Pearl millet cysteine protease inhibitor

Evidence for the presence of two distinct sites responsible for anti-fungal and anti-feedent activities


P. K. Ranjekar, Plant Molecular Biology Unit, Division of Biochemical Sciences, National Chemical Laboratory, Pune-411 008 India. Fax: + 91 020 284032, Tel.: + 91 020 393400, E-mail:


Recently, pearl millet cysteine protease inhibitor (CPI) was, for the first time, shown to possess anti-fungal activity in addition to its anti-feedent (protease inhibitory) activity [Joshi, B.N. et al. (1998) Biochem. Biophys. Res. Commun. 246, 382–387]. Characterization of CPI revealed that it has a reversible mode of action for protease inhibition. The CD spectrum exhibited a 35% α helix and 65% random coil structure. The intrinsic fluorescence spectrum was typical of a protein devoid of tryptophan residues. Demetallation of Zn2+ resulted in a substantial change in the secondary and tertiary structure of CPI accompanied by the complete loss of anti-fungal and inhibitory activity indicating that Zn2+ plays an important role in maintaining both structural integrity and biological function. The differential response of anti-fungal and inhibitory activities to specific modifiers showed that there are two different reactive sites associated with anti-fungal and anti-feedent activity in CPI located on a single protein as revealed from its N-terminal sequence data (AGVCYGVLGNNLP). Modification of cysteine, glutamic/aspartic acid or argnine resulted in abolition of the anti-fungal activity of CPI, whereas modification of arginine led to an enhancement of the inhibitory activity in solution. Modification of histidine resulted in a twofold increase in the protease inhibitory activity without affecting the anti-fungal activity, whereas modification of serine led to selective inhibition of the protease inhibitory activity. The differential nature of the two activities was further supported by differences in the temperature stabilities of the anti-fungal (60 °C) and inhibitory (40 °C) activities. Binding of papain to CPI did not abolish the anti-fungal activity of CPI, supporting the presence of two active sites on CPI. The differential behavior of CPI towards anti-fungal and anti-feedent activity cannot be attributed to changes in conformation, as assessed by their CD and fluorescence spectra. The interaction of CPI modified for arginine or histidine with papain resulted in an enhancement of CPI activity accompanied by a slight decrease in fluorescence intensity of 15–20% at 343 nm. In contrast, modification of serine resulted in inhibition of CPI activity with a concomitant increase of 20% in the fluorescence intensity when complexed by the enzyme. This implies the involvement of enzyme-based tryptophan in the formation of a biologically active enzyme–inhibitor complex. The presence of anti-fungal and anti-feedent activity on a single protein, as evidenced in pearl millet CPI, opens up a new possibility of raising a transgenic plant resistant to pathogens, as well as pests, by transfer of a single CPI gene.


cysteine protease inhibitor






5,5′-(2-nitrobenzoic acid)




poly(vinylidene difluoride)


ribosome-inactivating proteins


Woodward’s reagent K

Pests and pathogens are major constraints to plant growth and development, resulting in heavy losses in crop yield and quality. Since the use of chemical fungicides has a deleterious effect on human health, a recent trend is to use other, safer strategies to enhance the defense mechanisms of crops. Anti-fungal proteins such as chitinases, glucanases, ribosome-inactivating proteins (RIPs), protease inhibitors and permatins play an important role in the defense of a plant against pathogen invasion. They have vast potential either for developing transgenic plants or for direct use as fungicides. Protease inhibitors, in particular, have attracted a lot of attention since they have been shown to inhibit proteases that occur in many species of herbivorous insect as well as fungal pathogens [1,2]. Serine protease inhibitors have been shown to inhibit growth of Botrytis cinerae, a polyphagous fungus, Fusarium solani f. sp. pisi, a pea pathogen, and Alternaria brassicicola, a cabbage pathogen [2]. Expression of a cowpea trypsin inhibitor gene in tobacco to increase its resistance to herbivorous insects was the first example of the involvement of inhibitors in a plant’s defense action [3], this was followed by many such extensive studies in various plant species [4–7]. The role of the cysteine protease inhibitor (CPI) as an insecticidal protein has been reported previously by Wolfson and Murdock [8]. Oryzacystatin (rice CPI) I and II have been found to be responsible for the growth retardation of different species of bean insect pest [9]. Low doses of 4-[2S,3S)-3-carboxyoxiran-2-ylcarbonyl-l-leucylamido]butylguanidine (E-64), given through the diet, have been shown to cause late oviposition and few eggs in females, and to delay completion of larval development in the Mexican bean beetle [8]. The high sulfhydryl content of many serine protease inhibitors makes them susceptible to alkylation by plant-derived quinones. Hence proteins with a low potential for alkylative inactivation may be appropriate candidate proteins for developing host-plant resistance [10]. CPI, with its low disulphide bridge content, would seem to be a good candidate for exploring its potential as a defense protein.

