Zein of maize grain: I — isolation by gel filtration and characterization of monomeric and dimeric species

BIOCHIMIE, 1984, ~1~,~'~;1-A60

Zein of maize grain" I -- isolation by gel filtration and characterization of monomeric and dimeric species. J a c q u e s L A N D R Y * a n d Pierre G U Y O N * * .

Laboratoire d'dtude des Protdines, INRA, CNRA, 78000 Versailles. (Refu le 16-12-1984, accept~ apr@srevision ie 30-5-1984).

R ~ s u m ~ - - L a chromatographie de la zdine non rdduite sur Sephadex G 200 en prdsence d'ur@e

8 M, sur G 100 en pt'dsence de 1,5 ou 2,5 % de dodecyl sulfate de sodium (SDS) et sur G 100 hydroxypropyld en prdsence d'dthanol ~ 70 % met en dvidence deux fractions mineures A et B, et deux majeures D et M quel que soit le milieu. L'importance quantitative de la fraction M d~pend des conditions d'isolement de la z6ine. Elle ddcroft de 53 % des protdines contenues dans un extrait dthanolique et chromatographides telles que, ~t 40 % de la zdine purifide. Les valeurs de poids moldculaire obtenues ~t partir de l'@lectrophor@se en gel de polyacrylamide en prdsence de SDS et les donndes rehitives aux compositions en acides amines indiquent que les fractions D et M, telles qu'elles sont isoldes en prdsence d'@thanol, reprdsentent respectivement les formes dim@res et monom@res d'un mdlange de polypeptides de 22 000 et 24 000 comme M, et avec la thrdonine ou la ph~nylalanine comme r~sidu N-terminal. L'analyse dlectrophordtique de ia fraction M sdlectivement carbamyi6e en gel d'amidon p H 3,5 montre que les sous-unitds de zdine comprennent plusieurs polypeptides se diffdrenciant par le nombre et la nature des acides aminds basiques. Au moins un de ces polypeptides contient un r,;~i,t,, ,.h, !ysy!e. Mots-cl~s : mais / z@ine / filtration sat gel / electrophor/~e / masse molaire / acides amin/s. - - Unreduced zein chromatographed on Sephadex G 200 in 8 M urea, on G 100 in 1.5 or 2.5 % sodium dodecyl sulfate (SDS) and on hydroxypropylated G 100 in 70 % ethanol was resolved into two minor fractions A and B and two major ones D and M irrespective o f the medium. The quantitative importance o f the fraction M was dependent on the isolation conditions o f zein. It decreased from 53 % of the proteins contained in ethanolic extract and chromatographed as they were extracted, to 40 % o f the purified zein. The molecular weight vaf~es obtained fi'om SDS-polyacrylamide gel electrophoresis and amino acid compositional data indicated that fractions D and M, as isolated from purified zein in the presence o f ethanol, represented respectively dimeric and monomeric forms o f a mixture of M, 22 000 and 24 000 polypeptides with threonine or phenylalanine as NH2-terminal residue. Electrophoretic analysis o f selectively carbamylated fraction M on starch gel at p H 3.5 revealed that zein subunits comprised several polypeptides differing in the number and the nature o f basic amino acids. At least one o f these polypeptides contained one lysyl residue. Summary

Key-words : maize / zein / gel filtration / electrophoresis / molecular weight / amino acids.

Present addresses * Laboratoire de Chimie biologique et de Photophysiologie (INRA). lnslitut National Agronomique Paris-Grignon. Centre de Grignon. 78850 Thiverval-Grignon.

**Groupe de Recherchessur les Interactionsentre Microorganismes et Plantes(GRIMP). Unirersit#Paris-Sud. Institut de Microbiologie. 91405 Orsay.

452

J. Landrv a n d P. Gu.von

Introduction The most abundant product of protein accumulation in maize endosperm is zein which is made up of entities extractable from grain by aqueous ethanol or 2-propanol. These proteins, in reduced state, are resolved by polyacrylamide gel electrophoresis in the presence of sodium dodecyl-sulfate (SDS-PAGE) into two predominant sets of polypeptides with apparent molecular weight (Mr) ranging from 19 000 to 23 000 for the first one and from 21 000 to 25 000 for the second one [I-5]. In their native state (as it is found in the absence of reducing agent), they exist as monomers, dimers or small sized oligomers [5-11]. Zein polypeptides also display some variations in their primary structure, as judged by amino acid analyses, electrophoresis at acid pH, isoelectric focusing (IEF) and N-terminal amino acid sequence determination of samples originating from several genotypes [5, 12-16]. A greater multiplicity of zein proteins was elicited by subsequent characterization (gel electrophoresis or peptide mapping) of fractions isolated by IEF or gel filtration or chromatography on ion exchanger or reversed-phase support [17-22]. Other evidence of zein complexity was supplied by studies of zein mRNAs and their cloned complementary D N A (eDNA). Thus, the existence of more than 100 genes encoding zein proteins was postulated from hybridization studies between these e D N A and maize D N A [23-25]. On the other hand, analyses of cloned e D N A fragments enable to elucidate the coding nucleotide sequence of a dozen of mRNAs and accordingly the amino acid sequences of zein polypeptides, providing evidence that all the sequences are distinguished by unique structures [26-30]. In the present work the size and charge heterogeneity of zein is further assessed by fractionating unreduced samples of different degrees of purification via gel filtration. The two major resultant fractions were characterized by their molecular weight, subunit structure, amino acid composition and electrophoretic mobility in gel at pH 3.5.

