sequence of three bronze alleles of maize and correlation with the genetic fine … · ·...

Copyright 0 1988 by the Genetics Society of America

Sequence of Three bronze Alleles of Maize and Correlation With the Genetic Fine Structure

E. J. Ralston, J. J. English and H. K. Dooner Advanced Genetic Sciences, 6701 San Pablo Avenue, Oakland, California 94608

Manuscript received October 7 , 1987 Accepted February 1, 1988

ABSTRACT The genomic sequences of three bronze alleles from Zea mays, Bz-McC, Bz-W22 and bz-R, are

presented together with their flanking sequences. The bronze locus encodes UDPglucose flavonoid glucosyl-transferase (UFGT), an anthocyanin biosynthetic enzyme. The wild-type alleles Bz-McC and Bz-W22 condition purple phenotypes in the seed and plant, while bz-R conditions a bronze color. A full length cDNA corresponding to the Bz-McC allele was cloned and sequenced. Primer extension and RNase protection experiments were used to verify the 5' end of the bronze transcript. The Bz- McC allele has a 1416-bp coding region, a 100-bp intron and an approximately 83-bp 5' leader. Upstream of the message initiation site the sequences CTAACT and AATAAA occupy the positions where the eukaryotic consensus CCAAT and TATA boxes are normally found. The alleles Bz-McC and bz-R each have different large insertions with characteristics of transposable elements in their 5' flanking regions. The bz-R allele is distinguished by a 340-bp deletion starting within the intron and including 285 bp of the second exon. The Bz-McC and Bz-kV22 isoalleles are known to differ in two genetically defined locations. The uts and uqv sites from the Bz-McC allele condition, respectively, lowered thermostability for the UFGT enzyme and increased amount of UFGT activity when compared with the corresponding sites in the Bz-W22 allele. The uts site maps to a region of the gene encoding two adjacent amino acid differences, either or both of which might alter the thermostability of the UFGT enzyme. The difference in UFGT levels conditioned by the uqu site is shown here to be correlated with variation in the bronze mRNA level. A likely cause of this decreased bronze mRNA level in Bz-W22 is a 6-bp duplication near the sequence CTAACT located 74 bp upstream of the bronze message iniation site. This region is therefore tentatively identified as the uqv site.

T HE bronze (bz) locus on the short arm of chromosome 9 of Zea mays encodes the enzyme

UDPglucose flavonoid glucosyl-transferase (UFGT), which catalyzes one of the last steps in anthocyanin biosynthesis: the addition of a glucose moiety to the 3 hydroxyl group of a flavonoid (LARSON and COE 1977; DOONER and NELSON 1977a). The highly tissue- specific expression of the bronze gene is regulated primarily by the duplicate genetic loci R and B (DOONER and NELSON 1977a, 1979a; GERATS et al. 1984). Alleles of R control anthocyanin pigmentation in the aleurone layer and scutellum of the seed, the coleoptile and radicle of the seedling, and the leaf, nodes, pericarp, silks, and anthers of the mature plant (STADLER and FOCEL 1945). Alleles of B control expression predominantly in the leaf sheaths of the mature plant (including the husks and glumes, which are modified sheaths), but also in the leaf blade and stem (EMERSON 192 1 ; COE 1966). Normally only tissue that has been exposed to light becomes pigmented. Two other regulators, C and P1, make anthocyanin

The sequence data presented in this article have been submitted to the EMBLiCenBank Data Libraries under the accession number Y00616.

Genetics 119: 185-197 (May, 1988).

production in the seed and in the plant, respectively, less light dependent (EMERSON 1921; KIRBY and STYLES 1970; CHEN and COE 1973). C has been shown to regulate expression in the seed of other structural genes for anthocyanin biosynthesis besides bz (DOONER 1981, 1983; CONE, BURR and BURR 1986).

The bronze locus has also been the subject of extensive genetic research in the past because its phenotype is easy to score, the regulation of its expression is tissue specific and varied, and, since bz is not involved in a vital pathway, many different types of mutations can be obtained. The bronze locus has been studied extensively as a reporter locus of transposable element induced variation (MCCLIN- TOCK 1951, 1955, 1962; NELSON and KLEIN 1984; ROBERTSON 1985). An attractive feature of this system is that the variegated phenotype typical of transposable elements can be scored in all tissues competent to form pigment, e.g. , aleurone, anthers, husks, cob, and roots) (MCCLINTOCK 1967).

The UFGT enzyme has been purified and studied biochemically and immunologically. The three alleles of interest in this paper, Bz-McC, Bz-W22, and bz-R, have been characterized enzymatically and antigeni-

186 E. J. Ralston, J. J. English and H. K. Dooner

cally for material that would cross-react with UFGT- specific antibodies. The mutation, bz-R, the standard reference allele for the bronze locus (RHOADES 1952) produces no enzymatic activity and no cross-reacting material (CRM-) (DOONER and NELSON 1977b). We have found that it is associated with a 340-bp deletion of the transcribed region (RALSTON, ENGLISH and DOONER 1987). The UFGT enzymes in Bz-W22, the allele from the inbred line W22, and in Bz-McC, a wild-type isoallele from B. MCCLINTOCK'S stocks and the progenitor of the transposable element mutation bz -mZ(Ac) , differ in their thermal stability. In addition, the total amount of UFGT activity in Bz-McC and Bz- W22 differs by two- to threefold. In a large scale fine structure mapping experiment, DOONER (1986) mapped genetically the position of the protein thermostability difference in these alleles to the proximal side of the insertion site of D s in bz -mZ(DI ) , a deletion derivitive of bz-m2 (Ac) (DOONER et al. 1986), and the site responsible for the level of UFGT activity to the centromere-proximal end of the bz locus.

Since UFGT is not an abundant enzyme, traditional methods for cloning the gene involving cDNA or immunological methods are not readily applicable. However, the well characterized transposable element mutations of bronze made it the obvious first plant gene of choice for cloning by transposable element tagging (FEDEROFF, FURTEK and NELSON 1984; DOONER et al. 1985).

To determine the exact 3' and 3' boundaries of the bronze transcript and its intron-exon structure we have cloned and sequenced the genomic DNA corresponding to Bz-McC and a full length cDNA from this allele. In addition we have cloned and sequenced two other bronze alleles, Bz-W.22 and bz-R. As found for other genes in maize (Adh: SACHS et al. 1986; Sh: WERR et al. 1985), there is a large amount of sequence diversity between alleles especially in the region 5' of the transcription initiation site. The natural diversity can be used to help define regions of DNA important for bronze transcript regulation and for the stability of the UFGT protein.

Recently, several other genes in the anthocyanin pathway have been cloned using transposon tagging (a: O'REILLY et al. 1985; c: PAZ-ARES et al. 1986; CONE, BURR and BURR et al. 1986) c2: WIENAND et al. 1986; bz2: THERES, SCHEELE and STARLINGER 1986; MCLALGHLIN and WALBOT 1987). Since c 2 , and probably a and br2, are expressed coordinately with UFGT (DOONER 1983), a comparison of their promoter

"- pAGS 526 pAGS 528 pAGS 551

5' - 3' - 1 kb

FIGURE 1 .-Restriction map of the Bz-McC allele showing location of the subclones used as Southern and Northern probes and of the transcribed region. The sequenced region is demarcated by asterisks.

regions might define common regulatory regions. As a step toward this end, the sequence of the 5' region flanking the bronze gene is presented.

MATERIALS AND METHODS

Maize stocks: All of the alleles used in this study were derived from or incorporated into the common genetic background of the inbred W22. The Bz-McC allele was extracted from a C-I sh Bz chromosome obtained from a maize stock of Barbara McClintock by selecting C Sh Bz crossovers from C-I sh BziC Sh bz-R heterozygotes. The Bz allele resulting from this selection was thereafter designated Bz-McC (DOONER and NELSON 1977b). Other wild-type isoalleles derived from MCCLINTOCK stocks may differ from Bz-McC. The bz-R allele is derived from stocks obtained from RHOADES and was designated bz-R by DOONER and NELSON (1977b). This is presumably the mutation that originally defined the bronze locus (RHOADES 1952; RALSTON, ENGLISH and DOONER 1987). The Bz-W22 allele is the normal allele carried in the W22 inbred.

Molecular analysis: DNA extractions, Southern analysis, and genomic library constructions were performed as described previously (DOONER et al. 1983). DNA from a maize stock carrying the Bz-W22 allele was digested with BarnHI, size-fractionated by centrifugation through a 10% to 40% glycerol gradient, purified, ligated to EMBL3 arms and packaged. DNA from a maize stock carrying the bronze allele bz-R was partially digested with Sau3AI treated as above and ligated to EMBL4 arms. Phage libraries were screened with pAGS528, the KpnI to PstI fragment internal to the transcribed region of Bz-McC (Figure 1). Restriction fragments from phage hybridizing to the bronze probe were subcloned into pUC19 (YANISCH-PERRON, VIEIRA and MESS- I N G 1983).

