A Kazal-Type Serine Protease Inhibitor from the Defense Gland Secretion of the Subterranean Termite Coptotermes formosanus Shiraki

Abstract

Coptotermes formosanus is an imported, subterranean termite species with the largest economic impact in the United States. The frontal glands of the soldier caste termites comprising one third of the body mass, contain a secretion expelled through a foramen in defense. The small molecule composition of the frontal gland secretion is well-characterized, but the proteins remain to be identified. Herein is reported the structure and function of one of several proteins found in the termite defense gland secretion. TFP4 is a 6.9 kDa, non-classical group 1 Kazal-type serine protease inhibitor with activity towards chymotrypsin and elastase, but not trypsin. The 3-dimensional solution structure of TFP4 was solved with nuclear magnetic resonance spectroscopy, and represents the first structure from the taxonomic family, Rhinotermitidae. Based on the structure of TFP4, the protease inhibitor active loop (Cys(8) to Cys(16)) was identified.

Fig 1. The (A) DNA and (B) primary amino acid sequence of TFP4 cloned from the defense gland secretion of C. formosanus termite soldiers.

The Kazal-type serine protease inhibitor consensus sequence is highlighted in grey with the active site residue, P₁, indicated with the asterisk at Met¹⁰. Laskowski and Kato’s nomenclature for the Kazal-type active loop is shown above the amino acid sequence [21]. Cystines and disulfide bonds are indicated by Roman numerals and lines.

Fig 2. Elastase inhibition assays.

(A) Baseline elastase (0.2 μg/mL) activity with N-succinyl-(l-alanine)₃-p-nitroanilide (0.27 mM). Inhibition of elastase with (B) Roche protease inhibitor cocktail (60 μg/mL), (C) TFP4 (2 μg/mL), and (D) crude protein extract (1.2 μg/mL) from the defense gland secretion of C. formosanus. Additions are indicated by arrows. The hydrolysis rate and correlation coefficient (R²) are shown for each reaction.

Fig 3. Chymotrypsin inhibition assays.

(A) Baseline chymotrypsin (0.1 μg/mL) activity with N-benzoyl-L-tyrosine ethyl ester (0.5 mM). Inhibition of chymotrypsin with (B) Roche protease inhibitor cocktail (60 μg/mL), (C) TFP4 (2 μg/mL), and (D) crude protein extract (1.2 μg/mL) from the defense gland secretion of C. formosanus. Additions are indicated by arrows. The hydrolysis rate and correlation coefficient (R²) are shown for each reaction.

Fig 4. Trypsin inhibition assays.

(A) Baseline trypsin (0.3 μg/mL) activity with Nα-benzoyl-L-arginine ethyl ester (75 μM). Inhibition of trypsin with (B) Roche protease inhibitor cocktail (60 μg/mL) and (C) TFP4 (2 μg/mL). Additions are indicated by arrows. The hydrolysis rate and correlation coefficient (R²) are shown for each reaction. The assay with crude extract was not performed since no inhibition was observed with the recombinant protein.

Fig 5. Enzyme kinetic analysis of TFP4 inhibition of BTEE hydrolysis by chymotrypsin.

The reaction velocities of chymotrypsin hydrolysis of BTEE at varying concentrations of BTEE and TFP4 are plotted. The points are the observed velocities, and the lines represent the best global fit of the data to Eq 1, a competitive inhibition model. The K_i of TFP4 determined from the fit (91.3 ± 8.9 nM) and the correlation coefficient (R² = 0.9884) are displayed on the plot.

Fig 6. NMR solution structure of TFP4.

Two ribbon views of TFP4 are shown to illustrate the Kazal-type consensus loop. The amino acids are numbered based on the TFP4 sequence in Fig 1B. The asterisks indicate the active site amino acid, Met¹⁰. The secondary structure is colored from the N- to C-terminus, blue to red. The disulfide bonds are labeled with Roman numerals to indicate the cystine connectivity. Note that partial transparency of the second β-sheet is used in the structure on the right to show the aft III-VI disulfide bond.

Fig 7. TFP4 structure aligned with structural homologs.

(A) The top five, non-redundant, structural homologs as determined by the Dali server [45] aligned with the average TFP4 structure. TFP4 is shown in black and the structural homologs in grey (PDB ID: 1Y1B chain A, 1YU6 chain B, 1IY6 chain A, 1TBQ chain S, 2F3C chain I) [–48,51,52]. The extra loops in TFP4 that are not found in the homologous structures are labeled, Loop 1 and Loop 2, with their corresponding amino acid sequences. (B) Sequence alignments based on the structural alignment of TFP4 and the homologous proteins were generated using the Dali server [45]. The Kazal-type consensus sequence is highlighted in bold with the active site residue indicated by the asterisk. Loops 1 and 2 are labeled with the corresponding residues are in bold.