In a second article, the charge heterogeneity of zein will be further investigated by fractionating free subunits on ion exchanger [31].

Material and Methods Maize cuitivar Inra 260, whose biochemical characteristics have already been described [32] was used for all investigations. Zein was extracted at room temperature by 60% ethanol or 55 % 2-propanol from ground grains deprived of lipids and salt-soluble proteins. Proteins of 2-propanol extract were purified according the procedure of Landry (1979) whose steps are reported in Figure I. The purified preparation was termed unreduced zein when it was used as it was, isolated or reduced zein when it was incubated with an excess of 2-mercaptoethanol (0.65 mi for l g protein) for 15 h at room temperature. Sephadex G 100, G 150, G200 and LH20 were obtained from Pharmacia Fine Chemicals. All other

Ground defatted grain meal freed of salt-solublenitrogen and salt Soluble --2 PrOH, 20°C 1. Petroleum ether

~

Light phase

Heavy Jhase

•.~ Supernatant

2. NaCI 1%

Pellets 3. EtOH 92%

~

Insoluble

Soluble

/ 4. EtOH 100% , ~ i

Supernatant

f

Pellets I

I

5. Tert. BuOH ~--~ Insoluble

55% Soluble i

6. Chromato

i

~-Retarded fraction

LH20 FIG. 1. -- Schematic outfine for pur(fication and isolation of zein from maize grain. 2-PrOH : 2-propanol: EtOH : ethanol; tert-BuOH : tertiary butyl alcohol; chromato LH20 : chromatography on Sephadex LH20 in the presence of tert-BuOH 55 %.

~ Insoluble

Excluded fraction I

I

Tert. BuOH

7. Lyophilization~ 1

F

Powdered Zein

Characterization o f zein subunits chemicals were analytical grade. In particular, urea RP grade came from Prolabo (Rh6ne Poulenc). 8 M solutions were purified by stirring with charcoal for 15 minutes and by filtering the mixture through a sintered glass disk and through 0.45 ~tm Millipore filters. METHODS

Hydroxypropylation o f Sephadex G 100 The method was essentially that described by Ellingboe et al. [33] for Sephadex G 50. Sephadex G 100 (30g) was swelled in 900 ml water at 100"C for 15 h. After cooling, 20 g NaOH pellets and I g NaH,B dissolved in 100 ml water, were added to the medium and the mixture was soaked for 2 h. The wet Sephadex, free of excess aqueous phase, was transferred into 3 neck flask immersed in a water bath (38 "C), suspended in I 000 ml propylene oxide added progressively (I h) to avoid gel shrinkage, and refluxed with stirring for 2 h. The modified gel was filtered free of solvent and reaction products, then washed consecutively with acetone, water (until filtrate was non longer alkaline), 92% ethanol and equilibrated with chromatographic solvent overnight. After hydroxypropylation i g Sephadex G 100 equilibrated in 70 % ethanol gave 25 ml unpacked bed volume. Gel filtration chromatography Gel filtration chromatography was performed on a bed of Sephadex clutcd downward. All samples were dissolved in a small volume of solvent to give a I-3 % solution and introduced at the same flow rate as that used for subsequent elutions. Protein concentrations of effluent were determined by a.--.tomatic monitoring of 255 nm absorbance (A_,~d and by reading the 280 nm absorbance of the content of tubes with a UV spectrophotometer. Molecular weights were determined from the elution volumes of protein by gel filtration of unreduced zein on a calibrated Sephadex G 150 column (2.6 x 60 cm), according to the procedure of Page and Godin [341, using SDS !.5 %, Tris-HCI pH 8 and sodium p-chloromercuribenzoate 10 4 M. Ovalbumin, pepsin, chymotrypsinogen A and cytochrome C, used as molecular weight standards, were not reduced.

453

Gels were stained overnight with Coomassie brilliant blue R 250 in water-ethanol-acetic acid (40/30/10, v/v/v) and destained with 7.5 % acetic acid.