DNA sequencing: Ordered deletions obtained by Bal 3 1 digestions (DEAN et al. 1985) for Bz-McC, Bz-W22, and bz- R were subcloned into mpl8 (MESSING, CREA and SEEBURG 1981). Single strand DNA was sequenced using the dideoxy method (SANGER, NICKLEN and COULSON 1977). Regions of high GC content, where compression of the bands on the sequencing gel made the sequence difficult to read, were resequenced using inosine nucleoside triphosphate in place of guanosine (MILLS and KRAMER 1979)

RNA analysis: RNA isolation and Northern analysis

FIGURE 2.-Comparison of three bronze alleles. The sequence of Bz-McC a l l~ l e (the center h e ) is compared with the sequences of the Bz- W22 ( t o p ) and bz-R (bottom) alleles. Identical bases are indicated by stars, single base differences are shown where they occur and insertions or deletions are shown as gaps. Dashes indicate regions that were not sequenced. Large insertions and deletions are also indicated on the right margin. Direct repeats (DR) and inverted repeats (IR) for the insertions in Bz-McC and bz-R are indicated above the sequence as are certain restriction sites, A mRNA initiation site, the stop codon and poly(A) addition sites. Possible regulatory sequences are indicated by boxes. Asterisks above the Bz-WZ2 sequence mark DNA differences that cause changes in the corresponding protein sequence.

Sequence of Maize bronze Alleles 187

-1112M CCTGCTCCAAGATTGACGTCTCTTATTAAAGCACGCCATTTTAGGGATGTGGTGGCCGTTTGGGCCGTACGCGCCGGCCAAAAATAGAGCTGTTTATCGG -883W . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

-1294R ***ttttttftttttt*tt*tttt*ttt******ttt********************************************************CG~*

-783W * * +.+.*.++tA'*******".****t***"'.*~*****T********~***G*T******A*******~G*~****~.~~~******* -1O12M GGGCTTCAACCTCGTTCAACGGGTGGCCACTGGACAACGGATTCATGTTCCTGTGGAACACGTTTGGTTACTCGACTTTTTATTTACACTTTGGAAACAA -1194R C +...++ rA*********.*~"""+.++*****~*~******~~~******T********~***~*T***~**~~**~****G*~****~*~.~*******

-684W ++t++**+++***+*+**++r+++t+.*AT*C**'ftf'.~~**************T*G~*******C***AGTAGGCCACG*~** DR I R 1

CTAAGGGGTGTTTGAGAG -912M ACAGTCAAACACCATCACAAAACTACCATGG CTCACGAGCGCACACGATTGTAACCTATATACTTATTCG - 1 0 9 8 ~ ++*+~~++++~+*+~******~+++++~AT*C***AT*C***********~************T*G*~***~**~***AGTAGGTCACG~

-598W -82414 TGAAGTTTTTTCACAGTTTTGGAAGAATAAAATACTACGGTATTATATACAGAATTGTGTTTGGCAAGTTGGCTAAAATCTCTGTTTTTAAAACTGAAGT E -1012R 1; IR2 IR3

PStI -598W

-1012R -724M ATTGCAAATACTATAGTTGTTTTAAGTTATTCTAAACTTAGGTCTGGACCTCAGTTTTCAAAATTGTGGTATTGTAAACTGCAGTATTGTCATAACTAAA $ (Y

IR3 IR2 IR1 DR -598W -624M GTATACTGTAGTATTTCAAAAACTGTGGTTTTCAAAAACTTTGTTCCCAAACAGGGC~~ATGGGTAGGCACTAAGCACAGTACGACCCGCGATGCTTT

-1012R

+ + * * A * C f * f * * * " * * ~ ' G A " " " ' * A T * " ' * * * * C

+rr+A'C""**f'~"*G****+r***+T".*"'C

-559W r**+T**T*T***".**""""t************C***C*ACACGAC******C****AC*G*T********G*CT**GTACG ****+*********G* -524M GACACGACACGGACACGATTTATATAATGTCATACGGC -973R **+*+*+TtT****'***t*****'*f"C"*C**C*~AA~

ACAGGTTTAGTCACATCTTTGGACCACGACCTCCGCACAATGGACTGACATGAAC * + G ~ * * c * * A ' A c * * * * * * * A * * * * G ' + * . " . " * * * * * ~ * ~ * * * G G G ****~*c**GT

-462W **"tfCGATTAGGCACA**********ACTGG***********T*+***~ T'TTTG""A******f**"""""*'."********~**** -4 3 1M ACGACA GTTTAGACCAGA TCCCGTGACAA TCCGGCACAA TCCAGTGCCCCATCCCCTGGTCCACAGCCTTCACAGCTCTGG -883R *TAGT*CAATT GGCAC t+*****+r*ACTGG* ********t ~*~*~+***TTTG*"'A'**fttf"""""*.'*"**************~

-363W * * .+** *G*"*~*** * . * * " * * * " * *AG** '~~CCACT"*** * * * * * * *~*** * * *~ DR ?R

-350M CCCGGCCTTGCCATATCGTGCACGGG GGTGT GTCCAGAATGTACCTACTGG -787R T***'**G*~****************AG*****CCACT******************~~TATAGATGTCCAATAAGCACGAGGCCCGCGAGCCCGGCACGA

-305W -299M -687R AGCCCGCTGTTTGGGCCTGTCGAGCCGCACGCCCGGTATATGCCTGGGCGGCCCGGCACGAATAAGCGGGCCGGGCTCGGACAGGAAATTAGGCACGGTG

-305W -299M -587R G G C T A G ~ C C G C A C ~ G C C C G T T T A G ~ ~ ~ T A A G ~ ~ ~ G T T A A G ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ /I

IP

-305W If -299M -487R TTTCGTGCTAAACGGGCCGGCCCGACCCGTTTACGCTTGCGGGCCGGGCTTGGACAAGAAATTGAGCACGTGTGCTTAGCCGCCCGGCCCGGTTTTCTAA

I R DR -305W -299M

t*t*.

- 3 8 7 ~ TCGTGCCTGTCGGGCCGGCCCCAAAACGGGCCGGGCTTCACCGGGCCCGGGCCGGACCGGACCGGGCGGCCCGTTTGGACATCTCTACCTACTGG~.**' AGCGC

- 3 0 0 ~ t t * * * * f t t * t * ~ * t t t t t t * * ~ ~ ~ * * * * * * * ~ ~ * ~ ~ ~ * ~ ~ ~ ~ * * * * * * * * * * * * * * * * ~ * * * * * * * * * * * * * * * * * * ~ * ~ * * * * * ~ * * * * . * b ~ ~ ~ * * C

-294M GCTGCACGCTGGCGAGCGATCCACTCACCGTTGATCCCAACAAACTCCGGTCCCCAAACTCCACGGCACCAACAGCTAAGCCCGATGCGCTGCGTGCGCG -287R ***r*rt*t+rt+******+*****t+*tT**A**A**********************~****~*~*********~~~***~****~* **+****C

-194M GCGA CAAC CCGGCTCACCTAAAAATTTCGGCAC G TAAC CGACTGGCAGGTGCGCACGCGTGGTCGCGCG ATAA CGGACACGT *.**..* . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ***** ********.* -2OOW *t*ttttttt*tttt*tttt******~**********GTCTAA*~****~ .**t**t*t**t****t**t*+********** e * * * * * e * * * * * * * *

-194R **tt et*** t * * * * * * * t * * * * * * * * * * C t t t * t * $3 0 b -1OOW tt*ttttt.****tlttttt******~~~***********,,************G*******************.C*~~*+~~*******..******+*

mRNA initiation

-100M TGCGCCCCCAGCGAAGCCCGCACGCATCGCATTCGCATCGCATCGCAGGTCGCATCCGACGCTAGCGGCTAGCCTAGCCGAACAGCCTGAGCGCGCGAAG -100R ********..********tt******~*********************************************~***.~*********~************

1w . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1M ATGGCGCCCGCCGACGGCGAGTCCTCCCCGCCGCCGCACGTGGCCGTGGTCGCCTTCCCGTTCAGCTCCCACGCGGCGGTGCTGCTCTCCATCGCGCGCG 1R . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

oligo

101W * T " + * + ' + * * 4 * * * * * * * * * * * * * + + + * * * * * + * + * . + * * * * * * * * * * * * * * * * * * * * * 4 * * * * * * * * * * ~ * ~ * * * * * * * * * * * * * * ~ ~ * ~ * ~ * ~ * * * lOlM CCCTGGCTGCCGCCGCGGCGCCGTCCGGGGCCACGCTCTCGTTCCTCTCCACCGCGTCCTCCCTCGCGCAGCTCCGCAAGGCCAGCAGCGCCTCCGCCGG lOlR rT***ff*f************r*******************************~***********************************************

* 201W .................................................................................................... Kpnh

201M GCACGGGCTCCCGGGGAACCTGCGCTTCGTCGAGGTACCGGACGGCGCGCCCGCGGCCGAGGAGACCGTGCCGGTGCCGCGGCAGATGCAGCTGTTCATG 201R **t..tt**.*******Clt****************C******~********************T************~~*********************

3 0 1 ~ *ttttttt**tt****tt********t+**********~*************************~*********************** 301M GAGGCCGCGGAGGCCGGCGGGGTGAAGGCCTGGCTGGAGGCGGCCCGCGCCGCGGCGGGCGGCGCCAGGGTGACCTGCGTGGTGGGCGACGCGTTCGTGT 3 0 1 ~ +*tr++*t++rt+rr+r~tttt+++t*tt++t*t*++rr+**********~***~**~ t**tt~tttt*tttt***tt*****.*~~*********


