Evaluating the role of a trypsin inhibitor from soap nut (Sapindus trifoliatus L. Var. Emarginatus) seeds against larval gut proteases, its purification and characterization

Extraction and Purification

of SNTI The Soap Nut Trypsin Inhibitor was isolated and purified from soap nut seeds
(Sapindus trifoliatus L.) according to the procedure adopted by Annapurna and Siva
Prasad 42] and the results are shown in the Table- 2.

The extraction procedure was carried out maintaining physiological conditions and
ice cold acetone was used to remove lipids. The endosperm was collected from the seeds
after the removal of the hard seed coat and 25 g of the endosperm was homogenized
with 200 ml of 0.1M sodium phosphate buffer, pH 7.6 and then made up to 250 ml with
the same buffer. The extract was then centrifuged at 2500 rpm for 15 minutes at 4
ºC and the supernatant (230 ml) was used in further steps.

Table 2. Summary of purification of soap nut seed protease inhibitor

The supernatant (230 ml) was treated with 50 % ice cold acetone (1:5 V) and the resultant
mixture was centrifuged at 2500 rpm for 15 minutes at 4 ºC to remove lipids. The resultant
defatted solution was subjected to ammonium sulphate precipitation.

To the supernatant (200 ml) from acetone fractionation, solid ammonium sulphate (62.6
g) was added gradually with constant stirring at 4 ºC to obtain 50 % saturation. The
mixture was allowed to stand overnight at 4 ºC. The precipitate was collected by centrifugation
at 2500 rpm for 15 minutes at 4 ºC, then dissolved in 30 ml of 0.1 M sodium phosphate
buffer pH 7.6 and dialyzed against the same buffer.

Proteins have numerous functional groups that can have both positive and negative
charges. Ion exchange chromatography separates proteins with regards to their net
charge. If a protein has a net positive charge at pH 7, then it will bind to a column
of negatively charged beads, whereas a negatively charged protein would not. By changing
the pH so that the net charge on the protein is negative, it too will be eluted.

The dialyzed sample (172 mg) was loaded on a CM-Cellulose column (2×80cm) previously
equilibrated with 0.1M sodium phosphate buffer pH 7.6. After washing with 250 ml of
the equilibration buffer, the following stepwise elution was performed with 200 ml
each of 0.1M, 0.2M, 0.3M, 0.4M and 1.0 M NaCl in 0.1 M phosphate buffer pH 7.6. Fractions
of 5 ml were collected at a flow rate of 60 ml per hour. These fractions were assayed
for protein by measuring their absorbance at 280 nm as well as the inhibitory activity
against trypsin using BAPNA as the substrate. The elution profile of CM-Cellulose
chromatography for the inhibitor is shown in Fig. – 1. The fractions containing trypsin inhibitory activity (fractions 42-48) were pooled,
dialyzed against distilled water at 4 ºC and lyophilized. The protein yield from ion
exchange chromatography was 112 mg.

The sample from ion exchange chromatography (110 mg) was dissolved in 0.1 M phosphate
buffer pH 7.6 and was loaded on Sephadex G-100 column (1.8 × 30 cm) which was previously
equilibrated with 0.1 M phosphate buffer, pH 7.6. The inhibitor was eluted with the
same buffer. 2 ml fractions were collected at a flow rate of 12 ml per hour and the
protein was monitored by measuring the absorbance at 280 nm. The trypsin inhibitory
activity of the fractions was assayed using BAPNA as the substrate.

Fig. 1. Ion exchange chromatography of SNTI on CM-Cellulose. One hundred seventy two milligram
of the ammonium sulphate fractionated sample (0–50 %) was applied on to the column
(2?×?80 cm) in 0.1 M sodium phosphate buffer (pH 5.8) and the adsorbed proteins were
eluted with stepwise gradient in the buffer. Fractions of 5 ml were collected at a
flow rate of 60 ml per hour. The protein was monitored by absorbance at 280 nm. *
When the elution was done with a gradient of 0.1 to 1.0 M NaCl a single but broad
peak was obtained (Results not shown. To obtain a sharp peak, the elution was performed
using stepwise gradient

The elution profile of the gel permeation chromatography is shown in Fig. – 2. A single protein peak with corresponding trypsin inhibitory activity was observed.
The fractions (8 – 12) containing the trypsin inhibitory activity were pooled, dialyzed
against distilled water at 4 ºC and lyophilized. The yield of protein after gel permeation
chromatography was 52 mg. This preparation was stored at 0 ºC. The preparation thus
stored, showed full activity even after three months. By this procedure about 52 mg
of the inhibitor was obtained and the final yield was about 20.9 %.