Starch gel electrophoresis Electrophoreses were carried out in 12 % gels in the presence of 6 M urea and i 5 m M aluminium lactate at pH 3.5 as previously described [51.

Chemical analyses Amino acid analyses were performed after hydrolysis of the sample with 5.7 N HCI at l l0"C in tubes seailed under vacuum for 24 and 48 hours. Sulfur amino acids were determined after performic acid oxydation and hydrolysis for 18 hours. The dinitrofluorobenzene method of FraenkeI-Conrat et al. [351 was employed for determination of NH_,-terminal amino acids. The DNP amino acids, after hydrolysis, were identified by thin layer chromatography. Carbohydrate components were assayed as glucose by the phenol-sulfuric acid method of Dubois et al. [361. Protein and glucose samples were dissolved in 52 °0 ethanol.

Results

Gel filtration chromatography A typical c h r o m a t o g r a p h i c separation (Fig. 2) of u n r e d u c e d zein on h v.1 d. . .r. .o x v. i nr - r o- i nt - . ,v, " l.a. l .e.d. . .Kenha. . . . . r---0.7t"

o.5oE i-

0.25"

Polyacrylamide gel electrophoresis in the presence o f sodium dodecyl sulfate ( S D S - P A G E ) Samples were prepared by dissolving 0.25 mg of protein in 1 ml of solution containing I % S D S , 10%glycerol, 0.002% bromophenol blue in 0.01 M tris-glycine buffer pH 8.4. Separation gels contained 12.5 % acrylamide and 0.033 % N, N'-methylene-bis-acrylamide. The electrophoresis was carried out in glass tubes (i.d. 5 mm) using 0.375M tris and 0.06NHCI buffer, pH8.9 and 0.1% SDS, at ! mA per gel for 15 min, and 3 mA per gel for 2 hours.

.// 0.2

0.3

0.4

0.5

Ve / Vt

FiG. 2. -- Gel .fi.hrafion of unredm'ed -_era r200 rag) on a hvdroxvpropylawd Sephadex G !00 column (3 x 48 c m ) equifib'rawd and developed with 70 % ethanol and 0.01 N HCI. Abcissa : ratio of elution volume (VA to the total volume of gel bed (V,). Ordinate : absorbance (A) at 280 nm. A and B = minor fractions, oligomeric forms: D and M -- major fractions, dimeric and monomeric forms, respectively.

454

J. Landry and P. Guyon

dex G 100 equilibrated in 70% ethanol and 0.01 N HCI revealed four protein fractions, as detected by their absorbance at 280 nm (A2s0), their ratio A.,s0 to A255, and their nitrogen content • two minor ones A and B, and two major ones D and M. A fifth fraction (not shown) representing non-protein ultra-violet absorbing material appeared in a volume close to that corresponding to the complete retention. The elution profile depicted in Figure 2 is similar to that previously observed for a purified sample of unreduced zein chromatographed on Sephadex G 200 equilibrated in 8 M urea [6]. Five protein fractions, referred to as A, B, C, D and E in the order of their emergence have been found. Compared to the new nomenclature which is related to the degree of oligomerization, A and B represent the same fractions whereas C and D correspond to D and M, E lacking in the preparation studied in Figure 2. The ~verlapping of fractions A, B and D together with the clear separation of fraction M led us to follow the influence of some experimental conditions upon the size polymorphism of zein by the relative contribution made by fraction M to all the protein fractions including M, evaluated on the basis of their absorbance at 280 nm. As shown in Table I, the percentage of fraction M isolated from hydropropylated G 100 was independent of concentration and nature of salt or acid included in the alcohol (however, their absence resulted in ill-defined separation)° It remained unchanged when hydroxypropylated G 100 and 70 % ethanol were replaced by G 100 and 1.5 or 2.5 % SDS. But, it was lower when zein TABLE I Percentage of fraction M. according to the conditions used for gel filtration of purified zein'a'.. Sephadex G I00 H '~' G 100 G 200

Solvent ,b, 70% EtOH 70 % EtOH 70% EtOH 2.5 % SDS 1.5 % SDS 8 M urea + 7 M urea + 7 M urea +

+ 0.01 M LiCl + 0.05 M LiC! + 0.01 N HC!

0.01 N HCI 0.01 N HCI 0.01 N HCI, 4°C

{a) On the basis of absorbance at 280 nm; (b) At room temperature (othervise specified); (c) HydroxypropylatedG 100.

