Purification of a Zn-Binding Phloem Protein with Sequence ... · Purification of a Zn-Binding...

Plant Physiol. (1 996) 11 o: 657-664

Purification of a Zn-Binding Phloem Protein with Sequence ldentity to Chitin-Binding Proteins’

Kathryn C. Taylor*, 1. Cene Albrigo, and Christine D. Chase

Department of Plant Sciences, University of Arizona, Tucson, Arizona 8572 1 (K.C.T.); Citrus Research and Education Center, University of Florida, Lake Alfred, Florida 33850 (L.G.A.); and Department of Horticultural

Sciences, University of Florida, Gainesville, Florida 3261 1 (C.D.C.)

In citrus blight, a decline disorder of unknown etiology, the tree canopy exhibits symptoms of Zn deficiency while Zn accumulates in the trunk phloem. We have purified a Zn-binding protein (ZBP) from phloem tissue of healthy and blight-affected citrus (Citrus sinensis [L.] Osbeck on Citrus jambhiri [L.]). The molecular weight of the ZBP was estimated to be 5000 by size-exclusion chromatography and sodium dodecyl sulfate-polyacrylamide gel electrophoresis. lon-exchange chromatography at pH 8.0 demonstrated the 5-kD ZBP to be anionic. A partia1 N-terminal amino acid sequence revealed a cysteine-, glycine-rich domain with 45 to 80% identity with the chitin-binding domain of hevein, wheat germ agglutinin, and several class I chitinases. That the abundance of this protein increased 2.5-fold in association with Zn accumulation in the phloem is characteristic of citrus blight. Tissue mass changes of the phloem suggests that altered tissue structure accompanies blight. Phloem accumulation of the 5-kD ZBP may be in response to wounding or other stress of blight-affected citrus.

Citrus blight is a unique system for the study of Zn metabolism. Blight is a tree decline disorder that results in a reduction in water conductivity to the extent that the fruit load is appreciably reduced. Early in citrus blight etiology, the so-called predecline stage is characterized by Zn accumulation in phloem tissue, particularly in the trunk (Young et al., 1980; Albrigo and Young, 1981; Albrigo et al., 1986). This accumulation of Zn in trunk phloem is usually accompanied by Zn-deficiency symptoms in the leaves (Childs, 1953). The abundance of Zn in phloem tissue is correlated with the presence of Zn-binding factors (Taylor, et al., 1988). Zn redistribution occurs prior to later decline symptoms (decline stage) and is currently the single identifiable physiological change in blight-affected trees in the predecline stage. The dysfunction in Zn metabolism associated with blight allows the elucidation of the participation of Zn in normal plant processes.

One to three years after the onset of Zn redistribution, xylem vessels in the inner trunk wood become plugged (Brlansky et al., 1984), water conductivity is decreased

Supported in part by U.S. Department of Agriculture-National Research Initiative Competitive Grants Program grant No. 89-37- 264-4752. Florida Agricultura1 Experiment Station Journal Series No. R-04710.

* Corresponding author; e-mail [email protected]; fax 1-520-621-7186.

(Young and Garnsey, 1977), and water-stress symptoms develop (Albrigo et al., 1986). Altered Zn distribution may be directly associated with subsequent drought symptoms. Zn deficiency is correlated with disruption of xylem ele- ments and abnormal xylem function in cabbage (Sharma et al., 1984; Sharma and Sharma, 1987). In addition, IAA oxidase activity is reduced in blight-affected citrus (Bausher, 1982). This symptom may be associated with a disruption of IAA turnover and activity, which affects xylem development and function. The visible symptoms of blight are accompanied by nutrient changes in the leaf phloem and xylem (Wutscher et al., 1983a, 198310; Williams and Albrigo, 1984), including a shift in nitrogen to the amino acid pool, primarily as Pro (Hanks and Feldman, 1974; Syvertsen and Albrigo, 1984). Pro accumulation is associated with water stress in plants (Levy, 1980; Aspinall and Paleg, 1981).

Many enzymes in higher plants require the presence of Zn as a cofactor for optimum activity (Keilin and Mann, 1940; Vallee, 1983). Carbonic anhydrase (Bausher, 1979) and IAA oxidase activities (Bausher, 1982) are decreased in leaf tissues of trees in the later stages of blight, consistent with the known Zn requirement of these activities (Randall and Bouma, 1973; Lindskog, 1983; Vallee, 1983). In addition, ATP content is decreased in blight-affected trees at this stage. Zn may be important to the balance of ATP/Pi by affecting the activity of several phosphatases (Vallee, 1983). Such effects of Zn depletion may be secondary re- sponses to phloem Zn accumulation within blight-affected citrus trees. Zn redistribution, characteristic of the early stage of citrus blight, may cause a metabolic deficiency of Zn throughout the tree, creating alterations in tree metabolism.

