A mechanistic analysis of bacterial recognition and serine protease cascade initiation in larval hemolymph of Manduca sexta

Abstract

Serine protease cascades have evolved in vertebrates and invertebrates to mediate rapid defense responses. Previous biochemical studies showed that in hemolymph of a caterpillar, Manduca sexta, recognition of fungi by β-1,3-glucan recognition proteins (βGRP1 and βGRP2) or recognition of bacteria by peptidoglycan recognition protein-1 (PGRP1) and microbe binding protein (MBP) results in autoactivation of hemolymph protease-14 precursor (proHP14). HP14 then activates downstream members of a protease cascade leading to the melanization immune response. ProHP14 has a complex domain architecture, with five low-density lipoprotein receptor class A repeats at its amino terminus, followed by a Sushi domain, a Sushi domain variant called Wonton, and a carboxyl-terminal serine protease catalytic domain. Its zymogen form is activated by specific proteolytic cleavage at the amino-terminal end of the protease domain. While a molecular mechanism for recognition and triggering the response to β-1,3-glucan has been delineated, it is unclear how bacterial recognition stimulates proHP14 activation. To fill this knowledge gap, we expressed the two domains of M. sexta MBP and found that the amino-terminal domain binds to diaminopimelic acid-peptidoglycan (DAP-PG). ProHP14 bound to both the carboxyl-terminal domain (MBP-C) and amino-terminal domain (MBP-N) of MBP. In the mixture of DAP-PG, MBP, and larval plasma, inclusion of an HP14 fragment composed of LDLa repeats 2-5 (LDLa_2-5) or MBP-C significantly reduced prophenoloxidase activation, likely by competing with the interactions of the full-length proteins, and suggesting that molecular interactions involving these regions of proHP14 and MBP take part in proHP14 activation in response to peptidoglycan. Using a series of N-terminally truncated versions of proHP14, we found that autoactivation required LDLa_2-5. The optimal ratio of PGRP1, MBP, and proHP14 is close to 3:2:1. In summary, proHP14 autoactivation by DAP-type peptidoglycan requires binding of DAP-PG by PGRP1 and the MBP N-terminal domain and association of the LDLa_2-5 region of proHP14 with the MBP C-terminal domain. These interactions may concentrate the proHP14 zymogen at the bacterial cell wall surface and promote autoactivation.

Keywords: Hemolymph proteins; Insect immunity; Melanization; Microbe binding protein; Pattern recognition; Peptidoglycan; Peptidoglycan recognition protein; Toll pathway.

Fig. 1.

SDS-polyacrylamide gel electrophoretic analysis of the purified domains or regions of M. sexta MBP and proHP14 produced by baculovirus-infected insect cells. The amino- and carboxyl-terminal domains of MBP, LDLa_2–5, Sushi (S) and Wonton (W) domains, and truncated proHP14 (ΔL₁*, ΔL₁, ΔL_1,2, ΔL_1–3, ΔL_1–4, SWP and WP, where P stands for the protease domain) were separated by 15% SDS-PAGE under reducing conditions and stained with Coomassie brilliant blue. proHP14ΔL₁*, nearly identical to proHP14ΔL₁, resulted from the full-length proHP14 whose first LDLa repeat was removed by intracellular processing enzymes that cleave between R⁷⁵SRR⁷⁸ and Q⁷⁹ (Wang and Jiang, 2005). proHP14ΔL₁, proHP14ΔL_1,2, proHP14ΔL_1–3, and proHP14ΔL_1–4 were purified in the previous study (Wang and Jiang, 2010).

Fig. 2.

Association of soluble microbial cell wall components (A) or proHP14 (B) with purified M. sexta MBP or its domains. (A) The full-length MBP, MBP-N, MBP-C, and BSA were separately incubated with immobilized LTA, LPS, soluble E. coli DAP-PG, S. aureus Lys-PG, or laminarin. Total bindings were measured using diluted MBP antiserum, GAR-AP, and the AP substrate. Specific binding sites in MBP-fl, -N and -C, if any, were occupied by pre-incubating the proteins with excess amounts of the microbial polysaccharides before incubation with the same ligands coating the well surface. The AP activities (mean ± SEM, n = 3), representing non-specific or unsaturable bindings, are compared with those of the total bindings by Tukey HSD test using SPSS 27.0. (B) BSA, proHP14ΔL₁*, and proHP14ΔL₁* pre-incubated with excess amounts of MBP-N, MBP-C or PGRP1 were added to the wells coated with MBP-N, MBP-C and PGRP1, respectively. Total and non-specific bindings were determined in ELISA using diluted HP14 antiserum as primary antibody, GAR-AP as secondary antibody, and the substrate solution. The statistical significance (p >0.05, <0.05, 0.01, and 0.001) from the Tukey test is marked with “ns” (not significant), “*”, “**”, and “***”, respectively.

Fig. 3.