M 38 39 41 39 43 25 27 27

was chromatographed on G 200 in the presence of urea. Note that more concentrated solutions of zein were ebtained with aqeuous ethanol than with 8 M urea. The data presented in Table If show how the preparation conditions of zein influenced the yield of fraction M, as isolated from hydroxypropylated Sephadex G 100. Thus, about 54 % of zein proteins extracted by 70% EtOH appeared in fraction M when they were chromatographed just as they were extracted or after precipitation with 1% NaCl in water or after gel filtration on Sephadex LH 20 equilibrated in 70% ethanol. This percentage was the same for proteins extracted by 55 % 2-propanol and precipitated by 1% NaCl. But, it was smaller for more purified samples. TABLEII Influence of treatment of alcohol-soluble proteins upon the percentage of fraction M ta~isolated from hydroxypropylated Sephadex ~. Proteins Extracted by 70 % EtOH and a) submitted such as they were b) precipitated in 1% NaCl c) filtrated on LH 20 (70 % EtOH) Extracted by 55 % 2-PrOH and a) precipitated in 1% NaCI b) precipitated in 1% NaCl, then in 100 % EtOH c) purified zein (Figure I)

M

53 54 54 52 44 39

(a) On the basis of absorbance at 280 nm; (b) equilibrated in 70 % EtOH and 0.01 N HCI.

Since salting-out of alcohol soluble proteins by 1% NaCI had no noteworthy effect on the percentage of M proteins, fractions D and M originating from pooled tubes of purified zein chromatography on hydroxypropylated G 100 were isolated by a procedure involving the steps 3, 5, 6 and 7 of Figure 1. Rechromatography of these fractions on the same column, yielded 82 % of D for fraction D and 92 % of M for fraction M, indicating that they remained essentially under their original size after their recovery. Finally, fraction M levelled off to 93 % when it was isolated from proteins of alcoholic extract, reduced by 2-ME and alkylated by methyl iodide or acrylonitrile.

Characterization of zein subunits

455

Characterization o f fractions D and M Features of fractions D and M were determined on sample isolated after c h r o m a t o g r a p h y of purified zein through hydroxypropylated G 100.

TABLE I11

Composition of zein and its chromatographicfractions Amino acids '~

Molecular weights

Asx Thr Ser Glx Pro Gly Ala Cys Val Met lie Leu Tyr Phe Lys His Arg NH, (°~ Terminal Carbohydratd ~}

On the basis of behaviour of unreduced zein on a calibrated column of Sephadex G 150, the a p p a r e n t molecular weights of fractions D and M were evaluated to be about 45 000 and 22 500, respectively. These values comp,~red with those obtained from gel S D S - P A G E analysis of purified zein and its fractions D and M. As shown in Figure 3, unreduced D resolved as 4 bands with M~ around 45 000. U n r e d u c e d fraction M consisted of two minor b a n d s of Mr = 45 000, a minor band of M, = 2 4 0 0 0 and a major one of Mr = 22 000. After reduction D and M a p p e a r e d as 2 major oands of M~ = 22000 and 2 4 0 0 0 with minor bands o f higher Mr (45 000 and over). However, tae reduced fraction D, with respect to the reduced fraction M, comprised a larger amount of Mr --- 24 000 proteins and traces of Mr = 16 000 component, only detected when gels were overloaded.

Amino acid composition Table III lists the amino acid content of fraction D, M and two subfractions M~ and M_~ c o m p a r e d with those of purified zein and one zein subunit whose amino acid sequence was deduced from nucleotide sequence of the related e D N A clone A 30 [26, 281. The subfractions M~ and M,,

1

2

3

4

5

6

7

8

Zein

M

M,'"' M 2"' A30Z'~'

49 49 50 50 50 30 30 28 28 26 64 67 64 67 68 213 208 207 213 208 101 104 103 104 107 18 21 19 20 19 132 131 139 139 139 I0 7 9 9 !0 41 44 36 36 30 8 12 8 12 4 39 36 36 31 33 189 185 193 188 197 34 35 35 35 35 52 51 55 50 56 1 1 1 t ..... 0.4 9 8 9 7 7 10 10 10 11 1! Thr Thr Thr nd ('' nd Phe Phe Phe 0.28 0.17 0.18 nd nd

47.0 23.5 70.5 197.0 108.0 23.5 136.0 9.4 23.5 0 42 202 37.5 61 0 9.4 9.4 Thr nd

(a) number of residues per I 000 recovered residues (b) proteins eluted at the beginning of the peak M (c) proteins eluted at the end of the peak M (d) amino acid composition of polypeptide calculated from its nucleotide sequence (Geraghty et al., 1981) (e) Threonine was preeminent in all cases, ~l

!

zl~li

"~l.,t'.l

H|zil~,J

(g) expressed as glucose equivalent and as percent on dry

matter basis.

