Residue Histidine 669 Is Essential for the Catalytic Activity of Bacillus anthracis Lethal Factor

Abstract

The lethal factor (LF) of Bacillus anthracis is a Zn²⁺-dependent metalloprotease which plays an important role in anthrax virulence. This study was aimed at identifying the histidine residues that are essential to the catalytic activities of LF. The site-directed mutagenesis was employed to replace the 10 histidine residues in domains II, III, and IV of LF with alanine residues, respectively. The cytotoxicity of these mutants was tested, and the results revealed that the alanine substitution for His-669 completely abolished toxicity to the lethal toxin (LT)-sensitive RAW264.7 cells. The reason for the toxicity loss was further explored. The zinc content of this LF mutant was the same as that of the wild type. Also this LF mutant retained its protective antigan (PA)-binding activity. Finally, the catalytic cleavage activity of this mutant was demonstrated to be drastically reduced. Thus, we conclude that residue His-669 is crucial to the proteolytic activity of LF.

Anthrax is a zoonotic disease caused by toxigenic strains of the Gram-positive bacterium Bacillus anthracis (24). Because infections are highly fatal, the organisms are easily produced, and the spores spread easily, B. anthracis has been used as a bioweapon in biological war and biological terrorism (38). If inhaled, the spores are phagocytosed by alveolar macrophages, where they germinate to produce vegetative bacteria (10, 24). The vegetative bacteria further release anthrax toxins, which inhibit the innate and adaptive immune responses of the hosts. This enables the capsulated bacteria to escape the lymph node defense barrier to reach the blood system, causing bacteremia and toxemia, which can rapidly kill the hosts (24, 26). The great threat posed by anthrax to the public is not only due to the highly lethal rate of inhaled anthrax, but also is due to the social panic caused by the lethality. Therefore, efficient ways to defend against anthrax infection and spreading are greatly needed. This mostly depends on a full understanding of the mechanisms of anthrax infection and toxicities.

Anthrax toxins are the dominant virulence factors of Bacillus anthracis (6, 33, 37). They consist of three proteins: protective antigen (PA; 83 kDa), lethal factor (LF; 90 kDa), and edema factor (EF; 89 kDa). The 83-kDa PA (PA₈₃) directly binds to cellular membrane receptors and was cleaved to an active fragment of 63-kDa PA (PA₆₃) by cellular proteases of the furin family or by serum proteases. The receptor-bound portion of PA₆₃ self-assembles into either ring-shaped heptamers, which bind to three molecules of LF and/or EF, resulting in (PA₆₃)₇(LF/EF)₃ (21), or octamers which bind up to four molecules of these moieties, resulting in (PA₆₃)₈(LF/EF)₄ complexes (16, 17). The catalytic partners (EF and/or LF) are subsequently transported across the membrane to the cell cytosol (24, 27). EF is a Ca²⁺- and calmodulin-dependent adenylate cyclase that, together with PA, forms edema toxin. EF causes a rapid increase in intracellular cyclic AMP (cAMP) levels in host cells and alters the elaborate balance of intracellular signaling pathways (20, 23). LF is a Zn²⁺-dependent protease that, together with PA, forms lethal toxin (LT). It is a dominant virulence factor and the major cause of death for the B. anthracis-infected animals (1, 29, 30). LF specifically cleaves the N-terminal domain of mitogen-activated protein kinase kinases (MAPKKs) (11, 35). Because the N-terminal domain of MAPKKs is essential for the interaction between MAPKKs and MAPKs, the cleavage of this domain impairs the activation of MAPKs (8, 11, 15) and leads to the inhibition of three major cellular signaling pathways—the ERK (extracellular signal-regulated kinase), p38, and JNK (c-Jun N-terminal kinase) pathways (29, 31)—and thus induces the lysis of the host cells in an unknown mechanism.

The crystal structure of LF with the N-terminal domain of MEK2 has been reported (28). LF has 776 amino acids and comprises four different domains. Domain I (residues 1 to 254) is a PA-binding domain which delivers the remaining domains of the LF to the cell cytoplasm (3). The interface among domains II, III, and IV creates long, deep, 40-Å-long catalytic grooves into which the N terminus of MEK fits and forms an active site complex (28). Domain IV is central to catalytic activities of LF, containing two zinc-binding motifs (residues 686 to 690 and residues E735 to E739) and bound to a single Zn ion (18). However, which residues of LF are critical for efficient catalytic activities and execute the substrate cleavage remains unclear.

