Plant Physiol. (1985) 78, 178-1820032-0889/85/78/01 78/05/$0 1.00/0

Analysis of Leaf Proteins by Two-Dimensional GelElectrophoresisPROTEASE ACTION AS EXEMPLIFIED BY RIBULOSE BISPHOSPHATE CARBOXYLASE/OXYGENASE DEGRADATION AND PROCEDURE TO AVOID PROTEOLYSIS DURINGEXTRACTION

Received for publication September 27, 1984 and in revised form February 3, 1985

CATHERINE COLAS DES FRANCS*, HERVfE THIELLEMENT, AND DOMINIQUE DE VIENNELaboratoire de Genetique des Systemes Vegetaux, G.LS. Moulon, 91190 Gif-sur- Yvette, France

ABSTRACT

A previous genetic analysis by two-dimensional electrophoresis ofwheat leaf proteins led us to hypothesize that several polypeptides weredegradation products of the large subunit of ribulose bisphosphate car-boxylase/oxygenase (EC 4.1.139).

Various extraction procedures including the use of proteinase inhibi-tors allowed us to: (a) confirm that the suspected polypeptides are indeeddegradation products; (b) find out under which conditions the proteinasesact; (c) find a method which prevents degradation, by boiling the extractin the presence of sodium dodecyl sulfate and 2-mercaptoethanol.

During protein extraction, cell compartments are broken: pro-teinases mix with the cell content and degradation products arereleased in the extract. Thus, additional bands or spots appearon electrophoresis gels, that have no direct physiological orgenetic significance. Such a phenomenon has been suggestedfrom a previous work in our laboratory (13) comparing theprotein patterns obtained by 2D' gel electrophoresis of leavesfrom various wheat lines. The analysis of the same nucleargenome on different cytoplasms revealed two distinct patterns:one for the euplasmic lines (Triticum aestivum cytoplasm) andone for the alloplasmic lines (five closely related Aegilops cyto-plasms). Among the differences scored between those two cyto-plasmic patterns, we noticed that 24 spots, including the cyto-plasmically encoded LS of Rubisco (EC 4.1.1.39), exhibited asimilar shift for isoelectric point and for mol wt. Genetic argu-ments (13) led to the suggestion that all these 24 polypeptideswere products of the same Rubisco LS gene. Nineteen of thosepolypeptides had a mol wt lower than the LS. The questionremained whether they were prematurely terminated translationintermediaries or degradation products of the LS occurring invivo or during protein extraction.

In this paper, we report the effect of various extraction exper-iments on 24 spots related to the LS: the 19 spots of mol wtlower than the LS and five further spots that were not consideredby Zivy et al. (13). All but one are indeed products of a degra-dation which occurs during extraction, mainly in 9.5 M urea. A

'Abbreviations: 2D, two-dimensional; LS, large subunit; PMSF, phen-ylmethylsulfonyl fluoride; Rubisco, ribulose bisphosphate carboxylase/oxygenase.

method avoiding proteolysis and thus production of additionalspots in electrophoresis is proposed, which is essential for anygenetic or physiological study.

MATERIALS AND METHODS

Plant Material. Wheat seeds from the euplasmic lines ChineseSpring and Selkirk were sown in Petri dishes in the dark andtransferred after 4 d to small pots in the greenhouse (16 hphotoperiod). The expanded first leaf of an 8-d-old plant consti-tuted one sample.

Tris Extraction. Proteins were extracted according to Zivy etal. (13). One hundred mg of leaf pieces were dry crushed in amortar cooled by liquid N2. The powder was resuspended at 0°Cin 2 ml of 30 mm Tris-Cl1 buffer (pH 8.7) with 1 mm DTT, 1mm ascorbic acid, 1 mM EDTA-Na2, 5 mM MgCI2, and 12 mgof insoluble PVP. After two centrifugations of the sample(35,300g for 15 and 10 min), ice-cold acetone (8 ml) with 10mM 2-mercaptoethanol was added to the supernatant, and theproteins were allowed to precipitate for 1 h at -18C. After 10min of centrifugation at 35,300g, the pellet was air dried andresuspended in 200 Al of O'Farrell 'lysis buffer' (9.5 M urea, 5%2-mercaptoethanol, 2% Nonidet P 40, 2% ampholytes) (8). Thesamples were stored at -80°C. This procedure is further referredto as control conditions.

