Mycobacterium tuberculosis Hip1 modulates macrophage responses through proteolysis of GroEL2

Abstract

Mycobacterium tuberculosis (Mtb) employs multiple strategies to evade host immune responses and persist within macrophages. We have previously shown that the cell envelope-associated Mtb serine hydrolase, Hip1, prevents robust macrophage activation and dampens host pro-inflammatory responses, allowing Mtb to delay immune detection and accelerate disease progression. We now provide key mechanistic insights into the molecular and biochemical basis of Hip1 function. We establish that Hip1 is a serine protease with activity against protein and peptide substrates. Further, we show that the Mtb GroEL2 protein is a direct substrate of Hip1 protease activity. Cleavage of GroEL2 is specifically inhibited by serine protease inhibitors. We mapped the cleavage site within the N-terminus of GroEL2 and confirmed that this site is required for proteolysis of GroEL2 during Mtb growth. Interestingly, we discovered that Hip1-mediated cleavage of GroEL2 converts the protein from a multimeric to a monomeric form. Moreover, ectopic expression of cleaved GroEL2 monomers into the hip1 mutant complemented the hyperinflammatory phenotype of the hip1 mutant and restored wild type levels of cytokine responses in infected macrophages. Our studies point to Hip1-dependent proteolysis as a novel regulatory mechanism that helps Mtb respond rapidly to changing host immune environments during infection. These findings position Hip1 as an attractive target for inhibition for developing immunomodulatory therapeutics against Mtb.

Figure 1. Purification of Hip1 and Hip1(S228A) by ion exchange chromatography.

(A) and (B) Hip1 and Hip1(S228A) proteins were purified by gravity column chromatography and anion exchange chromatography. Top and bottom panels show the anion exchange column elution peak profiles of recombinant Hip1 and Hip1(S228A) respectively. The purity of the eluted protein was checked by SDS-PAGE analysis after each purification procedure. The arrows indicate the elution fractions used in subsequent assays: B6 for Hip1 and A12 for Hip1(S228A). (C) CD spectra of Hip1 and Hip1(S228A) mutant.

Figure 2. Analysis of the enzymatic activity of Hip1.

(A) Hip1 esterase activity with p-nitrophenylbutyrate. P-nitrophenylbutyrate was incubated alone for a negative control reaction (-control). PreScission protease was used in a positive control reaction (+ control). (B) Azocasein proteolysis assay showing that Hip1 is a protease. Azocasein was incubated alone (- control), with the protease subtilisin (+control), with Hip1 (0.05 mg/ml), or Hip1(S228A) (0.05 mg/ml) in 25 mM Tris pH 7.4, 150 mM NaCl. The enzyme activities are expressed as units of enzyme/mg protein (one enzyme unit is the quantity of enzyme required to increase absorbance by 0.01 units at 440 nm). (C) Endpoint assay showing proteolytic activity of Hip1. Hip1 (7.5 µM) was incubated with each peptide substrate (1.5 mM) or alone (-control) in 50 mM Tris pH 8.0 for 18 hr at 25°C. Elastase was used as a positive control (+ control). Hydrolysis of the peptide substrates was detected by monitoring an increase in absorbance at 410 nm. (D) Inhibition of Hip1 with various classes of protease inhibitors. Hip1 (4 µM) was pre-incubated with inhibitor for 30 min in 50 mM Tris, pH 8.0 at 25°C. Then, protease activity was measured by the addition of 1.5 mM Ala-Pro-Ala-pNa. The specific activity of Hip1 against Ala-Pro-Ala-pNa was defined as 100% (no inhibitor). Data are shown as one representative experiment from three independent experiments.

Figure 3. Mtb GroEL2 is a physiological substrate of Hip1 protease activity.

(A) Recombinant GroEL2 is cleaved by recombinant Hip1 but not by Hip1(S228A). Samples from the cleavage reactions were taken at 0 hours and 24 hours, separated by 10% SDS-PAGE gel and analyzed by Western blot with anti-S-tag antibody to detect GroEL2 and GroEL2(cl). (B) Hip1 mediated cleavage of GroEL2 is inhibited by the serine protease inhibitor AEBSF. Recombinant GroEL2 was incubated with recombinant Hip1 for 24 hours at 37°C either alone or in the presence of inhibitors AEBSF, PMSF, or bestatin. Samples were taken after 24 hours, separated by 10% SDS-PAGE gel and analyzed by Western blot with anti-S-tag antibody to detect GroEL2 and GroEL2(cl). (C) Optimal pH range for GroEL2 cleavage. Recombinant GroEL2 was incubated with recombinant Hip1 for 24 hours at 37°C under varying pH conditions. (D) Protein-protein interaction between GroEL2 and Hip1. Mycobacterium protein fragment complementation (M-PFC) assay demonstrates interaction between Mtb GroEL2 and Hip1 expressed in M. smegmatis as shown by growth on plates containing trimethoprim. M. smegmatis strain expressed either GCN4 homo-dimerization domains of Saccharomyces cerevisiae (positive control); GroEL2 and Hip1; GroEL2 and Hip1(S228A) or negative controls: vector and Hip1; vector and Hip1(S228A); GroEL2 alone; GroEL2 and KdpE; GroEL2 and SigA; GroEL2 and InhA. Data (A–D) are shown as one representative experiment from three to five independent experiments.