Fig. 2. Gel filtration of SNTI on Sephadex G100. One hundred ten milligram of lyophilized
preparation was applied to the Sephadex G-100 column 2 × 80 cm in 0.1 M phosphate
buffer pH 7.6 and eluted with the same buffer. 2 ml fractions were collected at a
flow rate of 12 ml per hour. The protein was monitored at 280 nm. Protease inhibitory
activity was followed using BAPNA as the substrate

SNTI was analyzed using reverse phase HPLC to confirm its purity. HPLC analysis revealed
a single peak (result not shown). The methodological procedure resulted in high purification
with a 20.92 % yield.

A sharp band was obtained on 12 % slab gel at pH 8.3 signifying the homogeneity of
the purified SNTI (Fig. – 3). SNTI did not respond to PAS (Periodic Acid Schiff’s) stain suggesting it to be
a non-glycoprotein.

Fig. 3. Polyacrylamide Gel Electrophoresis. 1 – Crude Extract. 2 – Acetone Fractionated. 3–
Dialysate form Ammonium Sulphate precipitation. 4 – Sephadex Purified Sample

Characterization of SNTI

Figure – 4a shows the protein band pattern of the inhibitor on 12 % SDS slab gels when stained
with coomassie brilliant blue. Silver staining of SNTI showed a sharper band on SDS-PAGE.
From the plot of distance migrated in cm versus log molecular weight for standard
proteins (Fig. – 4b), the inhibitor showed a molecular weight of 29 kDa. When subjected to Gel filtration
on Sephadex G-150, SNTI eluted out as a single protein with a corresponding activity
peak (Fig. – 5a). The plot of elution volume versus log molecular weight of the calibrating proteins
is shown in Fig. – 5b. The molecular weight of SNTI calculated from the plot was 28.5 kDa.

Enzyme inhibition studies were carried out to identify the specificity of the inhibitor
towards the mechanistic classes of proteases. SNTI was tested for its inhibiting capacity
against bovine trypsin using both BAPNA and casein as the substrates. The inhibition
patterns of the amidolytic activity of bovine trypsin by SNTI was linear up to 80
% inhibition (Fig. – 6). On extrapolation, it was found that 12 µg of the inhibitor can totally inhibit
amidase activity of 30 µg of trypsin.

The activity of the SNTI against chymotrypsin, elastase and pronase (Streptomyces
griseus protease) subtilisin, papain, pepsin, thermolysin and ?-amylase was tested.
Except pronase, the rest of the enzymes were not affected by SNTI.

Fig. 4. a SDS – PAGE. Direction of migration is from top (cathode) to bottom (anode). (1) Molecular
weight markers: Phosphorylase b (97.4 kDa), Bovine serum albumin (66 kDa), Ovalbumin
(43 kDa), Carbonic anhydrase (29 kDa), Lysozyme (14.3 kDa). (2). Purified SNTI (coomassie
brilliant blue stained). (3) Purified SNTI (silver stained). *(1) to (3) were kept
at 100 °C for 3 min with SDS and 2 mercaptoethanol. b Molecular weight determination of SNTI by SDS PAGE on 12 % slab gel. Plot of distance
migrated against log molecular weight of standard proteins. BSA ?66 K.Da Ovalbumin
?43 K.Da Carbonicanhydrase ?29 K. DaLysozyme ?14.3 K.Da

Fig. 5. a Gel filtration of SNTI on Sephadex G-150. Elution profile of SNTI on a calibrated
column of Sephadex G-150. 10 mg of purified SNTI was applied to the column in phosphate
buffer pH 7.6 containing 20 mM NaCl and eluted with same buffer. Fractions of 4 ml
each were collected at flow rate of 12 ml per hour. Protein was monitored at 280 nm.
b Molecular weight determination of SNTI by gel filtration on Sephadex G-150. Plot
of elution volume against log molecular weight of standard proteins. BSA ?66 K.Da
Ovalbumin ?43 K.Da. Carbonicanhydrase ?29 K.Da Lysozyme ?14.3 K.Da

Fig. 6. Activity of SNTI towards Bovine Trypsin. Thirty microgram of trypsin was incubated
with varying amounts of SNTI for 10 min at 37 °C. The percentage residual enzyme activity
was assayed using BAPNA as the substrate. The concentration of the inhibitor required
to cause 50 % inhibition of the enzyme activity was determined from the graph

The serine proteases trypsin and pronase were inhibited by SNTI. Majority of plant
protease inhibitors isolated so far have been found to be specific for serine proteases
and there are some reports of these inhibitors inhibiting other classes of proteases.
SNTI specifically inhibited serine proteases trypsin and pronase and it has no effect
on thiol, acidic, metalloproteases and ?-amylase.

