{2... {In}?! D? {Evil}. ($.37. J. .. (\L ) .< :1. r .v\- . . I i «I: .313: nu. .5253}... 33.3.5. $33.53. 1 53.3.1. \ {a (4‘: .{Lu 6.. Ii»: :3 .4 1... (14.23:. 2.7.2.. 3 .. 77. 3... .1 (set: x .3. 4 AI 41...). ..I \171 t .l: a .p. .3. .7.‘ 1...! ((1,. 1/.<.vrr..<:.\ . .1 x. (.v 4. 7.! 4. 1.23;: I.» x (a. t\ .7 E: 1;. .;r F . \c... . if..iJ.-.-.<.a>.r r :3 «1‘. 3:3,. ‘ . 2.) $5.91,. . (.4, Q 4.. x {is 42.5.... 3...: it! )1... (i. .. (. r . .57.. .7. : .7 .42. a: 1..il'4..1...u. 3.). . 5. {I {as . .. .1... r: E 1,? $2 1 . i... .1; 2.. .4 .r...s.sx§ {,9 1.. 3..i.5..<..§.fi Inii‘l’l.tul.l.(.“\ 2- ~ 5.. (5.1.1.1); 217).?! I); ti? 1.0. 1.1.1:: .523}. i. 25.3: I... .25‘ ‘ £52, 5......3... 7.. .1: tr >1: s .3 3. .. ..x «c... 4...)... {six .23.... 3.5:... 2:43.. -32 < ..I.e...<\.\........l:. \ 3.... 4. It ;»r\1o ¢Ib‘K.v . .\ .3 . .Lr. .!:\x .2119 .A :53 (t. ...\I.\ .. v.31 \ .. i. . \(1 x. < nvr II 7.3.34? {(5}. ...(...t,1\..!.( 5).} 5 «u «1.x: §s(..(n :( .2»... .Ii .5 €.:r...7....(:1<:t. ( e . 4.27! C {‘4' 127.1,... .4 1.. 37..) . \. (:Y.(\ .. \r. I c. \x 35.1.} Lr ( . I; 1.1.1 (1.}! .33. .v ()2. .1 r < tail... . {12: ._ Line L<<$..r. .15.! r(c....\c< EC. mEom 05:8 $0 $5 new diam How 62 ASE «.13 com .Ef ~88an $02 S_==oombxm ow 8:: S_==ooE§m a. 8:: =9» :00 cosmos 0528 beam 0528 beam oBEom\oB£8£ 32328 19.5804 anooomaflom ”258on :fiom_mwo:38< Emcotnm .mfioaoaonobmu gauczohgofif mo magma ooh; .N 2nt 5 well as hydroxyproline, serine, valine, tyrosine (Lamport, 1969), and sometimes histidine (Stuart & Varner, 1980)(Table 3). Extensin is a structural protein. Structural proteins frequently exhibit short term amino acid sequence periodicities (North, 1968) which, for the recently isolated soluble monomers of extensin, is reflected in their peptide profiles: they are simple and dominated by a few major repeating peptides (Figure 1). Although the insolubility of wall-bound extensin initially precluded obtaining a truly representative peptide map, enzymic digestion of tomato cell walls released only 5 peptides, accounting for 1/3 of the wall hydroxyproline (Lamport, 1969; 1977); thus those peptides occur repeatedly in extensin. Further inspection of the peptide sequences (Table 4) reveals a "sub-periodicity" within the already periodic peptides as Ser-Hyp-Hyp-Hyp-Hyp occurs at least once in each peptide. P1 004-. 5 20 0.03— ('3 rs N < 0.02- 19 1.: 16‘ 1 9 1‘ ‘3/ W 67 10\12|1s a 0.00“ 40‘ 45. so 55 so 65 7o 75 so as 90 Minutes Figure 1. Tryptic peptide map of deglycosylated tomato extensin precursor P1 prior to insolubilization into the cell wall (Smith et al., 1986). #0me £0 “0 586 80E _ wwmfi £0 00 0082 Boom . 300 00520000: £08004 0% Em30Nm=0UH . $3 £085» 0% 50: 00> 80E 2 $3 :3 0 £0004 805 v 90 208 8 00mm0axm . oz oz oz 8 oz oz 8 EH 8 8 8 8 8 8 8 ma. 3 8 8 8 8 S 0.: WE 8m 8. 0.8 on 8 0i 8 3 8 8 8 8 3 8 8 2: q: 8 8o 8 8 8 q: E. 8 3 8 8 8 2 8 =3 8 3 8 8 8 E 8 a: 8 8 8 8 8 8 8 a: 8 3 8 8 8 8 8 a0 8 8 8 8 8 8 8 a> 8 8 S 8 8. S 8 m2 8 S 8 3 0.0 8 8 :0 8 8 8 8 8 8 8 so 8 2 8 E E Z 8 :0 8o 8 8 8 8o 3. 8.: am 3 8 8 8 8 8 8 EH 8 E 8 8 8 8 8 8. 0.: 8m 84 0:. 8.9. 8m 84 no: .8 .E .0833 .9900 .802 .802 .650 23.. 03805 05505 053 3080002 ommo0§m 00 82:89:00 20¢. 0303 .m 030% Table 4. Amino Acid Sequences of Tomato Cell Wall Tryptic Peptides‘ Ser-Hyp-Hyp-Hyp-Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-("U"-Tyr)-Lys" Ser-Hyp-Hyp-Hyp-Hyp-Ser-Hyp-Lys Ser-Hyp-Hyp-Hyp-Hyp-Thr-Hyp-Val-Tyr-Lysc Ser-Hyp-Hyp-Hyp-Hyp-Lys Ser-Hyp-Hyp-I-Iyp-Hyp-Val-("U"-Lys)-Lysd 9:599)!“ From Lamport, 1973 " "U" is Lamport's "unknown tyrosine derivative", isodityrosine (Fry, 1982) that occurs as an intramolecular crosslink (Epstein and Lamport, 1984). This peptide defines Type 3 extensins (Showalter & Rumeau, 1989) ° This peptide occurs as a major repeating peptide in tomato extensin precursor P1 (Smith et al., 1986). This peptide occurs as a major repeating peptide in tomato extensin precursor P2 (Smith et al., 1986). Extensin is a glycoprotein having oligoarabinosides O-linked to hydroxyproline (Lamport, 1967), and a-O-linked galactosyl-serine (Allen et al., 1978; Lamport et al., 1973). In total, carbohydrate accounts for 40-65% of extensin's weight. The demonstration that hydroxyproline is O-glycosylated with arabinose oligosaccharides (1-5 residues)(Lamport & Miller, 1971; Mazau & Esquerre-Tugaye, 1986) is unique to plants, as hydroxyproline is not glycosylated in animal systems. The configuration of the tetra-arabinoside is: a-L-Araf(1-3)-B— L-Araf(1-2)-B- L-Araf(1-2)-B-L-Araf(1-4)-Hyp (Akiyama et al., 1980), while the configuration of the other arabinosides is unknown (Figure 2). Figure 2. Hydroxyproline tetra-arabinoside The general conformation of glycosylated extensin is a polyproline-II helix. Circular dichroic spectra, first of tomato extensin peptides, and later of carrot extensin monomers, is consistent with poly-L-hydroxyproline and poly-L-proline (a minimum at 205-206 nm and a maximum at 225-228 nm), which are both in a polyproline-II helix (3 residues /left-handed turn and 9.4 A pitch) (Lamport, 1977; van Holst & Varner, 1984). Extensin contains the unique amino acid isodityrosine (IDT) as an intramolecular crosslink. Lamport initially identified "an unusual modified tyrosine residue" in two extensin peptides isolated from tomato cell walls (Table 4)(Lamport, 1973). This observation eventually led to two elegant papers describing the chemical identity, localization and function of IDT. First, Stephen Fry demonstrated that cell wall hydrolysates contained a new amino acid, the diphenyl ether isodityrosine (Fig 3), and he speculated that it occurred in extensin as Lamport's "unknown" (Fry, 1982). Then, Epstein and Lamport (1984) showed that 9 the earlier extensin peptide "unknown" was indeed IDT (Table 4), and, surprisingly, that it served as an intramolecular crosslink rather than as an intermolecular crosslink as supposed. OH 0 3H2 qHNH2 3H2 COOH qHNH2 COOH Figure 3. Isodityrosine (IDT) Here the characterization of extensin may have stopped due to its unyielding insolubility, were it not for Maarten Chrispeels' interest in protein secretion. Employing [1‘C] proline pulse-chase experiments, Chrispeels dubiously demonstrated that precursors of wall-bound extensin occur in the wall first as a salt soluble pool before insolubilization (Brysk & Chrispeels, 1972). Lamport initially dismissed Chrispeel's observation for three reasons: first, the pulse-chase kinetics, although suggestive, did not conclusively demonstrate turnover in the ionically-bound putative precursor; secondly, the amino acid and carbohydrate compositions of the putative precursor differed considerably from those of extensin wall peptides; thirdly, David 10 Pope, working with Lamport and (unfortunately) sycamore cell suspensions‘ could not repeat Chrispeels' experiments (Pope, 1977). However, three years later Joseph Varner entered the cell wall protein field by purifying and partially characterizing the first extensin monomeric wall precursor (from carrot; Stuart & Varner, 1980). In a series of papers his laboratory not only corroborated much of Lamport's earlier wall peptide data but also made significant new contributions, most notably the cloning of the first extensin cDNA (a partial sequence from carrot)(Chen & Varner, 1985a) and genomic clone (carrot)(Chen & Varner, 1985b), thus providing information about extensin that is virtually impossible to obtain by protein sequencing: the complete sequence of the protein backbone (sans hydroxylation and IDT crosslinks). Furthermore, the clones proved that extensin is indeed a genuine protein. Some of Varner's other original contributions were the first visualization by transmission electron microscopy of extensin monomers as flexible 80 nm rods (van Holst & Varner, 1984), the demonstration that extensin mRNA levels rise in response to wounding (Chen & Varner, 1985a), and the observation that extensin monomer is insolubilized into walls concomitant with IDT formation (Cooper & Varner, 1984). Furthermore, Cassab et al. (1985) isolated the first putative extensin monomer from soybean seed coat and demonstrated via immunolocalization, its developmental and tissue-specific expression in the pallisade epidermal cells and hour glass cells of the soybean seed coat; and they developed a new method, tissue printing on nitrocellulose, for the ‘ Frequently, the choice of a system can decide an experiment, as it did here. Sycamore was a poor system, as the cells have a very small pool of extensin precursors that is difficult to extract (Heckman et al., 1987). 11 general immunological localization of soluble extensin in tissues (Cassab & Varner, 1987) Varner's purification of a salt-soluble extensin monomer was a significant advance that prompted Lamport to reinvestigate his method for salt-elution of intact cell suspensions (Lamport, 1965), but this time using an optimal system, tomato suspensions, and (initially) an undergraduate lab aide‘. In 1984 and 1986, Smith et al. rigorously demonstrated the precursor status of salt-elutable extensin monomers: 3H-proline pulse-chase data clearly indicated turnover, restoration kinetics indicated precursor status, and the chemical compositions (Smith et al., 1984) and, most convincingly, amino acid sequences of major peptides (Smith et al., 1986) coincided with those isolated earlier from the insoluble tomato cell walls (Table 4)(Lamport, 1973). This was also the first demonstration that multiple extensins exist in a single plant system, implying a multigene family and subtle differences in function from one extensin to another. Thus Smith et al. (1984, 1986) suggested that "types" of extensin occur which are defined by their major repetitive sequence motifs. Since Chen & Varner's cloning of the first extensin, numerous extensins from diverse species and tissues have been cloned, providing more details about the primary structure of the extensin polypeptides, the regulation of their expression, and therefore clues about their putative role(s) in the wall. ’ Combining extensin peptide and clone sequences, Showalter & Rumeau (1989) recently proposed two structural types of extensin. The types are defined, 5 Nathan Krupp, an undergraduate employee, did the initial precursor work. 12 as originally suggested by Smith et al. (1984,1986), by their major repeating peptides. One extensin type is predominantly composed of: Ser-Pro-Pro-Pro-Pro—Ser-Pro-Ser-Pro-Pro-Pro-Pro-Tyr-Tyr-Tyr-Lys6 repeats (called "P3-type" extensin by Smith et al., 1984, 1986); and the other type is composed of Ser-Pro-Pro-Pro-Pro-Thr-Pro-Val-Tyr-Lys repeats (called "Pl-type" extensins by Smith et al., 1984, 1986). Furthermore, extensins are organized into three domains distinguishable by their sequence motifs, as well as by their positions in the polypeptide: the C-terminus, the N-terminus, and the central polypeptide which is made up of the major repeating motifs (Showalter & Rumeau, 1989). B. Regulation of Extensin Expression Extensins and their mRNAs accumulate in tissues undergoing various stresses, such as: cell culture (Ecker & Davis, 1987; Lamport, 1965), infection (Esquerre- Tugaye & Mazau, 1974; Esquerre-Tugaye & Lamport, 1979; Showalter et al., 1985), wounding (Chen & Varner, 1985a; Chrispeels et al., 1974; ), heat shock (Stermer & Hammerschmidt, 1987); and they accumulate in response to elicitors (Roby et al., 1985, Showalter et al.,1985, Tierney et al.,1988), ethylene (Ridge & Osborne, 1970; Ecker & Davis, 1987), red light (Pike et al.,1979), gravity (Prassad & Cline, 1987), glutathione (Wingate et al., 1988) and development (Cassab & Varner, 1985; Hong et al., 1987; Franssen et al., 1987; Keller & Lamb, 1989). Apparently, many of these responses are differentially regulated at a transcriptional level, however we lack the precise details about the regulation of extensin genes (Lawton & Lamb, ‘ These sequences are derived from clones; actually, the Pro residues are Hyp as demonstrated by direct peptide sequencing (see Table 3). 13 1987; Wingate et al., 1988; Keller & Lamb, 1989). C. The Function of Extensin, and a Current Cell Wall Model While structural in a general sense, the exact functions of extensin are unknown despite numerous clues pointing to their involvement in wall architecture, disease resistance, and development. The location of extensin in the extracellular matrix, its relative abundance in the wall, its insolubility, periodicity, hydroxyproline content’, and apparent lack of enzymic activity led to a suggested role for extensin in cell wall architecture, and as an extension growth regulator via protein- carbohydrate and/or protein-protein crosslinks (Lamport, 1970). Several groups have demonstrated a positive correlation between HRGP levels and the cessation of cell elongation (Cleland & Karlsnes, 1967; Sadava et al., 1973; Monro et al., 1974; Klis & Eeltink, 1979). A role for extensin in plant defense (first suggested by Esquerre-Tugaye & Mazau, 1974) is indicated by the differential accumulation of sets of extensins and their mRNAs in response to infection, elicitors, and wounding (Showalter et al., 1985; Corbin et al., 1987; Wingate et al., 1988). Extensins are also known to agglutinate bacteria, probably due to their positive charge (Leach et al.,1982a), and may enhance the cell wall barrier and/or act as nucleation sites for lignin deposition (Whitmore, 1978). Extensins may be involved in differentiation and tissue development. The distribution of extensin in different cells of the seed coat changes during seed 7 Hydroxyproline is a rare amino acid that usually only occurs in proteins of the extracellular matrix. 14 development, and extensin ultimately accumulates in the walls of the epidermal palisade and hourglass cells (Cassab & Varner, 1987). Furthermore, Hong et al. (1987) noted the differential expression of an extensin in the mature section of soybean hypocotyl tissue; and Keller & Lamb (1989) demonstrated the specific expression of a cell wall HRGP, probably a class of extensin, that is transiently induced in a small set of cells involved in initiation of lateral roots, and as such, may have a specialized structural function. The periodic placement of active groups along the extensin peptide backbone raises possibilities of ionic and covalent interactions of extensin with extensin and other wall components, thus providing plausible mechanisms whereby extensin might fulfill its roles in wall architecture, disease resistance and differentiation. For example, the lysine 6 -amino groups probably ionically pair with the pectic uronic acids, or they may form Schiff bases with the reducing ends of wall polysaccharides (but very few). Histidine imidazole rings, when protonated, may also ionically pair to negatively charged wall components, or form covalent crosslinkages, as is suggested to occur in collagen (Fujimoto et al., 1982). Tyrosine residues and ferulate could oxidatively couple pectic polysaccharides (attached to ferulate by an ester linkage) to extensin (Neukom & Markwalder, 1978); and Whitmore (1978) hypothesized the in viva formation of covalent bonds between coniferyl alcohol and the free hydroxyl of unsubstituted hydroxyproline. Fry's 1982 discovery of the diphenyl ether crosslinked amino acid IDT in cell wall hydrolysates, and Epstein & Lamport's demonstration of IDT in extensin 15 (1984) provided a possible mechanism for extensin's insolubilization in the wall via peroxidase catalyzed IDT formation between two extensin molecules, making an intermolecular crosslink (Lamport, 1970; Fry, 1982). Thus Cooper et al., (1984) and Lamport & Epstein (1983) proposed cell wall models wherein an extensin network is independently crosslinked by IDT, around cellulose microfibrils. Lamport & Epstein (1983) elaborated on the concept by likening the wall to a woven fabric; cellulose (anticlinal) constituting the fabric "warp" interpenetrating a transmural extensin (periclinal) "weft". This "warp-weft" model suggested that extensin couples the major load-bearing cellulose polymers into a defined network by IDT intermolecular crosslinks. So far, the evidence for this model is indirect: 1. Wall-bound extensin remains intact and insoluble after treatment with anhydrous HF, a reagent which specifically cleaves glycosidic linkages, leaving peptide bonds intact (Mort & Lamport, 1977). 2. IDT, extensin's unique crosslink amino acid, was discovered in cell wall hydrolysates (Fry, 1982) and extensin wall tryptides (Epstein & Lamport, 1984). 3. The demonstration of highly soluble salt-elutable extensin precursors to wall-bound extensin which contain little or no lIDT (Cooper & Varner, 1984; Smith et al., 1984, 1986). 4. Acid sodium chlorite oxidation of cell walls destroys phenolic rings, releasing fragments of wall-bound extensin (Mort, 1978; O'Neil & Selvendran, 1980) 16 5. Wall IDT content suggests a crosslink frequency sufficient to create pores large enough to enclose a cellulose microfibril (Lamport, 1986). Thus, highly soluble extensin precursors containing virtually no IDT become insolubilized in the wall concomitant with the appearance of cell-wall IDT (Cooper & Varner, 1984). The possibility that extensin insolubility is due to covalent crosslinkage with wall polysaccharides is unlikely because extensin remains insoluble after all wall sugars are removed by anhydrous HF solvolysis. Finally, insoluble wall-bound extensin becomes soluble only after phenolic rings, such as those of IDT (and tyrosine), are destroyed by acid chlorite. It is requisite that a cell wall model be generally applicable, at least to the primary cell walls of higher plants. Thus the question arises that begins to test the general applicability of the warp-weft model: Is extensin ubiquitous to higher plants? D. Extensin in the Graminaceous Monocots So far, the information known about extensin has been gathered exclusively from dicot systems with walls rich in hydroxyproline. The role of extensin (or of any HRGP) in the monocots has been largely ignored, probably because the walls of monocots, at least those of the graminaceous monocots, are hydroxyproline-poor (Burke et al., 1974; Darvill et al., 1980a; Lamport, 1965; Showalter & Varner, 1989). Graminaceous walls are, in some respects, markedly different from, and yet in other respects, similar to, those of dicots. Both contain cellulose, pectins, hemicelluloses, and hydroxyproline-containing proteins; however, the quantities of 17 some of these components differ dramatically between the two groups (Figure 4)(Darvill et al., 1980a; McNeil et al., 1984; Fry, 1985). One major difference is in the acidic wall polysaccharides. The dicot wall is rich in the polyanionic gel-forming rhamnopolygalacturonans, or pectins, while graminaceous cell walls contain very little pectin (Talmadge et al., 1973; Burke et al., 1974). Instead, they contain the acidic polysaccharide glucuronoarabinoxylan, technically a hemicellulose, which however may have the same function in graminaceous walls as pectin does in the dicot wall (Darvill et al., 1980b; Carpita, 1983, 1984; Carpita et al., 1985; Fry, 1985). The graminaceous wall is also rich in "mixed-linkage" glucans (6-3 and B- 4-linked), but xyloglucan-poor (1-5%), while the dicot wall contains little "mixed- linkage" glucans, but is 20-30% xyloglucan (McNeil et al., 1984; Fry, 1985). Another major difference is in the protein component of the wall (Lamport, 1965; Burke et al., 1974; Showalter & Varner, 1989). The dicot wall generally contains 1020 times more hydroxyproline than the graminaceous wall“. However, evidence existed that the seed coat and pericarp of some monocots contained proteins possessing hydroxyproline (Van Etten et al., 1963), hydroxyproline arabinosides (Lamport & Miller, 1971) and having vaguely extensin-like compositions (i.e. 11% Hyp, 10% Thr; 6% Lys)(Boundy et al.,1967). Also in contrast to dicot systems, hydroxyproline levels in the graminaceous monocots are apparently not related to disease resistence (Clarke et al., 1981; Mazau & Esquerre-Tugaye, 1986). Thus, low levels of Hyp in ‘ Some dicot walls are Hyp-poor: for example, rose (Lamport, 1965) and sugar beet (Li et al., 1989). However, dicot tissues are generally much more Hyp-rich than graminaceous tissues. 18 graminaceous walls, and its apparent non-involvement in disease resistance suggest, at best, graminaceous extensins which are probably structurally and functionally different from dicot extensins. This led me to ask the following questions: 1. Does extensin occur in the cell walls of the graminaceous monocots, specifically Zea mays? 2. Does the warp-weft cell wall model extend to the primary cell walls of the graminaceous monocots, specifically Zea mays? 01 O l Percent of Wall (dry weight) M 01 Hemicellulose Figure 4. Dicot and graminaceous wall components compared. MATERIALS AND METHODS 1. Methods for the Preparation of Cell Walls and Cell Wall Proteins A. Cell Suspension Cultures I grew maize cell suspensions (variety Black Mexican, a gift from Dr. Tom Hodges, Purdue University), in 1 L flasks containing 50 mL Murashige and Skoog medium (Murashige and Skoog, 1962) (2 mg/L, 2,4-D). They were shaken at 120 rpm on a gyrotary shaker at 27° C under 900 lux of constant fluorescent lighting, and subcultured every 11 days to an initial packed cell volume of 3%. B. Intact Cell Elution I prepared batches of crude HRGP from 11 day cultures (500 mL/ 1 L flask; 17 flasks /batch) harvested on a 2 L coarsely sintered funnel followed by a water wash, then gentle agitation of the cell pad (about 600 g fresh weight) in 1 L of 100 mM AlCl3 (a nonplasmolyzing concentration) for 3 min before final suction. The eluate was reduced in volume to 100 mL at 32° C. After adding TCA (final concentration 10% w/v) to the concentrated eluate (18 hr, 4° C), centrifugation (13,000 g, 45 min, SS-34 rotor head) yielded a hydroxyproline-poor pellet (0.05% Hyp dw, discarded) and hydroxyproline-rich supernatant, which was dialyzed 72 hr at 4° C and then freeze-dried. I designate this TCA-soluble fraction 'crude HRGP'. 