9

FI(; 3. --

-45.0

O i W

D

-25.7

Q

-12.5

Polyat'o/unlidc gel ch'ctr~qTh(~rc.~is

ol zein, and lkactims.~ D and Jl in both unre
Lanes : I. 40 lag of unreduced zein: 2. standard proteins (unreduced). ovalbumin (molecular weigh! = 45.000). ct-chymotrypsinogen (mol. wt. 25,700) and cytochrome C (mol. ~t. 12.500): 3, 25 lag of reduced zein: 4. 40 lag of unreduced D: 5 and 6. 25 and 100 lag of reduced D: 7.40 lag of unreduced M: 8 and 9. 25 and 100lag of reduced M. Numbers in right margin refer to M, ( x l0 ') of the standards. Arrow indicates the presence of faint band in the lane 6.

456

J. Landry and P. Guyon some of which were spaced regularly. For convenience, they were identified by a first number representing the regular interspace !, followed by a serial number within this length [5]. Thus, zein electroforms were observed, in order of decreasing mobility, a s : i) a faint band, the fastest moving designated by 02; 2) two major closely spaced bands (doublet) I1 and 12; 3) a prominent band 21 and 4) a minor band 31. The interspace ! therefore, was the length which separated bands 11 and 21, or bands 21 and 31. When electrophoresed under the same conditions, the unreduced and reduced fraction M, as well as reduced zein, resolved as these typical bands. For the latter sample an extra band, which could be designated by 32, was also observed. Unreduced fraction D exhibited trailing and a series of regularly spaced bands, hardly visible between bands 12 and 31. Reduced fraction D, which was made up of bands 11, 21 and probably 41, resembled reduced fraction M, but differed from it by containing extra bands located near band 31.

were isolated on the basis of difference in their absorbance at 280 and 255 nm. From previous data [29], the subfraction MI, which comprised the early-eluting proteins of meak M, would be a mixture of Mr = 45 000, 24 000 and 22 000 species, whereas the subfraction M2 would mainly be made up of Mr = 22 000 polypeptides. The overall amino acid patterns of chromatographic fractions exhibited no special feature when compared between them or with that of total zein. However, fraction D was richer than fraction M in methionine. The same held for subfraction MI when compared to subfraction M2. On the other hand, the amino acid composition of subfraction M2 is very close to that of polypeptide A 30 Z, one of Mr = 22 000 proteins [26]. The NH2-terminii of purified zein and fraction D and M were determined to be threonine and phenylalanine with a predominance of the former one. Purified zein and fraction D and M were found to contain no appreciable amount of carbohydrate. Indeed, one subunit out of five would be contaminated with a glucose equivalent per polypeptide chain.

Starch gel electrophoresis at pH 3.5

Electrophoretic analysis of carbamylated components of fraction M

Starch gel electrophoresis at pH 3.5 (Figure 4) resolved unreduced zein into six typical bands

The charge heterogeneity of zein polypeptides was further assessed by studying the alteration in

!

2

3

4

5

..~

7

8

9

"",v

"1|

12

4-

L FIG. 4. -- Starch gel electrophoresis (6 M urea, pH 3.5) of :ein, and fractions D and M in unreduced and reduced state. Samples were loaded on 0.5 x 0.8 cm paper wick. Lanes : 1, 150 I.tg of unreduced zein: 2 and 3, 200 and 1501.tg of unreduced D; 4 and 5, 40 and 20 I.tg of unreduced M (preparation n ° I); 6, 40 I.tg of unreduced M (preparation no2); 7, 100 lxg of reduced zein; 8 and 9, 200 and 100 l.tg of reduced D; 10 and 11, 40 and 201,tg of reduced M (preparation n ° 1); 12, 40lag of reduced M (preparation no 2). On the left is indicated the numbering of main bands of unreduced zein.

31 22 21 12 II 02'

O g

O

Q

Characterization of zein subunits mobility of M electroforms following their incubation with potassium cyanate. This salt carbamylates amino groups of polypeptides, causing a net loss of one positive charge (at pH. 3.5) per each altered amino group. It is important to note that u-NH_+ groups are carbamylated 100 times as faster as ~-NH2 groups at pH 7, whereas both groups are substituted at the same velocity at pH 9 [371. Figure 5 compares acid-urea gel patterns of fraction M before and after a 4 days incubation with 0.01 M potassium cyanate in 8 M urea or 70 % ethanol or a mixture of equal volume of 8 M urea and 70 % ethanol. For each tested medium, samples were non treated proteins, and material

I

I

KCNO

0

8M urea 0 + + + +

HCI

0

0

22

33

+ +'0

0

incubated with cyanate in the presence or absence of HCl. Irrespective of the sample medium, patterns of material treated with cyanate in the presence of HC! were indistinguishable from those of non treated proteins, indicating that proteins were not affected by cyanate at acid pH. By contrast, patterns of material incubated without HCI, when referred to those of non treated proteins were dependent on carbamylation. So, the pattern of proteins treated with cyanate in 8 M urea (apparent pH of medium: 7.2), was characterized by an anodic translation of ! for all the electroforms such that their relative proportions were essentially the same as those obtained prior to cyanate treatment. Since NH,-terminus

2 233 22 3311 urea + EtOH 60% EtOH + + + + + + 0 0 0 0 + + + +00 0 0 + + 0000 11

-'31C

.~-31C

s

.~12C' ,;~21C

." .-'"

21 " "

12 e . f . ~It

.--31C

f

s

31 - - " " ' "

.-e21C

._!12C

~'" I

s

..