The sequestration of Zn in citrus trunk phloem was studied in blight-affected and healthy trees to understand how aberrant Zn distribution may be involved in the development of citrus blight. To undertake such studies it was first necessary to purify and characterize the Zn- sequestering agent(s) from blight-affected citrus. Zn-binding assays indicated that similar Zn-sequestering agents are present in the trunk phloem tissue extracts of healthy and predecline and decline stage blight-affected citrus trees (Taylor et al., 1988). However, blight trunk phloem extracts contain more Zn-binding activity than do healthy trunk

Abbreviation: ZBP, Zn-binding protein. 65 7

www.plantphysiol.orgon May 27, 2018 - Published by Downloaded from Copyright © 1996 American Society of Plant Biologists. All rights reserved.

http://www.plantphysiol.org

658 Taylor et al. Plant Physiol. Vol. 110, 1996

phloem extracts (Taylor et al., 1989). We report here the purification and characterization of a ZBP from phloem tissue. This ZBP appears to be responsible for the sequestration of Zn, removing it from routine metabolic processes of the blight-affected tree. Partia1 N-terminal sequence of the ZBP indicates a relationship to chitin-binding proteins. In addition, we suggest that changes in phloem structure are associated with the onset of citrus blight and the accumulation of this ZBP.

MATERIALS AND METHODS

Plant Material

Phloem tissue samples were collected from fully mature (>40 years old) Valencia sweet orange (Citrus sinensis [L.] Osbeck) scions on rough lemon rootstocks (Citrus jambkiri [L.]), a blight-susceptible rootstock common in Florida. Blight-affected trees sampled for this study were positive for Zn redistribution and reduced water conductivity as described by Albrigo et al. (1986) and for the presence of the 12-kD blight-specific protein (Derrick et al., 1990). Healthy trees lacked these symptoms. Predecline trees dis- played the Zn redistribution without the reduction in water conductivity. In the field, the outer bark was scraped away from the surface at 30 to 45 cm above the graft union with the edge of a stainless steel knife blade and the phloem tissue was scored down to the cambium. Two 10-cm2 patches of phloem were sampled from each tree. Phloem was then separated at the cambium and stripped away from the tree and immediately placed into liquid nitrogen. Sampling was accomplished during periods of cambial growth (Williams and Albrigo, 1984).

Phloem Tissue Extraction

To quantitate the recovery of ZBP, six independent purifications were performed, starting with phloem tissue samples collected from three healthy and three blight- affected trees. Frozen samples were ground in an electric coffee grinder for approximately 15 to 20 s. Phloem tissue samples (7.5 g/sample) from healthy or blight-affected trees were extracted three times each by stirring on ice for 20 min in 40 mL of cold 50 mM Tris-C1 buffer, pH 8.0, with 50 m h p-mercaptoethanol and 5% (w/w) polyvinylpolypyrrolidone. The 120-mL homogenate from each sample was strained through six layers of cheesecloth and centri- fuged for 20 min at 20,000gmax. The resulting supernatant was filtered through a 0.22-pm cellulosic membrane and degassed.

Chromatographic lsolation of the ZBPs from C. sinensis

Phloem tissue extracts were fractionated by ion-exchange chromatography, by applying the 120 mL of extract to a 2.5- X 14-cm QAE-Sepharose fast flow column (approximately 100 mL of bed volume). The column was washed with 300 mL of equilibration buffer (50 mM Tris-C1 buffer, pH 8.0) and eluted with 400 mL of 0.0 to 2.0 M NaCl in 50 mM Tris-C1, pH 8.0, at 4°C. Aliquots of each fraction were assayed for Zn by atomic absorption spectrometry

(Perkin-Elmer model3100) and for by UV spectropho- tometry. Fractions with elevations of A,,, and coincident elevated Zn were pooled. The volume of pooled samples was reduced in a centrifugal vacuum concentrator (model RC 10.10; Jouan, Winchester, VA). Concentrated Zn containing QAE-Sepharose eluant was chromatographed on a 2.5- X 14-cm size-exclusion column of Bio-Gel P6 (Bio-Rad) in 10 mM Tris-C1 buffer, pH 8.0 (4°C). Fractions within the peaks of A,,, and Zn content were pooled and further purified by 4 to 20% gradient SDS-PAGE.

Zn and Protein Assays

Total Zn was determined as a measure of A,,,., in atomic absorption spectrometry. Total protein was determined at A,,, with Quantigold (Diversified Biotech, Newton Centre, MA) or with the Bradford assay (Bio-Rad), depending on protein concentration with each stage of purification.

SDS-PACE

Purified ZBP (50 pg) was applied to 4 to 20% gradient gels in a 30-pL final volume containing 10% sample buffer (50 mM Tris-C1 [pH 6.81, 10% glycerol, 2% SDS, 5% P-mercaptoethanol, 100 mM DTT, 0.1% bromphenol blue). Elec- trophoresis was carried out for 1.5 h at 150 V, constant voltage, in a pH 8.3 Laemmli buffer system (Laemmli, 1970). Gels were silver stained using some modifications of Morrissey (1981).

Cel Fractionation

Following SDS-PAGE of samples containing 200 pg of protein, similar unstained gels were sliced across their width into 0.5-cm fractions down the length of the gel. Each gel fraction was eluted in 1.0 mL of 0.5% HC1 in HPLC- grade water. The total Zn content eluted from each gel slice was determined by atomic absorption spectrometry.