598W 59 3M 579R

698W 693M 579R

798W 793M 579R

898W 89 3M 574R

***l+***&t**+*t*t+t****~~*****~*C***+****+******4A*bC*+*+**+**+****+****++*******++**+***********~~~ * .

eC4ATQOCTCTTGC~~CCTCGT~C~~~CAGCAAACAGGGTGGACGGGCTACTGATCTCCCACCCGGGCCTCGCCAGCTACCGCGTCCGTGACCTCC

E 0

+ t . t t * t t + t t t * + t + + + + t t t t t * t t * * + + + + + * * + ~ * + * + + + + + + + + ~ ~ + + + + + * + * * * + * * * * * * * ~ * ~ ~ ~ * ~ ~ * * ~ * * b * * ~ + * + ~ ~ + ~ ~ + * ~ + * + + ~ i AC insertion site in bz-m2 - "

CAGACGGCGTCGTCTCCGGCGACTTCAACTACGTCATCAACCTCCTCGTCCACCGCATGGGGCAGTGCCTCCCGCGCTCTGCCGCCGCCGTGGCACTCAA r, 5: I A

+ + * + + + + * * * + + * + + + * * * . * * + * * + * * * + + ~ * * ~ * * ~ * + ~ * * ~ * * * * * * ~ * ~ ~ ~ ~ + * ~ * + + * * * + + + * * + * * * + * * * * * ~ * * ~ * * * ~ ~ * * ~ ~ * * * * ~ * * CACGTTCCCAGGCCTGGACCCGCCCGACGTCACCGCGGCGCTCGCGGAGATCCTGCCCAACTGCGTCCCGTTCGGCCCCTACCACCTCCTCCTCGCCGAG

*++*+*+***+**+**r+*+++r+*+Tff******f*f**+***+*+*+************+*****++T*+**+*****~~~~4*~*******~+*~****** GACGACGCCGACACCGCCGCACCAGCCGACCCGCACGGCTGCCTCGCCTGGCTGGGCCGCCAACCCGCGCGCGGCGTCGCGTACGTCAGCTTCGGCACGG

GGCGT*~+++*+*+***+*."""""'."".."~**********++++++***~~*~**+**+**+**+**~~*+~*

* 998W t*****ttt****+*++**+**********.********+*+***+****+**C**+*****+*****++*+++*******++*++*********+*A+**T

SStI * *

993M TGGCGTGCCCGCGGCCCGACGAGCTCCGCGAGCTGGCGGCCGGGCTGGAGGACTCGGGCGCGCCGTTCCTGTGGTCGCTGCGCGAGGACTCGTGGCCGCA 6581 **.+*******++*++T'+**+**++*+**+*****+*****+++*****~~C~~~**C***+**+*+++++*+++********************A~**T

1098W *************+**+*T".++++++++++++++++++**T******~*****+****++*++*A+*+++******++**+*+*+************* 1093M CCTCCCGCCGGGTTTCCTGGACCGCGCCGCGGGCACCGGGTCCGGGCTCGTGGTGCCCTGGGCGCCGCAGGTGGCCGTGCTGCGCCACCCTTCCGTGGGC 758R * . * * . . + + * * * + . + t . + + T * ' f f ' f " " ' f f ' . " . . " + + * + + ~ * * + * * * * * + * + * * * * * * * * * * * * * * * + ~ * + * * ~ + * T * * + * * ~ * * ~ * ~ * * ~ ~ ~ * ~

1198W ***++*****+++++*+T****+**r*******++**++***********+G****++****~~******+*******++******************+*+**+***** 1193M GCGTTCGTGACGCACGCCGGGTGGGCGTCGGTGCTGGAGGGCTTGTCCAGCGGGGTGCCCATGGCGTGCCGCCCCTTCTTCGGCGACCAGCGGATGAACG 858R * + t + + t + + * t t t + t t t . t t * ~ + * + * ~ * + ~ * * * ~ * ~ + * ~ * ~ * * ~ * * ~ * * * * * * * + * * * ~ + * + + * + * * + + * * ~ * ~ ~ ~ ~ * * * * * * * * * * * * ~ * ~ + * ~ ~ ~ * ~ ~ +

1298W +******~*++++**lt++++***+**+****++++**++++++*++**~*+***++*+*++***+~*~+*~+***+*+**++++++*+**++++*++*+

1293M CGCGGTCCGTGGCGCACGTGTGGGGGTTCGGCGCCGCGTTCGAGGGCGCTATGACGAGCGCCGGAGTGGCCACGGCCGTGGAGGAGCTGCTGCGCGGGGA 958R t~ttt*tttttttt..tt.tt.lttttt***+++++++***+++++*******C****++*****+*****++**~*******************+++.+++++

* PStI 1398W +**~~*+**.*+*+*t+t.~~~*******++*~+*++**+****+******~**+~***~***++**+****+++***~**~**~*+*+**+*+++++*+

1393M GGAAGGGGCGCGGATGAGGGCAAGGGCCAAGGAGCTGCAGGCCTTGGTGGCCGAGGCGTTCGGGCCAGGCGGTGAGTGCAGGAAGAACTTCGACAGGTTC 1058R *++*+t++**+**+ttt.+t****~****~++**-+~+*+*+**********~***********************+******+**+*****+**

stop 1498W +**++***+*+**+*** . . * * *~**~***+*++***+**++++**+*~*** *+*~*+*** * * * * *~+ 1493M GTCGAGATAGTCTGTCGCGCGTGAAAGGTCGTCTTGCTGTTCAGAGGTTTTACCAACAGAAGAACATAATGAATTGGATGGCATGCTACGTCGTATTCTC

e

8 1158R * * t * + t * * * * + t + * t t t t . * + ~ + * * + * ~ ~ + * * * + * * + + * * * * * * * ~ * * * * * * * + * ~ + * * ~ * + + * + ~ * * * * * + + ~ + * * * * * * * * + + * * + * * * * * + * + ~ ~ t

Poly (A) C

1565W tttt.tt.+t+++**+*ttt****++++**T +++++*+*C**,*.T+**+f**+********t*+~~ 2 1593M TTTTTTTGTTGATCCCTGAGTTGATACATTTTGTACTTGATACATGAGTTGCAGCAGCAGCAGCAACAGCCTTCTGTACCTTGGCTTTGGATCTGTATTC a n -

1258R tt*+t***+ttt+t+tttt*.t)tt.ttlt**+**~**~++++++********+*++*********~**~~+=*****+***+******+~+~****+*+++**~

1632w +.++++++.~++G+*++*+*******+~**~~+**++*+**+*****~***+****++*+++**A*****+***+*++*+*~C+~~~*T****T++*+~~ 1693M TTGTCACCAGTTATCTGAAAGCATCAATAACCTTCTGTCTTCTAGCAGTTGCCTCTCCAGATTGCCAAAATAGCATTTATTATAAGGTCTTATGCAATGT

Poly (A) Poly (A)

1358R *+.+t.++..+t++*++++++*+++++G."tff*++*~++*******~*****+*+************+*~~*~***~~~~~*~**~**~~~~~

1732W **++r.++..A."+*"*"*.*."---------------+*+*******~~****A~C+A---------------------------------------- 1793M TTTCAGATTGTTCCGATTAAATCTACGATTAGCATTTTAGCCCAGCAGTCCAGCCCATTG~GGCTTATTCAGTTATTTTTAATCCATATAAATCAAAAA 1458R +tt+tt++*****+t*tt++tttttttt.*t.**.*********+**+********~*+*****+***+**********+******~*********~************

1775W ---_---------------_---------------------------------------------------------------** *A**+ ******** 1893M AGATTGATATAGATTAGAAAATATTTTAGTTTACTAGGAATTAAAACCCCTCAATTTTTCTTAATCCATATAAATTGTGGCAGAACTGAACAATCCGGTG 15581 .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

179OW +++.+t.+++*++++*++**+*+G*+*T+**++*+T+*****************+***********++*T*+****~*****++*++~~~+**+++***+ 1993M CCCATAAGCAACACGGCTGCCAG CCTGCCAGACCGCCCAGCCCAGGTGCCCACACCCAGCTGCCACTCCGCAAGCGACCGCCGTGAGCCGAACTGAGGA 1658R *tttt+***+*++*tt.t....t +++**tt++t**tt..tt+***+***+b**~~******++*+*****+**++*+***~~~***++*+****~++**

1890W +****** 2092M ACCGGCGACGGATACAGCGGCAGCTCCCTCCTCCTCCGTCCCGGCGCACGTGCGCCCGGTTGATTTGGTGACAATAGGATCATAGGGGATTAGAGAAGAT 1 7 5 7 ~ + + * t + t + . . . . t * t + * t * + . * t t t . . * t t t * * . * * t + t . t * * * * * * + * * + + + * C ~ *********+* .***+*++******* t t * . t t . * *~+ 1890W +****** 2092M ACCGGCGACGGATACAGCGGCAGCTCCCTCCTCCTCCGTCCCGGCGCACGTGCGCCCGGTTGATTTGGTGACAATAGGATCATAGGGGATTAGAGAAGAT 1 7 5 7 ~ + + * t + t + . . . . t * t + * t * + . * t t t . . * t t t * * . * * t + t . t * * * * * * + * * + + + * C ~ *********+* .***+*++******* t t * . t t . * *~+

PStI

2192M TGAGGAAAAATAAACT AATTTTTCCTTTAATCCTTTCTAATCTTCCCATGATACCGAATCACCAAATCAGCTCTGATGC~CAGGTAGGGCTCCCGGGC 1897W +C..**Af*f***+*+*C.*Gff*""'

1856R *+++*++****+*++*T*.'~+***~******

1926W +**++*+*++++++.C+*C*4*4+**+*****+**+**********GGA*AAAAATTTGTTGCAGCC*~*+*****+~**********~**+G****~~~ 2291M TCCACGCCTCCACCTTTTTTCCCTAGTCCCCACTCCCCAATACACGTTCA 1888R

TCTCTTCTAGTGCTAGGTAACTGTGATTGTGGC


were performed as previously described (DOONER et al. 1985). A cDNA library was constructed in lambda gtlO from 4 pg of poly(A) RNA isolated from husks of Bz-McC; B ; PI plants by the procedure of HUYNH, YOUNG and DAVIS (1984). RNase protection experiments were performed as described by the Promega Biochemicals’s publication “Riboprobe Gene Analysis System” using either 20 kg of total RNA or 1 kg of poly(A) RNA. Primer extension analysis was performed as described by MCKNIGHT et al. (198 1). The oligonucleotide GGCGGGCAGGACTCG was synthized with a Pharmacia DNA systhesizer and purified on a 20% polyacrylamide gel.