19 20 C. Cell Wall Preparation Seeds of Sweet Sue sweet corn9 (Harris Seeds) germinated for 4 days on germination paper (Anchor Paper) soaked in tap H20. Root tips, roots, coleoptiles with their primary leaf (henceforth I refer to this preparation as coleoptile), and whole cells of Black Mexican cell suspensions were separately frozen in N2 (lq), ground to fine powder with a mortar and pestle, suspended in 1 M NaCl and sonicated for 10 min. The pelleted walls were washed with 5% w/v SDS followed by acetone washes (and pelleting) to remove SDS. Finally the walls were rinsed repeatedly with distilled H,O until the walls were free of salt and cell debri, judging by microscopic examination, then the walls were freeze-dried. D. Phosphocellulose Cation Exchange Chromatography I dissolved crude HRGP (10 mg/mL) in 12 mM (pH 3.0) McIlvaine buffer (McIlvaine, 1921), and applied a maximum of 170 mg to a Bio-Rad Cellex-P phospho-cellulose column (15 mm i.d. x 200 mm) equilibrated with 12 mM McIlvaine buffer (pH 3.0). I eluted with a 3.0 to 6.8 pH gradient (100 mL of Mchaine buffer) followed by a 200 mL 0 to 1 M NaCl gradient (200 mL) in 12 mM (pH 6.8) McIlvaine buffer at a flow rate of 19 mL/hr, monitoring the absorbance at 220 nm. Collected peaks were dialyzed 2 days against distilled H,O and freeze-dried. I determined Hyp across Cellex-P peak 2 (CP2) and Cellex-P peak 3 (CP3) by collecting, concentrating and desalting 4 mL fractions in Amicon 9 Sweet Sue was chosen because the seeds happened to be gardening "leftovers" and because they were NOT Black Mexican seeds. 21 centricon“ microconcentrators. Aliquots were blown down to dryness under N2, hydrolyzed in 6 N HCl for 18 hr at 110° C, and then assayed for Hyp. E. BioRex-70 Cation Exchange Chromatography I dissolved freeze-dried Cellex-P Peak 2 in 2 mL 30 mM sodium phosphate buffer (pH 7.4) and applied a maximum of 20 mg to a Biorex-70 (100-200 mesh) column (8 mm i.d. x 100 mm) equilibrated with 30 mM sodium phosphate buffer (pH 7.4), and eluted with a 200 mL buffered 0 to 1 M NaCl gradient at a flow rate of 19 mL/hr, monitoring the absorbancy at 220 nm. F. Superose-6 FPLC Gel Filtration I injected 0.1-1 mg of dHHRGP or dTHRGP, or semi-purified HRGP in 250 ,uL 200 mM (pH 7.0), 0.02% azide-sodium phosphate buffer onto a Pharmacia Superose-6 FPLC gel filtration column, and eluted at a flow rate of 14 mL/hr, monitoring the absorbancy at 220 mm. H. Methods for the Chemical Characterization of Cell Walls and Cell Wall Proteins A. Neutral Sugar Analysis I analyzed sugars as their alditol acetates (Albersheim et al., 1967) on a Perkin-Elmer 910 Gas Chromatograph using a 6 foot x 2 mm i.d. PEG8224 column programmed from 130° to 180° at 4°/min. for neutral sugars and a 6 foot x 2 mm i.d. OV 275 column programmed from 130° to 230° at 2°/min for amino sugars, using an SP4100 computing integrator for data capture. 22 B. Anhydrous HF Deglycosylation I deglycosylated 13 to 43 mg maize cell wall preparations (dw), or 3 to 13 mg maize HRGP, in a micro-apparatus containing 1 mL anhydrous I-IF/ 20 mg cell wall, and 10% (v/v) anhydrous methanol for 1 hr at 0° C (Sanger and Lamport, 1983). The reaction was quenched by pouring into stirred distilled H20 at 2° C to a final concentration of 10% (v/v) HF, and then dialyzed for 48 hr at 4° C and freeze-dried. The HF-treated cell wall material was then resuspended in distilled H20 and microcentrifuged to separate particulate (the HF-insoluble cell wall) from remaining HF-soluble (and H20 soluble) large molecular weight wall components. Both the HF-insoluble wall pellet and the HF-soluble wall supernatant were then freeze-dried separately. C. Hydroxyproline Assay After hydrolysis (6 N HCl, 110° C, 18 h) of wall preparations, and of the HF - soluble’0 and HF-insoluble wall, I assayed Hyp content by Kivirikko's method (Kivirikko & Liesma, 1959) which involves alkaline hypobromite oxidation, subsequent coupling with acidic Ehrlich's reagent and monitoring at 560 mm. D. Hydroxyproline Arabinoside Profile After alkaline hydrolysis (0.44 N Ba(OH)2 18 hr, 105° C) of maize cell walls or HRGP, careful neutralization with concentrated H2804, followed by centrifugation and freeze-drying of the supernatant fraction, I determined ‘° Henceforth, HF-soluble wall refers to the I-IF-soluble non-dialyzable large molecular weight wall components, and the HF-insoluble wall refers to the HF and H,O-insoluble particulate that remains after dialysis. 23 hydroxyproline arabinosides (Lamport, 1967) by redissolving the lyophilate in distilled H20 and applying 0.5 mL containing 100 to 200 ug hydroxyproline to a 75 x 0.6 cm column (H+ form) of Technicon Chromobeads C washed with water, eluting with a 0 to 0.5 N HCl gradient, and monitoring the post-column hydroxyproline reaction at 560 nm. E. Amino Acid Analysis We used a Pickering High Speed Na+ cation exchange column (3 mm i.d. x 150 mm) in series with a BX-8 cation exchange column (3.7 mm i.d. x 70 mm, Benson Co.) eluted by Pickering Buffers A (titrated to pH 3.1 with 6 N HCl), B and Sodium Regenerant. Post column fluorometric detection involved NaOCl oxidation and OPA coupling allowing Hyp and Pro detection (Yokotsuka & Kushida, 1983). I replaced B-mercaptoethanol (reductant) with 22.7 mM N,N- dimethyl-B-mercaptoethylamine HCl (Frister et al.,1988), and data capture was by a Compaq 386 with Nelson Turbochrom software. F. Cell Wall Isodityrosine Estimation I estimated maize cell wall (Black Mexican) IDT, after hydrolysis of the HF - insoluble wall in constant boiling HCl at 110° C for 20 hr, by reversed-phase on Hamilton PRP-1 column. Solvent A was 0.13% HFBA, and Solvent B was 0.13% HFBA in 80% CH,CN(aq). The programmed gradient elution was: 0-30% Solvent B for 15 min, a 5 min hold at 30% Solvent B, then returned to 100% A in 5 min. Flow rate was 0.3 mL/min. Standards were 2 ug each of tyrosine, dityrosine, and isodityrosine chromatographed with and without the HF-insoluble maize wall 24 hydrolysate. Absorbance was monitored at 214 nm, and data recorded via IBM 9001 and CAPS. Sugar degradation products [interfered with the assay therefore deglycosylation of the wall before IDT estimation was imperative. G. SDS Gel Electrophoresis This method is based on that of Laemmli & Favre (1973). I loaded 20 to 40 rag glycosylated THRGP or I-IHRGPs, and 10 to 15 ,ug deglycosylated THRGP (dTHRGP) or deglycosylated HHRGP (dHHRGP) in 10 to 25 ,uL sample buffer (Trizma base, 0.01 M; SDS, 1%; EDTA, 0.001 M; B-mercaptoethanol, 5%) onto a 15% polyacrylamide gel and ran the gels in pH 8.3 Tris-Gly buffer (Trizma base, 0.025 M; glycine, 0.192 M; SDS, 0.1%). Proteins were stained with 0.2% Coomassie Brilliant Blue R-250 in water:ethanol:acetic acid (25:25:10, v/v). Molecular weight standards (BRL) were: myosin H chain, 200 kD; phosphorylase b, 97.4 kD; BSA, 68 kD; ovalbumin, 43 kD; carbonic anhydrase, 29 kD; fi-lactoglobulin, 18.4 kD; and lysozyme, 14.3 kD. H. Sample Preparation for Transmission Electron Microscopy (TEM) Dr. John Heckman (from Michigan State University's Center for Electron Optics) prepared the THRGP and HHRGP for TEM following the general methods of Tyler and Branton (1980): highly diluted samples (1-30 ug/mL) dissolved in 50% v/v aqueous glycerol were sprayed onto freshly cleaved mica chips with a modified airbrush. After drying the chips in vacuo on a rotary stage in a Balzers BAE 080 high vacuum evaporator, he shadowed the molecules at an angle of 5 to 6°, with Pt/C from a modified Balzers 052 twin-mantle electron-beam 25 evaporator. After backing with carbon at 90° C he floated the replicas on distilled H,O, collected them on 300 mesh copper grids, and examined them in a JEOL 100 CX II transmission electron microscope operated at 100 kV. I. Circular Dichroic Spectra Mark Prairie and Dr. Bill Kreuger (of Upjohn Company, Kalamazoo, MI) recorded CD spectra of the THRGP, dTHRGP, HHRGP, poly-proline II and poly- hydroxyproline (1 mg protein /mL in 100 mM sodium phosphate buffer) on a J asco J -600 spectropolarimeter in a 0.086 mm pathlength quartz cell. They scanned each sample from 260 to 178 nm four times. Molecular ellipticity [6 ], has the dimensions of (degree-cmz/dmol). J. Precfiitation with B-Glucosyl Yariv Antigen I reacted 400 ,ug THRGP, HHRGP or sycamore arabinogalactan protein in distilled H20 (1 mg/mL) with an equal volume of fi-glucosyl Yariv Antigen (Jermeyn & Yeow, 1975)(1 mg/mL in 2% NaCl) for 1 hr at 27° C followed by pelleting in a bench-top centrifuge. The pellet was washed twice with 2 mL 2% NaCl, each wash was followed by pelleting in a bench-top centrifuge, and then dissolved in 2 mL 0.02 N NaOH. Absorbancy was read at 420 nm. K. Assay of Agglutination I assayed the agglutinating effect of a serial dilution of THRGP and HHRGP (100 to 10 ng/mL) on a 1% suspension of trypsinized rabbit erythrocytes in phosphate-buffered saline according to the method of Allen and Neuberger (1978) 26 L. Partial Acid Hydrolysis of HHRGP I hydrolysed 4 samples of 0.5 mg HHRGP in 250 ILL 0.1 N HCl (pH 1), for 0, 15, 30, or 60 min at 100° C, followed by microdialysis for 2 days, freeze-drying, and then alditol acetate derivatization after complete acid hydrolysis of the HHRGP. Then I assayed for neutral and amino sugars by gas chromatography. III. Methods for Generatmgéeparating and SequencingPeptides A. Proteolysis with Trypsin and Chymotrypsin I incubated 2 to 6 mg deglycosylated THRGP, HHRGP, or THRGP Chymotryptic peptide (10 mg/mL) in freshly prepared 1% (w/v) NH4HCO, (aq) containing 10 mM CaCl2 with TPCK-trypsin or Chymotrypsin (Worthington)(substrate:enzyme ratio was 100-20021, w/w) at 27° C overnight. B. Peptide Fractionation via Sephadex G-25 Superfine Gel Filtration I injected 0.1 to 1 mg freeze-dried peptides (in 0.5 mL 0.1 N acetic acid) onto a 27.5 mm x 1.25 mm i.d. column of Sephadex G-25-80 (superfine) eluted with 0.1 N acetic acid at 10 mL/hr. Absorbancy was monitored at 220 nm on a Hewlett- Packard Photodiodearray spectrophotometer. C. HPLC Peptide Mapping I obtained peptide maps via reversed phase HPLC of tryptic/Chymotryptic digests on a Hamilton PRP-1 (4.1 mm i.d. x 150 mm) column using programmed gradient elution (0.5 mL/min) with the following mobile phase solvents: A = 0.13% HFBA, and B=0.13% HFBA in 80% (v/v) acetonitrile (aq). For resolution of peptides, the gradient began at 100% A and increased (0.5% /min) from 0 to 50% 27 B in 100 min. Absorbancy was monitored at 220 um. I manually collected fractions for analyses. All peptides were rerun through the Hamilton column to assure purity before sequencing. D. Automated Edman Degradation Joe Leykam and Melanie Corlew (Michigan State University Macromolecular Facility) sequenced 2-200 nm HPLC-purified peptide via Edman degradation (Edman, 1970) on a 477A Applied Biosystems, Inc. gas phase sequencer. IV. Methods for Immunological Characterization of THRGP and HHRGP A. Generation of Rabbit Polyclonal Antibodies Three 5—lb New Zealand white rabbits were from the Small Animal Care Facility, Michigan State University. The primary injection was 70 to 120 fig THRGP or dHHRGP in distilled H,O emulsified in Freund's Complete Adjuvant (Cappel Laboratories), and injected subcutaneously into each shoulder and hip of the rabbit. When antibody titers fell to control serum levels, booster injections of antigen contained 50 to 100 u g protein in 500 ,uL water-in-oil emulsion of Freund's incomplete adjuvant injected subcutaneously over each hip. B. Enzyme-Linked Immunoassay (ELISA) Essentially, this is based on that of Engvall & Perlmann (1972). I coated each test well of 96 well polystyrene plates (Nunc, Thomas Scientific) with 20 or 200 ng antigen (antigens: THRGP, dTHRGP, HHRGP, dHHRGP, tomato extensin monomers P1 and P2, dP1 and dP2, tomato lectin, sycamore AGP) in 200 uL pH 28 9.6, 50 mM NaHCO3 buffer, for 15 hr at 4° C; washed the plate once in H20 and briefly dried it before blocking all remaining protein binding sites by addition of 200 ch BSA in PBS (final pH 7.5), for 30 min at 37° C; followed by washing twice with H,O. Then I added 25 uL of the diluted sera (immune and control preimmune sera were diluted 1:400 to 1:256,000 in PBS) to the antigen-coated wells already containing 25 ,uL 1% BSA-Tween-20 (1 uL/mL) PBS at pH 7.5. After 1 hr at 37° C we washed the plate with distilled HZO, added 50 ,uL goat anti-rabbit serum coupled to peroxidase (Cappel Laboratories) diluted 1:2000 in BSA/Tween-20/PBS to each well, incubated at 37° C for 30 min, washed the plate five times with distilled H,O, then added 100 yL substrate to each well (11 mg ABTS and 15 uL 30% H202 in 50 mL of 50 mM citrate buffer, pH 4). After 30 min. incubation at 23° C, I added 100 uL NaF/EDTA stopping reagent (0.04% EDTA, 6 mM NaF, 2.5 mM HF) to each well, and then determined absorbance at 405 nm. C. Immunoblotting These methods are based on that of Laemmli & Favre (1974) and Towbin et al. (1979). I loaded 20 to 40 ug THRGP, 10 to 15 ug dTHRGP, and 50 ,ug HF- soluble maize cell wall preparations in 10 to 25 ,uL sample buffer onto a 15% polyacrylamide gel and ran the gels in pH 8.3 Tris-Glycine buffer as described earlier. After electrophoresis, the proteins were transferred to nitrocellulose sheets. The sheets were blocked with incubation buffer (10 mM Tris, pH 7.4, 0.9% NaCl, 3% BSA) for 18 hr at 4°C, then incubated 1 hr at 37° C with immune or preimmune sera diluted 1:5000 in incubation buffer. The nitrocellulose sheets 29 were washed three times with wash buffer (10 mM Tris [pH 7.4], 0.9% NaCl, 0.1% BSA) then incubated 1 h at 37°C with alkaline phosphatase conjugated goat anti- rabbit IgG diluted 1:2000 in AP buffer (100 mM Tris, pH 7.5, 100 mM NaCl, 2 mM MgCl,) before washing three times in AP buffer, pH 7.5 and twice with AP buffer pH 9.5 (100 mM Tris, pH 9.5, 100 mM NaCl, 5 mM MgCl,). I added substrate (50 mL AP buffer, pH 9.5, with 16.5 mg nitroblue tetrazolium and 8.35 mg 5-bromo- 4-chloro-3-indolyl phosphate) to cover the sheet and incubated in darkness at 27° C until color developed. The reaction was stopped with 10 mM Tris, pH 7.5, 1 mM EDTA. RESULTS 1. Isolation of Maize HRGPs A. Intact Cell Elution The amount (mg/g cells dw) of elutable crude HRGP rose as a function of culture age. After subculture, total soluble eluate fell to a minimum at 2 days and rose to level off at 11 days (Figure 5). I used 100 mM AlCl3 (a non-plasmolyzing concentration) to elute the cells. Thus, for bulk preparations I used 11 day cells (optimal age for recovery of maize HRGPs, judging by recoveries of HRGP from elutions of different aged cells) and 100 mM AlC1,, the cells yielding 5.4 mg crude eluate / g cells dw. The crude eluate was 1.7% Hyp on a dw basis (i.e. 92 ug soluble Hyp/g dw cells). N O I ~4OO ~3OO .A U! I .3 O l -1 200 % Packed Cell Volume (on-c) (II I #100 mg Crude HRGP/g Cells dw (o—o) l 1 l 1 1 1 1 1 1 1 l 1 2 3 4 5 6 7 8 9 1O 11 12 Days after Subculture Figure 5. Yield of crude HRGP as a function of culture age. 30 31 B. Treatment of the Crude Eluate with TCA Overnight treatment with 10% (w/v) TCA at 4° C followed by centrifugation precipitated 50% by weight of the crude eluate. The TCA pellet was 72% protein and 0.5% Hyp dw, but the HRGP remained soluble. The yield of TCA-soluble crude HRGP was 2.7 mg/ g cells dw. The crude HRGP was 60% protein and 3.5% Hyp dw, representing a 60-fold Hyp enrichment over the whole cell Hyp content of 0.06%. The crude HRGP was fractionated as shown in the flow chart (Figure 6). [ Maize Cell Suspension‘sfil AICI3 Cell Pad LSalt Eluate 1 10% TCA TCA Pellet [Crude HRGP] Cellex-P chromatography LHyp-rlch Peak 2 ] BiorexJO chromatography [ Hyp-rich Peak1 I [ Hyp-rich Void I J/ Superose-6 gel filtration j] Figure 6. Maize HRGP fractionation flow chart 32 C. Phosphocellulose Chromatography of Maize Crude HRGP Chromatography on phosphocellulose (Cellex-P, Biorad) yielded a void and three major protein peaks designated CPl, CP2 and CP3 respectively (Figure 7). CPl eluted at pH 3.8 in the pH gradient, CP2 with 200 mM NaCl and CP3 at 450 mM NaCl. The void volume of the eluate and CPl contained a trace of Hyp while CP2 contained 8.4% Hyp dw and CP3 contained 2.7% Hyp dw. The dry weight yields of the void volume plus CPl, CP2, and CP3 accounted for a 40% recovery of the material loaded on the column, and accounted for 76% recovery on a dry weight basis, of the starting hydroxyproline (Figure 8a). Figure 9 quantifies the pg of Hyp per each 4 mL fraction, beginning with fraction 35 of CP2 through fraction 53 of CP3. CP2, the shoulder following CP2, and CP3 contained significant amounts of Hyp. 1.5- E CP2 -, 5“ Void / '5 51°“ 033/” «052 8 CP‘ //’ -o.4 § 2 /’I “0.3 g / 0-5‘ ,/ -o.2 2 ,/ ~O.1 g . -- . -o.o 2 ‘___pH Gradient—4p <___Salt Gradient__, 2 4 i 6 8 Hours Figure 7. Phosphocellulose chromatography of maize crude HRGP 33 a) Hydroxyproline recovered from Cellex-P/dialysis' 0.1% .CP Void, CP1 ICP2 ICP3 Ilost b) Hydroxyproline recovered from BioRex-70/dialysis Peak 1 42% IVoid lost 46%; c) Hydroxyproline recovered from Superose—S/u ltra filtration 32°/o TH RGP IHHRGP IARGP lost 2% 66 % 1 00% Figure 8. Recovery of hydroxyproline from a) Cellex-P b) BioRex-70 and c) Superose-6. 34 1 000 200 -180 .——-_-________ _ .. l .5 .5 N b O O 500 - 1 00 mAU Abs 220 nm o—Ouonaeu “up/dim 611 H / / ,3 9 l I 1 35 40 45 50 Fraction number (4 ml) Figure 9. The hydroxyproline distribution in Cellex-P peaks CP2 and CP3 D. Biorex-70 Cation Exchange Chromatography of CP2 Further chromatography of CP2 on Biorex-70 gave a void (4% Hyp dw) and a Hyp-rich fraction (12% Hyp dw) designated Peak 1, which eluted at 200 mM NaCl (Figure 9). The dry weight yield of the void plus Peak 1 accounted for 54% of the weight and 54% of the Hyp (dw) loaded on the Biorex column (Figure 8b). 35 I) Void ' Peak1 1.5 - c s :2 m g 1.0 — z N C < 8 c O 0 U to 2 0.0 Hours Figure 10. Cellex-P Peak 2 (CP2) on a Biorex-70 cation exchange column. E. Superose-6 FPLC Gel Filtration of Glycosylated and Deglycosylated THRGP and HHRGP Gel filtration of the Biorex Void gave a major Hyp-rich (8% Hyp dw), histidine-rich peak, designated HHRGP, at 1.8 V, (Figure 11a). Gel filtration of Biorex Peak 1 on Superose-6 gave a major Hyp-rich (18% Hyp dw), threonine- rich peak, designated THRGP, at 2 x V,, and a minor Hyp-containing peak (3% Hyp dw) that was also rich in alanine (16 mole% Ala, designated alanine-rich glycoprotein, or ARGP) eluted at 2.5 V0 (Figure 11b). The dry weight recovery of HHRGP was 52% of the starting Biorex void starting weight, and accounted for 100% recovery of Hyp loaded on the Superose-6 column (Figure 8c). Thus, after losses due to chromatography and dialysis, about 3.6% of the original TCA-soluble 36 crude HRGP dry weight (i.e. 3-4 mg HHRGP/100 mg crude eluate), and 8.4% (dw) of the starting Hyp was recovered. THRGP accounted for 1.8% and 9.4% of the crude HRGP dry weight and Hyp content respectively. Deglycosylated HHRGP eluted at 2.2 V0, while the dTHRGP was retained on the column, probably due to ionic interactions with the Superose agarose matrix (data not shown). A B HHRGP THRGP 1.0 — 1.0- g E o c H a O < N «1 < 0.5 —- 0.5-— Vold J/j J 1 o 44 O 50 Minutes Minutes Figure 11. Superose-6 gel filtration of a) HHRGP and b) THRGP II. Chemical and Structural Characterization of Maize HRGPs A. Amino Acid Analyses of the THRGP and HHRGP THRGP contained about 25 mole% Thr and Hyp, and was rich in Pro and Lys. HHRGP contained 35 mole% Hyp, and was rich in His and Ala (Table 5). 37 Table 5. Amino Acid Compositions of THRGP, HHRGP and Tomato Extensin P1 Tomato Extensin Amino Acid‘ THRGP HHRGP P1” Hyp 24.8 34.9 32.7 Asx 0.3 1.3 1.4 Thr 25.3 7.9 6.2 Ser 7.3 7.3 9.8 Glx 2.3 2.1 1.5 Pro 14.5 6.8 9.6 Gly 2.4 3.1 1.7 Ala 1.7 8.9 2.9 Val 0.7 1.5 8.3 Cys 0.0 0.0 0.0 Met 0.0 0.0 0.0 Ilu 0.1 0.0 1.0 Leu 0.2 0.0 1.0 Tyr 3.9 4.4 7.7 Phe 0.1 3.5 0.0 Lys 13.5 3.5 9.5 His 2.4 13.4 6.1 Arg 0.1 1.3 0.7 ' Represented as Mole % ° From Smith et al., 1984 B. HF-Deglycosylation of HRGPs The THRGP lost 27-33% of its weight, and the HHRGP 60-70% of its weight, after HF-deglycosylation. C. Neutral Sugar Analyses Hydrolysis of THRGP in 2 N TFA followed by reduction with NaBH,, alditol acetate derivatization, then gas chromatography showed arabinose as the only sugar substituent (Table 6). The arabinosezHyp molar ratio was 1.44:1 and accounted for 27% by weight of the THRGP. Alditol acetate derivatization of the HHRGP showed galactose and arabinose as the only sugar substituents (Table 6). 38 The arabinose:I-Iyp molar ratio was 2.4:1, the galactose:Ser and galactosezThr molar ratios were 5:1. Thus, arabinose and galactose accounted for 38% and 27% respectively, of HHRGP dry weight. Table 6. Neutral Sugar Compositions of THRGP, HHRGP and Tomato Extensin P1 Tomato Extensin Neutral Sugar° THRGP HHRGP P1b Arabinose 100 63 91 Galactose 0 37 9 ‘ Represented as Mole % ° From Smith et al., 1984 D. Partial Acid Hydrolysis of the HHRGP Treatment of HHRGP at pH 1 for 1 hr at 100° C removed all arabinose oligosaccharide substituents from the hydroxyproline residues (Figure 12). Very little galactose was hydrolysed, judging by alditol acetate derivatization and GLC of the HHRGP before and after hydrolysis. E. Hydroxyproline Arabinoside Profiles of the THRGP and HHRGP Hydroxyproline-arabinoside profiles of the THRGP (Figure 13) showed 48% nonglycosylated Hyp and Hyp-arabinoside 3 (Hyp-Ara,) as the major glycosylated component (Table 7). The double peaks correspond to the trans and cis Hyp-arabinosides, the result of alkaline hydrolysis (Lamport & Miller, 1971). The Hyp-arabinoside arabinoside profile of the HHRGP showed 20% non-glycosylated Hyp and Hyp-arabinoside 3 as the major arabinoside component (Table 7). 39 -03 OnM Arabinose/11.9 HHRGP 011M Galactose/p9 HHRGP nM Arabinose or Galactose 1 1 1 - O 1 5 30 45 60 Minutes Hydrolysed Figure 12. Partial acid hydrolysis of HHRGP at 100°C, pH 1 A560 nm 0.5 Free Hydroxyproline 3 1 4 2 0 w AA w‘: J L—+ 1 1 1 1 1 1 2 3 4 5 Hours Figure 13. Hydroxyproline arabinoside Profile of THRGP. Numbers correspond to the Hyp arabinoside 1, 2, 3 or 4. 40 Table 7. THRGP, HHRGP and Tomato Extensin P1 Hydroxyproline Arabinosides‘ Tomato Extensinb Hyp Arabinoside THRGP HHRGP P1 Hyp-Ara 15 8 9 Hyp-Ara,c 6 9 8 Hyp-Ara, 25 42 33 Hyp-Ara4 6 21 38 Free Hyp 48 20 12 ' Represented as percent of total Hyp ° Hydroxyproline di-arabinoside " From Smith et al., 1984 F. SDS-PAGE of the THRGP and HHRGP The THRGP migrated as a fuzzy band with M, range of 72-90 kD (Figure 14; lane 2). A high molecular weight band, probably a THRGP aggregate, migrated at about 200 kD; however the same preparation, after HF-deglycosylation (dTHRGP), lost the large molecular weight band and migrated with a M, of 50 kD (Fig 14, lane 3), which was a 22 kD loss in Mr somewhat greater than predicted from the 30% arabinose content. The deglycosylated HHRGP (dHHRGP) at 5 lug/lane ran as two discreet bands with M, 68 and 70 kD (Figure 15, lane 2) and at 15 pg, as a 68-70 kD smear (Figure 15, lane 3). Glycosylated HHRGP did not enter the SDS gel. G. TEM Visualization of the THRGP and HHRGP TEM shadowed preparations of the THRGP and HHRGP both showed rod- like molecules averaging 70 i 3 nm in length (Figure 16 a & b, respectively). 41 kD 1 2 3 200— ; 97.4— —— 1—THRGP 68— - 43_ . ‘ *‘dTHRGP 29— -- 18.4— - 14.3— - Figure 14. SDS-PAGE of THRGP (lane 2), dTHRGP (lane 3) and molecular weight standards (lane 1). 42 . 200— 1 . 