Histidine is the only naturally occurring amino acid to contain an imidazole residue as a side chain. The catalytic activity of histidine mostly depends on the special features of the imidazole residue. The logarithm of the proton dissociation constant of imidazolyl in the histidine residue is about 6.5; thus, under the physiological condition, it tends to form hydrogen bonds and shares donor and acceptor properties that can take part in either nucleophilic or base catalysis. The speed of the imidazole residue to give or accept protons is very fast, with a half-life of less than 10 s. So in the process of natural selection, histidine was chosen as the catalytic structure, indicating that it plays an important role in the catalysis process of enzymes (9, 12, 14). There are 21 histidines in LF, with 9 of them in LF domain I and 12 of them in domains II, III, and IV. The histidine residues important to LF activities in domain I have been identified (2, 22). The other 12 histidine residues in the remaining three domains include His-277, His-280, and His-424 in domain II; His-309 in domain III; and His-588, His-645, His-654, His-669, His-686, His-690, His-745, and His-749 in domain IV (28). His-686 and His-690 in domain IV were demonstrated to form a zinc binding site constituting a thermolysin-like zinc metalloprotease motif, HEXXH (18). The activities of the remaining 10 histidine residues in domains II, III, and IV have not been explored yet. In this study, we replaced these 10 histidine residues separately with alanine residues by site-directed mutagenesis. By the cytotoxicity assay of all these mutants, the H669A mutant was found to lose cell toxicity completely. Further assay revealed that residue His-669 was involved in neither zinc stabilization nor PA binding but participated in the substrate proteolytic activity of LF.

MATERIALS AND METHODS

Chemicals and reagents.

The Ni-nitrilotriacetic acid (NTA) His-binding column was a product of Merck Biosciences. Lethal factor protease substrate 2 (LFPS-2; colorimetric) was purchased from Alpha Diagnostic International. The enzymes used for DNA manipulation were purchased from TaKaRa Biotechnology Co., Ltd. The plasmid pGEX-6p-1-LF containing the LF coding gene was constructed previously (4). All other chemicals and reagents, unless otherwise noted, were purchased from Sigma.

Site-directed mutagenesis.

The LF gene, which derived from pGEX-6p-LF through double digestion with BamHI and NotI, was inserted into the expression vector pET-28a(+) (Merck), and the resulting plasmid, pET-LF, was used as the template for site-directed mutagenesis. All of the oligonucleotide primers used for site-directed mutagenesis are given in Table 1.

TABLE 1.

Sequence of oligonucleotides used for constructing site-direct mutations in the LF gene

Mutation	Primer orientationa	Primer sequenceb
H277A	F	5′-ATGGGAAAAGATAAAACAGGCCTATCAACACTGG-3′
	R	5′-TAAAGAATCGCTCCAGTGTTGATAGGCCTGTTTTATC-3′
H280A	F	5′-AAGATAAAACAGCACTATCAAGCCTGGAGCGATTC-3′
	R	5′-TTCAGATAAAGAATCGCTCCAGGCTTGATAGTGC-3′
H309A	F	5′-AGAAAGATGACATAATTGCTTCTTTATCTCAAGAAGA-3′
	R	5′-TCTTCTTGAGATAAAGAAGCAATTATGTCATCTTTCT-3′
H424A	F	5′-ATTGATGCTTTATTAGCTCAATCCATTGGAAGTACCTTG −3′
	R	5′-CAAGGTACTTCCAATGGATTGAGCTAATAA AGCATCAAT-3′
H588A	F	5′-GCTTATTACATTCAACGTGGCTAATAGATATGCATCC-3′
	R	5′-GGATGCATATCTATTAGCCACGTTGAATGTAATAAGC-3′
H645A	F	5′-GCTGAACAATATACAGCTCAAGATGAGATATATGAGC-3′
	R	5′-GCTCATATATCTCATCTTGAGCTGTATATTGTTCAGC-3′
H654A	F	5′-GATGAGATATATGAGCAAGTTGCTTCAAAAGGGT-3′
	R	5′-CATATAACCCTTTTGAAGCAACTTGCTCATATATCTC-3′
H669A	F	5′-GAATCCCGTTCTATATTACTCGCTGGACCTTCA-3′
	R	5′-CTACA CCTTTTGAAGGTCCAGCGAGTAATATAG-3′
H669R	F	5′-GAATCCCGTTCTATATTACTCAGAGGACCTTCA-3′
	R	5′-CTACA CCTTTTGAAGGTCCTCTGAGTAATATAG-3′
H745A	F	5′-GCAGAAGCCT TTAGGTTAATGGCTTCTACGGACCAT-3′
	R	5′-AACGTTCAGCATGGTCCGTAGAAGCCATTAACCTAA-3′
H749A	F	5′-AGGTTAATGCATTCTACGGACGCTGCTG AACGTTTA-3′
	R	5′-TTGAACTTTTAAACGTTCAG CAGCGTCCGTAGAATGC-3′

^aF, forward; R, reverse.

^bMutated residues are shown in boldface.

Long PCR amplification was used for site-directed mutagenesis by using the TaqPlus long PCR system (Stratagene). The PCR products were treated with DpnI disposal, transformed into freshly prepared competent DH5α cells. The amino acid replacements were identified by sequencing at Shanghai Sangon Biological Engineering Technology & Services Co., Ltd.

Screening the activities of LF mutants.