IEF

The pI of the ZBP was determined by preparative IEF with a Rotofor (Bio-Rad). An IEF gradient (53 mL) was established from pH 3.0 to 10.0 with ampholyte at a concentration of 2% (Servalyt 3-10, Serva Biochernicals, Para- mus, NJ) in Tris buffer (1 mM Tris-C1, 8 mM 3-[(cholami- dopropyl)dimethylammonio]-1-propanesulfonic acid, 5 mM P-mercaptoethanol) and up to 20 mg of protein of the Bio-Gel P6-purified ZBP. IEF was performed at 12 W of constant power, with generally 500 to 1500 V. Proteins were focused for 5 h. The pH gradient for the 20 fractions collected after preparative IEF was assessed by direct pH measurement of each fraction. Zn determinations were made for each fraction. The proteins in these fractions were further separated by SDS-PAGE (20% polyacrylamide, with pH 8.3 Laemmli buffer [Laemmli, 19701) to determine sizes of proteins in high Zn-containing fractions.



Zn-Binding Phloem Protein 659

Preparation of Polyclonal Antisera

ZBP purified by ion-exchange chromatography and size- exclusion chromatography was fractionated by preparative SDS-PAGE. The gels were stained in Coomassie blue for 5 min to identify the location of the band of interest. After the gels were destained for 30 to 45 min, ZBP bands were cut out of the gels and equilibrated in H,O (three washes of 45 min each). Gel pieces were forced through a 21-gauge needle. The shredded gel (containing approximately 200 pg of protein) was used for inoculation of New Zealand White rabbits with a boost after 21 d. A test bleed was taken at 5 weeks postinoculation. A third boost was given at 5 weeks and the sera were collected 8 weeks after the last inoculations.

Determination of Amino Acid Sequence

The amino acid sequence of the purified 5-kD ZBP was analyzed at the University of Arizona macromolecular structures facility after 4-vinyl pyridine derivatization of Cys residues under a nitrogen atmosphere (Fullmer, 1984).

Immunochemistry

Proteins were extracted from leaves, roots, and phloem of healthy and blight-affected trees as indicated above. Xylem exudate, obtained from roots of blight-positive rough lemon seedlings, was donated by Dr. Kenneth Der- rick (Lake Alfred, FL). In addition, leaves were dissected into midveins and remaining lamina to determine whether the proteins were specific to the vascular tissue in the leaf. Species distribution of the 5-kD ZBP was assessed by immunoblot analysis of protein extracts from leaves and phloem of castor bean, peach, plum, apple, corn, squash, tomato, bean, and cucumber. Chl was removed from leaf samples by incubating the extract (extraction in 50 mM Tris-C1 buffer, pH 8.0, with 50 mM P-mercaptoethanol and 5% [w/w] polyvinylpolypyrrolidone) in an equal volume of chloroform for 15 min on ice. The organic phase was removed following a 10-s centrifugation. For a11 immunoblot analyses, the proteins were separated by SDS-PAGE in a 20% polyacrylamide gel, followed by transfer to a 0.45-pm poly(viny1idene diflouride) membrane in the buffer of Towbin et al. (1979). Electrotransfer was accomplished with the Bio-Rad Mini Trans-Blot apparatus within 45 min. Protein blots were developed as described by King et al. (1985).

Determination of Phloem Fresh and Dry Weights

Observation of sampled phloem tissue suggested that blight-affected trees had thicker bark (phloem) than did healthy trees. A No. 9 cork borer was used to sample 1.28-cm circles from the trunk surface of 20-year-old citrus trees, approximately 20 cm above the graft union. Six discs were removed from each of three trees in the healthy, predecline, and decline stages. The discs were sealed in plastic film and stored on ice in the field. Upon returning to the laboratory the discs were weighed and then dried in a forced air oven at 60°C prior to dry weight determination.

These measurements were made with and without the rhytidome, which is the nontranslocating portion of the bark. Additionally, the starch contents of phloem tissues above the graft union of blight-affected and healthy trees were analyzed by staining with iodine-potassium iodide (Molnar and Parups, 1977). After the tissues were stained, fresh longitudinal and cross-sections of these tissues were viewed by bright-field microscopy.

RESULTS

Purification of the ZBP

Previous results (Taylor et al., 1988, 1989) indicated that the Zn-binding factors in healthy and blight-affected citrus phloem extracts were anionic at pH 8.0 when separated by ion-exchange chromatography using DEAE-Sephadex as the anion exchanger. In this system, a Zn mercaptide was eluted with 0.4 M NaCl. In the present study, we found that the primary Zn-binding activity was weakly bound by the quaternary chemistry of QAE-Sepharose at pH 8.0. The Zn-binding component apparently interacted with the quaternary chemistry in so far as it was isolated from QAE- Sepharose when the isocratic equilibration buffer (50 mM Tris-C1, 0.0 M NaC1) was fractionated. This fractionation gave a Zn peak that contained 74% of the Zn applied to the column in the crude phloem extract (Fig. ZA). When the isocratic elution was followed by a O to 2.0 M NaCl gradient, there were additional losses of Zn at a base line leve1 throughout the elution profile, with the exception of a small peak eluted by 0.43 M NaCl that accounted for 6% of the total Zn applied to the column. Subsequent purification by size-exclusion chromatography of the QAE-Sepharose- excluded Zn peak (Fig. 18) demonstrated that the Zn activity was complexed, with an apparent molecular mass of 5000 D (Fig. 2). SDS-PAGE confirmed that Zn was associated with a 5-kD protein (Fig. 3). In addition, separation of Zn-containing fractions by preparative IEF demonstrated that the 5-kD ZBP had a pI of 6.8 (Fig. 4). Based on specific Zn content of the 5-kD ZBP isolated from the trunk phloem tissues of blight-affected and healthy citrus, we achieved 4100- and 790-fold purifications, respectively.