Extraction in the Presence of Proteinase Inhibitors. Proteinaseinhibitors used are antipain, leupeptin, pepstatin A, chymostatin,and elastatinal (Peninsula laboratories), a2-macroglobulin (Boeh-ringer Mannheim) and PMSF (Sigma).The following solutions were used: solution A: 1 mM PMSF,

0.1 M pepstatin A, 2 mM EDTA-Na2 (10); solution B: 1.38 jsMa2-macroglobulin; solution C: 1 mm antipain; solution D: 1 mMantipain, 1 mM leupeptin, I mM pepstatin; solution E: 1 mMchymostatin, 1 mM elastatinal, 1 mM pepstatin.

Antipain, leupeptin, and elastatinal were solubilized in distilledH20, chymostatin in glacial acetic acid, pepstatin A in methanol,and PMSF in acetone. These solutions were added either to theTris-Cl1 extraction buffer (all solutions) or to the O'Farrell lysisbuffer (solution D only).

Degradation of the Extracts at Room Temperature. Two dif-ferent experiments were performed: (a) an extract in Tris-Clextraction buffer (pH 8.7) was kept at room temperature for 2 hbefore the first centrifugation; (b) after standard extraction inTris-Cl buffer (pH 8.7) and resolubilization of the final proteinpellet in the O'Farrell lysis buffer, the extract remained at roomtemperature for 2 h before storage at -80°C.

178 www.plantphysiol.orgon November 11, 2017 - Published by Downloaded from Copyright © 1985 American Society of Plant Biologists. All rights reserved.

LEAF PROTEIN EXTRACTION FOR 2D-GEL ELECTROPHORESIS

Acidic~~~~~~~~~~~~~~,-W

X0 # t R 5

M

A"no

U_ ...

AN. / .g~ ~ ~ ~ ~ _

*. _ --,

0-_ot :zf..* ..^e

*,.p.... / se

FIG. 1. Two-dimensional pattern of 8-d-oldwheat leaf proteins (cv Selkirk). The extraction(Tris procedure) was performed according to Zivyet al. (13). Arrows indicate the 24 spots related tothe large subunit of Rubisco (LS spot).

^ t , A.O

Extraction in the Presence of SDS. After grinding, the leafpowder was submitted to two kinds of extraction procedures.The SDS Procedure. The powder was resuspended at 0C in 2

ml SDS solution containing 4% SDS, 5% 2-mercaptoethanol,5% sucrose (5) with 12 mg of insoluble PVP. The mixture was

heated in boiling water for 3 min and then twice centrifuged at35,300g for 15 and 10 min (1). The subsequent procedure is asdescribed for Tris extraction.The Tris-Cl- Procedure followed by the SDS Procedure. The

powder was resuspended first in 2 ml of Tris-Cl- extractionbuffer with 12 mg of insoluble PVP. After the second centrifu-gation, 4% SDS, 5% 2-mercaptoethanol, and 5% sucrose wereadded to the supernatant which was boiled for 3 min. Thesubsequent procedure is as above mentioned.

Electrophoresis. The first dimension isoelectric focusing (IEF)was done according to Zivy el al. (13) in 120-mm-long glasstubes with 2 mm i.d. Gel composition was 4% acrylamide, 9.2M urea, 4% carrier ampholytes, and 2% NP40. The seconddimension electrophoresis was performed as described by O'Far-

rell (8) except for the acrylamide concentration of the gels. Weused uniform 11% acrylamide running gel (0.15% SDS) and 5%acrylamide stacking gel (0.1% SDS).The 2D gels were silver stained according to Oakley et al. (7)

except for the developer solution that was half-concentrated.Method ofComparing Gels. Gels were compared to the control

by superposing transparent films. The reported results wereobtained after observation of at least two well defined gels foreach extract.