RESULTS

Genomic sequences of bronze alleles The sequence of the bronze allele Bz-McC and its

flanking regions was obtained by sequencing 60 over- lapping subclones for one strand and 40 for the opposite strand. The alleles bz-R and Bz-W22 were sequenced in one direction, but any ambiguous regions were redone and any differences with the Bz- McC allele were rechecked in all three alleles. In addition, the Bz-McC allele was independently sequenced by D. FURTEK, J. SCHIEFELBEIN and 0. E. NELSON (personal communication) who generously exchanged sequences with us before publication. A comparison of these two sequences showed no differences. The sequences of Bz-McC, bz-R and Bz-W22 are presented in Figure 2. The Bz-McC sequence is given as the center line, M, and the sequences for Bz-W22 and bz-R are above, W, and below, R. In regions where one of the alleles has an insertion relative to the other alleles, the sequence of the insertion is given, and a gap is present in the other alleles. The numbers at the left reflect the number of nucleotides from the initial ATG of the bronze message for each allele. The restriction maps in this paper and the nucleotide sequence presented in Figure 1 follow the molecular convention of placing the 5‘ end of the gene on the left. Since the direction of transcription of the bronze gene is away from the centromere, the sequence is inverted relative to our earlier restriction map of the gene (DOONER et al. 1985) which was oriented according to the genetic convention of placing the centromere of 9s to the right.

Some of the major features of the Bz locus and its flanking sequence are indicated in Figure 2, and will be discussed subsequently. The A in the initial ATG of the message is assigned a position of 1, sequences 5’ to this are assigned negative numbers. A message initiation site at - 83, the TGA stop codon at 1514 bp, and three poly(A) addition sites are marked. The single intron is located from 524 to 624 in the Bz- McC allele. The Bz-W22 allele has a slightly longer intron (5 bp). The 5‘ flanking regions of the three bronze alleles have segments of high homology. Bz- McC and bz-R each have different large insertions at

different positions in their 5’ flanking sequences which are marked along the (right) side of Figure 2. The region between these insertions shows a greater amount of mismatch than the rest of the sequence. Bz-McC and bz-R also have identical 62-bp insertions relative to Bz-W22 in their transcribed but untranslated 3’ tails and further downstream, Bz-McC has 163 bp not found in Bz-W22. The bz-R sequence does not extend to the end of this insertion, but its sequence appears to be similar to that of Bz-McC. Experimental verification of these placements and their significance are discussed below.

The bronze transcript cDNA analysis: The genomic region homologous

to the bronze transcript and the direction of transcription have been determined by Northern analysis (DOONER et al. 1985). T o determine the exact start and stop positions and the positions of any introns, further RNA and cDNA analysis was undertaken.

Husk material from plants with the anthocyanin regulatory factors B and P1 is deeply pigmented and is the best source of bronze RNA. When these studies were initiated the only B ; P1 corn lines available with a characterized Bz allele carried the isoallele from the inbred line W22 (Bz-W22). A cDNA library was made from this material and several bronze hybridizing cDNAs were recovered. From these clones it was discovered that the bronze mRNA had several polyadenalation sites (DEAN et al. 1985). The sequence of these clones also revealed substantial polymorphism between the Bz-McC genomic DNA and the Bz- W 2 2 cDNA. These differences include a 62-bp insertion in Bz-McC relative to Bz-W22 and approximately 2% base changes.

Experiments to determine the 5‘ end of the Bz- McC transcript using Bz-W22 RNA would be incon- clusive if the same level of polymorphism existed at the 5’ end of the message as is present at the 3’ end. Therefore the Bz-McC allele was introduced into a B ; P1 background. RNA from these plants was used to synthesize a cDNA library and to determine the 5’ end of the transcript by primer extension and by RNase protection experiments.

Approximately 140,000 phage from a lambda gtlO cDNA library made with poly(A) RNA from B ; P1; Bz-McC husk material were screened with probes for the middle and the 5’ end of the bronze message (pAGS528 and pAGS526, respectively, Figure 1). Approximately 100 phage (0.07%) were detected with the 3’ probe and 8 of these also hybridized with the 5‘ probe. These 8 potentially full length cDNA clones were purified, subcloned, and characterized. The two longest ones were sequenced and the 5’ ends of these were found to be within 20 bases of each other. Further analysis, described below, confirmed that one of these was a full length message.


A comparison of the Bz-McC cDNA sequence with the genomic sequence revealed a single intron of 100 bp located at +523 bp (Figure 1). The sequence around the splice sites, AGGGTGCGT- and TCGTTCGCAG}C have 7/9 and 811 1 bases (under-

"- -" lined) in common with the 5' and 3' consensus splice junction sequences, respectively (BREATHNACH and CHAMBON 198 1). Similar agreements are obtained by comparison with the more recently derived consensus sequences from animal (MOUNT 1982) and plant (BROWN, FEIX and FRENDEWAY 1986) introns.

Primer extension and riboprobe protection: A 15- base oligonucleotide was synthesized which would hybridize to the RNA sequence 17 bases downstream of the translational start site. The oligonucleotide position is shown in Figure 2. T o find the start of the bronze message, this primer was hybridized to RNA from B ; P1; Br-McC husks and extended with reverse transcriptase. In addition, the oligonucleotide was used as a primer for a sequencing reaction that was used as a calibration ladder. The length of the extended fragment was 1 12 ? 3 bp (Figure 3). This corresponds to the 5' end of the longest cDNA clone. The family of bands in Figure 3 could be due to multiple transcriptional start sites (LEBOWITZ and GHOSH 1982). A faint 460-bp fragment was also present (not shown), possibly representing a second start site for the bronze message. However, when the oligonucleotide primer hybridization was carried out at different temperatures, the 1 12-bp and the 460- bp products had different optimal hybridizing temperatures, suggesting that the high molecular weight band was an artifact.

T o determine whether the transcription start sites are the same in aleurone (where the bronze locus is under R and C regulation) and husk (under B and PI regulation), RNA was isolated from these two different tissues. A RNase protection experiment was performed using this RNA and a riboprobe (corresponding to pAGS526 in Figure 1) complementary to the bronze message. Riboprobes can be labeled to a high specific activity and thus the low amounts of bronze message isolated from aleurone can be detected. Figure 4 shows that the protected fragment is the same for aleurone (lane A) and husk (lanes B and C). The size of the protected fragment agrees with the expected fragment size (329 bp) predicted from the primer extension experiment above. No high molecular weight band was detected in this experiment. The results from both the primer extension and the riboprobe protection confirm that the longest cDNA clone is essentially full length.

The first ATG in the message, located 83 bp 3' to the transcription start identified above, is the likely translation initiation site. The sequence around this site, AAG ATGG, shows high homology with the consensus eukaryotic translation initiation site de-

a b

- " C A

FIGURE 3.-Determination of the 5' end of the bronze message by primer extension form Ez-McC; E ; P1 plants. An oligonucleotide complementary to the bronze message was end-labeled and hybridized to either 20 pg of total RNA (lane a) or 1 pg of poly(A) RNA (lane b). Reverse transcriptase was used to extend this primer (MCKNIGHT et al. 1981) and the extended fragment was run on a sequencing gel along with the adenine, A, and cytosine, C, lanes of a sequencing reaction carried out with the same primer on single stranded DNA from a bronze subclone.

0 1.63

0 0 . 5 1

0.40

0 0.34 - - e 0 0.30

A B C FIGURE 4.-Verification of the 5' end of the bronze message by

riboprobe protection. RNase protection experiments were carried out with a labelled riboprobe made from the pAGS 526 KfmI fragment subcloned behind an SP6 promoter and hybridized to: (A) 20 pg of total aleurone RNA from plants homozygous for Ez- McC, (B) 20 pg of total husk RNA from plants heterozygous for Br-McClbz-R, or (C) 1 pg of poly(A) husk RNA from plants homozygous for Ez-McC. End-labeled, Hinfl pBR322 fragments were used as m. w. markers.

scribed by KOZAK (1981). Since the UFGT enzyme is present in low amounts, verification of this assign- ment by amino acid sequencing has not been attempted.