Account of Zn and Protein through the Purification of 5-kD ZBP

The presence of the 5-kD protein was further correlated with the redistribution of Zn noted in the citrus blight disorder. The levels of Zn and protein in each purification step were accounted for on a total protein and fresh weight basis. The abundance of 5-kD ZBP was higher in phloem tissue from blight-affected trees. On a protein basis, there was 2.5-fold more 5-kD ZBP in the phloem of blight- affected citrus trees relative to healthy trees (Table I). Fur- thermore, the blight-affected phloem tissue contained almost 6-fold more ZBP than did healthy tissue on a fresh weight basis. The Zn content per microgram of protein after ion-exchange chromatography and gel filtration was similar in healthy and blight-affected phloem tissue, which is further indication of the similar binding characteristic of the Zn ligand isolated from each tree. Zn and protein




A 181614

1210u

Ro> 8

642

0

1

0.8

0.6

0.4

0.2

Zn-Healthy

Zn-Blight

A254-Healthy

, A254-Blight

Fraction Number Fraction Number

Figure 1. A, Isocratic elution profile from QAE-Sepharose ion-exchange column containing 5-kD ZBP. B, Elution profilefrom Bio-Gel P6 column containing 5-kD ZBP. A and B, Zn-containing peaks from blight-affected and healthy citrus trunkphloem tissue each contained chromophores that absorbed UV at 254 nm.

determinations for three separate purifications of 5-kD ZBPwere similar.

ZBP Amino Acid Sequence Analysis

Once the 5-kD ZBP was purified, its amino acid sequencewas determined from the N terminus. The sequence anal-ysis of the 5-kD ZBP represented about one-half of theprotein (22 amino acid residues) (Fig. 5). This sequence wascharacterized by six Cys's and five Gly's. The Cys's werepresent in a repetitive pattern: Cys2-X5-Cys-X4-Cys2-X5-Cys. The sequence had 45 to 80% identity with the chitin-binding domains of several classes of wound-inducible

10-

5-

MolecularWeight(Kda)

2.

Blue Dextran, Vo

Bacitracin

VolumeFigure 2. Elution of the 5-kD ZBP from the Bio-Gel P6. The size-exclusion elution profile was calibrated for molecular mass with 3mg each of adrenocorticotropic hormone (ACTH, 4.5 kD) and bac-itracin (1.3 kD). The void volume (Vo) was determined by bluedextran (2000 kD) elution.

1 2

^^H^m

46.0^

30.0» ,

14.3k

6.0k5kDZBP»

MW (kDa)1.0 2.0

Zn (ug/ml)3.0

Figure 3. Association of ZBP with Zn. SDS-PAGE gels (4-20%) wereloaded with purified 5-kD ZBP. The graph adjacent to lane 2 indi-cates the level of Zn eluted from 0.5-cm gel slices of the 11-cm gel.Lane 1 contains molecular mass markers. www.plantphysiol.orgon May 27, 2018 - Published by Downloaded from

Copyright © 1996 American Society of Plant Biologists. All rights reserved.



Blig ht-aff ected

Healthy - pH Gradient -5 kDa ZBP

12

9

6

3 PH

12

9

6

3

1 10 20 Fraclion Number

Figure 4. Gel-filtered, Zn-containing fractions were further purified by preparative IEF. Fractions (2 mL) were assayed for Zn content and pH. Zinc peaks with pl 1 2 were associated with an ampholyte. The pl for the 5-kD ZBP was 6.8.

proteins. Hevein is a member of this group of chitin-binding proteins. It is a 5-kD protein that is present in the latex of the rubber tree and has been shown to have antifungal activity against several plant pathogenic fungi (Van Parijs et al., 1991). These proteins may be induced by a number of environmental and pathogenic stress conditions, such as drought, heat, heavy metals, vira1 infection, insect feeding, and funga1 infection (Raikhel and Broekaert, 1993). The presence of this type of protein is only correlated with plant stress, with no indication of the source of that stress. The sequence had 71 to 80% identity with several chitinases, 45 to 57% identity with the four chitin-binding domains of the wheat germ agglutinin isolectin A, and 60% identity with mature hevein from Heveu brusiliensis (Fig. 5).

lmmunochemistry

The ubiquity of the 5-kD ZBP was investigated by as- sessing the cross-reactivity of the 5-kD ZBP with tissues from a number of plant species. Polyclonal antiserum for the 5-kD ZBP was used in immunoblot analyses. We demonstrated that this antiserum was specific to the 5-kD ZBP, since it did not cross-react with proteins of similar molecular mass in leaf or phloem extracts from a range of woody

and herbaceous species, genera, and families (see "Materi- als and Methods" for species screened; data not shown).