97.4— 68- .. - ‘ thHRGPs 43— - 29— - 18.4— D g 14.3— Figure 15. SDS-PAGE of dHHRGP (lanes 2 & 3) and molecular weight standards (lane 1). 43 H. Circular Dichroic Spectra of THRGP, dTHRGP and HHRGP The CD spectra of the THRGP and dTHRGP showed no maximum and single broad minimums at 200 to 205 nm (Table 8)(Figure 17a), similar to that of a fibrous protein in an unordered, or "random coil" conformation (Tiffany & Krimm, 1969). The CD spectra of HHRGP showed a maximum at 224 nm (which indicates a more ordered structure for HHRGP than the THRGP (Tiffany & Krimm, 1969), possibly due to some polyproline II conformation; and HHRGP showed a minimum at 202 nm (Table 8) (Figure 17b). Table 8. Features in the CD of THRGP, dTHRGP, HHRGP, Poly-L-Proline and Poly-4-Hydroxy-L-Proline in Order of Decreasing km Polymer km; A Em‘ kg Am A Em HHRGP 224 0.6 218 202 -3 THRGP - - - 205 -3 dTHRGP - - - 202 -3 Poly-L-Proline 228 1 222 206 -14 Poly(4-L-hydroxyproline) 225 2.2 219 205 -13 Random Coil" - - - 190-200 varies B-Sheet° 190-200 varies 205 210-220 varies oz-Helix° 190-195 varies 205 208 + 222 varies ' The subscripts max, crs, and min refer to observed maximum, croSsover, and minimumAE is the average CD/peptide bond; >\ is the wavelength. ° From Tiffany & Krimm, 1969 ° From Johnson, 1988 I. Assay of Agglutination Serial dilutions of THRGP or HHRGP (100-10 ng/mL) did not agglutinate a 1% suspension of trypsinized rabbit erythrocytes. 44 Figure 16. Visualization of THRGP and HHRGP by transmission electron microscopy. Rotary shadowed A) HHRGP and B) THRGP were flexible rods about 70 nm long. 46 Figure 17. Circular dichroic spectroscopy of a) THRGP, deglycosylated THRGP (dTHRGP), and b) HHRGP. a) CD spectra of poly-L-proline II (Cl—D), poly-4-hydroxy- L-proline (H), THRGP (o—o), and dTHRGP (H), showed the THRGP secondary structure is not a polyproline H helix. b) CD spectra of polly-L-proline II (D—fi), poly-4-hydroyx- L-proline (H), and HHRGP (o—o) showed that HHRGP secondary structure was more ordered than the THRGP, possibly due to some polyproline II conformation, however, still mainly random coil. Molar Ellipticityx10—3 Molar Ellipticity x 10—3 15 47 l l l 210 2:30 Wavelength (nm) 250 270 15 l l l l J 1 190 210 230 Wavelength (nm) 250 270 48 J. Reaction. of THRPG and HHRGP with B-Glucosyl Yariv Antijgrl Neither THRGP nor the HHRGP reacted with Yariv Antigen, even at the relatively high level of 0.5 mg/mL where a standard AGP (sycamore) gave a absorbance of 2.36 at 420 nm. K. Proteolysis of dTHRGP with Chymotrypsin and/or Trypsin, Sephadex G-25 Gel Filtration, HPLC Peptide Mapping and Edman Degradation Chymotryptic digestion of dTHRGP yielded peptides of four size classes, judging by Sephadex G-25 gel filtration (Figure 18). Fractionation of the chymotryptic digest by HPLC gave a peptide map (Figure 19) consisting of a relatively few peptides dominated by a single major component, TC5 (See Table 9 for the amino acid molar ratios of the 8 peptides). Automated Edman degradation gave a 16- residue sequence for TC5: Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys-Pro-Thr- Hyp-Hyp-Thr-Tyr which, as the dominant peptide, therefore exists as a repetitive unit of the THRGP. Furthermore, residues 3 through 10 of TC5 and TC4 constitute 8-residue sequences highly homologous with tryptic peptide H5 from tomato extensin P1: Ser-Hyp-Hyp-Hyp-Hyp-Thr-Hyp-Val-Tyr-Lys (Smith et al., 1986), and tryptic peptide H5 from sugar beet extensin P1: Ser-Hyp-Hyp-Val-His-Glu-Tyr-Pro-Hyp-Hyp-Thr-Hyp-Val-Tyr-Lys (Figure 29) (Li et al., 1990). The minor maize peptides sequenced were also related, including TC1 which contained the only Ser-Hyp-Hyp-Hyp-Hyp sequence in the THRGP (Table~10). Peptides TC1, TC2, TC4 and TC5 were sequenced a minimum of twice (from different peptide preparations), and peptides TC6, TC7 and 'I'I‘l, the 49 tryptic peptide, were sequenced once. The typical peptide repetitive yield, i.e. for TC5, was 77%, vs. 97% for standard proteins. The weight percent distribution of recovered peptide was: TC1, 10%; TC2 and TC3 combined for 11%; TC4, 18%; TC5, 43%; TC6, TC7 and TC8 combined for 18%. Thus the total dTHRGP peptide recovered from the PRP-1 column was approximately 57% of the material loaded. 1000 - (D O O l 600 - 400 - mAU 220 nm 200 L- 0.1:.._ l l I l J L l l l 10 20 30 4O 50 60 7O 80 90 Time(min.) Figure 18. Sephadex G-25 gel filtration of dTHRGP chymotryptic peptides. 50 TC5 aoo - E 250 -— ‘ TC7 g 200 —- TC4 TC N D 150 — TC3 g 'rc2 1'03 1oo — TC1 50 — I l l I so so 7o 80 Time (min.) Figure 19. Chymotryptic peptide map of dTHRGP. Table 9. Amino Acid Compositions of dTHRGP Chymotryptic Peptides‘ Amino Acid TC1 TC2 TC3 TC4 TC5 TC6 TC7 TC8 Hyp 7.0 4.1 6.8 4.5 5.2 5.7 7.0 6.9 Thr 5.2 3.5 5.6 3.8 4.4 5.0 6.8 6.7 Ser 3.4 0.8 1.3 1.1 1.1 1.3 1.5 Pro 1.7 1.0 2.2 2.3 2.9 3.2 4.4 4.6 Ala 1.0 1.0 Val 1.1 Ilu Leu Tyr 1.0 1.0 1.0 1.0 1.0 1.0 1.0 1.0 Phe Lys 0.7 1.9 1.9 2.0 2.9 2.6 1.2 His 1.2 0.9 Arg ‘ Represented as Molar Ratios 51 Table 10. Amino Acid Sequences of dTHRGP Tryptic and Chymotryptic Peptides Chymotryptic Peptide Sequence TC1 Thr-Hyp-Thr-Hyp-Val-Ser-His-Thr-Hyp-Ser-Hyp-Hyp-Hyp—Hyp-Tyr TC2 Thr-Hyp-Ser-Hyp-Lys-Pro-Thr-Hyp-Hyp-Thr-Tyr TC3 ND TC4 Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Hyp-Lys-Pro-Pro- TC5 Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys-Pro-Thr-Hyp-Hyp-Thr-Tyr TC6 Thr-Hyp-Thr-Hyp-Lys-Pro-Hyp-Ala-Thr-Lys-Pro-Pro-Thr-Tyr TC7 Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-I-Iis-Pro-Thr-(Pro)- TC8 ND Tryptic Peptide Sequence TT1 Pro-Thr-Hyp-Hyp-Thr-Tyr-Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys L. Tryptic Digestion of dTHRGP and Chymotryptic Peptide TC5, HPLC Peptide Mappig, and Edman Degradation A major peptide, TT1, recovered from tryptic digestion of dTHRGP (See Figure 20 for tryptic peptide map) showed an apparently anomalous N-terminal proline residue (conventionally, trypsin does not cleave lysylproline), and whose sequence overlapped with chymotrytic peptide TC5 (Table 10). Figure 21a shows the amino acid sequence and chromatographic profile (220 and 280 nm) of dTHRGP major chymotryptic peptide TC5. Although this 16-residue peptide has two Lys-Pro peptide bonds (Ser-Hyp-Lys-Pro-Hyp and Thr-Pro-Lys-Pro-Thr), only the latter was cleaved by trypsin. Thus, in tryptic digestions of chymotryptic peptides of the deglycosylated THRGP, or in tryptic digestions of deglycosylated THRGP, the Lys-Pro linkage in the Thr-Pro-Lys-Pro-Thr sequence was slowly 52 I cleaved. There was no evidence of other Lys-Pro cleavage in the THRGP. Figure 21b shows the peptide profile of the partially digested chymotryptic peptide after a 12-hr incubation with TPCK trypsin. Two new peptides appeared. The first peptide (at 50 min on the map) was a 6-residue peptide beginning with the proline of the Lys-Pro cleavage and ending with the tyrosine residue that terminated TC5, Pro-Thr-Hyp-Hyp-Thr-Tyr. The second peptide (at 58 min) was a 10-residue peptide ending with the lysine residue of the lysyl-proline bond from TC5, Thr- Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys. The third peptide (67 min) was intact TC5. The presence or absence of tyrosine in the peptides was corroborated by absorbance, or lack of it, at 280 nm. mAU 220nm -20 1 ' 1 1 1 20 40 60 80 1 00 Time(min.) Figure 20. Tryptic peptide map of dTHRGP“. ‘1 The tryptic peptide map was done only once on impure THRGP therefore many minor peptides appear. 53 (a) ThrHyp—SenHyp- Lys-Pro-Hyp-Thr-Pro-Lys-Pro-Thr-Hyp—Hyp-Thr-Tyr ) mAU 220 nm( mAU 280 nm (——.—) Time (min.) (0) Th r-Hyp-Ser- Hyp-Lys- Pro-Hyp-Thr-Pro-Lys-Pro-Thr-Hyp-Hyp-Thr-Tyr 500 '1 Th r-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys ) 400 - Pro-Thr-Hyp- Hyp-Th r-Tyr 300? 200 4 mAU 220 nm( mAU 280nm (..__.) 100‘ 04 Time (min.) Figure 21. Chymotryptic peptide TC5 a) before and b) after tryptic hydrolysis 54 i M. Chymotryptic Digestion of dHHRGP, Sephadex G-25 Gel Filtration, HPLC Peptide Mapping, and Edman Degradation of dHHRGP Judging by Sephadex G-25 Gel Filtration (Figure 22), chymotryptic digestion of dHHRGP yielded peptides of several size classes, from large peptides that eluted immediately behind the G-25 void, to very small peptides (i.e about 6 residues; Table 12) eluting at 90 min. HPLC fractionation of the complete chymotryptic dHHRGP digest gave a peptide map consisting of 4-5 minor peptides and 7-8 major peptides (Figure 23) The amino acid molar ratios of the dHHRGP peptides do not match exactly the amino acid sequences of the peptides, especially HC1, whose amino acid composition (Table 11) predicts a serine residue which does not appear in the sequence (Table 12). However, many of the sequences are incomplete (HC1, HC4, HC6, and HC12); furthermore, the dHHRGP peptide map, amino acid compositions and sequences were determined only once for the HHRGPs, and therefore should be considered preliminary data. Judging by Edman degradation, the tetrapeptide Ala-Hyp-Hyp-Hyp occured in peptides HC3, HC4, HC10, HC11, and HC12; and comprised a "sub-periodicity" in the decapeptide: Ala-Hyp-Hyp-Hyp-His-Phe-Pro-Ser-Hyp-Hyp... that occurred in HC10, HC11 and HC12 and as a nine-residue variant in HC4 (a Tyr for His substitution at residue 5 and deletion of the serine residue). Thus, both the Ala- (Hyp)3 tetramer and the decamer: Ala-Hyp-Hyp-Hyp-His-Phe-Pro-Ser-Hyp-Hyp are repetitive units in at least one of the HHRGPs. Edman degradation of the major peptide HC6 did not give a complete sequence (Table 12), however, HC6 was 55 I homologous (89%) with the first 9 amino acids of "PB-type" extensin peptide (Ser-Hyp-Hyp-I-flp-Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-Tyr-Tyr-Tyr-Lys)(Smith et al., 1986; Showalter & Rumeau,1989), except for the single substitution of His for Hyp at residue 5. Furthermore, peptides HC10, HC11, and HC12 are 66% homologous with the first 10 amino acids of P3 type extensin, having Ala and Phe substitutions for Ser at residues 1 and 6, and a His for Hyp substition at residue 5. HC10 contained an asparagine-centered 9 residue palindrome: Hyp-Hyp-Ala- Ala-Asn-Ala-Ala-Hyp-Hyp. Chymotryptic peptide HC2 showed a C-terminal histidine residue, indicating that Chymotrypsin may be selectively cleaving after some histidine residues. Chymotryptic cleavage of His-Thr and His-Ser bonds occurs in Azurin (Ambler & Brown, 1967). 1 100 — 1000 - 900 — 800 — 700 - 600 - 500 r- 400 — 300 - v0 200 - 1 00 - . 1’ 0&4 I 1 l I 20 40 60 _80 100 Time (min.) mAU 220nm Figure 22. Sephadex G-25 gel filtration of dHHRGP chymotryptic peptides. 56’ HcesHC7 HC8 14o- Hcs 1 H09 HC10 HC11 E C o .01 N HC12 3 < E o .— 1 1 1 1 40 60 80 100 Time (min) Figure 23. Chymotryptic HPLC peptide map of dHHRGP III. Immunolggical Characterization of THRGP A. Cross-Reactivity of Anti-Tomato Extensin Polyclonal Antibodies with THRGP and HHRGP Rabbit polyclonal antibodies raised against tomato extensin monomers P1 and against the protein backbone of HIP-deglycosylated P1 and P2 cross-reacted ca. 40%,and 18% respectively with the THRGP and HHRGP (Figure 24), indicating common antigenic epitopes between the tomato extensins and the maize HRGPs. B. Generation of Antibodies against THRGP and Characterization via ELISA Three weeks after challenging 2 rabbits with THRGP, an immunogenic reaction was apparent as determined by ELISA assays. Titer rose until week 5 and remained high for more than 17 weeks after the primary injection. Figure 25 57 5:309:00 acumen 05 noon wofioooon .25 5 HS. xi :0 women 8:8 868 8 twosomoaom . we... 8 8 S. 8 8 8. 8 8 8 8 a: 93 8 8 8 8 23 8 8 8 8 8 8 E. 8o a: a> 8 8 S 8 8 8 8 8 8 8 a2 8 be 8 8 8 8 em so 8 8 8 .8 8 8 8 8 8 am 8 8 2:. 8 8 one... 8 8 2: 8 oz oz 8 8 8 E 8 8 of 80m 20m 20: 8m 8m Bo com mom .0: . 83 8m 6m 23.. 055 .888 2:56:80 8988 8 88880 2% 058... .2 2.8. 58 A5858sweméfiémém-&m-&m-m2-m2 Sum 5-%m-&m-%m-mfi-m£25am:fiémdféméabfi¢f§$§¥§¥§ :0: awaaZEEQmEmégm2&8.“2-8-928-95-8-5-0fi¢£§$§$§2§< 20: Q2 83 mz 8m Dz 5m -5Em;o858-5maifimémémém 8m oz 83 Em-Eb$8-056iéeémémémé< Em 5-fiéfifiQOEméfififiz 8m améEaZEmEWBm Sm -oa-2<-o£-m2-&m-&m-m2 8m ooaoscom oEEom ouRCSEEU 82:5 2330520 mommmo Ho 3888 2% came .2 28; 59 100 '- Threonine-rioh HRGP Bop-‘E‘ 60- 40 - + 20 .. ii % Cross-reactivity P1 dP1 P2 dP2 Anfibody Figure 24. Reactivity of THRGP with antibodies raised against tomato extensin precursors. a detectable reaction to 20 ng THRGP/microtiter well and antisera dilutions as high as 1:64,000. We routinely worked with primary antisera dilutions of 1:5000. Figure 26 shows the cross-reactivity of THRGP antibodies with antigens: ' dTHRGP, HHRGP, dHHRGP, tomato extensin monomers P1 and P2, deglycosylated P1 and P2 (dP1, dP2). The antibodies did not react with tomato lectin or sycamore AGP (data not shown). Pre-immune control serum did not react with any of the antigens. 6O 2 1.8 1.6 1.4 1.2 1- 0.8- I 0.6- ' 0.4 0.2- ' . . 0 l JJLLLIII 1 1 lllLlll l 1 111.1111. LLIIIL 1 00 1 000 1 0000 1 00000 1000000 THRGP Crude Serum Dilution Figure 25. Serial dilutiOn of anti-THRGP antiserum with THRGP antigen at 20 ng/ELISA microtiter well. 1 00% — —T— 90% l 80% 70% 60% L- 50% 40% 30% 20% ‘ 1 0% ” ' z ['57 [j I 0% h" . o" o? o? o? v‘ Q“ 9‘1 468‘ 5198‘ 958% 8693‘ 6 6d) Antigen (200 ng/well) I I I I r I I Absorbance 405 nm I I I I I Cross Reactivity T .1. Figure 26. Cross-reactivities of anti-THRGP antibodies with dTHRGP, HHRGP, dHHRGP and tomato extensins P1, dP1, P2, dP2. 61 I IV. Hydroxyproline Arabinoside and Protein Profile of the Maize Cell Wall A. Hydroxyproline Arabinoside Profile of the Maize Cell Wall Hydroxyproline-arabinoside profiles of the maize cell wall showed 24% nonglycosylated Hyp, and hydroxyproline tri-arabinoside as the major hydroxyproline arabinoside component (Table 13). Table 13. Hydroxyproline Arabinoside Profiles‘ of Maize and Tomato Cell Walls Hyp Arabinoside Black Mexican Maize Pericarp" Tomato: Hyp-Ara 9 15 10 Hyp-Ara;l 6 2 9 Hyp-Ara3 41 13 28 Hyp-Ara4 10 4 48 Free Hyp 24 66 5 ' Expressed as percent of total Hyp ° From Smith et al., 1984 " From Lamport & Miller, 1971 ‘ Hydroxyproline di-arabinoside B. HF Deglycosylation and Hydroxyproline Content of the Maize Cell Wall Cell walls prepared from maize (Black Mexican) cell suspensions and maize (Sweet Sue) coleoptile, root, and root tip contained bound hydroxyproline, mostly HF-soluble, although some (from a trace to 17.4%) remained associated with the HF-insoluble residual wall fraction (Table 14). Significant amounts of Hyp ("40%) and protein (”25%) were lost during dialysis, possibly as dialyzable molecules, or by adsorption to dialysis membranes. 62 i C. Amino Acid Analyses of the Black Mexican Maize Cell Wall, the HF- Soluble Wall and the HF-Insoluble Wall Judging by recoveries after amino acid analyses, the Black Mexican cell wall before deglycosylation was 10% protein, the I-IF-solubilized wall 20% protein, and the I-IF-insoluble wall fraction was 50% protein. The protein component of the HF-solublized wall fraction was enriched in Hyp and His, while the HF-insoluble wall fraction retained a trace of Hyp (Table 15). Table 14. Hydroxyproline Distribution in the HF-Soluble and HF-Insoluble Black Mexican Maize Cell Wall Cell Wall HF -Soluble HF-Insoluble Dry Weight High Molecular Weight Residue mg #g Hyp mg #g Hyp mg ugHyp Root Tip 100 144 10 63 1 trace % Hyp distribution 100% 43.8% NDa Root 100 70 8.9 30.4 4.1 12.2 % Hyp distribution 100% 43.4% 17.4% Coleoptile 100 200 10.5 104.7 5.1 12.3 % Hyp distribution 100% 52.3% 6.2% Black Mexican Cultures (11 d) 100 150 9.5 76 5 15 % Hyp distribution 100% 44% 10% ' recovery too low for accurate assay 63 ' D. BlackMexican Cell Wall Isodityrosine The HF -insoluble Black Mexican maize cell wall contained tyrosine, but no dityrosine or IDT, as assayed on a Hamilton PRP-1 column; however, an unknown "phenolic" compound eluted at 21.7 min, between dityrosine (20.7 min) and IDT ' (23.6 min). Figure 27a shows the Tyr (13.7 min), Dityr, and IDT standards, and Figure 27b shows the hydrolyzed HF -Insoluble maize cell wall on PRP-1. IDT, dityrosine and tyrosine standards combined with the I-IF—Insoluble wall hydrolysate and chromatographed on PRP-1 showed that the maize unknown eluted between dityrosine and IDT (data not shown). Table 15. Amino Acid Compositions of the HF-Deglycosylated Maize Cell Wall Amino Acid“ Intact Cell Wall HF-Soluble Wall HF—Insoluble Wall Hyp 1.1 3.5 trace Asx 10.4 8.5 10.6 Thr 5.1 6.0 5.3 Ser 6.9 9.8 6.3 Glx 9.3 8.7 10.8 Pro 3.7 3.5 4.7 Gly 10.7 11.8 10.8 Ala 10.6 12.8 10.8 Cys 0.0 0.0 0.3 Val 6.4 4.9 7.6 Met 1.7 0.2 0.1 Ilu 4.2 3.1 0.3 Leu 10.3 6.5 9.0 Tyr 1.9 1.1 2.0 Phe 4.0 6.1 4.8 Lys 6.2 2.9 7.0 His 2.1 8.6 2.6 Arg 4.7 4.7 6.0 ‘ Represented as Mole % 64 1.0 - a I- 1 .0 r b 9 E C E ?-' E c .,_ 1: V U V v- v- N N < < LWW I o. 0.0 " 1 1 1 1 0-0 " 1 1 1 1 10 20 30 4O 10 20 30 40 Minutes Minutes Figure 27. Assay of IDT in the Black Mexican HF-insoluble cell wall. a) Tyrosine, dityrosine and IDT standards. b) The HF-insoluble cell wall contains tyrosine and and unknown (?). E. Immunoblot Analysis of the Maize Cell Wall Anti-THRGP antibodies detected an immunoreactive protein in SDS-PAGE immunoblots of the HF-solubilized wall fractions of Sweet Sue root, root tip, and coleoptile: in each instance, the antibodies detected a major band that migrated with the same Mr as the dTHRGP (Figure 28). Preimmune control serum did not react with the blots. All cell wall preparations showed a Coomassie Blue-stained smear of HF -soluble cell wall proteins whose sizes on SDS-PAGE ranged from 200 kD to very small components that migrated with the marker dye (data not shown). 65 Am 285 a: “ooh 28 Av 283 Hook Am 28: 05328 02m 8on Bob 808mb =8, :8 oBEEiE 2: Es .6 a N 88: momma. é a. 8 88 885. Mo 838288 .8 8ng ImN I”? Q Iv mum—1.59 lwm Q0”— I... l lfifim ICON homwmww 0x DISCUSSION Monocots are, by comparison with dicots, relatively poor in hydroxyproline, although it has been clear for some time that monocot Hyp-containing proteins do exist, both in the grasses and other monocot families (van Etten et al., 1963; Lamport, 1965; Boundy et al., 1967; Burke et al., 1974). Much of this work implicitly assumes that easily soluble HRGPs correspond to arabinogalactan proteins, while the insoluble HRGPs correspond to extensin. The latter hypothesis is difficult to test. However, recent reinvestigation of the 'intact cell elution' technique (Lamport, 1965) showed that under optimal conditions, soluble monomeric extensin precursors to wall-bound extensin can be ionically desorbed directly from the cell surface of intact tomato cells grown in suspension culture (Smith et al., 1984). Smith et al. (1984 & 1986) and others have characterized soluble extensin monomers chemically (Chrispeels, 1969; Stuart & Varner, 1980; van Holst & Varner, 1984; Cassab et al., 1985), immunologically (Leach et al., 1982b; Kieliszewski & Lamport, 1986; Cassab & Varner, 1987), and electron microscopically (Heckman et al., 1988; Stafstrom & Staehelin, 1986; van Holst & Varner, 1984), and thereby provided. the tools to determine whether or not extensin occurred in graminaceous monocots, a question relevant to current ideas. about the control of cell extension (cf. oat coleoptiles) and the proposed 66 67 1 model for the primary'cell wall of dicots, which invokes an extensin 'weft' to mechanically couple the load-bearing microfibrillar polymer 'warp'-cellulose (Lamport & Epstein, 1983). First however, we must summarize the criteria which enable us to classify an HRGP as a member of the extensin glycoprotein family localized in the primary cell wall. These criteria involve primary, secondary and tertiary structure and, therefore, include posttranslational modifications (for the extensins this can account for more than 40% of the amino acid residues) by hydroxylation and glycosylation, which dramatically alter properties of the unadomed polypeptide backbone. Generally, extensins are defined as hydroxyproline-rich glycoproteins that are insolubilized in the cell walls of higher plants. They are basic, rodlike macromolecules with a polyproline-H helical conformation arising in part from the characteristic repetitive pentapeptide Ser-(Hyp),. Many of the Hyp residues are glycosylated by short oligoarabinosides, while the Ser residues are often galactosylated by a single residue. However, one must remember that these criteria are based on knowledge of a very few dicot extensins, from only three of which (tomato, Smith et al., 1986;, sugar beet, Li et al., 1990; melon, Esquerre-Tugaye & Lamport, 1979) do we have direct, rather than cDNA-derived primary sequence information. It would therefore be dangerous to elevate any single characteristic of a dicot extensin to the level of dogma and demand that all extensins subscribe to a pattern which might simply represent extensins from advanced herbaceous dicots. Thus, peptide sequences from sugar beet extensin (Li et al., 1989) p 68 (Figure 29) show that the Ser-Hyp-Hyp-Hyp-Hyp pentameric motif, common in other dicots, can no longer be considered a diagnostic sine qua non of extensin (Franssen et al., 1988; Hood et al., 1988; Kieliszewski & Lamport, 1988). In addition, we must also realize that fibrous proteins have their own rules which frequently differ from those formulated for globular proteins (Tiffany & Krimm, 1969; Doolittle, 1986), where folding is of paramount importance. 1. The Isolation and Characterization of Zea mays Extensins A. Elution of Intact Maize Cells and Preparation of HRGPs My data show the successful application of the intact cell elution technique to maize cell suspension cultures. Here an HRGP monocot/dicot difference occurs. In Hyp—rich cultures of dicots, like tomato, the pool of monomeric extensin peaks during rapid cell grth (Smith et al.,1984), while in cultured maize cells the elutable crude HRGP pool rises only slowly during rapid cell growth, and peaks at day 11 after subculture, long after the cells have (abruptly) ceased expansion growth (Figure 5)(Kieliszewski & Lamport, 1987), pointing to possible functional differences between dicot and monocot extensins. More than 30 proteins appeared in the crude HRGP eluate (judging by SDS-PAGE and staining with Coomassie Blue), but only four or five occurred as major components, of which at least three were HRGPs, one of them being unusually rich in threonine (hence a threonine- rich HRGP, or THRGP), and the other two rich in histidine (HHRGPs)(Tab1e 5). 