Differences in the transcripts and predicted proteins encoded by the bronze alleles: The nucleotide sequences of the two functional bronze alleles in this

Sequence of Maize bronre Alleles 191

TABLE 1

Amino acid differences between the UFGT polypeptides encoded by Bz-McC (M) and Bz-W22 (W)

Position Allele Codon Amino acid

265 M ACC THR w TCC SER

642 M GGG GLY w GAG GLU

645 M CTA LEU W CCA PRO

1045 M GAC ASP W GCC ALA

1089 M CCG PRO w ACG THR

1094 M CAC HIS w CTC LEU

1236 M TTG LEU W GTG VAL

1364 M ACG THR w GCG ALA

1426 M GAG GLU w GTG VAL

The nucleotide position identifies the base in the Br-McC sequence that corresponds to the first position of the codons speci- fying amino acid differences between the two “wild-type” alleles.

study, Bz-McC and Bz-W22, differ from each other and from the mutant allele bz-R. The bz-R allele has a 340-bp deletion which begins in the intron 54 bp past the 5‘ splice site, removes the 3’ splice site and extends 285 bp into the second exon. There is an 8- bp insertion relative to Bz-McC in the portion of the intron left in bz-R. This “insertion” of TACTACAG may represent a deletion in the Bz-McC intron, since it is flanked by direct repeats of the sequence TAC- TAC. A related insert, TACAG, is also found at the same position in the Bz-W22 intron.

Within the 1413 bp that constitute the translated region of the bronze transcript there are 21 bp differences (1.5%) between the functional alleles Bz- McC and Bz-W22. Of these, 9 produce changes in the amino acid composition of UFGT. These changes are listed in Table 1 and are indicated with stars in Figure 2. Excluding the region deleted in bz-R, there are 8 bp changes between bz-R and Bz-McC and 3 bp changes between bz-R and Bz-W22 that would result in amino acid differences. Thus Bz-W22 and bz-R seem to be closer to each other than to Bz-McC.

A protein of 471 amino acids with the composition shown in Table 2 is predicted from the Bz-McC cDNA sequence. The molecular weight of this protein, 48.8 kD, and the PI calculated from the average PIS of the amino acids, 6.0, agree roughly with the measured values of 50 kD and 5.0 (DOONER 1981). The previously estimated acidic PI of the UFGT protein allows

TABLE 2

Amino acid composition of UFGT predicted from the Bz-McC sequence

A. Composition of polypeptide sequence from the Bz-McC sequence.

Amino acid NO. Amino acid NO.

Ala* 90 Met 8

Asp* 23 Pro 38 Glu* 24 Gln 9 Phe 20 Arg 30

His 12 T h r 14 Ile 6 Val* 43

Leu 45 Try 4

B. Base percentage in each codon position for Bz-McC.

CYS 9 Asn 8

Gly* 44 Ser 30

LYS 4 TrP 10

Position

Base 1 2 3

T 14.4 25.9 1.9 C 26.8 35.0 50.1 A 11.3 17.8 3.2 G 47.6 21.2 44.8

G + C 74.4 56.2 94.9

Forty-seven percent of the first bases in the codons for UFGT are G. Thus 47% of the amino acids in UFGT are restricted to the 5 amino acids specified by this codon class and are indicated by asterisks.

us to be confident of the correct reading frame of our sequence. The GC content of the three coding positions are 74% (47.6% are G), 56% and 94%, respectively. Due to this distribution of nucleotides, any artificial frame shift would result in a protein with a drastically higher PI.

Expression of different bronze alleles Expression of the bz-R allele: Analysis of the en-

zymatic and antigenic properties of endosperm proteins isolated from plants homozygous for bz-R has demonstrated that no UFGT activity or cross-reacting material to UFGT-specific antibodies is detectable, suggesting that no protein is made (DOONER and NELSON 197713; DOONER 1986). The sequence of bz- R reveals a 340-bp deletion starting 46 bp inside the intron and extending 285 bp beyond the 3’ splice site of the intron (Figure 2). While this deletion is probably sufficient to cause a loss of UFGT activity, differences between the sequence of bz-R and Bz- McC in the 5’ upstream regions raise the possibility that the promoter region of bz-R might also be altered in function. To determine bz-R transcript levels, crosses were made to derive bz-Rlbz-R homozygous and bz-RIBz-McC heterozygous plants with the regulatory factors B and PI. RNA from these plants was

192

a

E. J. Ralston, J. J. English and H. K. Dooner

b c d separable from the transcript of the functional alleles by gel electrophoresis. Thus the level of bz-R RNA can be used as an internal standard for the levels of Bz-McC and Bz-W22 RNA as measured by Northern analysis.

Northern blots of RNA from Bz-McClbz-R and Bz- W22lbz-R seedlings were probed with the insert from pAGS551 (Figure 1) cloned behind the SP6 promoter. The similarity in the intensity of the bz-R and Bz-

5, lane c), suggests that these two alleles have functionally similar promoters. On the other hand, the pronounced differences in intensity between the bz- R and Bz-W22 band in the respective heterozygote (Figure 5 , lane b) suggest that the Bz-W22 allele has a weaker promoter than bz-R. The difference in intensity between the Bz-McC and Bz-W22 bands

<

7

“- , I+ McC bands in the Bz-McClbz-R heterozygote (Figure

< relative to the bz-R band is at least as large as the FIGURE 3.-Expression of Bz-McC, Bz-W22 and bz-R analyzed

by Sorthern blots. Lane a contains 1 kg of poly(A) RNA from a Bz-McC homozygous plant. Lanes b, c, and d each contain 20 kg of total RSA extracted from Bz-W22Ibz-R (lane b), Bz-McClbz-R (lane c) heterozygous plants, or a bz-Rlbz-R (lane d) homozygous plant. The B z and bz-R transcripts were separated by gel electrophoresis. The Sorthern blots were hybridized to the bronze specific probe pAGS328. The UYYOWS indicate R S A m. w. markers of 3.0, 2.23, and 0.9 kb.

compared by Northern analysis (Figure 5) . The bz-R message is smaller than the Bz-McC message and the two messages can be separated in RNA from the heterozygous plant. The similarity in intensities of the bz-R and the Bz-McC bands in RNA from a heterozygous plant RNA suggests that similar amounts of RNA are being made from the promoters of these two alleles. In addition, riboprobe protection experiments with RNA from bz-RIBz-McC heterozygotes (Figure 4, lane B) shows only one band, suggesting that the mRNA initiation sites are identical. Thus, the promoter region of bz-R is functional, a smaller message is made, yet no cross-reacting material to the UFGT enzyme can be detected. Appar- ently, any protein made from this message is unstable and quickly degraded.

Expression of the Bz-W22 allele: DOONER (1986) has shown that Bz-McC and Bz-W22 stocks differ in their amounts of enzymatic activity and has mapped genetically the site responsible for this difference. Northern analysis was used to determine if the difference in enzyme levels reflects a difference in message levels. It would be difficult to quantify accurately the difference in expression between two plants homozygous for Bz-McC and Bz-W22 because the bz allele is only expressed in a small fraction of the cell types and its expression is influenced by light levels which differ from leaf to leaf even on the same plant. Instead each of these alleles was made heterozygous with bz-R which makes a smaller transcript

three fold difference in the enzymatic activity levels found in aleurone tissue (DOONER 1986). These ob- servations suggest that the promoters for Bz-McC and Bz-W22 are functionally different and that the difference between the promoters most likely correspond to the previously mapped uqv site.

Insertions in the sequences flanking the bronze locus: Comparison of the 3’ region of the 3 alleles showed that Bz-McC and bz-R each have independent insertions with the characteristics of transposable elements. The Bz-W22 allele has no insertion. The insertion in Bz-McC at - 834 to -563 has a 4-base direct repeat at its ends, CTAA, which is found in only one copy in bz-R and Bz-W22. Possibly these 4 bp were duplicated when the insertion occurred, a feature of transposable element insertions. As shown in Figure 2 and Figure 6, this element also has terminal inverted repeats which would allow a “stem and loop” structure postulated to play a role in excision (SAEDLER and NEVERS 1985). We have called this insertion Ins1 (see DISCUSSION).