In our earlier work on pearl millet, we reported the presence of an inhibitor having cysteine protease inhibitory (anti-feedent) activity, as well as an anti-fungal activity [11]. Because such inhibitors have the potential to improve plant resistance against pests and pathogens, a detailed characterization of the CPI from pearl millet was undertaken to understand the basis of anti-fungal and inhibitory activities. In this paper, we investigate the role of Zn2+ in anti-fungal and anti-feedent activities and provide evidence for the presence of two distinct sites for the two activities.

Materials and methods

Seed material and fungal isolate

Pearl millet seeds (cultivar 81B) were obtained form ICRISAT (Hyderabad, Andhra Pradesh, India). The fungal strain Trichoderma reesei (1186) was from in-house Culture Collection Unit, National Collection of Industrial Microorganisms (NCIM), NCL, Pune, India.

CPI assay

The anti-feedent activity of CPI was monitored by assaying the inhibition of papain activity by CPI. Papain activity was assayed by a modication of the method described by Murachi [12] using casein as a substrate. Papain (55 µg in 0.1 m Tris/HCl buffer, pH 7.8, 0.5 mm cysteine, 0.2 mm EDTA) and inhibitor extract (50 µg) were pre-incubated at room temperature for 15 min. The reaction mixture (150 µL) was then added to tubes containing 2.0 mL of 0.5% casein (casein prepared in 0.1 m Tris/HCl, pH 7.8 containing 0.2 mm EDTA and 0.5 mm cysteine) at 37 °C. The assay was carried out at 37 °C for 30 min and the reaction was terminated by the addition of 3.0 mL of 5% trichloroacetic acid. The absorbance of the supernatant was measured at 280 nm after 30 min. One unit of inhibitor activity was defined as the decrease by one unit of absorbance of trichloroacetic acid-soluble casein hydrolysis product liberated by protease action at 280 nm·h−1 at 37 °C in a given assay volume.

Purification of CPI

CPI was purified from defatted seed meal of pearl millet using ammonium sulfate precipitation, CM–Sephadex and SP–Sepharose, cation-exchange column chromatography procedures as described previously [11].


Anti-fungal activity was checked against Trichoderma reesei as described by Roberts and Selitrennikoff [13]. Essentially, the method consisted of growing the fungal spore suspension (1.0–2.5 × 106 spores·mL−1) on Petri dishes containing nutrient agar for 24 h, followed by application of the protein solution to be tested on sterile filter paper disks. Clearance zones were seen around the discs having anti-fungal protein solution. The anti-fungal assay was carried out for a 1 : 1 complex of CPI with papain where the clearance zones were observed after 24 h and native CPI and papain alone served as controls.