69 ' Like dicot extensins, the THRGP and HHRGPs are basic11 proteins, the THRGP being rich in lysine and the HHRGPs rich in histidine. The THRGP and HHRGPs co-chromatograph on phosphocellulose, as both THRGP and HHRGP are positively charged at pH 3 (Figure 7). However, BioRex-70 exploits the pK.'s of lysine (PK. " 10.5) and histidine (PK. " 6) to effect a separation of THRGP from HHRGP as HHRGP is mainly uncharged at pH 7 and voids the Biorex-7O column, while the THRGP at pH 7 is highly basic and binds to the matrix (Figure 10). B. Amino Acid Compositions of THRGP and HHRGP \ Ten amino acids accounted for 98 mole% of the THRGP amino acid residues, being richest in threonine and hydroxyproline, each accounting for about 25 mole%, with a high proline, lysine, and serine content, and lesser but significant amounts of tyrosine, histidine, alanine, and valine (Table 5)(Kieliszewski & Lamport, 1987). The THRGP amino acid composition is consistent with the amino acid compositions of a putative THRGP cDNA clone recently isolated from maize coleoptile (Stiefel et al., 1988) and a related glycine-rich THRGP isolated from maize pericarp (Hood et al., 1988) (Table 16). Eleven amino acids accounted for 95 mole % of HHRGP amino acid residues. The HHRGPs are richer in hydroxyproline (" 34 mole%) than either the THRGP ("25 mole%) or tomato extensin P1 (”'32 mole%), they are histidine-rich like carrot extensin, and they are relatively rich in alanine ( ”9 mole%)(Tables 3 & 5)(Kieliszewski & Lamport, 1988). Such biased compositions are typical of ’1 In the wall microenvironment the pH “'3, judging by the pH of the growth medium, therefore the HHRGPs are positively charged. 7O HRGPs in general, and extensin in particular, although the threonine-rich feature of the THRGP and the alanine-rich feature of the HHRGPs are novel. Table 16. Amino Acid Compositions of Three Maize THRGPs Amino Acid THRGP PC-l‘ MC56 Clone” Hyp 24.8 21.9 Asx 0.7 2.1 0.4 Thr 25 .3 17.5 22.8 Ser 7.3 5.5 6.7 Glx 2.3 2.5 0.0 Pro 14.5 13.5 45.7 Gly 2.4 7.1 0.0 Ala 1.7 5 .2 2.2 Val 0.7 2.7 0.4 Cys 0.0 n.d. 0.4 Met 0.0 n.d. 0.4 Ilu 0.1 0.4 0.4 Leu 0.0 0.2 0.4 Tyr 3.7 4.6 6.4 Phe 0.1 0.1 0.1 His 2.4 3.6 0.7 Lys 13.5 11.3 11.2 Arg 0.1 0.7 0.7 " From Hood et al., 1988 b From Stiefel et al., 1988 C. Glycosylation Profiles of THRGP, HHRGP, and the Maize Cell Wall The THRGP, HHRGP and maize cell wall Hyp arabinoside profiles are consistent with typical extensin glycosylation patterns, and corroborate earlier work (Lamport & Miller, 1971) which showed a high proportion of nonglycosylated Hyp residues in the monocots (Tables 7 & 13)(Kieliszewski & Lamport, 1987). However, the absence of galactose from the THRGP, as well as its high threonine 71' content, distinguish it from dicot extensins (Table 6). Although the HHRGP Hyp arabinoside profile (Table 7) is consistent with dicot extensins, its high galactose content (37 mole% of the total sugars) is unique (Table 6), and considering the HHRGPs are rich in alanine, it suggests that they may be graminaceous arabinogalactan proteins (AGPs). However, the very large heterooligosaccharide characteristic of AGPs (an arabinogalactan O-linked to hydroxyproline; Pope, 1977) did not appear in the Hyp arabinoside profile of the HHRGPs. Furthermore, a 60 min partial acid hydrolysis removed all arabinose from the HHRGPs, but left the galactose residues virtually intact (Figure 12), indicating that galactose exists as pyranosides probably independent of the arabinosides, i.e. directly attached to an amino acid other than hydroxyproline. Thus, as serine and threonine are the prime candidates for O-glycosylation (possibly tyrosine), and judging by the molar ratios of galactose to either serine or threonine (5:1) (the molar ratio of GalzTyr is about 9:1), it is likely that polygalactose occurs on one or both of the HHRGPs. Although there has been no demonstration of polygalactosyl O-serine/threonine or arabinosyl-polygalactosyl serine/threonine as a component of glycoproteins, Desai et al. (1981) presented evidence that digalactosyl-serine occurs in a Hyp-rich lectin from Datum stramom’um. D. SDS-PAGE of THRGP and HHRGP The THRGP's status as a monomer is suggested by its behavior on Superose-6 gel filtration and SDS gel electrophoresis (Figure 11b). The THRGP migrated as a smear on SDS-PAGE with M, of 72-90 kD (Figure 14, lane 2), an overestimate judging by its contour length and glycosylation profile, and probably , 72 due, in part, to its cationic nature and glycosylation (Segrest et al., 1971). SDS- PAGE indicated the apparent M, of THRGP as 50 kD after deglycosylation (Figure 14, lane 3), or roughly twice that predicted from the related MC56 THRGP cDNA clone (Stiefel et al., 1988). Yet this anomalous M, agrees with the M, of a putative MC56 THRGP protein (after HF deglycosylation) isolated from maize cell walls and identified by immunoblotting by Stiefel et al. (1988). SDS- PAGE also overestimates the M, of deglycosylated tomato extensins P1 and P2 (Smith et al., 1984). Possibly SDS-PAGE does not provide a reliable estimate of molecular weight for deglycosylated HRGPs due their high content of secondary amino acids and charged amino acids which may interfere with their ability to bind SDS (Takano et al., 1988). Like dicot extensins, glycosylated HHRGP did not enter an SDS-PAGE gel. After deglycosylation HHRGP, which had appeared homogeneous by Superose-6 gel filtration (Figure 11a), resolved into two bands (68 & 70 kD) on SDS-PAGE (Figure 15, lane 2). Both bands are HHRGPs judging by their identical behaviour on two cation exchange columns and gel filtration (Figures 7, 10 (the void) & 11a), their simple amino acid composition, carbohydrate profile (Tables 5 & 6), and peptide amino acid compositions and sequences (Tables 11 & 12). E. Immunological Characterization of THRGP and HHRGP I recently raised and characterized antibodies against tomato extensin monomers P1 and P2, and against the protein backbone of these two monomers after stripping the carbohydrate off the proteins via anhydrous HF (Kieliszewski & Lamport, 1986). The two antibodies raised against the glycosylated extensins 73 I P1 and P2 cross-reacted with the THRGP and I-IHRGPs indicating common antigenic epitopes between the glycosylated tomato extensins and the glycosylated maize HRGPs (Figure 24). Some of the common epitopes probably involve the Hyp arabinosides, which are common to both maize and tomato HRGPs. Furthermore, antibodies raised against deglycosylated extensin P1 crossreacted 27% with the THRGP (Figure 24), indicating common antigenic epitopes between tomato extensin backbone and the THRGP protein backbone. Thus, the polyclonal antibodies raised against tomato extensins proclaim the THRGP and HHRGPs at least 'extensin-like' (Kieliszewski & Lamport, 1988) The THRGP is much more immunogenic than either tomato extensin P1 (and dP1) or P2 (and dP2), judging by the titer of the anti-THRGP antibodies (Figure 25): quantitative ELISAs using anti-THRGP antibodies showed a positive response of 1.2 Abs”, to 20 ng THRGP antigen, with antiserum dilutions of 1:2000; whereas a comparable reaction with anti-tomato extensin antibodies required dilutions of 1:200 to 1:800 and 200 ng antigen/microtiter well (Kieliszewski & Lamport, 1986; Kieliszewski et al., 1990). Judging from the high cross-reactivity (97%) of anti-THRGP antibodies (from two rabbits) with deglycosylated THRGP, the peptide backbone is highly antigenic, while the epitopes contributed by the hydroxyproline arabinosides are much less (Figure 26). The low cross-reactivities with other HRGP antigens, including glycosylated and deglycosylated HHRGP, confirm that the antibodies are quite specific for the THRGP backbone. Consistent with the cross-reactivities of anti- tomato extensin antibodies with THRGP, antigens tomato P1 and dP1 showed the 74 I most cross-reactivity with the anti-THRGP antibodies (Fig 26), again indicating common antigenic epitopes, and implying some homology between the maize THRGP and tomato P1. F. Transmission Electron Microscopy and Circular Dichroic Spectra of THRGP and HHRGP Transmission electron microscopy of low-angle rotary-shadowed material visualized THRGP and HHRGP molecules as flexuous rods (Figures 16a & b) similar to those of dicot extensins (van Holst & Varner, 1984; Stafstrom & Staehelin, 1986; Heckman et al., 1988) although somewhat shorter (70 nm contour length vs 80 nm for dicot extensins) (Kieliszewski & Lamport, 1988), but consistent with their elution position on Superose-6 gel filtration close to dicot extensin monomers. Furthermore, TEM visualization of the THRGP and HHRGPs, combined with their Hyp and Pro content of 39-41% suggested a polyproline-II conformation (3 residues /turn with 9.4 A pitch) similar to extensin. However, CD data indicate that the THRGP and dTHRGP exist in an "unordered" or "random coil" conformation (Table 8; Figure 17a), although the HHRGPs show a broad positive band at 210 nm indicating a more ordered structure, possibly having some polyproline II conformation (Table 8; Figure 17b)(Tiffany & Krimm, 1969). Although this lack of polyproline H conformation seemed anomalous in view of the elevated proline/hydroxyproline content of THRGP and HHRGP, the dispersion of proline and hydroxyproline residues (Tables 10 & 12) partially resolves the anomaly, as CD spectra of synthetic polypeptides show that nucleation of the polyproline H helix usually requires at least four contiguous proline residues 75 ‘ (Okabayashi & Isemura, 1968; Deber et al., 1970). Indeed, a tetrahydroxyproline block occurs only in the minor THRGP peptide TC1 (Table 10), in agreement with a single occurrence of tetraproline at the C-terminus of the cDNA clone MC56 (Stiefel et al., 1988)(cf Figure 30), while the HHRGPs contain repeating tri- hydroxyproline blocks (Table 12). The anomaly may also be a product of the poorly defined term "random coil" which clearly does not preclude the THRGP or HHRGP secondary structure deduced by TEM (Tiffany & Krimm, 1969), or the presence of secondary structure in other "random coil" proteins. For example there is even a report of a monoclonal antibody that can recognize a conformational epitope in a random coil protein (Saad et al., 1988). In contrast to dicot extensin, wherein the carbohydrate apparently helps to maintain the secondary polyproline-H structure (van Holst & Varner, 1984), the Hyp arabinoside moieties of the THRGP have little influence on the secondary structure of the THRGP, judging by its CD spectra before and after HF deglycosylation (Table 8; Figure 17a). The minimum at 205 nm increased slightly in its intensity (AE1min increased from -3 to -3.6) and was blue shifted by 3 nm to 202 nm (Kieliszewski et al. 1990). At this point, there is sufficient structural and chemical similarity with the extensins to consider the THRGP and HHRGPs as analogous proteins, although their lack of polyproline-H helix might argue against homology. G. dHHRGP Peptides Share Homology with P3-Type Dicot Extensins Chymotryptic digestion of dHHRGPs generated peptides, none of which voided the G-25 gel filtration column (Figure 22), hence the absence of a protease- 76 ' resistant coreand an essentially complete digestion. HPLC of the digestion produced a peptide map with 12 peptides (Figure 23), all of them containing Hyp, Ala and His (except HC7 & HC8, whose compositions and sequences are not yet determined) and reflecting the simple amino acid analyses of the proteins (Table 11). Thus both bands resolved on SDS-PAGE are HHRGPs. Although no tetra(hydroxy)proline occurs in any of the sequenced dHHRGP peptides, four peptides (Table 12) do share significant homology with the N- terminus of the P3-type extensin repetitive hexadecamer: Ser-Hyp-Hyp—Hyp-HypoSer-Hyp-Ser-Hyp—ij-Hyp—Hyp-Tyr-Tyr-Tyr-Lys (Table 4), deduced as follows: Chymotryptic peptide HC6 has only a His for Hyp substitution at position 5, and therefore shows 89% homology with the above hexadecamer terminus: Ser-Hyp-Hyp-Hyp-His-Ser-Hyp-Ser-Hyp-Gly. Peptides HC10, HC11 & HC12 each have one repeat of the decameric sequence: Ala-Hyp- Hyp-Hyp-His-Phe-Pro-Ser-Hyp-Hyp which is 88% homologous with HC6 (Ala for Ser at residue 1; Phe for Ser at residue 6, and unhydroxylated Pro at residue 7), and therefore 66% homology with the N-terminal portion of P3 type extensin domain. Thus, at least one of the HHRGPs has a (repeating) decamer sharing 66% homology with P3-type extensin, thereby defining the HHRGP(s) as members of the extensin family (Doolittle, 1981; Dayhoff et al., 1983; Marchelonis et al., 1984). One can then trace a possible divergence from a Type P3 domain through the HC6 sequence to the common repeat of HC10, HC11 & HC12 as follows: 77 . Ser-Hypryp-Hyp;Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-Tyr-Tyr-Tyr-Lys l Ser-Hyp-Hyp-Hyp-_I-§-Ser-Hyp-Ser-Hyp-Gly-... l l fla-Hyp-Hyp-Hyp-His-fihg-Pro-Ser-Hyp-Hyp-... Alternatively, the Ala-Hyp-Hyp-Hyp repeat might be a primitive feature, as HRGPs from Chlamydomonas are frequently rich in alanine, and a recently cloned cDNA for a Chlamydomonas cell wall HRGP demonstrates X-Pro-Pro-Pro repeats, although only one is Ala-Pro-Pro-Pro (U. Goodenough to D. Lamport, personal communication). Two intriguing sequences occur in the N-terminal portions of peptides HC10 & HC11 (Table 12). HC10 contains a 9 residue palindrome (residues 9-17) centered around an asparagine residue: Hyp-Hyp-Ala-Ala-_A;s_r_1-Ala-Ala-Hyp-Hyp, that is reminiscent of an alanine-rich 7 residue palindrome (of unknown function) that occurs in Chlamydomonas (U. Goodenough to D. Lamport, personal communication). The other unusual HHRGP sequence occurs in the N-terminal portion of HC11: Thr(Hyp)-Hyp-Hyp-His-His-His-Hyp-Hyp-Hyp-. The occurrence of 3 consecutive histidine residues is a rare event, as histidine is one of the four least frequently occurring amino acids (Doolittle, 1981), and may be involved in HHRGP function in the maize cell wall. Furthermore, if residue 11 of TC11 is hydroxyproline rather than threonine, the histidine triplet exists as the center of another palindrome: Hyp-Hyp-Hyp-His-His-His-Hyp-Hyp-Hyp. Thus palindromes 78 sandwiched between a decameric repeating motif may define at least one of the HHRGPs. Likewise, P3-type extensins from tomato, petunia and bean (Showalter & Rumeau, 1989) are defined by an 11-residue repeating palindrome: Hyp-Hyp-Hyp-Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp followed by a tetramer containing a tyrosine triplet: Tyr-Tyr-Tyr-Lys in which the outer two tyrosines are ultimately enzymically modified to form isodityrosine as an intramolecular crosslink (Epstein & Lamport, 1984)(Table 4). However, there is no evidence to suggest the three consecutive histidines of HC11 undergo any modification analogous to crosslinking. H. dTHRGP Peptides Share Homology with P1 and P3-Type Dicot Extensins Chymotryptic digestion of the dTHRGP generated peptides, none of which voided the G-25 gel filtration column, indicating the absence of a protease-resistant core and an essentially complete digestion (Figure 18). HPLC of the complete chymotryptic digest produced a simple peptide map with only 8 peptides (Fig 19). Despite the occurrence of only a single tetrahydroxyproline in the THRGP (Table 10, TC1), homology exists between the THRGP repetitive sequences (Table 10, TC5 and TC4) and tomato P1 extensin as follows: the decameric motif Ser-Hyp-Hyp-Hyp-Hyp-Thr-Hyp-Val-Tyr-Lys occurs as tryptic peptides from tomato P1 (Smith et al., 1986) and tobacco (M. Kieliszewski and D. Lamport, unpublished data), as cDNA and DNA sequences from tobacco, Petunia and carrot (Chen & Varner, 1985a & b; Showalter & Rumeau, 1989), and also as a modified peptide sequence in sugar beet (Figure 29)(Li et al., 1990). However, the modified decamer is also discernible in THRGP 79 chymotryptic peptides TC5 and TC4 as the repetitive motif: Ser-Hyp-Lys-Pro- Hyp-Thr-Pro-Lys which differs from the tomato P1 decamer essentially by a Lys for Hyp substitution at residue 3 and a Val-Tyr deletion at residues 8 and 9 (Figure 29)(Kieliszewski et al., 1990). The corresponding derived sequence occurs seven times in cDNA clone MC56 (Stiefel et al., 1988) (Figure 30) pointing to m 8 Ser Hyp M Ser Hyp ‘Igglatgg Ser Hyp Mg Ser Hyp Maize: Ser Hyp Hyp [X] Hyp Hyp Thr Hyp Hyp Lys Lys [X] = m = [Z] = Hyp Hyp [Y] Thr Hyp Hyp Thr Pro Hyp Thr Pro Hyp [Z] Thr Hyp Hyp Hyp Pro Pro Val His Glu Tyr Pro Val Tyr Lys Val Tyr Lys Val Tyr Lys ------------ Lys ------------ Lys Val Lys Pro Tyr His Pro Ala Thr Lys Pro Pro Figure 29. The decameric motif of P1-type extensins. significant homology with the isolated THRGP glycoprotein. Futhermore, like a typical dicot extensin, MC56 THRGP is organized into three distinct domains: the C and N-termini and the central domain composed of the major repeating motifs (Figure 30). Further homology with dicot extensins from tomato (Smith et al., 1986), petunia and bean (Showalter & Rumeau, 1989) occurs in chymotryptic 80 peptide TC1 (Table 10) and the corresponding C-terminal sequence of clone MC56 (Figure 30). The TC1 octapeptide Thr-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-Tyr (Table 10), and MC56 cDNA sequence Thr-Pro-Ser-Pro-Pro-Pro-Pro-Tyr-Tyr (Stiefel et al., 1988) are homologous with the dicot extensin Type 3 domain: Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-Tyr-Tyr (except for a Thr for Ser substitution in position 1). The single substitution in maize of Thr for the dicot Ser in the first position is effected by a single nucleotide base change. Thus the maize THRGP is a fusion of two dicot general extensin types into one protein: a modified tomato Type 1 extensin peptide backbone, with a C-terminal tail homologous with the repeating peptide of tomato Type 3 extensin. From both peptide and cDNA sequences I conclude therefore, that in maize a Ser-Hyp-Lys-Pro-Hyp pentamer replaces the dicot Ser-(Hyp)4 pentamer throughout the THRGP molecule except for a single occurrence of Ser-(Hyp)4 in peptide TC1 (Table 10) which from MC56 (Stiefel et al., 1988) corresponds to the C-terminal tail (Figure 30). Thus, a single Ser-(Hyp), in an advanced graminaceous monocot is probably an evolutionary remnant, hence the C-terminal tail may, like some others, represent a vestigial condition. This retention of 70% homology (allowing for a Val-Tyr "gap") in a repetitive motif, and almost 90% homology in the C-terminal tail establishes membership of maize THRGP in the extensin family (Doolittle, 1981), albeit a member with some unique characteristics, especially a distinctive and quite selective proline hydroxylation pattern with rather subtle determinants. 81 Figure 30. Proposed primary amino acid sequence of a THRGP encoded by cDNA clone MC56 isolated from maize coleoptile. Peptides common to the Black Mexican THRGP and MC56-THRGP are labeled on the right as TC1, TC5, TC6 and TC7. Tryptic peptide TC1 is underlined. Residue 160 may be Hyp rather than Pro. 82 1 ThrHypHypThrTyr 21 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 22 42 Thr Hyp Ser Hyp Lys Pro Hyp Ala Ser Lys Pro Pro Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr 58 Thr ij Ser Hyp Lys Pro l-Iyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 59 79 Thr Hyp Ser Hyp Lys Pro Hyp Ala Thr Lys Pro Pro Thr Pro Lys Pro Thr Hyp I-Lyp Thr Tyr 95 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 113 116 Thr Hyp Ser Hyp Lys Pro Hyp Ala Thr Lys Pro Pro Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr 117 132 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 148 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 167 Thr Hyp Ser Hyp Lys Pro Hyp ------ Thr His Pro ----- Thr (Pro) Lys Pro Thr Hyp Hyp Thr Tyr (TC7) 183 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 199 Thr Hyp Ser Hyp Lys Pro Hyp Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr (TC5) 220 Thr Hyp Ser Hyp Lys Pro Hyp Ala Thr Lys Pro Pro Thr Pro Lys Pro Thr Hyp Hyp Thr Tyr . 234 _ . Thr Hyp Thr Hyp Lys Pro Hyp Ala Thr Lys Pro Pro Thr Tyr (TC6) 236 - 251 Thr Hyp Thr Hyp Hyp Val Ser His Thr Hyp Ser Hyp Hyp Hyp Hyp Tyr Tyr - (TC1) 83 I. The Specificity of Prolvlhydroxylase Combining the THRGP peptide and MC56 cDNA sequence information (Table 10 & Figure 30) we see that the MC56 THRGP contains about 121 proline residues of which approximately 74 are candidates for posttranslational hydroxylation and glycosylation. The major repetitive chymotryptic peptide TC5 contains five Hyp residues, at least two being glycosylated (on average) based on the hydroxyproline glycoside profile showing 48% Hyp glycosylation of the intact THRGP (Kieliszewski & Lamport, 1987). Thus TC5 corresponds to 14 potential glycosylated domains which alternate regularly with non-glycosylated domains (Figures 30 & 31), the latter occurring mainly as the pentapeptide palindrome Thr-Pro-Lys-Pro-Thr. But what determines which proline residues are hydroxylated and which are not? Despite the reported preference of plant prolylhydroxylases for (artificial) substrates with a polyproline-II secondary structure (Tanaka et al., 1981), the CD data indicate that the polyproline-II conformation may not be a requirement for hydroxylation. Because two thirds of the THRGP proline residues are hydroxylated (Table 5) with a high degree of specificity (Table 10), a hydroxylation code seems possible, although it is not immediately self-evident. "Windows" of one, two, or three contiguous residues do not yield an exclusive hydroxylation code; for example, Thr-Pro-Lys is generally unhydroxylated although Thr-Hyp-Lys also occurs in peptide TC6 (Table 10 and residues 223 to 225 in Figure 30). Assuming the predicted hydroxylations of MC56 THRGP are correct, there are five occurrences of Lys- Pro-Pro-Thr but seven Lys-Pro-Hyp-Thr (See Figure 30). Interestingly Lys-Pro 84 —- _ 1 S L 113 S S L 117 L236 251 w vi 7% w MIN-00°“ L _. 2 Figure 31. Schematic block diagram for a THRGP based on combined direct peptide sequencing and cDNA clone MC56 (Stiefel et al., 1988). The blocks represent the repetitive eleven-residue sequence: Hyp-Hyp-Thr-Tyr-Thr-Hyp-Ser/Thr-Hyp-Lys-Pro-Hyp in which 2 or 3 of the 5 Hyp residues are glycoslated. These domains are separated by short (8) or long (L) non-glycosylated domains. The last block corresponds to the C-terminus of the molecule. Superscripts correspond to the residue numbers of Figure 30. Subscripts denote the number of pattern repeat units. and His-Pro are probably never hydroxylated here or in any of the known dicot extensin peptides. However, a window of four contiguous residues does identify four exclusive sequences (Figure 32) which account for every hydroxyproline residue in the THRGP molecule, except for the final Pro residue in TC4. Each window contains two candidates for hydroxylation: I. X-Pro-Pro-X II. X-Pro-X-Pro III. Pro-X-X—Pro and IV. Pro-Pro-Pro-Pro, where X refers to specific residues (Table 10, Figures 30 & 32). A single prolylhydroxylase would have to recognize subtle differences in peptide conformation; for example, the major nonhydroxylated THRGP domain occurs as the repetitive pentapeptide palindrome: Thr-Pro-Lys-Pro-Thr whose conformation might not allow hydroxylation of its Thr-Pro, while Lys-Pro is never hydroxylated. 