Symptom expression and whole-tissue analysis of leaf and trunk phloem tissue in blight-affected citrus led inves- tigators to suggest that Zn was redistributed from the leaves to the trunk phloem above the graft union. We performed immunoblot analyses to investigate this sugges- tion. The 5-kD ZBP was localized in the vascular tissue of blight-affected and healthy citrus. The 5-kD ZBP appeared to be specific to phloem in assays of soluble proteins extracted from leaves, roots, phloem tissue, and xylem exudate (Fig. 6) and in assays of proteins extracted from whole leaf, leaf lamina (without midvein), and midvein (Fig. 7).

Phloem Fresh and Dry Weights

Earlier casual observation indicated that the thickness of phloem tissue of blight-affected citrus might be greater than that of healthy citrus. We determined the relative fresh and dry weights of bark (rhytidome plus translocating phloem) for blight-affected and healthy citrus. For both fresh and dry weights, the bark of decline-stage trees weighed 14% more than that of healthy trees (Table 11). When the translucent phloem tissue (translocating region) was separated from the rhytidome, differences in phloem mass between healthy and blight-affected citrus were more apparent. Fresh and dry weights for this portion of the phloem were 48 and 37% greater, respectively, in blight- affected than in healthy phloem. The fresh/dry weight also was 8.5% greater in the translocating region of phloem from healthy than from blight-affected trees. The starch content of phloem tissues above the graft union of blight- affected and healthy trees was also compared. No differences in starch content were detected when fresh sections were viewed by bright-field microscopy.

DISCUSSION

The 5-kD ZBP was purified from trunk phloem tissue of blight-affected and healthy citrus. Based on specific Zn content of the 5-kD ZBP from these tissues, we achieved 4100- and 790-fold purifications, respectively. This ZBP was increased in concentration 2.5-fold on a protein basis with the occurrence of citrus blight. We demonstrated a 6.5-fold increase in the ZBP on a fresh weight basis in an earlier report (Taylor et al., 1989). The presence of the 5-kD ZBP in healthy trees suggests a role in normal citrus tree metabolism. However, its relative abundance along with Zn accumulation in trunk phloem tissue of blight-affected citrus strongly suggests that the protein has some role in

Table 1. Account of Zn and protein contents of 5-kD ZBP with purification

Healthy Trees Blight-Affected Trees

Total Total Zn pg Zn/pg mg protein/g Total Total Zn pg W c L g mg proteink Stage of Purification

protein Protein fresh wt protein Protein fresh wt

mg cL6 mg cL6 Crude extract 30 72.8 0.0024 3.015 30 274.4 0.0091 5.87 lon-exchange chromatography peak 0.33 54.2 0.1 64 0.033 0.1 5 127.4 0.849 0.029 Gel filtration peak 0.0042 42.0 10 0.0004 0.0115 82.5 7.175 0.0023




CCGKSV--- VC PGGECCSRFGWCGLII: : I II I I I I : I I I I I

EQCGRQAGGALCPGGNCCSQFGWCGSTTDYCG

..•.

NQRGE

• •S••R....... . ."S• >E• V«

..•R•K* •

•G

....• D..

. * * ' *

* • • *

DDK-• "KA

.GSNME----GR

.AS""• **•••

• GNN«««SGN«...NN-.•AS"...NN>.••S.M.

..K....

..SY."• *•*•»*

...Y. ..

...w .

..K....

...Y.Y.

.....Y.

•T-V.....PE-• • « PEF. ..DE-«N>N..• MGG.«•KGPK«

..

..• s• s• s• •

• *

5 kD ZBP

CHIT PHAVUCHI5 PHAVUCHI1 PISSATCHI2 TOBACCHI POPTRWIN2 SOLTUWIN1 SOLTUHEVE HEVBRCHIB SOLTUAGI 1 WHEATAC AMP2

Figure 5. N-terminal amino acid sequence of 5-kD ZBP and itsalignment with related proteins. Vertical bars indicate direct identitywith related proteins; double dots indicate conserved changes, con-sidered as substitutions within the amino acid homology groupsFWY, MILV, RKH, ED, NQ, ST, and PAG. Caps introduced foroptimal alignment are represented by dashes.

the citrus disorder. With further characterization we maybe able to determine its role in the etiology of the citrusblight disorder as well as understand its significance to Znhomeostasis in general.

The binding of Zn by 5-kD ZBP was stable throughout anumber of purification steps, including nondenaturing anddenaturing electrophoresis. Earlier data suggested that theZBP was able to bind Zn in excess of that found to beassociated with the citrus phloem extracts. These charac-teristics suggest a function in metal binding, possibly formetal homeostasis under stress conditions. The 5-kD ZBPappears to be a suitable ligand for Zn as well as for severalother metals. Zn-binding fractions also bind Pb and Cd(Taylor, et al., 1988). Like the phytochelatin polypeptidesand metallothionein proteins (Robinson et al., 1990; Stef-fins, 1990), the citrus ZBPs are anionic at pH 8.0, and theirbinding is temperature stable and reversibly pH labile(Taylor et al., 1989).

