., Felix, O., Palme, K., Szczyglowski, K., and deBruijn, F.J. (1 993). lnteraction of a rhizobial DNA-binding protein with the promoter region of a plant leghemoglobin gene. Plant Physiol. 102, 1095-1 107.

Yarosh, O.K., Charles, T.C., and Finan, T.M. (1989). Analysis of C4-dicarboxylate transpor? genes in Rhizobium meliloti. MOI. Microbiol. 3, 813-823.

Zhang, Y.P., Lambert, M.A., Cairney, E.L., Willis, D., Ray, P.N., and Andrulis, I.L. (1989). Molecular structure of the human asparagine synthetase gene. Genomics 4, 259-265.

30,883-897.

Shi, L., Twary, S.N., Yoshioka, H., Gregerson, R.G., Miller, S.S. ...

Documents