85 Hydroxylation Hydroxylaled Non-hydroxylated Substrate Product Related Sequences Sequences Sequences Window: 1. Thr-Pro-Pro-Thr ----> Thr-Hyp-Hyp—Thr ll. Thr-Pro-Ser-Pro ---‘> Thr—Hyp-Ser-Hyp lll. Pro-Lys-Pro-Pro ----> Hyp-Lys—Pro-Hyp Pro-Lys-Pro-Thr Hyp-Lys-Pro-Pro Thr-Lys-Pro-Pro IV. Pro-Pro-Pro-Pro ",“> Hyp-Hyp-Hyp-Hyp Figure 32. Suggested four-residue windows recognized by THRGP prolyl hydroxylase. An alternative hypothesis postulates the existence of three prolyl hydroxylase isozymes, each specific for its own tetrapeptide window, I, II, or 111 (Figure 32) and requiring specific non-proline residues within that window, while IV might be a special case involving recognition by I and III, which is not unreasonable considering the multimeric character of the enzyme (Bolwell et al., 1985) and that different catalytic subunits occur in animal systems (Kivirikko, 1989). Compared to the THRGP hydroxylation profile, the HHRGP peptide hydroxylation profile is straightforward: all proline residues are hydroxylated with two general exceptions. First, Phe-Pro is not hydroxylated. Thus the bulky side chains of Phe-Pro, Lys—Pro and His-Pro in the HHRGP, THRGP and dicot extensins (Smith et al., 1986) may sterically hinder prolyl hydroxylase. Secondly, 86 the HHRGP sequences contain two instances of the THRGP hydroxylation window X-Pro-X-Pro, but in only one sequence are the prolines hydroxylated: Ala-Pro-Ala-Pro occurs in HC1,and Ser-Hyp-Ser-Hyp in HC6. Thus X must be a specific residue for hydroxylation to occur. Alternatively, the peptide conformation surrounding the sequence may dictate hydroxylation. Like the THRGP, a definite polyproline-H conformation is apparently not required for extensive hydroxylation of HHRGP proline residues“. This implies that the use of natural rather than artificial substrates to assay prolyl hydroxylase(s) would facilitate its isolation and characterization in higher plants. J. Further Comparison of Black Mexican THRGP and MC56 THRGP Further comparison of the Black Mexican THRGP with MC56 THRGP clone (Stiefel et al., 1988) shows that some differences exist between the two THRGPs: MC56 THRGP lacks at least 2 small minor peptides, TC2: Thr-Hyp-Ser-Hyp-Lys-Pro-Thr-Hyp-Hyp-Thr-Tyr, and TC4: Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Hyp-Lys-Pro-Pro..., which occur in the Black Mexican THRGP (Table 10), while a major 21-residue peptide(s) of MC56 THRGP is missing from the Black Mexican (the peptide represented by residues 22 to 42 in Figure 30: Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Ala-Ser-Lys-Pro-Pro-Thr- Pro-Lys-Pro-Thr-Hyp-Hyp-Thr-Tyr). However, Black Mexican peptide TC8 (Table 9) has an amino acid composition resembling that of the missing peptide, therefore the peptide may be present in lesser amounts in the Black Mexican THRGP, or ‘2 This assumes the THRGP and HHRGP in vitro random coil conformations, as assayedby CD, are identical with their in viva conformations. 87 in view ‘of its more hydrophobic nature, it is retained on the reversed phase column used for peptide separations. Alternatively, the two THRGPs may be encoded by separate but homologous genes. K. Tryptic Hydrolysis of Lysylproline in dTHRGP Although trypsin was the obvious choice for peptide mapping of the THRGP, which contained 12 mole% lysine, I recovered few peptides after tryptic digestion of THRGP (Figure 20). 'A peptide initially purified from the tryptic digest showed an N-terminal proline residue (Peptide TT1 of Table 10): Pro-Thr-Hyp-Hyp—Thr-Tyr-Thr-Hyp—Ser-Hyp—Lys-Pro-Hyp-Thr-Pro-Lys which was originally dismissed as an artifact, because it is generally believed that Lys- Pro bonds are not hydrolyzed (Hill, 1965; Kasper, 1975). However, chymotryptic digestion of the THRGP yielded the major peptide TC5 which overlapped with the tryptic peptide TT1, and contained two Lys-Pro bonds, thus strengthening the suspicion that trypsin cleaved a specific Lys-Pro bond. Further tryptic cleavage of chymotryptic peptide TC5 showed that only the latter of the two Lys-Pro linkages, Hyp-Lys-Pro-Hyp and Pro-Lys-Pro-Thr, was labile (Table 17). One other example of a trypsin-sensitive Lys-Pro was reported in proline-rich proteins from human saliva (Table 17) (Wong et al., 1979; Wong & Bennick, 1980; Schlesinger & Hay, 1986). Thus for Lys-Pro cleavage, an extended polypeptide backbone (characteristic of a proline-rich polypeptide) is probably a necessary, but not sufficient, condition, because trypsin-resistant Lys-Pro bonds occur in the same proteins and also in another HRGP isolated from tomato (Smith et al., 1986). Inspection of the available sequences around susceptible and resistant Lys-Pro ----- 88 bonds (Table ‘17) suggests a second condition for cleavage, namely, backbone flexibility flanking the susceptible Lys-Pro bonds. Table 17 shows that a susceptible Lys-Pro bond occurs in sequences which have N-terminal flexibility at residue -3 (where Lys-Pro = -1 and + 1) and C-terminal flexibility at residue + 2. Peptides with resistant Lys-Pro fail to meet the required flanking flexibility, because proline or hydroxyproline residues at -3 or +2 constrain rotation around the phi (peptide N to alpha C) bond. Thus the seemingly antithetical requirements for specific flexibility, together with the highly constrained but extended backbone conformation of a proline-rich protein, may account for the rarity of Lys-Pro cleavage (Kieliszewski et al., 1989). L. Some Evolutionary Implications of Graminaceous Extensins Characterization of the graminaceous THRGP & HHRGP extensins also have some interesting evolutionary implications, as we can now directly relate three widely different genera to two repetitive elements (typified by the tomato P1 and P3 decamers) which contain tetrahydroxyproline (tomato) or variants of tetra(hydroxy)proline: i) split by an insertion sequence (sugar beet) or ii) a Lys or His for Hyp substitution (maize)(Figure 29). While it is not clear which condition is primitive, the Hyp-Lys-Pro-Hyp and Hyp-Hyp-Hyp-His of maize is an advanced feature judging from the relatively recent origin of the graminaceous monocots, and the single vestigial Ser-(Hyp)4 of maize THRGP. On the other hand split tetrahydroxyproline (or proline) is widespread in advanced dicots (Franssen et al., 1987; Hong et al., 1987), but also occurs in primitive dicots represented by the chenopod, sugar beet (Li et al., 1990). Thus one possible . 89 Table 17. Trypsin Labile and Trypsin Stable Lys-Pro Bonds Threonine Hydroxyproline-Rich Glycoprotein: Trypsin labile: Pro-Hyp-Thr-Pro-Lys-Pro-Thr-Hyp-Hyp-Thr Trypsin stable: Thr-Hyp-Ser-Hyp-Lys-Pro-Hyp-Thr-Pro-Lys Position -5 -4 -3 -2 -1 +1 +2 +3 +4 +5 Tomato Hydroxyproline-Rich Glycoprotein": Trypsin stable: Hyp-Hyp-Hyp-Val-Lys-Pro-Tyr-His-Pro-Thr Position -5 -4 -3 -2 -1 +1 +2 +3 +4 +5 Praline-Rich Phosphoproteins": Trypsin labile: Pro-Pro-Gln-Gly-Lys-Pro-G1n-Gly-Pro-Pro Trypsin stable: Pro-Pro-Pro-Gly-Lys-Pro-Gln-Gly-Pro-Pro Position -5 -4 -3 -2 -1 +1 +2 +3 +4 +5 * From Smith et al., 1986 ** From Wong et al., 1979; Wong & Bennick, 1980; Schlesinger & Hay, 1986 evolutionary progression is: Hyp-Hyp-[X]-Hyp-Hyp ---> Hyp-Hyp-Hyp-Hyp ---> Hyp-Lys-Pro-Hyp and Hyp-Hyp-Hyp-His. However, divergence from tetrahydroxyproline seems more likely, especially as sequences of HRGPs from Volvox and Chlamydomonas reinhardtii show repeating clusters of X-Pro-Pro-Pro (Chlamydomonas cDNA clone, U. Goodenough, personal communication) and Ser-(Hyp)>6 (Volvox; Schlipfenbacher et al., 1986). Resolution of this problem clearly requires sequence information from non-graminaceous monocots, primitive dicots and pre- angiosperms. Because the wall is so intimately involved in the creation of plant form, the evolution of structural wall proteins, such as extensin, must be coupled to and should therefore parallel, the evolution of structures per se. 90 Finally, there is the question of extensin function. While structural in a general sense, no discrete function is assigned to any extensin, therefore it is not possible to discuss functional homologies between dicot and monocot extensins; however, numerous clues point to fundamental roles for dicot extensins in growth, development and stress response (Showalter & Varner, 1989; Showalter & Rumeau, 1989). Isolation of multiple extensins (Smith et al., 1984; Smith et al., 1986) and extensin cDNAs (Corbin et al., 1987) shows that a small glycoprotein family exists, which, by analogy with collagen, may be tailored to the tissue. For example, of the twelve collagen types, four occur exclusively in cartilage (Piez, 1987). Thus, a systematic approach to function demands a classification of extensin types, starting with the most highly expressed, best-characterized and most easily recognized. Two easily recognizable decameric motifs (and variants) identify P1 and P3 type extensins which include the maize THRGP and HHRGPs, whose functions probably differ from dicot P1 and P3. For example, the maize HRGPs are expressed at a significantly lower level than dicot extensins. Furthermore, HF treatment of salt-washed walls solubilized the bulk of the HRGPs (Figure 28 & Table 14); this is quite unlike the dicots where covalently bound extensin is generally HF-insoluble (Smith et al., 1984). Interestingly, MC56 mRNA is actively expressed in the root tip and coleoptile, although much less in the root, suggesting a possible tissue specificity for the HRGP (Stiefel et al., 1988); but this is only apparent, as significant amounts of THRGP occurred in all maize tissues examined, notably, 4 day coleoptile, root, root tip cell walls (Table 14), and also in maize pericarp from which a related glycine-rich THRGP was recently isolated 91 (Hood et al., 1988). This was corroborated by the cell wall hydroxyproline content for each tissue (Table 14, Hood et al., 1988). The presence of THRGP protein in the virtual absence of THRGP mRNA simply confirms that structural cell wall proteins do not turnover. Other features, such as the exceedingly high structural periodicity which even includes a repetitive seven residue palindrome in the THRGP, and at least one (repetitive?) palindrome in one or both HHRGPs, point to a special function for the maize HRGPs, perhaps involving self-assembly by specific interactions (e.g. the e-amino groups of lysine) with the major acidic polysaccharide components which are glucuronoarabinoxylans in the graminaceous monocots rather than the rhamnopolygalacturonans characteristic of the dicots and non-graminaceous monocots (Figure 4)(Burke et al., 1974; Darvill et al., 1980b; Carpita, 1983, 1984, 1985; Jarvis et al., 1988). Also, the unusual lability of the palindromic lysyl residue in TC5 (Table 17) could imply a cleavage site possibly enabling cell expansion by relaxing the network. Significantly, the THRGP and HHRGPs apparently do not contain potential IDT-forming sequences like those that occur in dicot extensins. For example, neither Tyr-Tyr-Tyr-Lys nor Tyr-Lys- Tyr-Lys occurred in any sequenced maize peptide (Tables 10 & 12). Furthermore the tyrosine-containing putative intermolecular crosslink sequence of P1 (the Val- Lys-Pro-Tyr-His-Pro insertion sequence of Figure 29) is absent from the maize HRGP sequences. And although THRGP contains a version of the P1 type extensin decamer Ser-Hyp-Hyp-Hyp-Hyp-Thr-Hyp-Val-Tyr-Lys, the deletion of Val-Tyr in the THRGP eliminates a putative intermolecular IDT crosslink site, which may explain its solubility in HF. Finally, THRGP structural periodicity 92 involving regularly alternating glycosylated and non-glycosylated regions (Figure 32), may, as previously suggested, be related to the insertion mechanism for a transmural protein whose reptation into the wall would be aided by the glycosylated "thread" of a molecular screw (Lamport, 1989). H. The Maize Cell Wall Primary cell walls of higher plants contain both ionically-bound and covalently- bound protein components, the bulk of which consists of structural glycoproteins. In dicots, many of these proteins are rich in hydroxyproline, most of it firmly associated with the wall matrix; however, the cell walls of the graminaceous monocots are Hyp-poor which implies that the graminaceous walls contain significantly less structural protein than dicots (Carpita & Kanabus, 1988), or that structural protein alternatives to extensin occur in the cell walls of some higher plants. Although the maize cell wall is like the dicot wall in that it accounts for 40% of the cell and is about 10% protein, judging by protein recoveries from amino acid analyses, it is only 0.7-0.2% hydroxyproline on a dry weight basis (vs. 1-2% in dicots)(Figure 4). Furthermore, another graminaceous monocot/dicot difference shows up on treatment of the wall with anhydrous HF. Although most of the wall- bound dicot extensin remains insoluble in HF, suggesting covalent extensin-extensin crosslinks, possibly by IDT, as a mechanism for insolubilization of extensin monomers into the dicot cell wall, the maize cell wall generally retains little Hyp after treatment with HF (Tables 14 & 15), and the HF-insoluble wall contains no 93 IDT (Figure 27b). Clearly, if there is an HRGP network in muro, then HF cleaves the intermolecular crosslinks, presumably because they differ from the dicot crosslinks. Nevertheless, an HF-insoluble residual wall protein remains that accounts for the bulk of the wall protein, and is therefore probably structural rather than enzymic protein. Furthermore, fractionation of maize cell wall hydrolysates yielded a UV-absorbing peak eluting (from the PRP-1 column; Figure 27b) between dityrosine and IDT, suggesting another tyrosine derivative (assuming it is indeed an amino acid), and raising the possibility of another protein crosslink. Clearly, extensin is not the only structural cell wall protein and any cell wall model must take that into account. Thus the framework of the warp-weft model could be the same but with some differences in detail. These important monocot/dicot differences could well reflect the essential dichotomy between these two groups of Angiosperms and their growth habit, which especially in the dicots, relies on turgidity and for support, while silica plays a large role in the grasses (Stebbins, 1974). This together with a radically different arrangement of meristems (Table 18) may reflect a possible fundamental difference in primary wall organization (in addition to the switch from pectin to glucuronoarabinoxylans) between dicots and graminaceous monocots, which diverged > 150 million years ago. Nevertheless, it is conceivable that some generally accepted monocot/dicot differences (Table 18) may not hold in some closely related groups. Stebbins (1974) considers the Chenopodiaceae as fairly close to the monocot line of evolution, noting that: "the first-formed vascular bundles may either form a circle of widely separated units or be scattered through the stem, giving a superficial 94 Table 18. Six Key Structural Characteristics of Dicots and Monocots' Dicots Monocots Leaf Veins Net-like Parallel Cotyledons two one Flower Parts In multiples of 4 or 5 In multiples of 3 Vascular Bundles Cylindrical arrangement Scattered Vascular Cambium Produced No Vascular Cambium Roots Tap roots No tap root ' cf. Figure 4 for a cell wall comparison of dicots vs. graminaceous monocots resemblance to monocotyledons." Furthermore, the remarkable resemblance between the amino acid profiles of sugar beet and maize cell walls (both from cultured cells)(Table 19) seem highly significant rather than superficial, especially as the hydroxyproline arabinoside profiles of the two walls are also similar (Li et al., 1990). If we assume that extracellular matrices are a priori network structures, then new "Hyp-less" structural proteins in monocots raise questions about the sort of network these structural proteins may create: the kind of crosslinks involved; how they are formed; and of course the role played by small amounts of HRGPs. 95 Table 19. Amino Acid Compositions of Black Mexican Maize and Sugar Beet Wall Amino Acid Maize Beet' Hyp 1.1 0.4 Asx 10.4 12.0 Thr 5.3 5.1 Ser 6.9 7.3 Glx 9.3 12.0 Pro 3.7 5.1 Gly 10.7 9.5 Ala 10.6 8.7 Cys 0.0 0.0 Val 6.4 6.5 Met 1.7 1.6 Ilu 4.2 4.8 Leu 10.3 9.7 Tyr 1.9 1.3 Phe 4.0 3.8 Lys 7.0 7.1 His 2.1 2.2 Arg 4.7 3.3 ' From Li et al., 1990 IDEAS FOR FUTURE WORK13 To ascribe a precise function to purely structural proteins, which lack enzymic activity, demands several different approaches; especially as the existence of several extensin types (in tomato, Pla, P1b, P2 & P3, Smith et al., 1986; and in maize, THRGP and multiple HHRGPs) probably reflects a diverse functionality. For example, the analogous Hyp-rich structural (glyco)protein of animal extracellular matrices, collagen (Types I-XII), serves very different functions which, according to Eyre (1980), include i. ropes (tendons and ligaments) ii. woven sheets (skin and facia) iii. filtration membranes (glomeruli) iv. supporting skeleton reinforced with mineral salts (bone and dentin) v. tissue organization and vi. mediation of the interactions between specific cell layers (Bornstein & Sage,1980) A few of the different approaches toward determining extensin function might involve: the determination of in situ tissue distributions of particular extensin types, manipulation of extensin's post-translational modifications, in vitro binding experiments involving particular extensins and other wall polymers (e.g. GAX or pectins), isolation of extensin/polysaccharide heteropolymers that interact in viva, 1’ The data presented in this section is very preliminary, therefore it is not presented in the Results section of this thesis. 96 97 crystallization. of extensins for X—ray diffraction, the isolation of extensin mutants, and finally, a comparison of diverse species to determine which features are rigorously conserved (e.g. Hyp-arabinosides, P1 and P3—type domains). But first we need a precise structural model of the THRGP, HHRGPs, and other structural wall proteins. This entails a detailed chemical and structural characterization of these proteins coupled with macromolecular modeling experiments in order to define domains that might confer function. I. Characterization of Maize Hyp-Containing Wall Proteins So far, the two HHRGPs have been characterized as one protein because they co-chromatograph on the ion exchangers Cellex-P, BioRex-7O and Superose- 6 gel filtration. One approach to characterizing the individual HHRGP polypeptide backbones” is to isolate and characterize cDNA clones (via oligonucleotide probes based on dHHRGP peptide sequences or antibodies raised against the dHHRGPs) from a Black Mexican Maize cDNA library constructed recently for me by Stratagene. Sequences of the HHRGP cDNA clones combined with the HHRGP peptide sequences to identify Hyp residues will allow the determination of the complete primary amino acid sequence of the HHRGPs (this is virtually impossible by peptide sequencing alone because of the repetitive nature of the proteins). Clones will also detail the differences in the HHRGP polypeptide backbones, thus providing clues for separating the intact HHRGPs from each other “ In vitro translation of Black Mexican mRNAs (using rabbit reticulocyte lysate) yielded two proteins (with M,~ 68 & 70 kD) identified by immunoblotting using anti-dHHRGP antibodies; thus the differences between the HHRGPs are in the polypeptide backbones, rather than post-translational modifications. 98 for further chemical characterization as individual proteins. Separation of the two HHRGPs is desirable (but not absolutely necessary) before characterization of their carbohydrate components, especially the galactose moieties. Once separated, mild acid treatment of each HHRGP at pH 1 for 1 h at 100° C will remove all arabinosides on the hydroxyproline hydroxyl groups, leaving galactopyranosyl residues mainly intact“ (Lamport et al., 1973). Then NaOH/borohydride catalyzed B-elimination, will yield alanine from O-glycosyl serine or a-aminobutenoic acid from O-glycosyl-threonine (Aminoff et al., 1980). The eliminated carbohydrate moieties can be detected by fluorometry after end labeling via pyridinylation with 2-aminopyridine and detection and characterization after HPLC by the methods of Maness & Mort (1989) or Seto & Shinohara (1989). As well as the THRGP and the two HHRGPs already characterized from the Black Mexican cell suspensions, the crude eluate also contains at least two other Hyp-containing proteins that are candidates for characterization; One is another HHRGP, but more basic than the other two judging by its elution on Cellex-P in CP3 (Figures 7 & 9) and a preliminary amino acid analysis. The other Hyp- containing protein in the crude eluate is about 4 mole% Hyp and rich in Ser, Gly and Ala (I designated it an alanine-rich glycoprotein, or ARGP)(Figure 8c), and separates from THRGP only during gel filtration on Superose-6 (Figure 11). I have purified this the ARGP to constant composition and have several mg set 1‘ In dicot extensins, the arabino-oligosides ionize at high pH and hence electrostatically shield the galactosylserine linkage from iii-elimination. Therefore the arabinosides must first be removed from the protein before proceeding. 