The ZBP is similar to other metal ligands in plants, suchas the phytochelatin polypeptides and metallothionein pro-teins (Robinson et al., 1990; Steffins, 1990). Based on aminoacid sequence data for the 5-kD ZBP, 27% of the sequencedamino acid residues were Cys's, 23% were Gly's, and 4.5%

were glutamate. Despite the high Cys percentages and theCys-x-Cys arrangements, the overall percentages and ar-rangements of these amino acids were more consistentwith the chitin-binding domain of class I chitinases thanphytochelatin or metallothionein sequences. Block analysis(Henikoff and Henikoff, 1994) of the sequence was in the100th percentile (with a score of 1645) for similarity withthe chitinase from Phaseolus vulgaris (Broglie et al., 1986).The overall sequence identity of the N-terminal sequencewith the chitin-binding domain of the P. vulgaris chitinaseis 80%. The sequence identity between 5-kD ZBP and otherclass I (hevein-like) chitinases was also high (71-78%).When structural fidelity was considered, the sequences hadeven greater similarity (up to 85%). Several Cys's (residues2, 8,13,14, and 20), a key Pro (residue 9), and a hydropho-bic region (residues 17-19) were conserved in the sequence.

The identity of the 5-kD ZBP with a number of proteinshaving chitin-binding domains was surprising, given theapparent metal-binding function of the ZBP. Such metal-binding function has not been reported for chitin-bindingproteins of the type found in class I chitinases. Perhapsdivalent cations are bound by the Cys residues that occupya large portion of the typical chitin-binding domain. Cer-tainly there are reports of Ca and Mn binding by theleguminous lectins as well as some mammalian membrane-bound lectins (Sharon and Lis, 1989). However, metal bind-ing is accomplished through aspartate residues rather thanCys's. The chitin-binding capacity by the Cys residues ofthe ZBP is under investigation. Such cross-linking effectsmay have been the antagonizing factor that interfered withthe antimicrobial activity of the two chitin-binding, antimi-crobial peptides (Ac-AMPl and Ac-AMP2) of Amaranthuscaudatus (Broekaert et al., 1992) when challenged with amonovalent versus divalent cation. An understanding ofpotentially antagonistic roles of metal and chitin in 5-kDZBP function may be useful for elucidating the etiology ofcitrus blight as well as improving the understanding ofchitinases and other wound-inducible proteins in plantresponse to stress and disease.

SKDaZBP

Figure 6. Immunoblot analysis of 5-kD ZBP distribution within thecitrus tree. Lane 1, Molecular mass marker; lane 2, blight, 50 /j,gof protein extracted from trunk phloem tissue; lane 3, healthy, 50 /j,gof protein extracted from trunk phloem tissue; lane 4, blight, 50 /j.g ofprotein extracted from fully expanded leaves; lane 5, blight, 50 /ngof protein extracted from fibrous roots; lane 6, blight, 20 /u.g ofprotein from a 500-fold concentrate of xylem exudate.

SKDaZBP

Figure 7. Immunoblot analysis of 5-kD ZBP distribution in leaves ofblight-affected citrus. Lane 1, Molecular mass marker; lane 2, 15 yu,gof protein extracted from trunk phloem tissue; lane 3, 35 /ng ofprotein extracted from leaf midvein; lane 4, 35 jag of protein ex-tracted from fully expanded whole leaves; lane 5, 35 jag of proteinextracted from fully expanded leaves without midvein. www.plantphysiol.orgon May 27, 2018 - Published by Downloaded from

Copyright © 1996 American Society of Plant Biologists. All rights reserved.



Table 11. Comparison of phloem tissue fresh weight and dry weight in healthy and blight-affected citrus Bark Discs Phloem

Phloem/Bark Dry wtf fresh wt

Fresh wt Dry wt Fresh wt Dry wt fresh wt Dry wt/ Status

g/disc g/disc g/disc g/disc Healthy 2.487 aa 1.225 a 0.493 a 0.562 a 0.244 a 0.432 a 0.225 a

Predecl ine 2.565 ab 1.274 ab 0.496 a 0.625 ab 0.265 a 0.424 a 0.243 ab

Decline 2.850 b 1.395 b 0.489 a 0.833 b 0.334 a 0.400 b 0.291 b

a Values in the same column with different letters are significantly different by at least the 5% level (Student's ttest).

(0.05 1 ) b (0.027) (0.002) (0.080) (0.038) (0.006) (0.027)

(0.041) (0.027) (0.002) (0.034) (0.01 9) (0.007) (0.009)

(0.1 35) (0.072) (0.003) (0.074) (0.032) (0.003) (0.01 2)

Values in parentheses are SE.

Large quantities of vascular tissues can be separated easily for study of their constituents. Therefore, we were able to use immunoblot analyses to localize the ZBP on a tissue level. Our western blot analyses indicated that the 5-kD ZBP was specific to the vascular tissue and may be phloem specific. Zn accumulation occurs first in the phloem of blight-affected citrus. ZBP specificity to the phloem would correlate with apparent changes in phloem tissue suggested by the changes in fresh and dry weights and bark thickness associated with the occurrence of citrus blight. In addition, there was no immunological evidence of 5-kD ZBP in root xylem exudate, despite documentation of eventual accumulation of Zn in the outermost trunk wood tissue of late-stage blight-affected trees (Smith, 1974; Wutscher, 1981). Immunocytological studies are underway to localize these ZBPs with respect to phloem cell type.