•

•

" ,l~p

.--12C

S

•

- ' " 2 IC ~'-''

• °

.fS.m'llC

~-02C

. ~

s

OS

---.

,,',-02C

• s •

~12C

11C d

~02C

~"-"

J

s J

4" p

02 " 1,2

3

457

1,2

3

1,2

3

FIG 5. -- Starch gel electrophoresis (6 M urea. pH 3.5) analysis of the carbano.lated products fronl the fraction M. Samples to be electrophoresed were dissolved in 8 M urea or in 60 % ethanol (EtOH) or in a mixture of equal volume of 8 M urea and 60% EtOH in the absence (0) or presence ( + ) of potassium cyanate and HCI. For each medium lanes are : 1, native sample: 2, cyanate treated samples in the presence of HCI: 3, cyanate treated samples in the absence of HC!. in lower portion the changes in the mobility are explicited. The electroforms are designated according to their mobility in the native state. C, electroform with carbamylated NH,-terminus. C'. electroform with carbamylated NH,-terminal and lysine residues.

J. Landry and P. Guyon

458

was the only amino group which was present in all the electroforms, the length ! represented the decrement in mobility of any M polypeptide relative to the loss of one positive charge. On the other hand, the absence of the fast moving doublet in conjunction with the presence of the slow moving one in the pattern of fraction M, after its treatment by cyanate in aqueous ethanol (apparent pH9.4), indicated that one electroform had undergone a 2 ! shift and lost two positive unit charges. This electroform contained one lysine and was identified as electroform 12. Indeed, in the pattern of proteins carbamylated in urea and ethanol (apparent p H : 7.9), the electroform 12 was present as mono-substituted state in the fast moving doublet and partially as di-substituted state in the slow moving doublet. Therefore, at least one among zein polypeptides contained one lysyl residue. One the other hand, since a distance equal to 4 ! separated band 41 from band 02, zein polypeptides at acid pH can display differences of four positive unit charges between them.

Discussion In this investigation zein was isolated from whole grain and was therefore made up of some alcoholic-soluble polypeptides released from oerm and containing ~nrn~ ly~in~ (~ r,~i,4,1,~ nor 1000). However, such polypeptides which amounted 1% of total zein [32] represented a too low contamination of zein proteins originating from endosperm to alter significantly the properties of this material and more specially its lysine content. In this regard, it is worth mentioning that small amounts of lysine were reported for endosperm zein isolated from the same variety [32] or others [24, 11, 14, 38, 39] and for alcohol-soluble proteins originating from free protein bodies [40]. On the other hand, SDS-PAGE and amino acid analyses revealed that our zein preparation was devoid of G~-glutelins, which are methionine-rich proteins with Mr ranging between 10 000 and 16 000 [1, 2, 4, 10, 41, 42]. They also emphasized the absence of G2-glutelins, which are histidinerich polypeptides with Mr = 28 000 [42, 43]. Therefore, our zein sample turned out to be more homogenous than most preparations isolated by other research groups. Thus it consisted of two sets of subunits with Mr of 22 000 and 24 000 corresponding to absolute molecular weights of 23 300 and 25 500, as calculated from amino acid .

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

.

i[.! ~,.