Characterization of CPI

Inhibition kinetics was studied using various concentrations of casein at a fixed concentration of CPI (40 µg). The experiment was repeated by taking one more concentration of CPI (60 µg). A Lineweaver–Burk curve was plotted and Ki and Vmax were calculated from the equation. A CPI solution assay was also carried out at eight different concentrations of CPI to check the concentration required for maximum inhibition at temperatures ranging from 30 to 60 °C. Similarly, optimum pH was studied by performing the assay in 0.1 mm sodium acetate buffer at pH 5.0–7.0 and 0.1 mm Tris/HCl buffer at pH 8.0–10.0 containing 0.2 mm EDTA and 0.5 mm cysteine. The graph of percentage inhibition against temperature and pH was plotted and the optimum temperature and pH were calculated. Aliquots of CPI (100 µL; 100 µg) were incubated at temperatures of 10–70 °C for 10 min and checked for anti-fungal and CPI activities using their respective assays. The graphs of anti-fungal and residual CPI activities were plotted. CPI was blotted onto a poly(vinylidene difluoride) (PVDF) membrane using 10 mm Caps buffer containing 10% methanol. Partial N-terminal sequence was carried out at Commonwealth Biotechnology Inc. (USA), using automated Edman degradation.

Metal chelation of CPI using EDTA

CPI (500 µg) was dialyzed extensively against 10 mm EDTA for chelation of Zn2+. The excess of EDTA was removed by dialyzing against deionized water. The Zn2+ chelation by EDTA was checked by taking atomic absorption spectra at 213 nm using Hitachi-Z-8000 polarized Zeeman spectrophotometer. Anti-fungal and CPI activities were checked after demetallation.

Circular dichroism

The CD spectra were recorded on a Jasco 715CD spectropolarimeter with a cell length of 0.1 cm at 25 °C. A 100 µg aliquot of each native and modified protein in sodium phosphate buffer, pH 8.0 was used for CD analysis. The CD data were the average of three spectra. CD spectra were analyzed using a program based on ‘Unintelligent search technique for minimum error’.

Intrinsic fluorescence spectroscopy

Fluorometry was performed on a Perkin-Elmer Spectrofluorometer LS-5B using a slit width of 5 mm and an excitation wavelength of 278 nm. Sodium phosphate buffer, pH 8.0, containing the respective modifiers served as controls. CPI having an A280 value < 0.1 was used for the emission spectra. In the enzyme–inhibitor complex a mixture of (1 : 1) CPI and papain was taken.


Biochemical properties of CPI

Protease inhibitors from plants and microorganisms are characterized by either a reversible or irreversible mechanism [21]. The kinetics of inhibition of papain by CPI revealed that it has a reversible mechanism of action (Fig. 1). Inhibition of casein degradation by papain occurs at a very low concentration of CPI and the apparent Ki was found to be 6.5 nm under the assay conditions. The low Ki value implies that CPI is a powerful inhibitor of papain. Maximum inhibition (55%) of papain was obtained by 50 µg of CPI and remained constant even after increasing the concentration of CPI to 200 µg. The effect of pH on CPI activity was examined at a pH range 5.0–10.0. The optimum pH for CPI activity was 8.0. CPI showed maximum activity at 37 °C, which decreased drastically at 40 °C. The optimum pH and temperature of CPI were found to be in accordance with those reported for other serine or cysteine protease inhibitors [22,23]. The inhibitor activity of CPI was stable up to 30 °C, whereas anti-fungal activity was stable up to 60 °C but lost at 70 °C. The sequence of first 13 amino acids at the N-terminal end of CPI was AGVCYGVLGNNLP and showed 80% homology with the N-terminal sequence of β-1,3-glucanase, a known anti-fungal protein from barley.

Figure 1.

Inhibitor kinetics of CPI. Inhibitor kinetics was studied as described in Materials and methods and Ki was calculated from the Lineweaver–Burk plot of CPI.