Increases in tissue weight and thickening of bark associated with blight suggest that structural change of the phloem tissue accompanies the disorder, possibly affecting the integrity of the tissue. Microscopy data (Taylor and He, 1994) demonstrated that phloem structure was adversely affected in blight-affected trees. Our measurements indicated that the greatest change in the phloem tissue of blight-affected trees was a gain in fresh weight with pro- gression to the decline stage of the disorder. This may be an indication of a relative decrease in cell wall (structural) components or an increase in hygroscopic components such as starch. However, the iodine-potassium iodide test for starch indicated no such accumulation in that tissue. Previously, we demonstrated that phloem tissue of blight- affected citrus had almost twice as much protein per gram fresh weight as phloem tissue from healthy citrus (Taylor et al., 1989). Yet there was almost a 6-fold increase in ZBP accumulation in that tissue on a fresh weight basis. The protein fraction that increased in blight-affected citrus was correlated with an apparent increase in hydrophilicity of that tissue, accounting for the increased fresh weight in blight-affected relative to healthy tissues. These vascular changes are in addition to others noted in the xylem of blight-affected trees (Cohen, 1978; Vasconcellos and Castle, 1994).

Phloem fresh weight and dry weight increases together with accumulation of the 5-kD ZBP indicate that both structural and compositional changes accompany the occurrence of citrus blight. The accumulation of the ZBPs

apparently is specific to the phloem tissue. Clearly, cumu- lative data concerning vascular changes in citrus blight and in other systems (Sharma et al., 1984; Sharma and Sharma, 1987) implicate Zn sequestration as an important factor in their occurrence. Additionally, Zn is essential for a number of processes in plant metabolism, severa1 of which could have an impact on vascular development in plants. There- fore, it appears that Zn redistribution can cause a metabolic deficiency of Zn throughout the tree, with consequent alterations in tree metabolism. This system is unique in pro- viding a vehicle for the study of Zn involvement in plant metabolism. We are beginning to elucidate the specific effects of Zn on vascular development.

With this work we have sought to investigate Zn redistribution, the earliest physiological change in blight- affected citrus. A ZBP associated with this early symptom was isolated and characterized. Characterization of ZBP as a chitin-binding protein opens additional avenues of ex- ploration in the study of citrus blight. Furthermore, the unique finding that this protein is similar to chitin-binding proteins and that it tenaciously sequesters metals may have revealed a potential area for increased understanding of wound-inducible protein function in plants.

Received July 20, 1995; accepted October 11, 1995. Copyright Clearance Center: 0032-0889/96/110/0657/08.

LITERATURE ClTED

Albrigo LG, Syvertsen JP, Young RH (1986) Stress symptoms of citrus trees in successive stages of decline due to blight. J Am Soc Hortic Sci 111: 465475

Albrigo LG, Young RH (1981) Phloem zinc accumulation in citrus trees affected with blight. Hortscience 16: 158-160

Aspinall D, Paleg LG (1981) Proline accumulation: physiological aspects In LG Paleg, D Aspinall, eds, The Physiology and Bio- chemistry of Drought Resistance in Plants. Academic Press, New York, pp 206-242

Bausher MG (1979) Changes in ATI' levels and carbonic anhydrase in the presence of citrus blight (abstract No. 640). Plant Physiol 63: S-115

Bausher MG (1982) Changes in IAA-oxidase activity and the severity of citrus blight. Hortscience 17: 497

Brlansky RH, Timmer LW, Lee RF, Graham JH (1984) Relation- ship of xylem plugging to reduced water uptake and symptom development in citrus trees with blight and blight like declines. Phytopathology 74: 1325-1328

Broekaert WF, Marien W, Terras FRG, DeBolle MFC, Proost P, Van Damme J, Dillen L, Claeys M, Rees SB, Vanderleyden J,



664 Taylor e t al. Plant Physiol. Vol. 11 O, 1996

Cammue BPA (1992) Antimicrobial peptides from Amaranthrrs caudatus seeds with sequence homology to the cysteine/glycine- rich domain of chitin-binding proteins. Biochemistry 31: 4308- 4314

Broglie KE, Gaynor JJ, Broglie RM (1986) Ethylene-regulated gene expression: molecular cloning of the genes encoding an endochitinase from Phaseolus vulgaris. Proc Natl Acad Sci USA

Childs JFL (1953) Observations on citrus blight. Proc Fla State Hortic SOC 66: 33-37

Cohen M (1978) Growth and longevity of trees with citrus blight in St. Lucie County. Proc Fla State Hortic SOC 91: 66-68

Derrick KS, Lee RF, Brlansky RH, Timmer LW, Hewitt BG, Barthe GA (1990) Proteins associated with citrus blight. Plant Dis 7 4 168

Fullmer CS (1984) Identification of cysteine-containing peptides in protein digests by high performance liquid chromatography. Anal Biochem 142 336-339

Hanks RW, Feldman AW (1974) Use of arginine and total amino acids to determine site of stress in citrus trees with young tree decline. In LG Weathers, M Cohen, eds, Proceedings of the 6th Conference of the International Organization of Citrus Virolo- gists, 1972, Swaziland. University of California, Riverside, Divi- sion of Agricultura1 Sciences, pp 184-190

Henikoff S, Henikoff JG (1991) Automated assembly of protein blocks for database searching. Nucleic Acids Res 19: 6565-6572