Perturbation of DAP-PG-elicited proPO activation in control plasma (CP) by domains of the MBP and proHP14. (A) Aliquots of 1:10 diluted CP (5 μl) from day 2, 5^th instar naïve larvae were incubated at 4°C for 60 min with MBP-fl, -N, -C, LDLa_2-5, Sushi or Wonton domain at the indicated concentrations in the presence or absence of E. coli DAP-PG (1 μg). PO activities were then determined to test possible roles of the protein fragments in DAP-PG recognition and proPO activation. The activity data (mean ± SEM, n = 3) were compared between paired samples using Student’s t-test and Tukey HSD test to reveal statistically significant difference. “ns”, not significant (p>0.05) in both tests; “●”, p = 0.035 in the t-test andp = 0.173 in the Tukey test for MBP-fl. (B) The enhanced proPO activation was perturbed by pre-incubating E. coli DAP-PG (1 μg) and MBP-fl (0.1 μg or 0.1 μM, left) or MBP-N (1 μg or 3.5 μM, right) with LDLa_2-5 (1.6 μM), Sushi (5.2 μM), Wonton (4.5 μM), or MBP-C (4.25 μM) on ice for 10 min. At 60 min after aliquots of 1:10 diluted CP (5 μl) had been incubated with the mixtures, PO activities were measured and compared to reveal possible involvement of the domains in interactions during recognition and system activation. PO activities were determined and shown as mean ± SEM (n = 3). The data without and with the protein fragments were compared using Tukey HSD test to show significant decreases (**: p <0.01, ***: p <0.001).

Fig. 4.

Conditions for the autoactivation of M. sexta proHP14 upon exposure to E. coli DAP-PG, PGRP1, and MBP. (A) Size of proHP14. The proHP14 from induced hemolymph (0.1 μg proHP14ΔL₁^N, LDL₁ processed, natural) and truncation mutants of proHP14 from baculovirus-infected insect cells (0.1 μg proHP14ΔL₁*, proHP14ΔL₁, proHP14ΔL_1,2, proHP14ΔL_1–3, proHP14ΔL_1–4, SWP and WP, see Fig. 1 legend), MBP-fl (300 ng, 250 nM), PGRP1 (300 ng, 625 nM), E. coli DAP-PG (1 μg), and buffer A to a final volume of 24 μl were incubated for 90 min at 37 °C. (B) Linkage of MBP-N and -C. The proHP14 (0.1 μg proHP14ΔL₁*), MBP (500 ng MBP-fl, -N, -C, or both -N and -C each at 250 ng), PGRP1 (300 ng, 625 nM), E. coli DAP-PG (1 μg), and buffer A were incubated. (C) Stoichiometry. Lanes 1–5: The proHP14 (0.1 μg proHP14ΔL₁^N), MBP-fl (0 to 125 nM), PGRP1 (300 ng), E. coli DAP-PG (1 μg), and buffer A were incubated; Lanes 6–9: The proHP14 (0.1 μg proHP14ΔL₁*), MBP-fl (300 ng), PGRP1 (39 to 312 nM), E. coli DAP-PG (1 μg), and buffer A were incubated. The reaction mixtures in (A–C) were subjected to 10% SDS-PAGE under reducing conditions, followed by immunoblot analysis using 1:2000 diluted HP14 antiserum as primary antibody and GAR-AP as secondary antibody to detect autoproteolysis.

Fig. 5.

Effects of MBP-N, -C, and HP14 fragments on proHP14 autoactivation caused by E. coli peptidoglycan. (A) DAP-PG (1 μg), PGRP1 (0.3 μg, 625 nM), MBP-fl (0.3 μg, 250 nM), and purified proHP14 (0.1 μg proHP14ΔL₁^N, 55 nM) were incubated with HP14 LDLa_2–5 (1 μg, 1.6 μM), Sushi (1 μg, 5.2 μM S), Wonton (1 μg, 4.5 μM W), MBP-N (0.9 μg, 3.12 μM), or MBP-C (1.7 μg, 1.7 μM), and buffer A to a final volume of 24 μl for 1.5 h at 37 °C. (B–D) DAP-PG (1 μg), PGRP1 (312 nM), MBP-fl (125 nM), proHP14 (55 nM), different concentrations of LDLa_2–5 (0, 50, 100, 200, 400, and 800 nM for (B), MBP-N (0, 0.49, 0.98, 1.95, 3.9, 7.8 μM for (C), or MBP-C (0, 0.27, 0.53, 1.06, 2.12, 4.25 μM for (D), and buffer A (to 24 μl) were incubated for 90 min at 37 °C. The reaction mixtures in (A–D) were subjected to 10% SDS-PAGE under reducing conditions, followed by immunoblot analysis using 1:2000 diluted HP14 antiserum as the primary antibody and GAR-AP as the secondary antibody to detect proHP14 autoactivation.

Fig. 6.

A model for proHP14 (proHP14ΔL₁) auto-proteolysis and protease cascade activation triggered by DAP-PG, PGRP1 and MBP interaction. A color difference between the same domain represents a conformational change upon binding or proteolytic cleavage. Shape changes in the protease domain depict activation after proteolysis between the regulatory and catalytic domains.