Zn2+ is essential for anti-fungal and protease inhibitory activity of CPI

The atomic absorption spectrum of CPI showed the presence of one atom of Zn2+ per mole of CPI. The Zn2+ atom of CPI was chelated by treatment with EDTA and its effect on CPI and anti-fungal activity was checked. Metal chelation led to the complete loss of both anti-fungal and protease inhibitory activities, indicating a key role for Zn2+ in these activities. Supplementation of Zn2+ did not restore the CPI activity. The secondary structure of native CPI in the far-UV region showed a positive band at 198 nm, a cross-over from positive to negative at 205 nm and negative maxima at 217 and 222 nm (Fig. 2). Structural analysis revealed a 35% α-helical and 65% random-coil structure for CPI. Analysis of metal-chelated CPI using CD spectroscopy indicated a significant change in the secondary structure as revealed by the altered structural configuration viz. 25% α helix, 65% β sheet and 15% random-coil structure (Fig. 2). The emission spectrum of CPI at an excitation wavelength of 278 nm revealed a major peak at 307 nm and a small shoulder at 334 nm, indicating that there is no tryptophan fluorescence and the emission spectrum reflects a contribution only from the tyrosine residues of CPI. The emission spectrum of metal-chelated CPI remained unchanged except for the increase in fluorescence intensity (Fig. 3). The emission spectrum of the complex of enzyme (papain) with native CPI showed a 53% decrease in fluorescence intensity at 343 nm compared with enzyme alone (Fig. 4). The intrinsic fluorescence spectroscopy of metal-chelated CPI when complexed with papain resulted in a spectrum having no peak at 307 nm and a 27% decrease in fluorescence intensity.

Figure 2.

Effect of metal chelation on the secondary structure of CPI. The CD spectra of native CPI (a) and metal-chelated CPI (b) were recorded in the far-UV region on a Jasco 715-CD spectropolarimeter.

Figure 3.

Fluorescence spectra of native andmetal-chelated CPI. Intrinsic fluorescence spectroscopy was carried out on a Perkin-Elmer spectrofluorometer. Fluorescence spectra of native CPI (a) and demetallated CPI (b) are shown.

Figure 4.

Emission spectra of papain complexedwith native and demetallated CPI. Fluorescence spectra of papain (a), complexed with native CPI (b) and demetallated CPI (c) were recorded at an excitation wavelength of 278 nm.

Two distinct sites for anti-fungal and protease inhibitory activity

Seven different amino acids were individually modified using specific chemical modifiers (Table 1) and the effect of modification on anti-fungal and CPI activities was determined. Protease inhibitor activity of the unmodified CPI was considered as 100% CPI activity and relative activities after chemical modifications were expressed accordingly (Table 2). With the exception of tyrosine and lysine residues, modification of other amino acids resulted in a change in either anti-fungal or CPI activity.

Table 1. Chemical modification of various amino acid residues in CPI. Chemical modifications of amino acids of CPI were carried out using different chemical modifiers under their respective optimum reaction conditions. Papain was also modified with respective modifiers. A 50 µg aliquot of purified CPI was used for each modification. Anti-fungal and cysteine protease inhibitor activities were estimated after modification of CPI with respective modifiers.
Amino acid
Reaction conditions Ref.
N-Ethylmaleimide (10 mm)Cysteine30 °C0.1 m Tris/HCl buffer (pH 7.8) 60 min[14]
Succinic anhydride (10 mm)Lysine30 °C0.1 m Sodium carbonate buffer (pH 8.0)120 min[15]
N-acetylimidazole (10 mm)Tyrosine25 °C0.05 m Sodium borate buffer (pH 7.5) 60 min[16]
Woodward’s reagent K (10 mm)Aspartic/glutamic acids25 °C0.1 m Tris/HCl buffer (pH 6.0) 20 min[17]
Diethylpyrocarbonate (10 mm)Histidine30 °C0.1 m Tris/HCl buffer (pH 7.5) 30 min[18]
p-Nitrophenylglyoxal (10 mm)Arginine30 °C0.05 m Sodium pyrophosphate buffer
containing 150 mm sodium ascorbate
(pH 8.0)
 30 min [19]
PhCH2SO2F (20 mm)Serine25 °C0.05 m Tris/HCl buffer (pH 7.8)120 min[20]
Table 2. Effect of chemical modification on anti-fungal and CPI activities. Purified CPI (50 µg) was modified with different modifiers under their respective reaction conditions. The anti-fungal and CPI activities were assayed as described in Materials and methods.
Amino acids modifiedAnti-fungal activity (%)CPI activity (%)
Unmodified CPI100100
Cysteine  0 94
Tyrosine100 90
Aspartic acid/Glutamic acid  0 90
Arginine  0158
Serine100  0