Intensity changes (increase, stability, decrease, or disappear-ance) of the 24 spots related to the LS were visually noted. Thespot nomenclature is as described by Zivy et al. ( 14).

RESULTS

The 24 spots examined are shown in Figure I and schematizedin Figure 2. Their variations according to the different extractionprocedures are listed in Table I.

Proteinase Inhibitors. The 24 spots remained unchanged whensolutions A or B are added to the Tris-C- extraction buffer.

Basic

I E F

u')0Cf)

179

-i AW.-d4w

6la; '...

op...iq f, ."w

-.-f,%1--i qw,A

I

N,

VI

www.plantphysiol.orgon November 11, 2017 - Published by Downloaded from Copyright © 1985 American Society of Plant Biologists. All rights reserved.

COLAS DES FRANCS ET AL. Plant Physiol. Vol. 78, 1985

Acidic

000C

0CoooO

coo°aC~~~~~~~~ O00 °00 so %O

0~~0 8K~~3~0o AOc?FOIO _0oa et°O0°

o oe, OPP v O

d_o O O000 900 00

0~~~~~~~~~~t O000 d

OcDO~~~ oOO° >g°O C) °OC.C

05 aa

0 0o900 O0 ° ° O ° ° °

0~~~60ov o8°O# dk87oto

o~ccI~ o 0 °Q16090

CD03 0 0 o

050

Cb40C)0 OO1602~

oCC 9 07 0

1507 175017191609 0 ~~~~~00 >c p

2201/2205'

0

CO

FIG. 2. Map of the reproducible spots retainedfrom the control gels. Numbers indicate the 24 spots

o analyzed. Spots are striped, black, or circled accord-ing to their behavior in the presence of proteinaseinhibitors (see text).

0

When solutions C, D, or E were added to this buffer, the spotsare not similarly altered (Fig. 2): six spots (striped spots) do notexhibit any intensity change and five spots (black spots) show an

intensity decrease. Three spots (circled spots) have a differentbehavior according to the proteinase inhibitors (solutions C, D,or E) present in this buffer: spots 1609 and 2201 decrease withD and E and remain unchanged with antipain alone; spot 2205remains unchanged with C and D and decreases with E (TableI).When solution D is added to the lysis buffer, all the spots but

one behave as when solution D is added to the Tris-Cl- buffer:spot 1606 only does not decrease but remains unchanged.

Degradation at Room Temperature. Gels obtained from an

extract kept for 2 h in Tris-CI- extraction buffer (pH 8.7) looklike control gels. Only spots 11 1 1 and 1503 seem to be more

intense.Gels obtained from an extract kept for 2 h in the O'Farrell

lysis buffer at room temperature are rather similar to the controlexcept for spot 1653 that shows increased intensity.The SDS Procedures. After extraction in the SDS solution,

three spots show noticeable intensity decrease (1609, 1807, 2201),and 20 spots disappear. Only spot 2205 seems to be more intense(Fig. 3).

After the Tris-Cl- procedure followed by the SDS procedure,23 spots among 24 are much fainter than in the control. Spot1501 seems not to be altered.

DISCUSSION

The intensity decrease of nearly all the studied spots afterextraction in the presence of some proteinase inhibitor mixturesis consistent with our hypothesis that they are degradation prod-ucts of the LS of Rubisco. It is surprising, however, that theseinhibitor mixtures are not more effective under the presentconditions: solution A (containing PMSF) that is considered byScopes (10) to be suitable to prevent proteinase action and a2-

macroglobulin (known to be efficient against all kinds of protein-ases) (4) have no noticeable effect. Pepstatin and leupeptin,although widely used and recommended (2, 9, 10), and antipain,chymostatin, and elastatinal, are not fully effective since none ofthe degradation spots disappears.The fact that gels from an extract kept for 2 h in Tris-Cl-

extraction buffer (pH 8.7) are not very different from the controlindicates that the proteinases act either very early in the Tris-Cl-buffer or in the urea lysis buffer. The latter hypothesis wasretained because after the Tris procedure followed by the SDSprocedure, the degradation products are present, but at a very

180

Basic

EF

C]

=n

10'

o

~00

www.plantphysiol.orgon November 11, 2017 - Published by Downloaded from Copyright © 1985 American Society of Plant Biologists. All rights reserved.