compositions of homologous polypeptides predicted from the nucleotide sequence of the related cDNA clones can be deduced from the nucleotide sequences [26-30]. In unreduced zein, the subunits are free or associated into oligomeric forms, predominantly into dimers. Gel filtration of this material on hydroxypropylated Sephadex G 100 enables us to isolate monomers and dimers, as fractions M and D, in the absence of denaturing agents. Previous findings [12, 39] with gel filtration of zein in the presence of SDS suggest that the technique described here could be the basis for isolating the Mr = 22 000 subunits from the Mr. 24 000 ones. Polypeptides originating from both subunits sets were very similar in the overall amino acid composition and in the quality of NH2 terminii. However, Mr = 24 000 subunits, with respect to Mr = 22 000 ones were richer in methionine and very prone to be under dimeric form. Moreover, they contained the extra electroforms detected at pH 3.5 in the reduced fraction D and not in the unreduced or reduced fraction M. Reduced whole zein therefore was more complex than the unreduced or reduced fraction M. This has led us to study the traits of fraction M rather than those of reduced whole zein, that is to mainly characterize M r = 22000 subunits since they constituted about 80 % of Coomassie blue stains on the SDS gel of fraction M, as estimated by densitometric scans. But it is evident that the separation of Mr --- 24 000 subunits from Mr = 22 000 ones will be needed for a more precise characterization. The data obtained with carbamylation, contrary to those reported by Righetti et al. [14] provided a basis for calibrating the electrophoretic pattern of M proteins at pH 3.5 by the determination of a migration increment related to a difference of one positive charge between polypeptides of similar size, and for supplying further evidence that at least one polypeptide chain contained one lysyl residue. Nevertheles~s, the presence of one lysine in several polypeptides can be established only after a further fractionation of M proteins, as shown in the following article [31]. Thus the six electroforms seen in the fraction M at pH 3.5 correspond to six distinct polypeptide chains differing by at least the number of the quality of basic amino acid, or both. Since starch gel exerts a low sieving effect on protein migration, as shown by comparing patterns of unreduced fractions D and M, any electrophoretic form detected at pH 3.5 in reduced zein must

459

Characterization o f zein subunits

indicate the occurrence of polypeptide chains with an unique primary structure. Regarding NH.,-terminal residues our results were in accordance with those of Bietz et al. [15] but slightly different from those of Drenska et al. [16] who detected a predominance of phenylalanine. It is interesting to note that NH2-terminal residues of Mr 22 000 zeins were found to be threonine for clones A2o, A30 and A20 or serine for clones A20 whereas those of Mr 24 000 zeins were identified to be phenylalanine for clone ~. ZG 99 and serine for clones ~. ZG 99 and Z4 [27-30]. In alcoholic extract the majority of zein subunits was in free form and remained in this state after the proteins were salted out in the presence of sodium chloride. In contrast, a notable oligomerization of subunits was quoted by many workers who isolated zein by an extensive dialysis of its alcoholic solution against water followed by lyophilisation of protein pellets or the entire content of dialysis bag, as described by Craine et al. [44]. This was revealed for unreduced preparation by a limited solubilization in 8 M urea and by the low level of fraction M present in samples when analyzed by gel filtration, it may be speculated that with the procedure of Craine et al., some ethanol solvating hydrophobic residues remains bound to zein in spite of exhaustive dialysis. Its elimination, more rapid than that of water during cryodessication, would favour hydrophobic interactions between polypeptides, causing their aggregation. For instance, ct-zein isolated from whole grain by extraction with 92 % ethanol, followed by dialysis and lyophilization has been found to be enriched in high molecular weight oligomers with respect to starting zein [1 I]. In contrast, ot-zein isolated by fractional precipitation on Sephadex LH 20 [45] and submitted, s~ch as it was isolated, to gel filtration on hyd~oxypropylated Sephadex G 100 has been proved to be free of material with a degree of oligomer~ation higher than 2 (data not shown). Therefore, ~yophilization of zein from an aqueou r solution of ~e~iary butanol, appears to be a convenient and perhaps the most gentle procedure for isolating zein. Indeed, the polymorphisme of such a p)'eparation, as detected by SDS-gel electrophoresi~ [5], is similar to that of proteins extracted by 70 % ethanol and dialysed againts 0.5 % SDS [10]. Finally, all these findings strongly suggest that oligomerlzation of zein is not necessarily related to the fora-~ation of interchain disulfide-bonds.

Acknowledgements The authors whish to thank J.C. Huet. S. Delhaye. C. Monget (amino acid analysis). M. Sallantin (electrophcresis) and M. Domin for their collaboration.