Reverse zymography: Substrate containing SDS- PAGE enables visualization of trypsin
inhibitor. The inhibitory activity produced by SNTI detected using trypsin and gelatin
substrate in the gel is shown in Fig. – 7. SNTI showed a single inhibitory band specific to trypsin and subjected to electrophoresis.

Mode of inhibition of Trypsin: Trypsin activity in the presence and absence of SNTI
was measured at different substrate concentrations. The double reciprocal plot of
the kinetic data is shown in Fig. – 8. In the presence of inhibitor, there was a decrease in the Vmax and the curves met
on the X –axis at a point equivalent to -1/km. The mode of inhibition of trypsin by
SNTI was non-competitive. The Ki value of trypsin for SNTI calculated from Dixon plot
was 0.75 + 0.05 × 10­-10 M.

Complex Studies: SNTI was treated with excess trypsin and the mixture was pre-incubated
at 37oC for 15 minutes. This mixture when applied onto a column of Sephadex G-150
at 5oC gave rise to two distinct at 280nm peaks (Fig. – 9). Peak-I had an elution volume of 20 ml which is higher than free SNTI 35 ml. The
binary complex of trypsin – SNTI did not show any trypsin activity or trypsin inhibitory
activity.

The molecular weight calculated for trypsin – SNTI complex on Peak-I based on the
calibration curve for standard proteins (Fig. – 4b) gave a value of 68.9 kDa. This would mean a mole/mole interaction of SNTI with
trypsin. Peak-II was small and represented uncomplexed SNTI with corresponding trypsin
inhibitory activity. The trypsin left over after the enzyme inhibitor complex formation,
was eluted out as peak-III with a corresponding elution volume of 51 ml.

Fig. 7. Reverse zymogram. 1. Crude fraction. 2. Ammonium Sulphate dialysate. 3. CM-Cellulose
sample. 4. Sephadex purified sample

Fig. 8. Mode of inhibition of trypsin by SNTI Line weaver –Burk plot. Inhibition of amidolytic
activity of trypsin by SNTI was done by incubating 30 ?g of trypsin and BAPNA solution
(0.8 to 5.0 ? mole) with reaction system containing 2.5 to 7.5 ?g of SNTI

Fig. 9. Elution patterns of SNTI, Trypsin and Trypsin-SNTI complex on Sephadex G-100 column.
Elution patterns of trypsin and trypsin – SNTI complex on Sephadex G – 100. Protein
was monitored at 280 nm

Fourier Transform Infra-Red Spectroscopy (FTIR): IR spectroscopic studies elucidates
functional groups in a molecule. The IR peak at 3399 (broad) and 2939 cm-1can be assigned
to OH of carboxylic group and asymmetric CH3 stretching. The over ton peak can be
observed at 2074 cm-1. The peak at 1642 and 1423 cm-1 could be due to amide C=O (CONH2)
and CH3 bending vibrations. The peaks observed at 995 and 925 could be attributed
to OH bending vibrations. The presence of amide and carboxylic groups are confirmed
by the above peaks (Fig. – 10).

Database and Sequence information: Protein sequences of gut proteases of H. armigera
and S. frugiperda were retrieved from protein NCBI database bearing the accession
number AHX25877.1 (Kazal-type serine protease from H. armigera) and ACR25157.1 (Trypsin
protease from S. frugiperda).

Fig. 10. FTIR. FTIR peaks representing presence of amide and carboxylic groups

Homology modeling of SNTI and Threading based modeling of insect gut proteases:

The protein structures of all the three were modeled using Prime module from Schrodinger
Suite. PDB BLAST provided a template 2C1X_A (UDP-Glucose Flavonoid 3-O Glycosyltransferase)
with 42 % identities, 55 % positives and score of 192.6 for SNTI. Secondary structure
of target SNTI sequence was identified using run SSP. After secondary structures for
target are identified, template and target sequences are aligned and then the structure
of SNTI is modelled based on the template 2C1X_A and the structure represents 11 helices
and 8 beta sheets (Fig. – 11).