181LEAF PROTEIN EXTRACTION FOR 2D-GEL ELECTROPHORESIS

Table I. Behavior ofthe 24 Spots Related to the LS ofRubisco According to the Extraction ProcedureArrows indicate unchanged ( ), increased (A), or decreased (X) intensities ofthe spots relative to the control. Symbol (-) represents disappearing

spots.Extraction Procedure

Tris procedureSpot Tr+ + + Lysis Tris 2 h at Lysis buffer 2 h SDS followed by

solution C solution D solution E solution D temp. at room temp. procedure SDSprocedure

110711091111 * -* -. -/ A X .-1002 N. N. N.X, -X1501 , - -1503 - ' ' ' A ' -

1505 - -' X1507 N. N. N. N. - -1602 X X X X X -1651 NUX X X --1652 X X X X X X,1653 NUA1604 X X \1654 X -1750 X X X X ' - -1704 N X X N ' - - N1708 NU X X1706 X X X X - -1606 X X X - - -1719 N NX X - - N.1609 NX N1807 N. X X2201 NX X X - N N2205 -X N -,A

faint intensity. This means that proteinases act a little in the Trisbuffer but are much more active in urea. When solution D isadded to the urea lysis buffer, it has a noticeable action whichconfirms that in the control the proteinases are effective in 9.5M urea. Such a result is not so surprising, as Doi et al. (2) pointedout that the hydrolytic rate for casein by pepstatin-sensitive acidprotease from resting rice seeds increases in the presence of 3.6M urea. Miller and Huffaker (6) also reported that pretreatmentwith 5 M urea does not affect the activity of barley leaf thiolproteinase.Then we can wonder why an extract kept for 2 h in the urea

lysis buffer does not seem to be more degraded than the control:it can be suggested that the proteinases act in urea only during ashort time before being either degraded by autolysis or denaturedby urea. Autolysis was indeed reported by Drivdahl and Thimann(3) for oat leaf proteinases.The reported results show that the SDS procedure is a drastic

method: many spots of low mol wt are lacking on the gels (Fig.3). In particular, all the spots (except 2205) related to the LS ofRubisco dramatically decrease or disappear (Table I). The heattreatment of the extract in a solution containing 4% SDS (astrong detergent) and 5% 2-mercaptoethanol (a powerful re-ducer) alters quaternary and tertiary structures of proteins. So itis likely that proteinases also undergo denaturation and lose theiractivity in this solution.Among the 24 spots analyzed, only one (spot 2205) has an

unexpected behavior. It is only affected by solution E and in-creases with the SDS procedure although the 23 other spotsdecrease or disappear. This spot appears to be unrelated to the

others and thus might be the product of another cytoplasmicgene, as previously discussed (13).

CONCLUSION

The experiments described above demonstrate that 23 amongthe 24 spots related to the LS ofRubisco are degradation productsof this polypeptide. This has recently been confirmed in ourlaboratory by the absence of these spots on 2D gels from anextract passed through an immuno-affinity column preparedwith Rubisco antibodies.The proteinases are active not only with the Tris-Cl- extraction

buffer but also and preferentially in the O'Farrell lysis bufferduring the resolubilization of the protein pellet at room temper-ature.The abundance of Rubisco in green leaves (about 50% of

soluble proteins) and its fast degradation (11, 12) may explainthe presence of such a high number of degradation products ongels. Since these proteinases are not Rubisco-specific, it is likelythat many other proteins are degraded during the Tris extractionprocedure. The gels obtained after extraction by heating in thepresence of SDS are consistent with this hypothesis as, even ifno exhaustive gel analysis has been performed, numerous otherspots are lacking (Fig. 3) that may also be degradation products.The gels produced by this method permit more accurate geneticor physiological analysis: a better relationship 'one gene prod-uct-one spot' can thus be obtained.