Effect of chemical modification on CPI activity. Modification of cysteine, lysine, tyrosine and carboxylic acids had no effect on the CPI activity, whereas modification of arginine and histidine resulted in the activation of CPI activity. Increasing concentrations of p-nitrophenylglyoxal (pNPG) or diethylpyrocarbonate (DEPC) resulted in an increase in inhibition of papain activity (Fig. 5A,B). The number of arginine residues modified was calculated as described by Yamasaki et al. [19], whereas hisidine modification by DEPC of CPI was monitored by the increase in absorbance at 240 nm. Complete inhibition was obtained at 25 mm of pNPG or DEPC. In contrast, serine modification by PhCH2SO2F resulted in the loss of only CPI activity at a higher concentration (20 mm) of the reagent.

Figure 5.

Effect of increasing concentrations of pNPG and DEPC on CPI activity. The CPI was treated with increasing concentrations of pNPG (A) and DEPC (B) under their respective reaction conditions.

In order to investigate whether the enhancement in CPI activity was due to the structural changes in the protein, CD and emission spectra of modified CPI were analyzed. Table 3 summarizes the effect of chemical modification on the structure of CPI. As seen from Table 3, the CD spectrum of CPI modified at arginine residues showed no change in the extent of the α helix and random-coil structure. Modification of histidine resulted in a 14% decrease in α helix and a subsequent increase in random coil. Emission spectra of CPI modified for arginine or histidine residues showed a slight decrease in fluorescence intensity compared with unmodified CPI (Fig. 6). Thus, the increase in CPI activity is due to modification of arginine or histidine residues and cannot be attributed to the change in the three-dimensional structure of the protein. The secondary structure of CPI modified for serine residues exhibited a 16% increase in α-helical content and a subsequent decrease in random coils, with the complete loss of fluorescence (Fig. 7). However, there were no noticeable differences in the structures of the CPIs modified for cysteine residue as revealed by its CD and fluorescence emission spectra (Fig. 6).

Table 3. CD analysis of chemically modified CPI. Amino acids were chemically modified under their respective reaction conditions and CD spectra were recorded.
Amino acid modified% α-helix% Random coil
Unmodified CPI3565
Figure 6.

Emission spectra of chemicallymodified CPI. Emission spectra of native CPI (a), hisidine-modified CPI (b), cysteine-modified CPI (c) and arginine-modified CPI (d).

Figure 7.

Fluorescence spectra of serine-modified CPI. Emission spectra of native CPI (a) and serine-modified CPI (b).

Because the reaction of CPI modified at arginine, histidine or serine residues showed variation in CPI activity, the interactions of CPI modified for these amino acids with papain were analyzed by monitoring their emission spectra at an excitation wavelength of 278 nm. The reaction of CPI modified for arginine and histidine with papain resulted in a slight decrease in fluorescence intensity of 15–20% (Fig. 8). In contrast, modification of serine resulted in a 20% increase in fluorescence intensity (Fig. 8) suggesting that during the formation of the enzyme–inhibitor complex, there is a change in the environment of the tryptophan residues making them inaccessible for casein degradation. The presence of trypophan residues near the active site of papain has been reported by Galley and Stryer [24].

Figure 8.

Fluorescence spectra of chemicallymodified CPI complexed with papain. Emmision spectra of the complex of papain with native CPI (a), histidine-modified CPI (b), arginine-modified CPI (c) and serine-modified CPI (d).

Effect of amino acid modification on anti-fungal activity. Modification of the cysteine, aspartic/glutamic acids or arginine residues resulted in a loss of anti-fungal activity of CPI as monitored by the inability to inhibit the growth of Trichoderma reesei. Modification of cysteine was confirmed by titration of modified CPI with 5,5′-(2-nitrobenzoic acid) (Nbs2). An increase in the absorbance at 250 and 340 nm of the CPI modified by WRK indicated modification of the carboxylic acids. CPI complexed with papain was also assessed for its anti-fungal activity. The CPI–papain complex showed the formation of clearance zone as observed for native CPI indicating that its anti-fungal activity is not affected, although the active site for protease inhibition is blocked.