The target proteins Kazal type serine protease from H. armigera and trypsin protease
from S. frugiperda are subjected to BLAST search for the identification of homologous
template. Template structures with very low identity were retrieved so, instead of
homology modeling threading or fold recognition approach was further used to model
these proteases. In threading first the secondary structure of target protein sequences
were predicted using run SSP option. Based on these secondary structures template
is identified from the fold library and the best templates identified were crystal
structure of insect derived kazal complex of serine protease (1TBQ) and crystal structure
of a non-psychrophilic trypsin (1A0J) respectively.

Fig. 11. Modeled Structure of SNTI by Homology modeling. Homology modeled structure of SNTI
representing 11 helices and eight sheets

These templates were further used for modeling Kazal type Serine and Trypsin proteases
by homology modeling approach. Now, again the same first step is repeated but instead
of finding the homologs, the template structure predicted by threading is used and
the sequence of template and targets are aligned. Finally the model was built for
Kazal type serine protease from H. armigera and was found to have 4 helices and 3
beta sheet (Fig. – 12). Similarly, Trypsin from S. frugiperda has 5 helices and 3 beta sheets (Fig. – 13). The 3D structure obtained is then validated using PROCHECK and ERRAT (Fig. – 14a, 14b). The protein structure that is modeled is satisfactory as evidenced by the validation
tools. Ramachandran plot derived from PROCHECK analysis represents about 99.2 % of
amino acids residues of SNTI are in favored region (Fig. – 14a) and ERRAT validates the overall structure quality to be 86.029 % (Figure – 14b). About 92.5 % of amino acids residues falling in favored regions for Kazal type
Serine protease (Figure – 15a) and 97.5 % for trypsin protease (Fig.- 16a). ERRAT validates the overall structure quality of Kazal type Serine protease to
be 86.96 % (Fig.- 15b) and 81.04 % for trypsin protease (Fig. – 16b).

Fig. 12. Modeled Structure of Kazal type Serine by Threading approach. Kazal type Serine protease
was modeled using threading approach representing 4 helices and 3 sheets

Fig. 13. Modeled Structure of Trypsin protease by Threading approach. Trypsin protease modeled
using threading approach represents 5 helices and 13 sheets

Fig. 14. Model validation of SNTI by PROCHECK and ERRAT. (a) Ramachandran plot of SNTI represents 99.2 % of amino acids in favored region. (b) The overall quality of structure is 86.029 %

Fig. 15. Model validation of Kazal type Serine protease by PROCHECK and ERRAT. (a) Ramachandran plot of Kazal type Serine protease represents 92.5 % of amino acids
in favored regions. (b) The overall quality of structure is 86.96 %

Fig. 16. Model validation of Trypsin protease by PROCHECK and ERRAT. (a) Ramachandran plot of Trypsin protease represents 97.5 % of amino acids in favored
regions. (b) The overall quality of structure is 81.04 %

 Binding Site prediction: The predicted structures were subjected to SiteMap for binding
site identification. The hydrophobic binding sites predicted by SiteMap on the surface
of SNTI and gut proteases are shown in the Table – 3.

Protein-Protein Docking: By following these combinations of SNTI x Kazal type Serine
protease and SNTI x Trypsin protease was performed using PIPER. All these proteins
prior to docking were prepared, optimized and energy minimized. From the resultant
set of 10 poses the hydrogen binding interactions and other interactions were identified
(Figs. – 17, 18). They are further checked whether these interactions are present in predicted binding
sites. Table – 4 represents interacting residues of SNTI x Kazal type Serine protease complex and
SNTI x Trypsin protease that are involved in binding site regions.

Table 3. Binding site surfaces predicted by SiteMap

Fig. 17. Protein-Protein interactions of SNTI with Kazal type Serine protease from H. armigera. Cyan color residues represents Kazal type Serine and Tan color residues represents
SNTI

Fig. 18. Protein-Protein interactions of Kazal type Trypsin. Cyan color residues represents
Trypsin and Tan color residues represents SNTI

Table 4. Protein-protein interaction analysis of SNTI with Kazal type Serine protease and SNTI
with Trypsin