99 aside for future characterization (including peptide mapping and sequence analysis). Others have reported Hyp-poor proteins as wall components (Kimmins & Brown, 1975), however, because the small amounts of Hyp in these proteins may result from contamination with Hyp-rich proteins, proof that such proteins exist requires the isolation of Hyp-poor peptides. A complete structural characterization of the THRGP and the HHRGPs also requires the assignment of carbohydrate to particular amino acid residues. Drs. Dan Kassel and Klaus Biemann (M.I.T.) have recently offered to do GC—Mass Spectroscopy of THRGP and HHRGP glycopeptides in order to identify the exact locations and nature of the glycosides on the peptide backbones. II. The Major Protein Component (non-HRGP) of the Maize Cell Wall Unlike the major protein component of many dicot walls, the major protein component of the maize cell wall is not an HRGP. Nor is this component extremely glycine-rich (i.e. > 60 mole%), and therefore it is not directly comparable to glycine-rich wall proteins such as those associated with seed coat walls (Varner & Cassab, 1986) or the glycine-rich proteins from petunia (see Appendix B)(Condit & Meagher, 1986, 1987) and bean (Keller et al.,1989 a,b). I have isolated three peptides from the firmly bound non-HRGP protein component of the wall. These peptides contain no Hyp, and together with amino acid analyses of the maize cell wall confirm that the major structural protein component of the maize cell wall is not an HRGP (Table 20). My goal is to purify and characterize a soluble precursor to the insoluble wall-bound protein, by analogy with work on dicot extensins from tomato, sycamore-maple, tobacco, sugar beet and maize. This 100 assumes that, like extensin, an elutable pool of soluble precursors to the insoluble protein exists in the maize cell wall. This is a reasonable assumption in the light of current views of the wall as consisting of subunits which are assembled in muro. As the the HF-insoluble wall protein residue contains no obvious "marker" like the hydroxyproline of extensin, how does one unequivocally identify a protein that has no known enzymic activity and apparently no unusual component? Peptide sequences of seven or more residues are generally considered as statistically unique and therefore provide absolute identification. I have two strategies for purification and characterization of the major maize wall protein component(s). First, fractionate tryptic or chymotryptic digests of all eluted proteins (after HF deglycosylation) and sequence peptides to identify peptides homologous to those isolated from wall digests. This "brute force" approach virturally assures results, but is labor intensive. A second potentially quicker method is to raise antibodies against major wall peptides and use the IgGs to identify a precursor via immunoblot analysis, immunoprecipitation, and / or immunoaffinity chromatography. Because trypsin solubilizes about 30% of the HF-insoluble wall residue (yielding at least three major peptides > 20 residues, with compositions which are consistent with the overall composition of the insoluble protein) antibodies raised against the tryptic peptides will facilitate identification of the soluble "precursor" non-HRGP in salt eluates of intact cells. This approach assumes the wall peptides are antigenic and the antibodies specific. 101 Table 20. Amino Acid Compositions of the Black Mexican Maize HF-Insoluble Cell Wall and Tryptic Peptides Isolated from the HF-Insoluble Wall Amino HF-Insoluble Acid Cell Wall Peptide 1 Peptide 2 Peptide 3 Hyp trace 0.0 0.0 0.0 Asx 10.6 8.0 8.1 13.2 Thr 5.3 7.3 3.4 5.6 Ser 6.8 15.5 9.0 9.9 Glx 11.3 12.1 7.0 12.9 Pro 4.7 3.4 6.3 2.2 Gly 9.6 17.6 14.9 13.3 Ala 10.4 4.0 7.3 9.0 Cys 0.0 0.0 0.0 0.0 Val 7.2 9.0 18.4 9.8 Met 1.5 0.0 0.0 0.0 Ilu 3.5 1.5 2.8 1.0 Leu 8.8 5.1 4.5 2.6 Tyr 1.9 2.8 0.0 0.8 Phe 3.9 3.1 1.4 1.2 His 2.5 1.4 3.2 3.0 Lys 6.4 3.8 8.6 10.4 Arg 5.6 5.4 5.1 5.0 Part of the characterization of the maize major structural (non-HRGP) wall protein is identification of crosslink amino acids. The presence of an "unknown phenolic" in the HF-insoluble maize wall hydrolysate (Figures 27 a & b) raises the possibility of another crosslink, possibly an amino acid, that may be analogous to IDT or dityrosine. HPLC of wall hydrolysates will separate the unknown from other wall components, after which it can be characterized (cf. Fry, 1982). Another approach to defining roles for the maize HRGPs and other structural proteins is to attempt to localize wall proteins to their respective tissues. Are they restricted? If so, to what tissues? And what properties do they impart to those 102 tissues? A first step is to achieve specific cytolocalization. HI. Immunocytochemical Localization of Maize HRGPs In 1985, I raised and characterized antibodies against two dicot extensins (P1 and P2 from tomato), including their deglycosylated forms (Kieliszewski & Lamport, 1986). The existence of these antibodies made it possible for the first time to determine whether cross—reactivities arise from carbohydrate or from polypeptide epitopes. This is important because antibodies must be specific to be useful for cytolocalization, i.e. show minimal cross-reactivities. Therefore I raised and characterized two sets of rabbit polyclonal antibodies, one set against the THRGP, whose polypeptide backbone is the major antigenic epitope (Figure 26), and the other set against the two deglycosylated HHRGPs (glycosylated HHRGP apparently is not antigenic, as two attempts to raise antibodies to HHRGP failed) (Figures 33 & 34) Although the cross-reactivity of the two sets of antibodies with ARGP and poly-L-proline still needs to be determined, the antibodies for each maize HRGP show little cross-reactivity with the other maize HRGP (Figures 26 & 34). Furthermore, the anti-THRGP antibodies do not cross-react with solanaceous lectins or AGPs (from dicots), and will probably therefore be specific for THRGP in intact tissues. However, only the anti-THRGP antibodies will be useful for cytolocalization because they react with glycosylated THRGP, while the anti-dHHRGP antibodies are not of immediate use for immunolabeling as they do not react with glycosylated HHRGP (Figure 34)(See section V for use of anti- dHHRGP antibodies to screen for cell wall mutants. Dr. Keith Roberts (John Innes Institute, UK.) has begun TEM immunogold labeling of maize sections using 103 the anti-THRGP antibodies; hopefully this work will provide details about the tissue and cell specificity (if any) of THRGP. 1.4 1.2— 51.6— 1.0 8 m 0.8+ O C 2 0.6— O 8 < 0.4— 0.2- 1 1111111' 1 1 111111] 1 1 1111111 1 1 111111 100 1000 10,000 100,000 dHHRGP Crude Serum Dilution Figure 33. Serial dilution of anti-dHHRGP antiserum with dHHRGP antigen at 20 ng/ELISA microtiter well. IV. Enzymes that Catalyze the Post-Translational Modifications of Extensin Although posttranslational modifications account for as much as 40% 0f extensin's amino acid residues (via hydroxylation, crosslink formation and glycosylation) and from 30-65% of extensin's dry weight (by glycosylation) we know very little about the roles the modifications play in the function of extensin, and the enzymes that catalyze those modifications. Nor are we likely to soon learn anything about those enzymes or the roles of extensin's post-translational 104 100" 90" 80F 70- 60- 50" 4:— 40" F—H 30" Percent Cross Reactivity 20" 10- mash , v v 9 v 9‘ 9‘ (n qt 6? e0 95’ v9 (*0 6 6 1s 0"» .86 <\'\ 64% Antigen (20ng/well) Figure 34. Cross-reactivities of anti-dHHRGP antibodies with HHRGP, THRGP, dTHRGP, P1, dP1, P2, dP2, and sycamore AGP. modifications, as most current extensin research is focused on the extensin polypeptide backbone as it is derived from clones. Yet the function of extensin within the wall matrix probably has as much to do with its posttranslational modifications as it does with the unmodified polypeptide backbone per se. For example, most of the extensin protein backbone is wrapped in carbohydrate (except for the THRGP which is about 30% carbohydrate, dw) with very few peptide sequences presented to the wall matrix. This is evidenced by the very limited digestibility of glycosylated extensin with proteolytic enzymes (Lamport, 1973; Smith et al., 1986). Furthermore, many extensins become crosslinked 105 intramolecularly by EDT, and are probably crosslinked intermolecularly by unidentified links. Therefore, one step toward determining the function of extensin as dictated by its posttranslational modifications, is to isolate and characterize the enzymes involved. Unfortunately, projects aimed at isolating and characterizing the arabinosyl and galactosyl transferases that catalyze the glycosylation of extensin are very difficult (and therefore high risk) because the enzymes are membrane- bound (Karr, 1972). Furthermore, obtaining substrate quantities of carbohydrate- stripped extensin peptides was, until recently, a problem. However, the generation of substrate for in vitro glycosylation is no longer an obstacle in view of the "intact cell elution" technique and HF-deglycosylation which allows the facile preparation of several milligrams of deglycosylated extensin monomer in a relatively short time. Two "posttranslational" enzymes that may be easier to isolate and characterize than the arabinosyl or galactosyl transferases are 1) the enzyme that catalyzes IDT formation“ and 2) the enzyme that hydroxylates peptidyl proline to form 4- hydroxyproline, prolyl 4-hydroxylase (Kivirikko, 1989). The only demonstration of the specific location of IDT is as an intramolecular crosslink in two tomato extensin peptides (Epstein & Lamport, 1984). The Type- 3 domain: Ser-Hyp-Hyp—Hyp-Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-Tyr-Tyr—Tyr-Lys occurs as a tomato wall tryptic peptide (with IDT)(Table 4), and as a major repeat in some extensin clones (sans IDT)(Showalter & Rumeau, 1989); however, an 1‘ The isolation of an IDT-forming enzyme applies only to dicot and monocot walls that contain IDT. The maize cell wall apparently contains no IDT. 106 extensin precursor containing this repeat has never been isolated. The other IDT- containing sequence: Ser-Hyp-Hyp-Hyp-Hyp-Val-Tyr-Lys-Tyr-Lys occurs as a major repeat in tomato wall peptides (Table 4) and in tomato extensin precursor P2 (Smith et al., 1986). The isolation, characterization, and manipulation of an IDT-forming enzyme will have a significant impact on our current thinking about mechanisms for extensin's wall insolubilization and cell wall models which hypothesize IDT as an intermolecular crosslink responsible for the incorporation of extensin into the wall. Thus there are two approaches to extensin crosslinking: a) isolate an enzyme that catalyzes an extensin intermolecular crosslink (which may or may not be IDT), and b) isolate the enzyme that catalyzes the formation of intramolecular IDT. The first approach is currently in progress by Derek Lamport (Everdeen et al., 1988). The second approach aims specifically at isolation of the IDT-forming enzyme. Here generation of substrate is crucial; specifically, the production of monomeric extensin or extensin peptides containing either of the two potential intramolecular crosslink sequences, but without IDT: Ser-Hyp-Hyp-Hyp-Hyp-Val-Tyr-Lys-Tyr-Lys or Ser-Hyp-Hyp-Hyp-Hyp-Ser-Hyp-Ser-Hyp-Hyp-Hyp-Hyp-1yr-Tyr-Tyr-Lys. Then the (p01y)peptides can be used as substrate to assay in vitro IDT formation using crude (and later, not-so-crude) wall enzyme preparations”. Supposedly, the ‘7 The IDT-forming enzyme is a wall enzyme, judging by increases in wall IDT concomitant with extensin precursor insolubilization (Cooper & Varner, 1985) 107 addition of peroxidase inhibitors and free radical scavengers inhibits HDT formation in the cell wall (Cooper & Varner, 1981), thus addition of peroxidase inhibitors and free radical scavengers to cell suspensions followed by isolation of P2, hopefully without HDT, may be one way to generate substrate for the HDT- forming enzyme. Alternatively, Joseph Leykam (Macromolecular Facility, M.S.U. Biochemistry Dept.) can synthesize substrate quantities of the substrate peptides. One drawback is the assay of DDT per se, as it requires an 18 hr acid hydrolysis followed by HPLC and spectrophotometric HDT detection. However, recent developments in microwave-driven acid hydrolysis (5 min /hydrolysis) followed by HPLC and spectrophotometric IDT detection, may expedite the assay (Choiu & Wang, 1989). Of the enzymes that catalyze the posttranslational modifications of algal and higher plant HRGPs, prolyl hydroxylase is the best characterized, although nothing is known about the active site of the plant enzyme. Futhermore, all of the "characterization" of the plant enzyme has been done using artificial substrate (i.e. polyproline H and protocollagen) (Andreae et al., 1988; Bolwell et al., 1985; Cohen et al., 1983; Kaska et al., 1987a & b). This has lead to the dubious assumption that prolyl hydroxylase of higher plants prefers a polyproline H substrate conformation (Tanaka et al., 1981) although the enzyme also hydroxylates protocollagen (also in a PPH conformation) but at a very low rate (Andreae et al., 1988; Cohen et al., 1983; Sadava & Chrispeels, 1971; Sauer & Robinson, 1985). The very specific hydroxylation of the THRGP and the extensive hydroxylation (34 mole%) of 108 HHRGP, neither of which are in an obvious polyproline H conformation, suggests that a) this assumed specificity of prolyl hydroxylase for polyproline H is incorrect, and b) that multiple forms of prolyl hydroxylase exist, each with a defined substrate specificity. Furthermore, the existence of the THRGP extensin, with 1/2 of its Hyp residues nonglycosylated, suggests that non-glycosylated Hyp, as well as glycosylated Hyp, may impart some function to extensin in general and to the THRGP in particular, possibly, by loose analogy with collagen“, in its association with other wall polymers. Therefore, Hyp may play a dual role in the function of extensins, both as the major site of glycosylation, and as a hydrogen bond donor; and the use of natural substrate (i.g. non-hydroxylated extensins generated and purified from cells treated with the suicide inhibitor, 3,4-dehydroproline; Fowden et al., 1963; Myllylla et al., 1979) should facilitate the purification and accurate characterization of plant prolyl hydroxylase(s). A complementary approach to isolating, characterizing and manipulating the enzymes that catalyze the post-translational modifications of extensin involves the generation of extensin mutants involving those enzymes. V. Extensin Mutants: A Positive Screen The isolation and characterization of cell wall mutants involving extensin is an obvious approach to defining the function of extensin in muro. The question 1‘ Hydroxyproline plays a critical role in the assembly of animal extracellular matrices, as its hydroxyl groups are essential for the folding of procollagen polypeptide chains into a triple helix at body temperature. Thus prolyl hydroxylase is a target enzyme for therapeutic intervention in fibrotic disorders (Kivirikko, 1989). 109 arises: How does one screen for mutants in a non-enzymic structural protein? A possible positive screen for "posttranslational" mutants (e.g. arabinosyl and galactosyl transferases and/0r prolyl hydroxylase(s)) could utilize the anti- dI-H-IRGP antibodies which only recognize the polypeptide backbone epitopes normally obscured by glycosylation. For example, dHHRGP antibodies cross-react 2% with HHRGP and THRGP, 1% with tomato extensin P1, 0% with tomato extensin P2 and AGP (Figure 34). Furthermore, in vitro translation of Black Mexican mRNA followed by SDS-PAGE and immunoblotting using anti-dI-H-IRGP antibodies” showed that the antibodies recognize both of the HHRGP polypeptide backbones without hydroxylation of the proline residues (data not shown). Thus, mutants in the glycosyl transferases or prolyl hydroxylase (resulting in underglycosylated or nonglycosylated extensins, respectively) should be detectable in both tomato and maize walls by using the anti-dHHRGP antibodies. For example, a mutagenized suspension culture could be plated and grown as a colonies on a solid medium. Then one could use a replica filter screening technique (applying Varner's tissue-blotting technique, Cassab & Varner, 1985), using CaCl,-s0aked nitrocellulose "lifts" to screen for non-glycosylated extensins by immunodetection with the dHHRGP antibodies. Mutants detected would likely be mutants in prolyl hydroxylase or the glycosyl transferases responsible for most of extensin's post-translational modifications. Controls would involve immunoblots ‘9 Immunoscreening for mutant glycoproteins has precedence in animal systems i.e. the detection of abberant underglycosylated breast mucins associated with breast cancer using antibodies raised against the HF-deglycosylated mucin (Gendler et al., 1987) 110 of cells grown ‘on medium supplemented with 3,4-dehydr0proline. A logistical drawback is that our suspension cultures are diploid, do not undergo meiosis, and therefore any recessive mutations would not be detected by this approach. On the other hand, the alternative of immunoscreening individual seedlings for extensin mutants is overwhelmingly labor intensive. One solution is to mutagenize and culture haploid cells (Catt, 1981), (e.g. pollen-derived cultures or anther cultures of maize, rye, Arabidopsis thaliana, tomato, tobacco, petunia or beet; Sangwan & Norreel, 1975; Bajaj, 1983). This approach is particularly attractive because frequently plants can be regenerated from such cultures. Alternatively, the immunoscreening of individual seedling might be feasible if one could first select a phenotype associated with extensin mutants. Thus, a crucial preliminary experiment would involve germinating "wild type" seeds on medium containing 3,4-dehydroproline which should inhibit virtually all prolyl hydroxylase activity (Cooper & Varner, 1983) and result in non-hydroxylated non- arabinosylated extensin "mutants." If such a "mutation" has profound effects on the cell wall it is likely to have profound effects on plant morphology, therefore a certain phenotypic subpopulation could be selected for immunoscreening. VI. Macromolecular Associations of Extensin It is likely that extracellular matrices must self-assemble in situ, the assembly being regulated primarily by the physical and chemical properties of the macromolecular monomers (Eyre, 1980). An elegant example is the alga Chlamydomonas reinhardtii whose lattice-like crystalline cell wall self-assembles 111 in vitro (Roberts, 1974; Goodenough et al., 1987). Significantly, three of the four molecules involved are hydroxyproline-rich glycoproteins, and the fourth is a glycine-rich species. Assuming that the cell wall of higher plants, like the Chlamydomonas cell wall, also largely self-assembles, the precise interactions and "structures" that occur between extensin and the other extracellular macromolecules, in part, define the function of extensin in the cell wall. An obvious extensin-polysaccharide interaction likely occurs between the positively-charged extensin lysine or histidine residues and the negatively-charged GAX (in maize) or pectins (in dicots). One method for isolating putative interacting GAX with THRGP or HHRGP involves using cross-linking agents in vitro and in vivo. For example, if Schiff bases form between the lysine residues of THRGP and aldehyde groups of other wall components, cyanoborohydride reduction should stabilize the Schiff base. Or, if the THRGP lysine residues ionically interact with the glucuronic acid carboxyl groups of GAX, a "zero length" isopeptide bond between the two might be formed by reaction with carbodiimide (Vandekerckhove et al.,1989). One should then be able to isolate crosslinked dimer fragments after degradation or fragmentation of the cell wall (for in vivo experiments) or of in vitro crosslinked polymers, i.e. to distinguish a random from an orderly pattern of ionic interactions. Another possibility is the use of extensin or extensin peptides coupled to a solid support, e.g. affinity chromatography of cell wall molecules (pectins, GAX) 0r fragments of molecules. These experiments are difficult because they require the isolation and assay of wall polysaccharides or their fragments (a miserable 112 job!). Controls for the binding experiments would include preincubation of the extensin-matrix with anti-extensin antibodies to inhibit binding, derivatization of extensin active R-groups (acetylation of lysines, destruction of histidine imidazole rings with diethylpyrocarbonate), and competing reactions with free lysine, histidine, uronic acids, or even Ca"”°. A biochemical approach to cell wall regeneration will also help to elucidate a possible role of cell wall proteins as "organizers" of cell wall assembly, as suggested by experiments with 3,4-dehydroproline, which inhibited dicot wall regeneration . Evidently dicot protoplasts cannot regenerate their walls using posttranslationally defective HRGPs (unhydroxylated) and therefore underglycosylated)(Cooper & Varner, 1984). This experiment bears repeating and can be extended to the monocot Zea ma ys, where only the Hyp-containing protein components should be affected by dehydroproline; the major non-HRGP structural components should remain unaffected as dehydroproline is highly specific for prolyl hydroxylase. This approach will therefore help distinguish between the roles of HRGPs versus the non-HRGPs. For example, dehydroproline inhibition of maize protoplast wall regeneration would suggest an "organizer" role for the minor HRGP components of a monocot cell wall. A critical control experiment will involve addition of exogenous maize HRGPs, which should overcome the inhibiting effect of 2° Calcium may have a dual function in pectin-extensin interactions. It probably forms salt bridges between pectic carboxyls, furthermore calcium "melts" polyproline H conformations (Tiffany & Krimm, 1969); thus calcium may alter the secondary structure of dicot extensins in the wall. Some potentially useful experiments would be CD of dicot extensins plus and minus Ca”, or of pectin and extensin plus or minus Ca”, and especially as a funtion of pH and ionic strength. 113 dehydroproline on protoplast wall regeneration. Presumably, the various types of extensin evolved to perform different functions in different tissues with different mechanical and physiological properties. For example, the maize cell wall has low levels of extensins which are also HF- insoluble; this contrasts dramatically with the large amounts of HF-insoluble extensins in dicot cell walls. These differences may be related to the very different growth habits of grasses (silica support) and herbaceous dicots (turgor support). Despite extensive data on the primary structure of extensins, (posttranslational modifications and regulation of expression, etc.) the discreet function of any extensin type remains unknown, probably because we don't know how extensin interacts specifically with other wall macromolecules. Yet its role in wall self— assembly and disease resistance, the roles of its post-translational modifications, and its function in morphogenesis, all depend on intermolecular reactions such as those suggested above. Future tests of these working hypotheses demands a combination of chemistry and molecular biology. For example, peptide sequences hypothetically involved as functional domains could be altered by site-directed mutagenesis with predictable (?) results. And highly expressed extensins could be down-regulated by antisense RNA. Ultimately, all these questions involve an understanding of cell wall chemistry, and the ability to model that chemistry in four dimensions. APPENDICES 00000800: 00000005 00 000mm :0 05000000: 05 000 0000300 000300: 5000 00000500 < 0000000 0. 