Figure 4. GroEL2 is cleaved in vitro and in vivo by Hip1.

(A) LC/MS/MS showing elution of the uncleaved GroEL2 N-terminal peptide (m/z = 2163). (B) LC peak profile of GroEL2 N-terminal peptide incubated with Hip1 for 18 hours at 25°C. Two abundant product fragments with m/z = 1122 and 757 indicate cleavages between Thre₃ and Ile₄, as well as between Arg₁₂ and Gly₁₃. (C) AEBSF inhibits Hip1 cleavage of the GroEL2 peptide. (D) Mass spectrometry analysis of the 7.3 min peak in (C) indicates the predominant species is uncleaved GroEL2. The second peak corresponds to substrate sulfonated by AEBSF at threonine. (E) Determining cleavage site of Mtb GroEL2 using M. smegmatis. Culture supernatants from M. smegmatis strains expressing Mtb GroEL2-FLAG, GroEL2 (R12P)-FLAG or GroEL2 (R12P G13P)-FLAG were subjected to cleavage by recombinant Hip1 for 24 hours at 37°C and analyzed by Western blot. (F) Cleavage site of GroEL2 in Mtb. GroEL2-FLAG or GroEL2 (R12P G13P)-FLAG mutant were expressed in Mtb H37Rv. Protein extracts corresponding to the pellet (P) and supernatant (S) fractions of those strains were prepared, and analyzed by Western blot with anti-FLAG antibody (to detect GroEL2) and anti-SigA antibody (to detect the sigma 70 subunit of RNA polymerase). Data (A–F) are shown as one representative experiment from three independent experiments.

Figure 5. GroEL2 is a multimer in vitro and is converted to a monomer following cleavage by Hip1.

(A) Size exclusion chromatograms of recombinant full length GroEL2, cleaved GroEL2 and Hip1. (B) Standard curve based on the elution profiles of a set of standard molecular weight marker proteins. The logarithms of the molecular weights (log M_r) were plotted as a function of K_av. Data are shown as one representative experiment from three independent experiments.

Figure 6. Expression of secreted GroEL2(cl) in hip1 mutant restores wild type levels of proinflammatory cytokine responses in macrophages.

Production of IL-6, IL-1β, and TNF-α by C57BL/6 bone marrow derived macrophages (BMM) 24 hours after infection with wild type, hip1 mutant, and hip1 mutant complemented with either Hip1 (comp) or GroEL2(cl). Data are shown as mean ±S.D. of one representative experiment from three independent experiments. *, P<0.05; **, P<0.01.

Figure 7. Differential stimulation of proinflammatory cytokine production from macrophages by GroEL2 and GroEL2(cl).

(A) Production of IL-6 and IL-1β by C57BL/6 bone marrow derived macrophages (BMM) 24 hours after stimulation with GroEL2 or GroEL2(cl). (B) Production of IL-6 and IL-1β in response to GroEL2 and GroEL2(cl) occurs in a partially TLR2-dependent manner. (C) Presence of GroEL2(cl) leads to diminished stimulatory capacity of GroEL2. Each form of GroEL2 was added to C57BL/6 BMM either alone (5 µM) or together (5 µM each) for 24 hours. The expected additive effect of GroEL2 and GroEL2(cl) is represented as a sum of the cytokine levels for each protein alone. Data are shown as mean ±S.D. of one representative experiment from three independent experiments. *, P<0.05; **, P<0.01.

Figure 8. Model of Hip1-GroEL2 interactions. Proteolysis of full length GroEL2 by Hip1 converts multimeric GroEL2 to monomers.

In the cell wall of wild type Mtb, GroEL2 multimer interacts with the Hip1 protease, which cleaves GroEL2 and leads to release of GroEL2 monomers extracellularly. In contrast, in the hip1 mutant, GroEL2 remains in its multimeric form and is released extracellularly as a multimer.