The experimental results therefore indicate that there are two distinct sites for anti-fungal and CPI activities.


Plant seeds possess specific defense proteins whose function is to protect them from pest and pathogen attack. These proteins are of two types, namely anti-feedent and anti-fungal. The role of CPIs as an insecticidal protein has been documented previously [8]. In addition, we demonstrated that CPI from pearl millet exhibits the novel property of possessing anti-fungal activity [11], and is thus an ideal candidate defense protein for conferring pest, as well as pathogen, resistance to a plant. The objective of this study was to deduce the structure–activity relationship of CPI with special reference to its protease inhibitory and anti-fungal activities.

The removal of a Zn2+ atom from CPI resulted in the loss of anti-fungal and protease inhibitor activity. Disruption in the secondary structure of the demetallated inhibitor implied the role of Zn2+ in maintaining the structural integrity of CPI. Demetallation is also followed by a significant increase in the fluorescence intensity of CPI, suggesting that binding of the metal ion results in the quenching of tyrosine fluorescence. The increase in the tryptophan fluorescence of papain at 343 nm indicates the inability of demetallated CPI to bind with the active site of the enzyme. Thus, the metal ion is important in maintaining the biologically active conformation of CPI to exhibit the anti-fungal as well as anti-feedent activity.

In proteins, the arrangement of functional groups in a particular manner is an essential requirement for their activity. In chemical modification, a chemical reagent binds covalently to specific amino acid side chains of a protein and may produce changes in the properties/activity of a protein. Attempts to correlate these changes with catalytic activity have been made previously [25–27]. Although, several anti-fungal proteins of plant origin have been identified, details of their mode of action are unexplored. In this study, based on the structure–function relationship of chemically modified CPI, we provide evidence for the presence of two distinct sites responsible for anti-fungal and anti-feedent activities. Anti-fungal activity was abolished after modification of cysteine, arginine or aspartic/glutamic acid residues, whereas CPI activity was selectively enhanced by modification of histidine or arginine residues. In spite of the chemical modifications the basic pattern of the CPI spectra remained unaltered as examined by CD and intrinsic fluorescence spectroscopy. The differential inhibition of CPI and anti-fungal activity cannot be attributed to major changes in conformation. In the majority of specific serine protease inhibitors, the inhibitory action is localized to a specific reactive site situated within a loop closed by a disulfide bridge [28]. Involvement of an arginine residue in α amylase inhibitor activity in the case of barley has been reported by Abe et al.[29]. An increase in CPI activity following histidine and arginine modification was associated with a decrease in fluorescence intensity. The binding of papain to CPI did not abolish the anti-fungal activity of CPI corroborating the presence of two active sites on CPI. The differential nature of these two activities is also reflected in the difference in their temperature stabilities. The two activities reside in two different regions of the same protein as indicated by the single N-terminal sequence for CPI. Small polypeptide regions, such as systemin, are known to play a broad role in signaling, the environmental stress response and regulating plant growth and development [30]. Chitinases are known to possess two different sites viz. the catalytic site and the chitin binding site on the same protein molecule [31]. The two functional domains, namely the recognition domain and the pore-forming domain, are also present in thaumatin [32].

The structural environment essential for the function of anti-fungal proteins is unknown. Detailed structure–function analysis of these proteins is important to facilitate their use in genetic engineering of plants for acquiring resistance against fungal diseases. Cysteine proteases are the evolutionary adaptations in coleopteran insects, which feed naturally on plant material high in serine protease inhibitors [1]. The presence of protease inhibitory activity and anti-fungal activity in a single protein makes CPI an ideal candidate for developing transgenic plants for both fungal and insect resistance by minimum genetic manipulations.


B. N. J. thanks the Council of Scientific and Industrial Research, New Delhi for the award of Senior Research Fellowship. The supply of pearl millet seed material from ICRISAT is gratefully acknowledged.