Sequences structurally related to Inr l are found elsewhere in the genome. When the Bz-McC lambda gt 10 cDNA library was screened for full length bronze messages with the pAGS526 probe (Figure 1) which contains Ins1 in addition to the 5’ end of the bronze message, two clones were isolated which did not hybridize to the bronze gene. The sequence of these clones revealed homology to Ins1 (Figure 6). One of these clones (Figure 6, #1) has the entire element flanked by a new 4-bp direct repeat and ends with a poly(A) tail beyond the right end of the element. The other (Figure 6, #2) terminates inside of the insertion sequence with a poly(A) tail close to the left end of the element (shown as a poly(T) in Figure 6). These three sequences are highly homologous and may have been derived from a common ancestor. Some of the differences in Figure 6 are in the inverted repeats which may have been more extensive in the


DR IR1 IR2 IR3 I R4 cDUA #I TCAG..........G................................CTCAGTATTCTC.......G.A............T...G............C. raw1 CDNA I2

CTAAGGGGTGTTTGAGAGTGAAGTTTTTTCACAGTTTTGGAAGAATA AAATACTACGGTATTATATACAGAATTGTGTT'NXCMGTT T T T m T

" " - cDUA 11 .................. ..... I n s 1 GGCTAAAATCTCTGTTTTTGAAGTATTGCAAATA~ATAGTTGTTTT~GTTA~TAAAC~AGGTCTGGACCTCAGT~AAAATTGTGGT

C.....................................G........................ C..........

cDUA t 2 A........A.T......C.....................................G..................A...T.......C............

cDNA #I ........... G..........G........................... .......... T..... ................. C.TCAG 1n.1 ATTGTAAACTGCAGTATTGTCATAACTAAAGTATACTGTAGTATTTCAAAAACTGTGGTTTT CAAAAACTTTGTXCCAAACAGGGCCTAA cDNA t 2 ............ G..........G............................. .................................. CTCA

IR4 IR3 I R2 IR1 DR

c" - 4" "

FIGURE 6.-Comparison of the 5' insert in Bz-McC, Insl, with sequences from two cDNA clones. The two cDNA clones are transcribed in opposite direction. cDNA # 1 is presented 5' to 3' and its sequence extends over 100 bp before the poly(A) tail is reached. cDNA #2 is presented 3' to 5'. In this orientation its poly(T) tail occurs within the insertional element. The four bp direct terminal repeats (DR) produced upon insertion and the internal inverted repeats (IR) are also indicated.

ancestral form. For instance, the 12-bp "insert" in cDNA #1 makes a fourth inverted repeat (IR4), with all three sequences. Genomic Southern analysis revealed approximately 50 bands or more that hybridize to Z n s l . Thus, the three elements that we have cloned and sequenced are members of a large family. At least two members of this family are inserted in regions of the genome that are transcribed.

The insertional element in the 5' flanking region of bz-R is located 299 bp upstream of the first ATG in the bronze message (Figure 2). The 10-bp direct repeats at its ends (shown underlined in Figure 2) are found in one copy in the other two alleles, suggesting again that the duplication occured as a consequence of the insertion. The inverted terminal repeats characteristic of transposable elements are also indicated in Figure 2. This element, designated Ins2 (see DISCUSSION), is found to hybridize with many bands when used as a probe for genomic Southern analysis. We have not isolated any other members of this family to see what level of divergence is present.

Both bz-R and Bz-McC are found to have a 62-bp insertion in their 3' transcribed, but untranslated, ends when their sequences are compared with the sequence of Bz-W22. While this sequence is flanked by a four base pair direct repeat ACAT it does not have the more extensive level of internal inverted and direct repeats usually found in transposable elements, and also found in the 5' insertions. Further downstream, at 2097 bp in Bz-McC, a 163-bp insert in Bz-McC or a deletion in Bz-W22 is found, again with no internal structures similar to those of a transposable element.

DISCUSSION

The bronze locus has been studied extensively both genetically and biochemically. Three bronze alleles have been sequenced in this investigation. In the following discussion we will attempt to integrate the sequence data with the existing genetic and biochem- ical data.

The transcribed region: The bronze transcribed region is simple. The Bz-McC allele has only one intron of 100 bp located 523 bases past the translation start codon. The intron in Bz-W22 and the portion of the intron left in bz-R are identical to the Bz-McC intron except for additional 5 and 8 bp, respectively. The intron length variation may be due to DNA replica- tion errors or to other deletogenic events in Bz-McC and Br-W22, since the region deleted is flanked by the direct repeat, TACTAC. We previously reported that multiple polyadenylation sites are used in the bronze message (DEAN et al. 1986). Multiple poly(A) addition sites have also been found in the Adh (SACHS et al. 1986) and Wx (KLOSGEN et al. 1986) loci in maize and in the petunia CAB and small subunit of RUB- ISCO gene families (DEAN et al. 1986). Perhaps the most unusual structural feature of the bronze sequence is its high GC content. Excluding the intron which has 50% GC, the coding sequence has an overall GC content of 75% and has long stretches of almost 100% GC.

The allele bz-R, the first or reference mutation described in the bronze locus (RHOADES 1952), has been shown to be CRM-, i e . , to accumulate no cross- reacting material to UFGT antibodies (DOONER and NELSON 1977a; DOONER 1986). The sequence of the bz-R allele shows a 340-bp deletion starting in the intron and extending 285 bp past the normal splice site for Bz-McC. Potential secondary 3' splice sites exist 9 and 117 bp beyond the deletion end point. If the bz-R allele was otherwise undamaged, a shorter spliced mRNA might be made. However, the deletion in bz-R may also include the internal intron splicing signal described for animal systems by KELLER and NOON (1984). An analysis of 36 introns from maize Adh (SACHS et al. 1986), Sh (WERR et al. 1985) and W x (KLOSGEN et al. 1986) shows a consensus sequence of C(21),T(34),G(27),A(36), C or T(26) 15 to 45 bases upstream of the 3' splice junction. The corresponding sequences in the bronze intron, ATGAT or TTGAT, are deleted in bz-R. An alternate internal splice signal,

194 E. J. Ralston, J. J. English and H . K. Dooner

" CCGAC, is found 10 bp past the deletion end point, however no 3' splice sites, i.e., AG, occur within 45 bp downstream. Further analysis of the bz-R message is needed to determine if an alternate splice site is used.

A comparison of the coding regions common to the bz-R, Bz-McC and Bz-W22 alleles shows a high degree of divergence, similar to that found in com- parisons of other maize alleles studied so far (Sh, Adh). Considering only nucleotide differences that do not result in amino acid changes, Bz-McC and Bz- W 2 2 differ at 12 positions, Bz-McC and bz-R differ at 9 positions, and Bz-W22 and bz-R differ at 8 positions. The percent difference in silent bases as calculated by the method of PERLER et al. (1980), and the rate of divergence for silent changes in plants estimated by WILSON, OCHMAN and PRAGER (1987) predict that these three bronze alleles separated from each other 2 to 3 X lo6 yr ago. WERR et al. (1985) have found a similar degree of divergence in two alleles of sh.

Unlike sh and adh, the bronze alleles sequenced here also have many nucleotide changes that would affect the amino acid composition of the UFGT enzyme. Considering the nucleotide differences that would result in amino acid changes, Bz-McC differs from Bz-W22 and bz-R at 9 and 8 positions, respectively, while Bz-W22 and bz-R differ at only 3 positions. The fact that almost half of the nucleotide differences between Bz-McC and Bz-W22 result in amino acid changes suggests that the UFGT enzyme can tolerate more change than sucrose synthase or ADH. The observation that the divergence in silent substitutions is approximately the same for all three alleles, yet the divergence in amino acid replacement substitutions is not, suggests that some of the amino acid changes are the result of selection pressure not experienced by all three alleles.

The UFGT enzymes found in Bz-McC and Bz-W22 have been studied biochemically, and found to differ in their thermal stability. The uts UFGT ihermal stability) site responsible for this difference has been mapped genetically to be 5' of the Ds insertion site in b z - d ( D Z ) (DOONER 1986). By cloning and sequencing the appropriate alleles, the position of this Ds insertion has been identified and found to be identical to the Ac insertion in bz-m2 which is located at + 749 bp in Figure 2 (DOONER et al. 1986; DOONER, RALSTON and ENGLISH 1988). The uts site has also been placed by Southern analysis of selected Bz intragenic crossovers to be 3' to the AatII site which differs in Bz- McC and Bz-W22 at +514 bp in Br-McC (H. K. DOONER, unpublished data). Between + 514 and +749 there are two amino acid differences in these alleles. These changes, a leucine to a proline and a glycine to a glutamine in Bz-McC and Bz-W22 respectively, are located 3 bases apart at +642 and

+645 and thus would behave essentially as one genetic site. Since either or both of these changes could cause a change in the protein structure and hence in thermal stability, we conclude that this location probably represents the uts site. The insertion site of Ds in b z - d ( D Z ) is also known to influence the thermal stability of UFGT since reversions that arise from excision of Ds result in proteins with large variation in their thermal stability (DOONER and NEL- SON 1979b).

The insertions outside the coding region: Outside of the coding regions, at both the 3' and the 5' ends, the bronze alleles differ from each other by the presence of large insertions or deletions. The insertions upstream of the 5' ends of the bz-R and Bz-McC alleles are flanked by 10- and 4-bp direct repeats, respectively, and have terminal inverted repeats characteristic of transposable elements. They are also members of separate large families as judged by genomic Southern hybridization. Thus, it is likely that these elements are derived from transposable element insertions. However, since they have not been observed to move, we cannot determine whether they are still mobile or remnants of an old transposable element system.

We have compared the sequences of the inverted terminal repeats in the insertions upstream of Bz- McC and bz-R with those of other maize insertions (DORING and STARLINGER 1984; SHEPHERD et al. 1984; ZACK, FERL and HANNAH 1986). The insertion upstream of the Bz-McC allele has a 6-bp inverted repeat near its termini and is flanked by a 4-bp target site direct repeat. The inverted repeat shares 5 of 6 bases with the inverted terminal repeat of c in l , an element that also causes a 4-bp target site duplication, but otherwise, the two elements have little internal homology. We have designated the family of elements homologous to the insertion upstream of Bz-McC as Ins1 (Insertion 1 ). The terminal inverted repeat in the insertion upstream of bz-R is not homologous with the terminal repeat of any sequenced maize insertion. We have designated this family of insertions as Ins2 (Insertion 2 ) .

The 3' transcribed but untranslated ends of bz-R and Bz-McC have a 62-bp insertion not present in the Bz-W22 sequence. This insertion does not have extensive characteristics of a transposable element. Further downstream, Bz-McC and Bz-W22 differ by a 163-bp insertion in Bz-McC which also seems to be present in the bz-R allele (our bz-R sequence terminates within this insertion). This insertion does not have terminal repeats or extensive internal repeats.