0m0mm 90: .A00050c 80:00000bm :03 :00 500050 05: 0000—05 50:50.: .03.: 00:09:03 00: 0050000: 5 9mm 5500:0030 0358 «OH. 05 005 000 00000000 00003 0 003 am :03 :00 5 0000.65 28-3 05 0030800 “580$ 050 0000 .0500 05 000055 50000050 000: 0000500 0505 505 =00550000:008:00: 350500-05: 0 05:50 3:000: 00 00 >000om :03 :00 35 0500555005 08 005000.: PO Z 5 500008 :0 0050030050000 0000000000 000 05000080»: 00:00:03 .53on 00:0»050 :0 0030800 05 5 0:00 0 0>0: :05 050000: w550~0000>$ 80:03: a. 00505 00:00:50.00505850000 “08:0: 58000000005058 300 0 08505 50050.: Sm: 0:00 000500 :0 0:03 :00 5 0000550000 am 50050.: n03 0900000005 00:00: 0:: :0 0005000350 05 8>_o>5 “5.: 5.5: 9 05 :0 000005000»: 05. 50:50.: 000: 500050 0050 Z .00 00:50:05 0050.000 300:: 5 50050.35: 05.: 50050..— 83 .805 SN :50505 .08500 000 $000 0000 5 «.5000 0mm .3 00 00am 00> $0: 50000: :03 :00 :0 0000800 00505 000 00008—000»: :03 :00 05000080 0005005 5 03.: 00008059 0805qu a. 50050.: .000 00:00:05: .8: 05:00 050000000 :00 000500.? 0300050 0 0000000: 00005.52 00 50050.: 00008—000»: :03 :00 00:00 000009 5 0mm 0000.60.05 50000553 0. :0m000 83 00003000.»: 050 «0:00 5 05.: 000005050 00350.5. 00 000305 03: 000083000»: :03 :00 00:00 5 50 000 00m .004 0050000: :0 00 05C. :3: 00000000000 :03 :00 5 5080 00352 00000m 0% 0005:; m8: .Eex 0.8.5 :0 0:03 :00 5 5000.5 00: 0000006 50.500005: 00030.5 8:005: 0% 30000000009 33 00300808508050 $38058 0000 VOOAOZHmZEXm Z: mfimMUZOO QZ< Emammmxm éémm .mO FOOAOZOMEU < < NHQZmmew 114 115 .0085 000000505000 .00 0:03 5000 0000005 00000.00 500008 00 000000000 00.0.”: 55000058 00w00000 00500.0 000 00m 00 00000050000 0003000 000000000000 00025 00300m 002935 080 0550 000 5 00m 0000 000 00000500: 00 00m000 0000 000 5 00295 5 3000000000 500000 0000 00000ww0m 0500000 :03 w5500000003 00 080000000 00> 000000 00000000000 0900000 500000 8000080000 0 00-5: 08.. 8005000 50 9 800 000280008 0 030088000 0000500500 0000000000 003000 "0000000 :03 :00 500008 00 0000 Q0 0000.”: 000000500 :0 $000000 0300500 00% 00. 500000 8% 5 :03 035005-00: 0:00:00 000 000000800000 :03 0003 mm 5 0.05005 5 500008 00000-:03 0000 00300m 05000000003 0000000000000 00 00.: 000000000 0000 0050 000000000 5000000008 500008 00 050 00.005 5 000000000 0306 0:03 :00 500.0 500008 000008 00 .0080 00.000 0005000 000 00.000 00000000 0000 0000000 00 0000000 0500.000 000D 00000005 0000005508 5 000000500 :03 808500 000 000 0.0005050000000535 00395 5000000 5 00000 0000 5000 00000000 000?. 5 500000 80000000 0000005 000 00035 .05.: 00 000050 000 000: v 000 0:03 0050 8 500000 :03 00 000050 0500 000 500000 0005508 0000000 050000000 00 0:03 :00 .0000000 000000 :03 :00 000500 5 =05005000 0500000 0305—00.. 00 0000 Z 000800 000:0 08000 .8508... :000000 0 00 0000005 0000300 "0:03 :00 00:05 008005-308 5 0.00-0: 080005 00000050000003 .00000m00 :03 000 000.0 00000 00000000» .202 0000 00000ww0m 500000000 00500.0 5 50000 00 00000 0000—0060 0.02 0000000 5 5008mm 500008 00 0050: 0000—0060 5 50000 0503 5 8005 :03 :00 0000500 0 0300mm0m 0000030 5 5505000000 £053-08 0000000 500008 00 000000000 005.”: 0000000 :03 500008 000500 5 000000 05800000000006 :03 000 5 000000 500008 00000-:03 00 0000000000 00 :000 0 00000ww0m 0:03 :00 000000 5000 00500000 500008 00000000 20000008000 0 0050000 0:00000 500055 000—0 000 5 005500 000 0000050000 0500000000002 0000050000 0000000000000 0.500008 00 0000000 0000008 8.0-0 00> 500008 00 00000m00 0005— 000 0005000005 0005:00 0503 5 00005 :03 :00 0 00000000 00005.05 00. 0005500000050 .00 00 00055000000005 00050.53 0000 000050.: 000050.: a. 0002 080000 0 080000 00000023 .00 00 0003: 000050..— 00002 00 0005500000080 .00 00 00002 .00 00 0000003: 000053 .00 00 000050.: 0000000 0 0000 00:02 0 000000 000055 000: mg: 03: mg: 3.0: m3: N00: :0: 00050000 ....< 000000000 116 .000 050 05:00: 505053 0850000 .5: m50m0mwsm 0005500330 00:0 F0505500 30— 3005? 5000500500 £05055?“ 0 an: 5 55500 0050005 0058 00 £00 DU $05 5: cm 000 0 8 0050005 58050 00:00 050003 508?me 00.5020 «53 590300000 5 0000000030 :03 005055 0005008005 05600 00 0050.55 .5000 500008 00 :83: 00303.6 .003: 0350 0 5 058050 25:05 00 0000580500 55m 5000555000 55050030 50553 00555 :00 05 505 0250 00 08 5050005 55008 00305m He 00 58000 >05 0:0 £00300: 58050 03585 05 00 00950000 5 55050 0380850030 50050005 3050 80055 0000000000 000 0350.05:& 5050.55 000008000 000 5030208 00505 00.5 >0 005555 5 :03 05 005 55008 00 0000050505 0000000 :03 500050 5 55—3000 850505055 00 mm 0020a Be 005 0000.38050Q 5003000 08.500-55.008 00000000008 0 «0 0055500 .0005 :03 :00 0 0000005 500 Baa 2: 85.3% 05 .95 053 a 85%; a? a .3858 0585 3225... 80:82 $3 5000000 0000583? 558mm .035 00000000 «0 0:03 :00 5 500050 .00 000050000— 305200055 «5..» A500008v 50000500 :03 :00 :83 005.30 a “5&0 0500550 050 005mm 08005000 >0 00508 8.5—808 En: 00500—050005 3 :03 5 E05 5 58050 0000033 000 050050 005500 an 5% 5 En: 00396 ._.0>00>t00 0500.5 03055... $000505 00 55050 5 005000 5 028%?” 000 000030005 :03 :00 5 65855005 .500 0550 55:88 30: 05 05.50005 .0008 539m 805mm: 00 «comma 5 35 5 55000 000050000 005 00002 08650850 055:0 5 3:055 :03 :00 08 00080 00 005500 0.0825 BERN 50305 083 000029» 250000000 5 >420 000:? 000005-308 00 0:03 :00 5 00.00005 £0>0_ gm .503 :00 00500 5 55050 03500 «0 55050 05 0080.85 08008000 .50 000555 000 «ONE 00 006500 05 00398 A505? 050 .0500 5000550 505 0050005 55008 05.“— 05 00ut=m 553000 500050 3 :03 5 20: 500050 00~m0ww=m 00000500000 00 000 .5805 00 000 5003000 500000005 030 00 >05 0005b ...0>00>t00 05805 85055:: 05 5> :03 05 5 005530000 m=000000800m 5 55050 00~8ww0w .5»: 000 00 00000000 5 :03 :00 5 0005050000 ARE 5000035905 550 “0 $0008 own—awe mmwmm 0000035 0053». a. 50$ 00> 000000 50 “0 555m 00:53 00 000000 588m 0 0353 .~U .3 £063 .NQ um :03 E :8 0 50083 :00 .3 00 90.20 0050.0» 00 000000 00053 0% 50% 0.000002% a. 002.0 000953 50 00 05m 0:03 00:50:00 0.me S3 50H 83 .4 5050?. 117 A805 00N: 88:05 0 50:: 80me :Omm :0 :0 8:855: :5: .525: 0: 080:8: 5 «0:055:00: :5: 50:50 5:56 :05: m5:::03 0: 080:8: 5 8:05:00 «.80» 580:3“: .0035: 52: 5 58050 03500 :0 8:55:82 50520555 :0: 6025:0005: :0 w5:5:: 0:05: 00::05 00:20>0Q .052: 5:5: 5 8000 858w? 000:0 55:00:35 :0:8wm:m .:0::0::0 :3 8055 0:: 00:0: :05]: :0 :00. 852:5: < 5:2: 32005.50: :0 0:03 :00 5 8:235:00: :5: 558050 0550:: 50:0::00:_w 32:80:30: 50:2:500 0 :0 0:08:00: 50:9: 0:: :85»: 8:02:50 :5: 350m 803 :00 380:5 85> 5 8:055:80 :5: ...8%=8.. 0: :00: 0:: 88:00 58050 2:050:05 :0 :05230053: 85:50 58050 2:050:05 5 be 3:00:00 805805 :0 8:0 0:: :0 w525:a8:0 03880:: 0:: 5:0: 505030 :8: 05: 0:050:05 0:5 :82: 80:00::5 Be 550205055 :0:8mw:m 5505050.. 58050 :0 8:050:55 :0::00_m 50:9: 58-05050 :0? :8 05:05: 0 :0 0:20 250:0: :5: 3:205 82::0: :03 50:: 5:50 85:80:20 80:02:00 0:: :0 050: 0058:0800 5:80:50: “50800:: 580:6 :0 80:050.: 200 0550 :005: :5: :85: :2:0::0=0 :0 875580.: 55:00:25 0.02:5: 00:0 0:02 5:2: 22:008.: 5 :2:000: 0:2 0 8 :50 3:2: 258008» :0: 5 080:8: 5:00 5: 8 22% :Cmm :0 :2555800 0800 :2:00.:5 :80 50:0:0 55:": :3: :03 :00 :50 82:50 :0 80:5»: 83:: 50:50 5:5: 52: 58050 500980 0 :0 :230550 0900:: 0:05: :5: 5505:280: 0:: 52:05:820855: 0:, 505580502 60:50: 0:20 250:0: 58050 :5: 00:20:: 220556803 0: a. :8me :0 :0 5800 225550555: 0:: :05:0:m 50:53 :0 2800800: 0:0»:508053 a. :82 5:0:00:m 0:. 505055 8:932 a. 50:00 :0 :0 55m :0 :0 :0::030:w :0 :0 30m :0 :0 9008“: :0:.a> :0 :0:U :0::0> 5 :06 :0 :0 555050555: :0::0> 0m 8:000 :00: 3:: 3:: 0005::8...< 550:3. 118 .5003 5% a 00 800000 §I-§I-&m-%m¢0w 500808 8000 0&3 .80 000000000 500 0550 85m d 0565500 800 5 003058.500 «5 8865 $05.? 00 8.00% CD 000 .8050 553 35050: 500050 03.8 0% E 305 mOmE 500050 0505 00 00000008 500 0550 35m 800500—050 5 5.0.0808 “0 28 08 00.80030 #00 E58055 «000 E85 5 00295 500 «0 80¢ :55 0 5 08:05 3:06:08 5 500050 558% < .mOmm 0505 5 8590305 0:05 50%.: 080800050Q @8580 0505 500... mOmE 0 08505 .3me 695 00Nv mOmm 8000005 0 08 0050 <20”. 85.8 08505 0&me w500000 800m 00 0585000508 085253 0055850 .633 58:00am 00. 508805 00 00538500 05 08 .0858 .800 0308: mm 00:00:03.5 EM 505 0000550 000 03050 «00 0.: 008020.— 3050005 500050 0803000330 «0 058035000m .3 N0 5 .3 80 Egonmuog 82 050..— um 00:0M .3 80 2033509 $3 40 “0 000m .3 00 805% .3 80 8&5? .3 0 $583 002 00:558....aq. 500009“ APPENDD{ B GLYCINE-RICH PROTEINS (GRPs) GRP Structure In 1965, Lamport surveyed the amino acid compositions of nine dicots, and a primitive gymnosperm (Gingko), and concluded from the wide ranges present in the value of any particular amino acid of one wall compared to another, that structural wall protein alternatives to the Hyp-rich extensins were likely (Lamport, 1965). Burke et al. (1974) reached a similar conclusion about graminacous monocot walls after analyzing four species. For several years, rumors circulated about glycine-rich hydroxyproline-poor cell walls and wall protein fractions in some plant species and tissues (Rackis et al., 1961; Melin et al.,1979; Dreker et al., 1980; Varner & Cassab, 1986). Finally, in 1986, Condit & Meagher serendipitously1 isolated a petunia gene encoding a glycine-rich (67 mole %) putative cell wall protein (GRP) composed of (Gly-X)n repeats, which make it structurally analogous to silk fibroin (Pauling & Corey, 1951; Condit & Meagher, 1986). Keller et al. 1 The proline residues in extensin are primarily encoded by CCA, while glycine in the GRPs is primarily encoded by GGT; thus extensin probes can be used to isolate GRP genes, because their mRNAs are encoded by opposite strands of similar sequences. In fact, Condit was originally probing for an extensin when she unexpectedly recovered her glycine-rich gene (C. Condit to M. Kieliszewski, personal communication). 119 120 (1988) have since used extensin clones to isolate clones for two distinct glycine- rich cell wall2 proteins from bean. The three GRP clones encode proteins with a range of sizes and having similar, but distinctive, amino acid compositions (Table 1). All three proteins are likely B-pleated sheets, the two larger GRPs (from petunia and bean GRP 1.8) consisting of 8 anti-parallel strands with charged residues along the edges of the sheet, bulky side chains occurring regularly along one face, while the opposite face holds the hydrogens of the glycine R-groups (Condit & Meagher, 1986; Keller et al., 1988). Like extensins, the GRPs are organized into three distinct domains: the C-terminus, the N-terminus, and the middle section composed of the major repeating polypeptides (Condit & Meagher, 1986,1987) Regulation of GRP Expression. Like some extensins, the GRPs are apparently developmentally regulated and induced by wounding (Condit & Meagher, 1986; Keller et al., 1988, 1989a). The two GRPs from bean, GRP1.8 and GRP1.0, although encoding proteins of similar 2 Keller et al. (1988) demonstrated GRP localization in the cell wall by immunolocalization, using polyclonal antibodies raised against a fusion protein, however, the data supporting their conclusion is not entirely convincing. The anti- GRP 1.8 antibodies also reacted with two proteins in the cytoplasm. The authors give three explanations: the cytoplasmic GRPs are GRP 1.8 in transit to the wall; the cytoplasmic GRPs are actually contaminants from the cell wall; or GRP 1.8 does indeed occur as a bone fide cytoplasmic component but with another function in the cytoplasm. Unfortunately, Keller presents no quantitative data about the titer, dilution, and specificity of their antibodies. For instance, do they cross-react with other GRPs such as collagens, the structurally analogous fibroins, or poly-glycine? These are important controls considering the very high background contributed by the control rabbit serum. 121 may accumulate in different cell types (Keller et al., 1989b). GRP 1.8 shows cell- type specificity, as it is specifically expressed in the protoxylem of bean hypocotyl vascular tissue; thus the GRP1.8 promoter is probably regulated by specific developmental and environmental signals (Keller et al., 1989a & b). Table 21. Glycine-Rich Proteins of Dicots Petunia GRP‘ Bean GRP 1.8" Bean GRP 1.0" Estimated Size: Protein by Clone 25 kD 36.7 kD 19.9 kD Protein by SDS-PAGE ND 53 kD ND Abundant Amino Acids‘: Gly 67 58 63 Ala 8 8 9 Tyr 0 7 5 p1: ND ND - ND Postranslation Modifications: ND None? ND ‘ From Condit & Meagher, 1986, 1987 ° Represented as Mole % b From Keller et al., 1988, 1989b Function of GRPs There exists an obvious relationship between GRPs and the development of vascular systems, implying a specific role for GRPs in the functional specialization of vascular tissue. Furthermore, GRP is insolubilized in the vascular tissue concomitant with the cessation of cell extension growth and lignification of the vascular tissue. Because the bean GRPs are fairly rich in tyrosine, the tyrosine residues may serve as a "substratum" for lignin deposition in protoxylem and / or be 122 the vehicle for GRP wall insolubilization via the formation of IDT intermolecular crosslinks (Keller et al., 1989b). GRPs in the Graminaceous Monocots GRPs occur in two graminaceous monocots: maize and rice; however, they, they are smaller than the dicot GRPs and, except for being Gly-rich, are compositionally distinct (Table 2). The maize GRP has a repeating peptide (Gly-Gly-Tyr-Gly-Gly) (Gomez et al., 1988), while the rice GRP has no definite repeating motif (Mundy & Chua, 1988). Furthermore, both monocot GRPs are rapidly induced by ABA and water stress (Gomez et al., 1988; Mundy & Chua, 1988) Table 22. Glycine-Rich Proteins of Graminaceous Monocots Maize GRPa Rice GRPb Estimated Size: Protein by Clone 15.4 kD 16.5 kD Protein by SDS-PAGE ND 21 kD Abundant Amino Acidsc Gly (37%), Arg (9%) Gly (26%), Thr (14%) Glu (6%), Ala (6%) Lys (8%), Gln (7%) pI: 5.7 9.4 Post-Translational Modifications: ND ND ' From Gomez et al., 1988 ° Represented as Mole % " From Mundy & Chua, 1988 REFERENCES Akiyama Y, M Mori, K Kato (1980) 13C-NMR Analysis of hydroxyproline arabinosides from Nicotiana tabacum. Agric. Biol. Chem. 44:2487-2489. Albersheim P, DJ Nevins, PD English, A Karr (1967) A method for the analysis of sugars in plant cell-wall polysaccharides by gas liquid chromatography. Carbohyd. Res. 5:340-345. Allen AK, NN Desai, A Neuberger, JM Creeth (1978) Properties of potato lectin and the nature of its glycoprotein linkages. Biochem. J. 171:665-674. Allen AK, A Neuberger (1973) The purification and properties of the lectin from potato tubers, a hydroxyproline-containing glycoprotein. Biochem. J. 135 :307- 314. Ambler RP, LH Brown (1967) The amino acid sequence of Pseudamonas fluorescens azurin. Biochem. J. 104:784-825. Aminoff D, WD Gathman, CM McLean, T Yodomae (1980) Quantitation of oligosaccharides released by the fi-elimination reaction. Anal. Biochem. 101:44-53. Andreae M, P Blankenstein, YH Zhang, DG Robinson (1988) Towards the subcellular localization of plant prolyl hydroxylase. Eur. J. Cell Biol. 47:181- 192. Bailey RW, H Kauss (1974) Extraction of hydroxyproline-containing proteins and pectic substances from cell walls of growing and non-growing Mung bean hypocotyl segments. Planta 119:233-245. Bajaj YPS (1983) In vitro production of haploids. In DA Evens, WR Sharp, PV Ammirato, Y Yamada, eds. Handbook of Plant Cell Culture Vol.1: Techniques for propagation and breeding. pp.228-287. Macmillan, Inc. Basile DV (197 9) Hydroxyproline-induced changes in form, apical development, and cell wall protein in the liverwort Plagiochila arctica. Amer. J. Bot. 66:776- 783. 123 BIBLIOGRAPHY 124 Basile DV, MR Basile (1989) Changes in cell wall-associated arabinogalactan proteins "correlated with experimentally-induced altered patterns of cell division and organogenesis. Fifth Cell Wall Meeting, Edinburgh, UK. Abst. 17. Bell AA (1981) Biochemical mechanisms of disease resistance. Ann. Rev. Plant Physio]. 32:21-81. Blaskek W, D Haass, H Hoehler, G Franz (1981) Cell wall regeneration by Nicotiana tabacum protoplasts: chemical and biochemical aspects. Plant Sci. Lett. 22:47-57. Bolwell GP, MP Robbins, RA Dixon (1985) Elicitor-induced prolylhydroxylase from French Bean, Phaseolus vulgaris. Biochem J. 229:693-699. Bomstein P, H Sage (1980) Structurally distinct collagen types. Ann. Rev. Biochem. 49:957-1003. Boundy JA, JS Wall, JE Turner, JH Woychik, RJ Dimler (1967) A mucopolysaccharide containing hydroxyproline from corn pericarp. Isolation and composition. J. Biol. Chem. 242, 2410-2415. Brysk MM, MJ Chrispeels (1972) Isolation and partial characterization of a hydroxyproline-rich cell wall glycoprotein and its cytoplasmic precursor. Biochem. Biophys. Acta 257: 421-432. Burke D, P Kaufman, M McNeil, P Albersheim (1974) The structure of plant cell walls VI. A survey of the walls of suspension-cultured monocots. Plant Physiol. 54:109-115. Caplan AI (1987) The extracellular matrix is instructive. Bioassays V015, No. 3:129- 132. Carpita NC (1983) Hemicellulosic polymers of cell walls of Zea coleoptiles. Plant Physiol. 72:515-521. Carpita NC (1984) Fractionation of hemicellulose from maize cell walls with increasing concentrations of alkali. Phytochem. 23:1089—1093. Carpita NC, J Kanabus (1988) Chemical structure of the cell walls of dwarf maize and changes mediated by gibberellin. Plant Physiol. 88:671-678. Carpita NC, D Whittern (1986) A highly substituted glucuronoarabinoxylan from developing maize coleoptiles. Carbohydr. Res. 146:129-140. 125 Carpita NC, JA Mulligan, JW Heyser (1985) Hemicelluloses of cell walls of a proso millet cell suspension culture. Plant Physiol. 85 :480-484. Catt J (1981) Cell wall mutants from higher plants. Phytochem. 20:2487-2488. Cassab G], J Nieto-Sotelo, JB Cooper, GJ van Holst, JE Varner (1985) A developmentally regulated hydroxyproline-rich glycoprotein from the cell walls of soybean seed coats. Plant Physio]. 77:532-535. Cassab GI, JE Varner (1987) Immunocytolocalization of extensin in developing soybean seed coats by immunogold staining and by tissue printing on nitrocellulose paper. J. Cell Biol. 105:2581-2588. Chen J, JE Varner (1985a) Isolation and characterization of cDNA clones for carrot extensin and a proline-rich 33-kDa protein. Proc. Natl. Acad. Sci. USA 82:4399-4403. Chen J, JE Varner (1985b) An extracellular matrix protein in plants: characterization of a genomic clone for carrot extensin. EMBO J. 4:2145- 2151. Choiu, SH, KT Wang (1989) Peptide and protein hydrolysis by microwave irradiation. J. Chromatogr. 491:424-431. Chrispeels MJ (1969) Synthesis and secretion of hydroxyproline-containing proteins in carrots. 1. Kinetic analysis. Plant Physio]. 44:1187-1193. Chrispeels MJ, D Sadava, YP Cho (1974) Enhancement of extensin biosynthesis in ageing disks of carrot storage tissue. J. Exp. Bot. 25:1157-1166. Christiansen GS, KV Thimann (1950) The metabolism of stem tissue during growth and its inhibition 1. Carbohydrates. Arch. Biochem. Biophys. 26:230-247. Clarke JA, N Lisker, DTA Lamport, AH Ellingboe (1981) Hydroxyproline enhancement as a primary event in the successful development of Erysiphe graminis in wheat. Plant Physio]. 67:188-189. Cleland R, A Karlsnes (1967) A possible role for hydroxyproline-containing proteins in the cessation of cell elongation. Plant Physiol. 42:669-671. Cohen PB, A Schibeci, GB Fincher (1983) Biosynthesis of arabinogalactan—protein in Lolium multiflorum (Ryegrass) endosperm cells. III. Subcellular distribution of prolyl hydroxylase. Plant Physiol.72:754-758. 126 Condit CM, RB Meagher (1986) A gene encoding a novel glycine-rich structural protein of petunia. Nature 323:178-181. Condit CM, RB Meagher (1987) Expression of a gene encoding a glycine-rich protein in petunia. Mo]. & Cell Biol. 7:4273-4279. Cooper JJ (1984) Hydroxyproline synthesis is required for cell wall regeneration. In WM Dugger & S Bartnicki-Garcia, eds. Structure, Function and Biosynthesis of Plant Cell Walls. Proc. 7‘h Annual Symp. in Botany, UC, Riverside. Cooper JJ, JA Chen, JE Varner (1984) The glycoprotein component of plant cell walls. In WM Dugger & S Bartnicki-Garcia, eds. Structure, Function and Biosynthesis of Plant Cell Walls. Proc, 7"“ Annual Symp. in Botany, UC, Riverside. Cooper JB, JE Varner (1983a) Selective inhibition of proline hydroxylation by 3,4 dehydroproline. Plant Physiol. 73:324-328. Cooper JB, JE Varner (1983b) Insolubilization of hydroxyproline-rich cell wall glycoprotein in aerated carrot root slices. Biochim. Biophys. Res. Comm. 112:161-167. Cooper JB, JE Varner (1984) Cross- linkin-g of soluble extensin in isolated cell walls. Plant Physiol. 77. 414- 417. _ Corbin DR, N Sauer, CJ Lamb (1987) Differential regulation of a hydroxyproline- rich glycoprotein gene family in wounded and infected plants. Mol. Cell. Biol. 7:4337-4344. Darvill AG, P Albersheim, M McNeil, JM Lau, WS York, 'I'I‘ Stevenson, J Thomas, S Doares, DH Gollin, P Chelf, K Davis (1985) Structure and function of plant cell wall polysaccharides. J. Cell Science Supp]. 2:203-217. Darvill AG, M McNeil, P Albersheim, DP Delmer (1980) The primary cel walls of flowering plants. In NE Tolbert, ed. The Biochemistry of Plants, Vo] 1,91- 161. Academic Press, Inc. Darvill JB, M McNeil, AG Darvill, P Albersheim (1980) Structure of plant cell walls. XI. Glucuronoarabinoxylan, a second hemicellulose in the primary cell walls of suspension-cultured sycamore cells. Plant Physiol 66:1135-1139. Dayhoff MO, WC Barker, LT Hunt (1983) Establishing homologies in protein sequences. Methods in Enzymo]. 91:524-545. Academic Press,Inc. 127 Deber CM, FA Bovey, JP Carver, ER Blout (1980) Nuclear magnetic resonance evidence for cis-peptide bonds in proline oligomers. J. Am. Chem. Soc. 292:6191-6198. Desai NN, AK Allen, A Neuberger (1981) Some properties of the lectin from Datum stramonium (thorn-apple) and the nature of its glycopeptide linkages. Biochem. J. 197:345-353. Doolittle RF (1981) Similar amino acid sequences: Chance or common ancestry? Science 214:149-159. Doolittle RF (1986) Synthetic peptide antigens. In R. Doolittle, ed. Of Urfs and Orfs: a primer on how to analyze derived amino acid sequences. University Science Books, California, pp. 63-81. Dougall DK, K Shimbayashi (1960) Factors affecting grth of tobacco callus tissue and its incorporation of tyrosine. Plant Physiol. 35:396-404. Dreher ML, CW Weber, WP Bemis, JW Berry (1980) Cucurbit seed coat composition. J. Agric. Food Chem. 28:364-366. Ecker JR, RW Davis (1987) Plant defense genes are regulated by ethylene. Proc. Natl. Acad. Sci. USA 84:5202-5206. Edman P (1970) Sequence Determination. In S Needleman, ed. Protein Sequence Determination, pp. 211-275. Springer, New York. Engvall E, P Perlmann (1972) Enzyme-linked immunosorbant assay ELISA III. Quantitation of protein specific antibodies by enzyme-linked anti- immunoglobin in antigen coated tubes. J. Immunology 109:109-135. Epstein L, DTA Lamport (1984) An intramolecular linkage involving isodityrosine in extensin. Phytochem 23, 1241-1246. Esquerre-Tugaye MT, C Lafitte, D Mazau, A Toppan, A Touze (1979) Cell surfaces in plant-microorganism interactions II. Evidence for the accumulation of hydroxyproline-rich glycoproteins in the cell wall of disease plants as a defense mechanism. Plant Physiol 64:320-326. Esquerre-Tugaye MT, DTA Lamport (1979) Cell surfaces in plant-microorganism interaction 1. A structural investigation of cell wall hydroxyproline-rich glycoproteins which accumulate in fungus infected plants. Plant Physio]. 