Keilin D, Mann T (1940) Carbonic anhydrase. Purification and nature of the enzyme. Biochem J 34: 1163-1177

King SM, Otter T, Witman GB (1985) Characterization of monoclo- na1 antibodies against Chlamydomonas flagellar dyneins by high resolution protein blotting. Proc Natl Acad Sci USA 8 2 4717

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227: 680-685

Levy Y (1980) Field determination of free proline accumulation and water-stress in lemon trees. Hortscience 1 5 302-303

Lindskog S (1983) Carbonic anhydrase In TG Spiro, ed, Zinc Enzymes. John Wiley and Sons, New York, pp 77-121

Molnar JM, Parups EV (1977) A histochemical study of starch, lipids, and certain enzymes in senescing rose stems. Can J Bot

Morrissey JH (1981) Silver stain for proteins in polyacrylamide gels: a modified procedure with enhanced uniform sensitivity. Anal Biochem 117 307-310

Raikhel NV, Broekaert WF (1993) Chitin binding proteins, their biology and possible function. In DPS Verma, ed, Control of Plant Gene Expression. CRC Press, Boca Raton, FL, pp 407423

Randall PJ, Bouma D (1973) Zinc deficiency, carbonic anhydrase, and photosynthesis in leaves of spinach. Plant Physiol52: 229-232

Robinson NJ, Gupta A, Fordham-Skelton AP, Croy RRD, Whit- ton BA, Huckle JW (1990) Prokaryotic metallothionein gene characterization and expression: chromosome crawling by liga- tion-mediated PCR. Proc R SOC Lond Biol Sci 242 241-248

83: 6820-6824

55: 617-624

Sharma CP, Mehrotra SC, Sharma PN, Bisht SS (:1984) Water stress induced by Zn deficiency in cabbage. Curr Sci 53: 4446

Sharma CP, Sharma PN (1987) Mineral nutrient deficiencies affect plant water relations. J Plant Nutr 1 0 1637-1641

Sharon N, Lis H (1989) Lectins. Chapman and Hall, London, pp

Smith PF (1974) History of citrus blight in Florida. Citrus Ind

Steffins JC (1990) The heavy metal binding peptides of plants. Annu Rev Plant Physiol Mo1 Biol 41: 553-575

Syvertsen JP, Albrigo LG (1984) Leaf water stress and water uptake of trunk wood in blight affected citrus trees In HWS Montenegro, CS Moreira, eds, Proceedings of the VI Interna- tional Citrus Congress, Sao Paulo, Brazil, Vol 2. International Society of Citriculture, pp 309-310

Taylor KC, Albrigo LG, Chase CD (1988) Zinc complexation in the phloem of blight affected citrus. J Am Soc Hortic Sci 113: 407-411

Taylor KC, Albrigo LG, Chase CD (1989) Characterization of a Zn binding peptide associated with a decline disorder in citrus. In D Winge, D Hamer, eds, UCLA Symposium on Metal Ion Ho- meostasis: Molecular Biology and Chemistry. Alan R Liss, New York, pp 385-394

Taylor KC, He X (1994) Evidence for structural and compositional change in phloem of blight-affected citrus. Hortscience 29: 760

Towbin H, Staehlin T, Gordon T (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets. Proc Natl Acad Sci USA 76: 43504354

Vallee BL (1983) Zinc biology and biochemistry In TG Spiro, ed, Zinc Enzymes. John Wiley and Sons, New York, pp 1-24

Van Parijs J, Broekaert WF, Goldstein IJ, Peumans WJ (1991) Hevein: an antifungal protein from rubber-tree (Hevea brasilien- sis) latex. Planta 183: 258-264

Vasconcellos LABC, Castle WS (1994) Trunk xylem anatomy of mature healthy and blighted grapefruit trees on severa1 rootstocks. J Am SOC Hortic Sci 119: 185-194

Williams GL, Albrigo LG (1984) Some inorganic eleinent changes in trunk phloem of healthy and blight affected citrus trees. J Am SOC Hortic Sci 109: 437-440

Wutscher HK (1981) Seasonal changes in zinc and water soluble phenolics in the outer trunk wood of healthy and blight affected sweet orange trees. Hortscience 16: 157-158

Wutscher HK, De1 Valle N, de Bernard A (1983a) Citrus blight and wood pH in Cuba and Florida. Hortscience 18: 486488

Wutscher HK, Youtsey CO, Smith PF, Cohen M (198310) Negative results in citrus blight transmission tests. Proc Fla State Hortic

Young RH, Albrigo LG, Tucker DPH, Williams G (1980) Inci- dente of citrus blight on carrizo and citrange and some other rootstocks. Proc Fla State Hortic Soc 93: 14-17

Young RH, Garnsey SM (1977) Water uptake patterns in blighted citrus trees. J Am SOC Hortic Sci 102: 751-756

50-61

55(9): 13-14, 16, 18-19; (10): 9-10, 13, 14; (11): 12-13

SOC 96: 48-53



Purification of a Zn-Binding Phloem Protein with Sequence ... · Purification of a Zn-Binding...

Documents

Transcript of Purification of a Zn-Binding Phloem Protein with Sequence ... · Purification of a Zn-Binding...