It is puzzling that these 3' insertions are present in Bz-McC and bz-R, but not in Bz-W22, since when the coding regions are compared, bz-R is closer to Bz-W22. While bz-R and Bz-McC differ by 1-2% through most of the coding region, the sequences


flanking the 62-bp insertion at the 3’ end, expected to be more divergent since they are outside of the coding region, are 100% homologous for over 300 bases. Possibly, a crossover event between the 3’ ends of bz-R and Bz-McC has occurred recently, leading to homogenization of the 3’ flanking sequences. In fact, a crossover event between sh and bz ( i e . , on the 3’ side of bz) was selected initially when a Sh Bz-McC chromosome was extracted from a sh Bz-McC lSh bz- R heterozygote (see MATERIALS AND METHODS). Other unselected crossover events could have occurred subsequently during the incorporation of Bz-McC into a W22 genetic background by repeated backcrossing to a W22 bz-R stock.

The bronze 5’ end: The 5‘ regions of the three alleles studied here can be divided into 3 parts separated by Insl and Ins2 (Figure 7). Region I, from - 1 to around - 300 is highly conserved. The first 193 bp 5’ to the first ATG are 100% conserved in bz-R and Bz-McC. This homology is broken by the deletion of 9 bp (GCGTGCGCG) in bz-R relative to Bz-McC. Further upstream lies a 100-bp region of 95% homology delimited by the insertion in bz-R. No consensus “TATA” box is found within 80 bases of the transcription start site. Instead, the sequence AATAAAA is found 32 bp before the start of transcription, a normal location for the “TATA” box (MESSING et al. 1983). The sequences CCAACC and CTAACT are found 68 and 34 bp upstream of this AATAAAA, respectively. Both of these are within the normal range for the consensus CCAAT box (NUSSINOV, OWENS and MAIZEL 1986). While the Bz- W22 promoter region has a few single base changes in this region compared to Bz-McC and bz-R, the only major difference is the duplication of the sequence GTCTAA two bases upstream of the CTAACT sequence, creating a near duplication of this potential CCAAT box, GTCTAACGTCTAACT. This insertion may correspond to the uqu site discussed below.

Region 11, which is flanked by Insl and Ins2, is characterized by a high degree of divergence between the three alleles. Insertions may have played some role in generating this divergence, possibly by inter- fering with crossing-over in this region (DOONER 1986). The high degree of divergence in this region suggests that it is not involved in the regulation of bronze expression, although there are stretches of up to 37 bp that are conserved.

Region I11 starts upstream of Insl (at -840 bp). The sequences here are again highly conserved. This region could represent a different gene (if so, it is not expressed in husks since we did not detect the corresponding cDNA in our husk cDNA library), or some distant cis-acting regulatory region for the bronze locus, or a nonfunctional sequence that has been conserved by chance.

The promoter regions of Bz-McC and bz-R are

Q . lkb -

R I I

FIGURE 7.-Composite structure of the Bz-McC, Br-W22, and br-R alleles. Insertions relative to the other sequences are shown as triangles: A, insertion in Br-McC only (Zm1); B, insertion in br- R only (Zns2); C and D, insertions in both Br-McC and br-R and not in Br-W22. The protein coding region is shown as a dark bar, the direction and length of the RNA transcript is indicated by the arrow, and the deletion in the br-R allele is marked. The KpnI site within the coding region is found only in the Br-McC allele. The sequenced region presented in Figure 2 is indicated by stars. Except for the region between inserts A and B the sequences of the three alleles are 98 to 99% homologous.

functionally equivalent for two reasons. Their mRNA initiation site is the same ( i e . , only one riboprobe protected band is observed with RNA from bz-RlBz- McC heterozygous plants, Figure 4, lane B), and they produce similar amounts of message as judged by the intensity of the bz-R and Bz-McC bands in North- ern blots of bz-RIBz-McC heterozygous plants (Figure 5). The Bz-W22 allele is shown above to have a functionally less efficient promoter. The difference in transcript levels between Bz-McC and Bz-W22 correlates with the difference in UFGT activities of these alleles. The genetic site responsible for this enzyme activity difference, uqu, has been mapped genetically to be 5’ of uts (DOONER 1986) which is located between + 642 and + 645 bp in Bz-McC. From genetic data it is also known that the uqu site lies close or 5‘ to the Ds insertion site in bz-ml. This site has been localized by Southern analysis to the region from + 120 to + 234 flanked by the KpnI and BglI sites (DOONER et al. 1985).

In an attempt to correlate the uqu site with a specific sequence, we can scan the 5’ upstream regions of the different bronze alleles for any significant differences starting at +642 bp. Of the 6-bp differences within the coding region between Bz-W22 and Bz-McC upstream of uts, only one difference results in a amino acid change, a minor change of a threonine to a serine. Most of these nucleotide differences are also found in bz-R which seems to make the same amount of RNA as does Bz-McC. Therefore the most likely location for the u p site is 5’ of the message. The sequence CTAACT, at - 157, has an upstream 6 bp duplication in Bz-W22, as discussed above. If this sequence is involved in protein binding, it is possible that the related sequence CTAAGT, generated in the duplication, would serve as a competing binding site. Alternately, the approximately one-half turn of the DNA helix resulting from a 6 base insertion could disrupt any protein binding sites that normally span the insertion site. A comparison of the sequences of the three bz alleles in regions I1 and I11 shows that


many of the differences between Bz-McC and Bz-W22 are also found in bz-R and are thus unlikely to be responsible for the low Bz-W22 message levels. There- fore it is likely that the UQV site corresponds to the presence versus absence of the 6-bp insertion at - 157. Similar structure-function relationships have been observed in the promoter region of the Pal gene of Antirrhinum mujus. Changes directly upstream of the sequence CCAACT due to excision of the transposable element Turn3 apparently affect the level of expression of that gene (COEN, CARPENTER and MARTIN 1986).

To date every allele of Bz that has been sequenced, has had insertions in different positions outside of the coding region, (this paper and D. FURTEK, J. SCHIEFELBEIN and 0. NELSON, personal communication). These insertions or deletions can be identified only when the sequences of several alleles are compared. A large insertion 5 ’ to the message of a wild- type allele helps define the boundary of the region involved in direct transcriptional control and, of course, allows one to discount the sequence of the insertion itself as playing a role in regulation. If the natural insertional diversity surrounding genes is a common feature in maize or plants in general, the utility of sequencing several different alleles to define the conserved region surrounding the message is apparent.

We would like to thank D. FURTEK, J. SCHIEFELBEIN, and 0. NELSON for prepublication exchanges of bronze sequences, R. JORGESSEN, W. TUCKER and M. RALSTOS for reviewing the manuscript, and C. STEPHENS for the artwork. This is Paper No. 5-7 from Advanced Genetic Sciences.

LITERATURE CITED

BREATHNACH, R., and P. CHAMBOS, 1981 Organization and expression of eukaryotic split genes coding for proteins. Annu. Rev. Biochem. 50: 349-383.

BROWN, J. W. S., G. FEIX and D. FRENDEWEY, 1986 Accurate in vitro splicing of two pre-mRNA plant introns in a HeLa cell nuclear extract. EMBO J. 5: 2749-2758.

CHEN, S.-M., and E. H. COE, JR., 1973 Control of anthocyanin synthesis by the C locus in maize. Biochem. Genet. 15: 333- 346.

COE, E. H., JR. 1966 The properties, origin and mechanism of conversion-type inheritance at the B locus in maize. Genetics

COEN, E. S., R. CARPENTER and C. MARTIN, 1986 Transposable elements generate novel spatial patterns of gene expression in Antirrhinum majus. Cell 47: 285-296.

CONE, K. C., F. A. BURR and B. BURR, 1986 Molecular analysis of the maize anthocyanin regulatory locus C I . Proc. Natl. Acad. Sci. USA 83: 9631-9635.

DEAN C., P. VAN DEN ELZEN, S. TAMAKI, P. DUNSMUIR and J. BEDBROOK, 1985 Differential expression of the eight genes of the petunia ribulose bisphosphate carboxlyase multi-gene family. EMBO J. 4: 3055-3061.

DEAN C., S. TAMAKI, P. DUNSMUIR, M. FAVREAU, C. KATAYAMA, H. DOONER and J. BEDBROOK, 1986 mRNA transcripts of several plant genes are polyadenylated at multiple sites in vivo. Nucleic Acid Res. 14: 2229-2240.

53: 1035-1063.

DOONER, H. K., 1981 Regulation of the enzyme UFGT by the controlling element Ds in bz-m4, an unstable mutant in maize. Cold Spring Harbor Symp. Quant. Biol. 45: 457-462.

DOONER, H. K., 1983 Coordinate genetic regulation of flavonoid biosynthetic enzymes in maize. Mol. Gen. Genet. 189: 136- 141.

DOOKER, H. K., 1986 Genetic fine structure of the bronze locus in maize. Genetics 113: 1021-1036.

DOONER, H. K., and 0. E. NELSON, 1977a Genetic control of UDPglucose: flavonol 3-0-glucosyltransferase in the endosperm of maize. Biochem. Genet. 15: 509-519.