64:314-319. 128 Esquerre-Tugaye MT, D Mazau (1974) Effect of a fungal disease on extensin, the plant cell wall glycoprotein. J Exp Botany 25 :509-5 13. Everdeen DS, S Kiefer, JJ Willard, EP Muldoon, PM Dey, XB Li, DTA Lamport (1988) Enzymic cross-linkage of monomeric extensin precursors in vitro. Plant Physiol. 87:616-621. Eyre DR (1980) Collagen: Molecular diversity in the body‘s protein scaffold. Science 207:1315-1322. Fowden L, S Meale, H Tristam (1963) Effect of 3,4-dehydroproline on grth and protein synthesis. Nature 199:35-38. Franssen HJ, JP Nap, T Gloudemans, W Stiekema, H van Dam, F Govers, J Louwerse, A van Kammen, T Bisseling (1987) Characterization of cDNA for nodulin-75 of soybean: A gene product involved in early stages of root nodule development. Proc. Natl. Acad. Sci. USA 84:4495-4499. Frister H, H Meisel, H Schlimme (1988) OPA method modified by use of N,N- dimethyl-Z-mercaptoethylammonium chloride as thiol component. Fresenius A. Anal. Chem. 330:631-633. Fry SC (1982) Isodityrosine, a new cross-linking amino acid from plant cell-wall glycoprotein. Biochem. J. 204:449-455. Fry SC (1985) Primary cell wall metabolism. In BJ Biflin, ed. Plant Mol. Cell Biol. 2:1-42, Oxford Univ. Press. Fry SC (1986) In-vivo formation of xyloglucan nonasaccharide: a possible biologically active cell-wall fragment. Plant 169:443-453. Fujimoto D, M Hirama, TI Washita (1982) Histidinoalanine, a new crosslinking amino acid in calcified tissue collagen. Biochem. Biophys. Res.Comm. 104:1102-1106. Gardiner M, MJ Chrispeels (1975) Involvement of the Golgi apparatus in the synthesis and secretion of hydroxyproline-rich cell wall glycoproteins. Plant Physiol. 55:536-540. Gendler SJ, JM Burchel], T Duhig, D Lamport, R White, M Parker, J Taylor- Papadimitriou (1987) Cloning of partial cDNA encoding differentiation and tumor-associated mucin glycoproteins expressed by human mammary epithelium. Proc. Nat]. Acad. Sci., USA 84:6060-6064. 129 Gomez J, D Sanchez-Martinez, V Stiefel, J Rigau, P Puigdomenech, Montserrat Pages (1988) A gene induced by the hormone abscisic acid in response to water stress encodes a glycine-rich protein. Nature 334:262-264. Goodenough UW, B Gebhart, RP Mecham, JE Heuser (1986) Crystals of the Chlamydomonas reinhardtii cell wall: Polymerization, depolymerization, and purification of glycoprotein monomers. J.Ce]1 Bio]. 103:405-417. Hammerschmidt R, DTA Lamport, EP Muldoon (1984) Cell wall hydroxyproline enhancement and lignin deposition as an early event in the resistance of cucumber to Cladosporium cucumen'num. Physiol. Plant Path. 24:43-47. Heckman JW, BT Terhune, DTA Lamport (1988) Characterization of native and modified extensin monomers and oligomers by electron microscopy and gel filtration. Plant Physio]. 86:848-856. Hill RL (1965) Hydrolysis of proteins. In CB Anfinsen, Jr., ML Anson, JT Edsal], FM Richards, eds., Advances in Protein Chemistry, pp.64-68; Academic Press, New York. Holleman J (1967) Direct incorporation of hydroxyproline into protein of sycamore cells incubated at growth-inhibitory levels of hydroxyproline. Proc. Natl. Acad. Sci. USA 57:50-54. Hong J C, RT Nagao, JL Key (1987) Characterization and sequence analysis of a developmentally regulated putative cell wall protein gene isolated from soybean. J .Biol. Chem. 262:8367-8376. Hood EE, QX Shen, JE Varner (1988) A developmentally regulated hydroxyproline-rich glycoprotein in maize pericarp cell walls. Plant Physiol 87 :138-142. Jarvis MC, W Forsyth, HJ Duncan (1988) A survey of the pectic content of nonlignified monocot cell walls. Plant Physiol 88:309-314. Jermyn MA, YM Yeow (1975) A class of lectins present in the tissues of seed plants. Aust. J. Plant Physiol. 2:501-531. Johnson WC, Jr. (1988) Secondary structure of proteins through circular dichroism spectroscopy. Ann. Rev. Biophys. Biophys. Chem. 17:145-166. Karr, AL Jr. (1972) Isolation of an enzyme system which will catalyze the glycosylation of extensin. Plant Physiol. 50:275—282. 130 Kaska DD, V. Gunzler, KI Kivirikko (1987a) Characterization of a low-relative molecular-mass prolyl 4-hydroxylase from the green alga Chlamydomonas reinhardii. Biochem. J. 241:483-490. Kaska DD, P Myllyla, V Gunzler, A Gibor, KI Kivirikko (1987b) Prolyl-4- hydroxylase from Volvox caten'. A low-Mr enzyme antigenically related to the oz subunit of the vertebrate enzyme. Biochem. J. 256:257-263. Kasper CB (1975) Fragmentation of proteins for sequence studies and separation of peptide mixtures. In, SB Needleman, ed. Protein Sequence Determination. pp.114-162; Springer-Verlag, New York. Keegstra K, KW Talmadge, WD Bauer, P Albersheim (1973) The structure of plant cell walls III. A model of the walls of suspension-cultured sycamore cells based on the interconnections of the macromolecular components. Plant Physiol. 51:188-196. Keller B, CJ Lamb (1989) Specific expression of a novel cell wall hydroxyproline- rich glycoprotein gene in lateral root initiation. Genes & Dev., in press. Keller B, N Sauer, CJ Lamb (1988) Glycine-rich cell wall protein in bean. Gene structure and association of the protein with the vascular system. EMBO J. 7:3625-3633. Keller B, J Schmidt, CJ Lamb (1989a) Vascular expression of a bean cell wall glycine-rich protein-B-glucuronidase gene fusion in transgenic tobacco. EMBO J. 8:1309-1314. Keller B, MD Templeton, CJ Lamb (1989b) Specific localization of a plant cell wall glycine-rich protein in protoxylem cells of the vascular system. Proc. Natl. Acad. Sci. USA :1529-1533. Kieliszewski M, DTA Lamport (1986) Cross-reactivities of polyclonal antibodies against extensin precursors determined via ELISA techniques. Phytochem 25: 673-677. Kieliszewski M, DTA Lamport (1987) Purification and partial characterization of a hydroxyproline-rich glycoprotein in a graminaceous monocot, Zea mays. Plant Physiol. 85: 823-827. Kieliszewski M, DTA Lamport (1988) Tying the knots in the extensin network. In JE Varner, ed. Self-Assembling Architecture. pp 61-76. Alan R. Liss, Inc. 131 Kieliszewski M, JF Leykam, DTA Lamport (1989) Trypsin cleaves lysylproline in a hydroxyproline-rich glycoprotein from Zea ma ys. Peptide Res. Vol. 2: 246- 248. Kieliszewski M, JF Leykam, DTA Lamport (1990) Structure of the threonine-rich extensin from Zea ma ys. Plant Physiol. in press. Kimmins WC, RG Brown (1975) Effect of a non-localized infection by southern bean mosaic virus on a cell wall glycoprotein from bean leaves. Phytopath. 65:1350-1351. Kivirikko KI (1989) Protein hydroxylation: Prolyl 4-hydroxylase, an enzyme with 4 cosubstrates and a multifuntional subunits. FASEB J. 3:1609-1617. Kivirikko KI, M Liesmaa (1959) A colorimetric method for determination of hydroxyproline in tissue hydrolysates. Scand. J. Clin. Lab. Invest. 11:128-131. Klis FM, H Eeltink (1979) Changing arabinosylation patterns in wall-bound hydroxyproline in bean cultures. Planta 144:479-484. Laemmli UK, M Favre (1973) Maturation of the head of bacteriophage T4. J. Mol. Biol. 80:575-599. Lamport DTA (1963) Oxygen fixation into hydroxyproline of plant cell wall protein. J. Biol. Chem. 238:1438-1440. Lamport DTA (1964) Hydroxyproline biosynthesis: loss of hydrogen during the hydroxylation of proline. Nature 202:293-294. Lamport DTA (1965) The protein component of primary cell walls. Adv. Bot. Res. 2:151-218. Lamport DTA (1967) Hydroxyproline-O-glycosidic linkage of the plant cell wall glycoprotein extensin. Nature 216:1322-1324. Lamport DTA (1969) The isolation and partial characterization of hydroxyproline- rich glycopeptides obtained by enzymic degradation of primary cell walls. Biochemistry 8:1155-1163. Lamport DTA (1970) Cell Wall Metabolism. Ann. Rev. Plant Physiol. 21:235-270. Lamport DTA (1973) The glycopeptide linkages of extensin: O-D-galactosyl serine and O-L-arabinosy] hydroxyproline. Biogenesis of Plant Cell Wall Polysaccharides Academic Press. pp. 149-164. Academic Press, Inc. New York and London. 132 Lamport DTA (1977) Structure, biosynthesis and significance of cell wall glycoproteins. In FA Loewus & VC Runeckles, eds, Recent Advances in Phytochemistry Vol 11. pp 79-115. Plenum Publishing, New York. Lamport DTA (1978) Cell wall carbohydrates in relation to structure and function. In TA Thorpe, ed. Frontiers of Plant Tissue Culture. pp235-244. Proceedings of the 4th International Congress- Plant Tissue and Cell Culture; Int Assoc Plant Tissue Culture, Univ of Calgary, Calgary Canada. Lamport DTA (1980) Structure and function of plant hydroxyproline-rich glycoproteins. In J Preiss, ed., The Biochemistry of Plants: A comprehensive treatise, Vol 3; Carbohydrates: structure and function. pp. 501-541, Academic New York. Lamport DTA (1986) The primary cell wall: A new mode]. In RA Young & RM Rowell, eds. Cellulose: Structure, Modification and Hydrolysis. pp. 77-89; Wiley & Sons, Inc. Lamport DTA (1989) Extensin peroxidase ties the knots in the extensin network. In D. Osborne, M Jackson, eds. Signals for Cell Separation in Plants NATO ASI SER, pp. 101-113. Lamport DTA, JW Catt (1981) Glycoproteins and enzymes of the cell wall. In W Tanner & FA Loewus, Encyclopedia of Plant Physiology (New Series) Vol. 13B; Plant Carbohydrateds II: Extracellular Carbohydrates. pp.133-165. Springer, New York. Lamport DTA, L Epstein (1983) A new model for the primary cell wall: A concatenated extensin-cellulose network In DD Randall, DG Blevins, RL Larson, BJ Rapp, eds. Current TOpics in Plant Biochemistry and Physiology., pp 73-83. Proceeding of the 2"“ Annual Plant Biochem and Physio] Symp. University of Missouri-Columbia. Lamport DTA, L Katona, S Roerig (1973) Galactosylserine in extensin. Biochem J 133, 125-131. Lamport DTA, DH Miller (1971) Hydroxyproline arabinosides in the plant kingdom. Plant Physiol 48, 454-456. Lamport DTA, D Northcote (1960) Hydroxyproline in primary cell walls of higher plants. Nature 188:665-666. Lawton MA, CJ Lamb (1987) Transcriptional activation of plant genes by fungal elicitor wounding and infection. Mo]. & Cell Biol. 7:335-341. 133 Leach JE, MA Cantrell, L Sequeira (1982a) Hydroxyproline-rich bacteria] agglutinin from potato. Plant Physiol. 70:1353-1358. Leach JE, MA Cantrell, L Sequeira (1982b) A hydroxyproline-rich bacteria] agglutinin from potato: its localization by immunofluorescence. Physiol. Plant Path. 21:319-325. Li XB, M Kieliszewski, DTA Lamport (1989) A chenopod extensin lacks repetitive tetrahydroxyproline blocks. Plant Physiol. in press. Maness NO, AJ Mort (1989) Separation and quantitation of galacturonic acid oligomers from 3 to over 25 residues in length by anion-exchange high- performance liquid chromatography. Anal. Biochem. 178:248-254. Marchalonis JJ, GR Vasta, GW Warr, WC Barker (1984) Probing the boundaries of the extended immunogloban family of recognition molecules; jumping domains, convergence and minigenes. Immunology Today 5:133-142. Mazau D, MT Esquerre-Tugaye (1986) Hydroxyproline-rich glycoprotein accumulation in the cell walls of plants infected by various pathogens. Physio]. & Mol. Plant Path. 29:147-157. Mazau D, D Rumeau, MT Esquerre-Tugaye (1988) Two different families of hydroxyproline-rich glycoproteins in mellon callus. Plant Physiol. 86:540-546. McIlvaine (1921) A buffer solution for colorimetric comparison. J. Biol. Chem. 49:183-186. McNeil M, AG Darvill, P Albersheim (1980) Structure of plant cell walls X. Rhamnogalacturnonan I. A structurally complex polysaccharide in the walls of suspension-cultured sycamore cells. Plant Physiol. 66:1128-1134. McNeil M, AG Darvill, SC Fry, P Albersheim (1984) Structure and function of the primary cell walls of plant. Ann. Rev. Biochem. 53:625-663. Melon D, JC Vallee, G Vansuyt, J Prevost (1979) Acides aminis libres et proteiques des divers types de rameaux du Peri placa graeca: Role dans la circumnutation des tiges volubiles compartmentation intracellulaire. Plant Physiol. 46:68-72. Monro JA, D Penny, RW Bailey (1976) The organization and grth of primary cell walls of Lupin hypocotyl. Phytochem. 15:1193-1198. Mort AJ (1978) Partial characterization of extensin by selective degradation of cell walls. Ph.D. Dissertation, Michigan State University, East Lansing, Michigan. 134 Mort AJ, DTA Lamport (1977) Anhydrous hydrogen fluoride deglycosylates glycoproteins. Anal. Biochem. 82:289-309. Mundy J, NH Chua (1988) Abscisic acid and water-stress induce the expression of a novel rice gene. EMBO J. 7:2279-2286. Murashige T, F Skoog (1962) A revised medium for rapid growth and bioassay with tobacco tissue culture. Plant Physiol. 15:473-497. Muray RHA, DH Northcote (1978) Oligoarabinosides of hydroxyproline isolated from potato lectin. Phytochem. 17:623-629. Myllylla R, LM Schuboltz, U Weser, KI Kivirikko (1979) Involvement of superoxide in the prolyl- and lysyl- hydroxylase reactions. Biochem. Biophys. Res. Comm. 89:98-102. Neukam H, HU Markwalder (1978) Oxidative gelation of wheat flour pentosans: A new way of cross-linking polymers. Cereal Foods World 23:374-376. North ACT (1968) The structure and activity of lysozyme. In WG Crewthers, ed. Symposium on Fibrous Proteins. pp. 13-21; New York, Plenum Press. Okabayashi H, T. Isemura, S Sakakibara (1968) Steric structure of L-proline oligopeptides II. Far-ultraviolet absorption spectra and optical rotations of L- proline oligopeptides. Biopolymers 6:323-330. O'Neil MA, RR Selvandran (1980) Glycoproteins from the cell wall of Phaseolus coccineus. Biochem. J. 187:53-63. Pauling L, RB Corey (1951) Configurations of polypeptide chains with favored orientations around single bonds: Two new pleated sheets. Proc. N at]. Acad. Sci. USA 37:729-740. Piez C (1987) Collagen Types: a review. In C Piez, ed. Development and Diseases of Cartilage and Bone Matrix, New Series, 46. UCLA Symposium on Molecular and Cellular Biology. Alan R. Liss, Inc. pp. 1-19. Pike CS, H Un, JC Lystash, AM Showalter (1979) Phytochromes control of cell wall-bound hydroxyproline content in etiolated pea epicotyls. Plant Physiol. 63:444-449. Pope DG (1977) Relationships between hydroxyproline-containing proteins secreted into the cell wall and medium by suspension-cultured Acer pseudo platanus cells. Plant Physiol. 59:894-900. 135 Prassad TK, MG Cline (1987) Shoot inversion inhibition of stem elongation in Pharbitis' nil. Plant Physiol. 85:104-108. Preston RD (1974) The Physical Biology of Plant Cell Walls. pp.1-50. Chapman & Hall. Preston RD (1979) Polysaccharide conformation and cell wall function. Ann. Rev. Plant Physiol. 30:55-78. Preston RD, JH Wardrop (1949) The submicroscopic organization of the walls of conifer cambium. Biochim. Biophys. Acta 3:549-559. Rackis JJ, RL Anderson, HA Sasame, AK Smith, CH van Etten (1961) Amino (acids in soybean hulls and oilmeal fractions. Agric. Food Chem. 9:409-412. Ridge 1, D Osborne (1970) Hydroxyproline and peroxidases in cell walls of Pisum sativum: Regulation by ethylene. J. Exp. Bot. 21:843-856. Roberts K (1974) Crystalline glycoprotein cell walls of algae: their structure, composition and assembly. Phil. Trans. R. Soc. B. 268:129-146. Roby D, A Toppan, MT Esquerre-Tugaye (1985) Cell surfaces in plant- microorganism interactions V. Elicitors of fungal and of plant origin trigger the synthesis of ethylene and of cell wall hydroxyproline-rich glycoproteins in plants. Plant Physiol. 77:700-704. Saad B, G Corradin, HR Bosshard (1988) A monoclonal antibody recognizes a conformational epitope in a random coil protein. Eur. J. Biochem. 178:219- 224. Sadava D, MJ Chrispeels (1971) Hydroxyproline biosynthesis in plant cells. Peptidylproline hydroxylated from carrot disks. Biochim.Biophys.Acta 227:278- Sadava D, MJ Chrispeels (1973) Hydroxyproline-rich cell wall protein (extensin): Role in the cessation of elongation in excised pea epicotyls. Dev. Biol. 30:49- 55. Sanger MP, DTA Lamport (1983) A microapparatus for liquid hydrogen fluoride solvolysis. Sugar and amino sugar composition of Erysiphe graminis and Triticum aestivum cell walls. Anal. Biochem. 128:66-70. Sauer A, DG Robinson (1985) Intracellular localization of posttranslational modifications in the synthesis of hydroxyproline-rich glycoproteins. Peptidyl proline hydroxylation in maize roots. Planta 164:287-291. 136 Schlesinger DH, DI Hay (1986) Complete covalent structure of a proline-rich phosphoprotein, PRP-1, an inhibitor of calcium phosphate crystal grth from human paratid saliva. Int. J. Peptide Res. 27:373-379. Schlipfenbacher R, S Wenz], F Lottospeich, M. Sumper (1986) An extremely hydroxyproline-rich glycoprotein is expressed in inverting Volvox embryos. FEBS Lett. 209:57-62. Segrest JP, RL Jackson, EP Andrews, VT Machesi (1971) Human erythrocyte membrane glycoprotein: a re-evaluation of the molecular weight as determined by SDS polyacrylamide gel electrophoresis. Biochem. Biophys. Res. Commun. 44:390-395. Selvandran RR (1975) Cell wall glycoproteins and polysaccharides of parenchyma of Phaseolus coccineus. Phytochem. 14:2175-2180. Seto Y, T Shinohara (1989) Size fractionation of oligosaccharides by liquid chromatography on a cation-exchange column. J. of Chrom. 464:323-331. Showalter AM, JN Bel], CL Cramer, JA Bailey, JE Varner, CJ Lamb (1985) Accumulation of hydroxyproline-rich glycoprotein mRNAs in response to fungal elicitors and infection. Proc. Natl. Acad. Sci. USA 82:6551-6555. Showalter AM, JE Varner (1989) The biology and molecular biology of plant hydroxyproline-rich glycoproteins. In A Marcus, ed. The Biochemistry of Plants: A Comprehensive Treatise Vol. 15, Molecular Biology. Academic Press. Showalter A , D Rumeau (1989) Molecular biology of plant cell wall hydroxyproline-rich glycoproteins. In WS Adair, RP Mecham, eds. Recognition and Assembly of Animal and Plant Extracellular Matrix, in press. Smith JJ, EP Muldoon, DTA Lamport (1984) Isolation of extensin precursors by direct elution of intact tomato cell suspension cultures. Phytochemistry 23:1233-1239. Smith JJ, EP Muldoon, JJ Willard, DTA Lamport (1986) Tomato extensin precursors P1 and P2 are highly periodic structures. Phytochemistry 25:1021- 1030. Stafstrom J, LA Staehelin (1986) Cross-linking patterns in salt-extractable extensin from carrot cell walls. Plant Physiol. 81:234-241. Stebbins GL (1974) Flowering plants. Evolution above the species level. Harvard Univ. Press. 137 Stermer BA, R Hammerschmidt (1987) Association of heat shock induced resistance to disease with increased accumulation of insoluble extensin and ethylene synthesis. Physio]. & Mol. Plant Path. 31:453-461. Steward FC, HW Israel, MM Salpeter (1967) The labeling of carrot cells with H- proline: Is there a cell wall protein? Proc. Natl. Acad. Sci. USA 58:541-544. Steward FC, JK Pollard (1958) 1‘C-Proline and hydroxyproline in the protein metabolism of plants. Nature 182:828-832. Steward FC, JF Thompson (1954) Proteins and protein metabolism in plants. In H Neurath, K Bailey, eds., The Proteins Vol. IIA; pp. 513-594. Academic, New York. Stiefel V, L Perez-Grau, F Albericio, E Geralt, L Ruiz-Avila, M Dolors Ludevid, P Puigdominech (1988) Molecular cloning of cDNAs encoding a putative cell wall protein from Zea ma ys and immunological identification of related polypeptides. Plant Mol. Biol. 11:483-493. Stuart DA, JE Varner (1980) Purification and characterization of a salt-extractable hydroxyproline-rich glycoprotein from aerated carrot discs. Plant Physiol. 66:787-792. Takano E, M Masatoshi, M Hirotaka, M Hatanaka, T Marti, K Titani, R Kannagi, T Ooi, T Murachi (1988) Pig heart calpastatin: Identification of repetitive domain structures and anomalous behavior in polyacrylamide gel electrophoresis. Biochem. 27:1964-1972. Tahnadge KW, K Keegstra, WD Bauer, P Albersheim (1973) The structure of plant cell walls I. The macromolecular components of the walls of suspension- cultured sycamore cells with a detailed analysis of the pectic polysaccharides. Plant Physiol. 51:158-173. Tanaka M, K Sato, T Uchida (1981) Plant prolyl hydroxylase recognizes poly(L- proline)II helix. J. Biol. Chem. 256:11397-11400. Tierney ML, J Wiechart, D Pluymers (1988) Analysis of the expression of extensin and p33-related cell wall proteins in carrot and soybean. Mol. Gen. Genet. 211:393-399. Tiffany ML, S Krimm (1969) Circular dichroism of the "random" polypeptide chain. Biopolymers 8:347-359. Thimann KV, J Bonner (1933) The mechanism of the action of the growth substances of plants. Proc. Roy. Soc. Ser. B 113:126-149. 138 Towbin H, T Staehlin, J Gordon (1979) Electrophorectic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc. Natl. Acad. Sci. USA 76:4350-4354. Tran Thanh Van K, P Toubart, A Cousson, AG Darvill, DJ Gollin, P Chelf, P Albersheim (1985) Manipulation of the morphogenetic pathways of tobacco explants by oligosaccharins. Nature 314:615-617. Tripp VW, AT Moore, ML Rollins (1951) Some observations on the constitution of the primary wall of the cotton fibre. Text. Res. J. 21:886-894. Tupper-Carey RM, JH Priestley (1924) The composition of the cell-wall at the apical meristem of stem and root. Proc. Royal Soc. Ser. B 95:109-131. Tyler JM, D Branton (1980) Rotary shadowing of extended molecules dried form glycerol. J. Ultrastruct. Res. 71:95-102. Vandekerckhove JS, DA Kaiser, TD Pollard (1989) Acanthamoeba actin and profilin can be cross-linked between glutamic acid 364 of actin and lysine 115 of profilin. J. Cell Biol. 109:619-626. van Etten CH, RW Miller, FR Earle, IA Wolff, Q Jones (1963) Hydroxyproline content of seed meals and distribution of the amino acid in kemal, seed coat and pericarp. Agric. Food Chem. 9:433-436. van Holst GJ, JE Varner (1984) Reinforced polyproline II conformation in a hydroxyproline-rich cell wall glycoprotein form carrot root. Plant Physiol. 74:247-251. Varner JE, GI Cassab (1986) A new protein in petunia. Nature 323:110. Whitmore FW (1978) Lignin-protein complex catalyzed by peroxidase. Plant Sci. Letters 13:241-245. Wingate VPM, MA Lawton, CJ Lamb (1988) Glutathione causes a massive and selective induction of plant defense genes. Plant Physiol. 87:206-210. Wong RSC, A Bennick (1980) The primary structure of a salivary calcium-binding proline-rich phosphoprotein (Protein C), a possible precursor of a related salivary Protein A. J. Biol. Chem. 255:5943-5948. Wong RSC, T Hoffman, A Bennick (1979) The complete primary structure of a proline-rich phosphoprotein from human saliva. J. Biol. Chem. 254:4800-4808. 139 Yamada KM (1983) Cell surface interactions with extracellular materials. Ann. Rev. Biochem. 52:761-799. Yokotsuka K, T Kushida (1983) An improved method for fluorometric amino acid analysis including proline and hydroxyproline.J. Ferment. Techno]. 61:1-6 ‘ ‘”‘llllllllllllllllS