Acknowledgments-We wish to thank Joel Blaisonneau, Marie-Pierre Digard,Fabienne Granier, and Martine le Guilloux for their technical assistance andEvelyne Ron for the English rewriting. We are grateful to Pr. P. Gadal for hiscritical reading of the manuscript.

www.plantphysiol.orgon November 11, 2017 - Published by Downloaded from Copyright © 1985 American Society of Plant Biologists. All rights reserved.

COLAS DES FRANCS ET AL.

Acidic

4,

A1...

// /'il

Plant Physiol. Vol. 78, 1985

FIG. 3. Two-dimensional pattern of 8-d-oldwheat loafproteins (cv Selkirk) obtained after SDSextraction procedure. All degradation products butone (spot 2205) are absent. Arrows indicate theirlocation (see Fig. 2).

.1

of

/..

LITERATURE CITED

1. BAHRMAN N, D DE VIENNE, H THIELLEMENT, JP HOFMANN 1985 Two-dimensional gel electrophoresis of proteins for genetic studies in Douglas fir(Pseudotsuga menziesii). Biochem Genet 23: 245-253

2. Doi E, D SHIBATA, T MATOBA, D YONEZAWA 1980 Characterization ofpepstatin-sensitive acid protease in resting rice seeds. Agric Biol Chem 44:74 1-744

3. DRIVDAHL RH, KV THIMANN 1978 Proteases of senescing oat leaves. II.Reaction to substrates and inhibitors. Plant Physiol 61: 501-505

4. HARPEL PC 1973 Studies on human plasma a2-macroglobulin-enzyme inter-actions. Evidence for proteolytic modification ofthe subunit chain structure.J Exp Med 138: 508-521

5. HARRISON PA, CC BLACK 1982 Two-dimensional electrophoretic mapping ofproteins of bundle sheath and mesophyll cells of the C4 grass Digitariasanguinalis (L) Scop. (crabgrass). Plant Physiol 70: 1359-1366

6. MILLER BL, RC HUFFAKER 1981 Partial purification and characterization ofendoproteinases from senescing barley leaves. Plant Physiol 68: 930-936

7. OAKLEY BR, DR KIRSCH, R MORRIS 1980 A simplified silver stain for detectingproteins in polyacrylamide gels. Anal Biochem 105: 361-363

8. O'FARRELL PH 1975 High resolution two-dimensional electrophoresis of pro-

teins. J Biol Chem 250: 4007-40219. ROSICHAN JL, RC HUFFAKER 1984 Source of endoproteolytic activity associ-

ated with purified ribulose bisphosphate carboxylase. Plant Physiol 75: 74-77

10. ScoPES RK 1972 Protein purification. Principles and Practice. Springer Verlag,New York

11. WITTENBACH VA 1978 Breakdown of ribulose bisphosphate carboxylase andchange in proteolytic activity during dark-induced senescence of wheatseedlings. Plant Physiol 62: 604-608

12. WITTENBACH VA 1979 Ribulose bisphosphate carboxylase and proteolyticactivity in wheat leaves from anthesis through senescence. Plant Physiol 64:884-887

13. Zivy M, H THIELLEMENT, D DE VIENNE, JP HOFMANN 1983 Study on nuclearand cytoplasmic genome expression in wheat by two-dimensional gel elec-trophoresis. I. First results on 18 alloplasmic lines. Theor Appl Genet 66: 1-7

14. Zivy M, H THIELLEMENT, D DE VIENNE, JP HOFMANN 1984 Study on nuclearand cytoplasmic genome expression in wheat by two-dimensional gel elec-trophoresis. II. Genetic differences between two lines and two groups ofcytoplasms at five developmental stages or organs. Theor Appl Genet 68:335-345

182

Basic

E F

ce

C)

(f)

'I,.. fiI

I.:, s?, i" ..

:,- .1 .:

i.

www.plantphysiol.orgon November 11, 2017 - Published by Downloaded from Copyright © 1985 American Society of Plant Biologists. All rights reserved.