DOOSER, H. K., and 0. E. NELSON, 1977b Controlling element- induced alterations in UDPglucose: flavonoid glucosyltransferase, the enzyme specified by the bronze locus in maize. Proc. Natl. Acad. Sci. USA 7 4 5623-5627.

DOOSER, H. K., and 0. E. NELSON, 1979a Interaction among C, R and V p in the control of the Bz glucosyltransferase during endosperm development in maize. Genetics 91: 309-315.

DOOSER, H. K., and 0. E. NELSOS, 1979b Heterogeneous flavonoid glucosyltransferases in purple derivatives from a controlling element-supressed bronze mutant in maize. Proc. Natl. Acad. Sci. USA 76: 2369-2371.

DOONER, H. K., E. RALSTOS, and J. ENGLISH 1988 Deletions and breaks involving the borders of the Ac element in the bz- mZ(Ac) allele of maize. In: Znternational Symposium on P b n t Transposable Elements, Edited by 0. E. NELSON and C. WILSON. Plenum Press, New York. In press.

DOOSER, H. K., E. WECK, S. ADAMS, E. RALSTON, M. FAVREAU and J. ENGLISH, 1985 A molecular genetic analysis of insertions in the bronze locus in maize. Mol. Gen. Genet. 200: 240-246.

DOOSER, H. K., J. ENGLISH, E. RALSTON and E. WECK, 1986 A single genetic unit specifies two transposition functions in the maize element Activator. Science 234: 210-21 1.

DORING, H. P., and P. STARLINGER, 1984 Barbara McClintock’s controling elements: now at the DNA level. Cell 39: 253-259.

EMERSON, R. A., 1921 The genetic relations of plant colors in maize. Cornell Agric. Exp. Stn. Mem. 39.

FEDEROFF, N., D. FURTEK and 0. E. NELSON, 1984 Cloning of the bronze locus in maize by a simple and generalizable procedure using the transposable element Activator (Ac) . Proc. Natl. Acad. Sci. USA 81: 3825-3829.

GERATS, A., J. BCSSARD, E. H. COE, JR., and R. LARSON, 1984 Influences of E and P1 on UDPG:flavonoid 3-0-glucosyltransferase in Zea mays L. Biochem. Genet. 22: 1161- 1169.

HUYNH, T. V, R. A. YOUNG and R. W. DAVIS, 1984 Constructing and screening cDNA libraries in lambda gtlO and lambda gtl 1. pp. 49-78. In: DNA Cloning Techniques: A Practical Ap- proach, edited by D. GLOVER. IRL Press, Oxford.

KELLER, E. B., and W. A. NOON, 1984 Intron splicing: a conserved internal signal in introns of animal pre-mRNAs. Proc. Natl. Acad. Sci. USA 81: 7417-7420.

KIRBY, L. Y., and E. D. STYLES, 1970 Flavonoids associated with specific gene action in maize aleurone and the role of light in substitution for the action of a gene. Can. J. Genet. Cytol. 12:

KLOSGEN, R. B., A. GIERL, Z. SCHWARZ-SOMMER and H. SAEDLER, 1986 Molecular analysis of the waxy locus of Zea mays. Mol. Gen. Genet. 203: 237-244.

KOZAK, M., 1981 Possible role of flanking nucleotides in recog- nition of the AUG initiator codon by eukaryotic ribosomes. Nucleic Acids Res. 9: 5233-5252.

LARSON, R. L., and E. H. COE, 1977 Gene-dependent flavonoid glucosyltransferase in maize. Biochem. Genet. 15: 153- 156.

LEBOWITZ, P., and P. K. GHOSH, 1982 Initiation and regulation of simian virus 40 early transcription in vitro. J. Virol. 41: 449-461.

MCCLISTOCK, B., 1951 Mutable loci in maize. Carnegie Inst. Wash. Year Book 50: 174-181.

934-940.


MCCLINTOCK, B., 1955 Controlled mutation in maize. Carnegie Inst. Wash. Year Book 5 4 245-255.

MCCLINTOCK, B., 1962 Topographical relations between elements of control systems in maize. Carnegie Inst. Wash. Year

MCCLINTOCK, B., 1967 Regulation of pattern of gene expression by controlling elements in maize. Carnegie Inst. Wash. Year

MCKNIGHT, S. L., E. R. GAVIS, R. KINGSBURY and R. AXEL, 1981 Analysis of transcriptional regulatory signals of the HSV thymidine kinase gene: identification of an upstream control region. Cell 25: 385-398.

MCLAUGHLIN, M., and V. WALBOT, 1987 Cloning of a mutable bz2 allele of maize by transposon tagging and differnetial hybridization. Genetics 117: 771-776.

MESSING, J., R. CREA and P. H. SEEBURG, 1981 A system for shotgun DNA sequencing. Nucleic Acid Res. 9: 309-321.

MESSING, J., D. GERAGHTY, G. HEIDECKER, N. Hu, J. KRIDL and I. RL-BENSTEIN, 1983 Plant gene structure. pp. 211-227. In: Genetic Engineering of Plants, Edited by A. HOLLAENDER. Plenum Press, New York.

MILLS, D. R., and F. R. KRAMER, 1979 Structure-independent nucleotide sequence analysis. Proc. Natl. Acad. Sci. USA 76:

MOL-ST, S. M., 1982 A catalogue of splice junction sequences. Nucleic Acids Res. 10: 459-472.

NELSOS, 0. E., and A. S. KLEIS, 1984 Characterization of an Spm-controlled bronze-mutable allele in maize. Genetics 106:

NLSSINOV, R., J. OWESS and J. V. MAIZEL, JR., 1986 Sequence signals in eukaryotic upstream regions. Biochim. Biophys Acta

OREILLY, C., N. S. SHEPHERD, A. PEREIRA, Z. SCHWARZ-SOMYER, I. BERTRAM, D. S. ROBERTSOS, P. A. PETERSOS and H. SAEDLER, 1985 Molecular cloning of the a1 locus of Zea mays using the transposable elements En and Mul. EMBO J. 4: 877-882.

PAZ-ARES, J., U. WIENAND, P. A. PETERSON and H. SAEDLER, 1986 Molecular cloning of the c locus of Zea mays: a locus regulating the anthocyanin pathway. EMBO J. 5: 829-833.

PERLER, F., A. EFSTRATIADIS, P. LOMEDICO, W. GILBERT, R. Ko- LODNER and J. DODGSON, 1980 The evolution of genes: the chicken preproinsulin gene. Cell 20: 555-566.

RALSTOS, E. J., J. ESGLISH and H. K. DOOSER, 1987 Stability of

Book 61: 448-461.

Book 65: 568-578.

2232-2235.

769-98 1.

866: 109- 1 19.

deletion, insertion and point mutations at the bronze locus in maize. Theor. Appl. Genet. 74: 471-475.

RHOADES, M. M., 1952 The effect of the bronze locus on anthocyanin formation in maize. Am. Nat. 86: 105-108.

ROBERTSON, D. S., 1985 Differential activity of the maize mutator M u at different loci and in different cell lineages. Mol. Gen. Genet. 200: 9-13.

SACHS, M. M., E. S. DENNIS, W. L. GERLACH and W. J. PEACOCK, 1986 Two alleles of maize alcohol dehydrogenase 1 have 3' structural and poly(a) addition polymorphisms. Genetics 113:

SAEDLER, H., and P. NEVERS, 1985 Transposition in plants: a molecular model. EMBO J. 4: 585-590.

SASGER, F., S. NICKLEN and A. R. COULSON, 1977 DNA sequencing with chain terminating inhibitors. Proc. Natl. Acad. Sci.

SHEPHERD, N. S., Z. SCHWARZ-SOMMER, J. BLL-MBERG VEL SPALVE, M. GUPTA, U. WIESASD and H. SAEDLER, 1984 Similarity of the Cinl repetitive family of Zea mays to eukaryotic transposable elements. Nature 307: 185-187.

STADLER, L. J., and S. FOGEL, 1945 Abstract. Gene variability in maize. 11. The action of certain R alleles. Genetics 30: 23-24.

THERES, N., T. SCHL-LE and P. STARLINGER, 1987 Cloning of the Br2 locus of Zea mays using the transposable element Ds as a gene tag. Mol. Gen. Genet. 209: 193-197.

WERR, W., W. B. FROMMER, C. MAAS and P. STARLISGER, 1985 Structure of the sucrose synthetase gene on chromosome 9 of Zea mays L. EMBO J. 4: 1373-1380.

WIESASD, U., U. WEYDEMASS, U. NIESBACH-KLOSGES, P. A. PE- TERSOS and H. SAEDLER, 1986 Molecular cloning of the c2 locus of Zea mays, the gene coding for chalcone synthase. Mol. Gen. Genet. 203: 202-207.

WILSOS, A. C., H. OCHMAS and E. M. PRAGER, 1987 Molecular time scale for evolution. Trends Genet. 3: 241-247.

YASISCH-PERROS, C., J. VIEIRA and J. MESSISG, 1985 Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUCI9 vectors. Gene 33: 103- 119.

ZACK, C. D., R. J. FERL and L. C. HASSAH, 1986 DNA sequence of a shrunken allele of maize: evidence for visitation by insertional sequences. Maydica 31: 5-16.

449-467.

USA 74: 5463-5467.

Communicating editor: W. F. SHERIDAS

sequence of three bronze alleles of maize and correlation with the genetic fine … · ·...

Documents