REFERENCES I. Misra, P.S., Mertz, E.T. & Giover, D.V. (1976) Cereal Chem.. 53, 705-711. 2. Gianazza, E., Righetti, P.G., Pioli, F., Galante, E. & Soave, C. (1976) Maydica. 21, 1-17. 3. Lee, K.H., Jones, R.A., Dalby, A. & Tsai, C.Y. (1976) Blochem. Genetics, 14, 641-650. 4. Paulis, J.W. & Wall, J.S. (1977) Cereal Chem.. 54, ! 223- !228. 5. Landry, J. (1979) Biochimie. 61, 549-558. 6. Landry, J. (1965) C.R. Acad. Sci., 261 C, 2775-2778. 7. Ganchev, K.D. & Stoyanova, R.Zh. (1972) C.R. Acad. Bulg. Sci., 32, 8. Ivanov, Ch.P., Ganchev, K.D. & Drenska, A.I. (1976) C.R. Acad. Buig. Sci., 29, 571-574. 9. Rewa, W.A. & Brfickern, C.C. (1979) Biochem. Physiol. Pflanz., 174, 451-461. 10. Tsai, C.Y. (1980) Cereal Chem., 57, 288-289. !i. Paulis, J.W. (198i) Cereal Chem., 58, 542-546. 12. Landry, J. & Sallantin, M. (1983) Cereal Chem.. 60, 242-245. 13. Aiexandrescu, V., Chihaia, C. & Cosmin, O. (1976) Rev. Roum. Biochim., 13, 169-172. 14. Righetti, P.G., Gianazza, E. Viotti, A. & Soave, C. (1977) Planta, 136, 115-123. 15. Bietz, J.A., Paulis, J.W. & Wall, J.S. (1979) Cereal. Chem.. 56, 327-332. 16. Drenska, A.I., Ganchev, K.D. & Ivanov, Ch.P. (1979) C.R. Acad. Bulg. Sci., ,67-70. 17. Valentini, G., Soave, C. & Ottaviano, E. (1979) Heredity. 42, 33-40. 18. Vitale, A., Soave, C. & Galante, E. (1980) Plant. Sci. Lett.. 18, 57-64. 19. Hagen, G. & Rubenstein, I. (1980) Plant Sci. Lett.. 19, 2 i 7-233. 20. Wilson, C.M., Shewry, P.R. & Miflin, B.J. (1981) Cereal Chem., 58, 275-281. 2 !. Esen, A. (1982) Cereal Chem., 59, 272-276. 22. Bietz, J.A. (i982) J. Chromatog.. 255, 219-238. 23. Viotti, A., Sala, E., Marotta, R., Aiberi, P. Balducci, C. & Soave, C. (1979) Eur. J. Biochem.. 102, 2 ! ! -222. 24. Wienand, V. & Feix, G. (1980) FEBS Lett.. 116, 14-16. 25. Hagen, G. & Rubenstein, I. (1981) Gene, 13, 239-249. 26. Geraghty, D., Peifer, M., Rubenstein, I. & Messing, J. (1981) Nucleic Acids Res.. 9, 5163-5174. 27. Marks, M.D. & Larkins, B.A. (1982) J. Biol. Chem., 257, 9976-9983. 28. Geraghty, D.E., Messing, J. & Rubenstein, I. (1982) EMBO J.. I, 1329-1335.

460

J. Landry and P. Guyon

29. Hu, N.T., Peifer, M.A., Heidecker, FG., Messing, J. & Rubenstein, 1. (1982) EMBOJ., 1, 1337-1342. 30. Pedersen, K., Devereux, J. Wilson, D.R., Sheldon, E. & Larkins, B.A. (1982) Cell., 29, 1015-1026. 31. Landry, J. & Guyon, P. (1984) Biochimie, 66,451-460, 461-469. 32. Landry, J. & Moureaux, T. (1980) J. Agric. Food Chem., 28, 1186-119!. 33. Eilingboe, J., Nystr6m, E. & SjOvall, J. (1968) Biochim. Biophys. Acta, 152, 803-805. 34. Page, M. & Godin, C. (1969) Can J. Biochem., 47, 401-403. 35. Fraenkel-Conrat, H., Harris, J.I. & Levy, A.L. (1955) in : Methods of biochemical analysis (D. Gluck) lnterscience, New York, 2, 370-371. 36. Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A. & Smith, F. (1956) Anal. Chem., 28, 350-358. 37. Stark, G.R. (1967) in : Methods in Enzymology

38. 39. 40. 41. 42. 43. 44. 45.

(S.P. Coiowick & N.O. Kaplan) Academic Press, New York & London, XI, 590-594. Misra, P.S., Mertz, E.T. & Glover, D.V. (1976) Cereal Chem., 53, 699-704. Abbe, M., Arai, S., Kato, H. & Fujimake, M. (1981) Agric. Biol. Chem., 45, 1467-i 472. Christianson, D.D., Khoo, U. Nielsen, H.C. & Wall, J.S. (1974) Plant. Physiol., 53, 851-855. Melcher, U., Fraij, B. (1980) J. Agric. Food Chem., 28, 1334-1336. Landry, J., Paulis, J.W., Fey, D.A. (1983) J. Agric. Food Chem., 31, 1317-1322. Vitabe, A., Smaniotto, E., Longhi, R. & Galante, E. (1982) J. Exp. Bot., 33, 439-44~'. Craine, E.M., Freimuth, D.V., Boundy, J.A. & Dimler, R.J. (1961) Cereal Chem., 38, 399-406. Landry, J., Guyon, P. & Moureaux, T. (1967) C.R. Acad. Sci., 265 D, 264-267.