The plasmids harboring the desired mutations were transformed into Escherichia coli BL21-CodonPlus(DE3)-RIL. After induction with IPTG (isopropyl-β-d-thiogalactopyranoside), whole-cell lysates were prepared and tested for their cytotoxicities by the methods reported before (25). Briefly, a single colony of each site mutant was picked and inoculated into 1.5 ml EP tube containing 200 μl LB supplemented with 50 μg/ml ampicillin. After overnight cultivation, 100 μl fresh LB containing 50 μg/ml ampicillin and 0.6 mM IPTG was added to each tube. The expression of mutant proteins was induced at 28°C for 3 h. Then 10 μl of bacteriophage T7 lysate was added to each tube. After shaking at 37°C for 1.5 h to release the target proteins, samples were centrifuged at 10,000 × g for 15 min to remove the cellular debris.

The toxicities of the LF mutants were evaluated on LT-sensitive cell line RAW264.7. The cells were cultured in Dulbecco's modified Eagle's medium (DMEM) containing 10% heat-inactivated fetal calf serum. Before the assay, the cells were seeded into 96-well culture plates and cultured to 90% cell confluence. Old medium was replaced with 100 μl fresh medium containing 1 μg/ml PA. Then 3 μl of LF mutant protein lysate was added to the cells. Each mutant lysate was tested in triplicate. The mixture was incubated with the cells at 37°C. At the same time, the lysate of wild-type (WT) LF was added as a positive control, and the lysate of inactive mutant LF-Y236F (Tyr→Phe) was added as the negative control (4). After 5 h of incubation, alamarBlue dye (AbD Serotec, Oxford, United Kingdom) was added to each well. Active cells will reduce the alamarBlue dye, and the reaction can be detected with a multidetection microplate reader (BioTek) at an excitation wavelength of 530 nm and an emission wavelength of 590 nm. The effective concentrations of the LT can be calculated based on the fluorescence intensity (FI) of the dye (5). Meanwhile, the dye was added in parallel to the media to measure the background FI of the dye and added to the cells as the reference of 100% dye reduction. The cell viabilities were calculated according to the following formula: (FI of the test wells − FI of the dye)/(FI of the active cells − FI of the dye) × 100%.

Protein expression and purification.

PA was purified from E. coli according to the methods described previously (5). WT LF and all of the LF mutants were purified by immobilized chelate affinity chromatography. Briefly, the cells harboring the plasmids were grown at 37°C in LB broth with 100 μg/ml ampicillin. When the optical density at 600 nm (OD₆₀₀) reached 0.8, IPTG was added to a final concentration of 0.2 mM. After 4 h of induction at 28°C, the cells were harvested, followed by being resuspended in the binding buffer (0.5 M NaCl, 20 mM Tris-Cl, pH 7.9, 5 mM imidazole). The proteins were released from the bacteria with a French cell press. The lysate was then centrifuged at 10,000 × g for 30 min. The supernatant was collected, and proteins were purified with an Ni-NTA His binding column according to the manufacturer's protocol. The purified proteins were stored at −80°C until use.

Cytotoxicity assay of LF mutants.

Activities of the LF mutants were further measured by toxicities of the mutant on LT-sensitive macrophage cell line RAW264.7. LF mutant proteins were added to the cells in various concentrations together with 1 μg/ml PA and incubated with cells at 37°C. At the same time, 1 μg/ml LT (both LF and PA at 1 μg/ml each) was added in parallel as a positive control. After 4 h of incubation, alamarBlue dye was added to the wells, and the cell viability was measured by the method described above.

Binding assay of LF mutants to PA.

To study the binding activities of LF mutants to PA in solution, the PA was nicked with trypsin according to a ratio of 1 ng trypsin per μg of PA for 30 min at room temperature in 25 mM HEPES, 1 mM CaCl₂, and 0.5 mM EDTA, pH 7.5. The digestion reactions were stopped by adding 1 mM phenylmethylsulfonyl fluoride (PMSF). The nicked PA then was incubated with LF (1 μg/ml) for 15 min in 20 mM Tris, pH 9.0, containing 2 mg/ml CHAPS {3-[(3-cholamidopropyl)-dimethylammonio]-1-propanesulfonate}. After incubation for 15 min at room temperature, the samples were loaded onto 4% to 20% native gradient gels (Genscript, Inc.).

Binding of LF mutants to PA was also examined on the macrophage cell line RAW264.7. The competitive inhibition of LT toxicity by LF mutants was examined by adding increasing concentrations of the mutant proteins in combination with constant concentration of LT components (1 μg/ml PA and 1 μg/ml WT LF) for 4 h at 37°C. The binding activity of LF mutants with PA was appraised by the viabilities of the cells. After incubation, alamarBlue dye was added and the viabilities of cells were calculated based on the fluorescence intensity of the dye.

Atomic absorption spectrophotometric analysis.

Chelex 100 resin was chosen to eliminate the free metal ions in protein buffer. Chelex 100 resin (2 ml) was added to a gravity column, followed by being washed with 6 ml deionized water. The purified proteins were then added to the resin. After 0.5 h of incubation at 4°C, the proteins were eluted with deionized water and collected for further zinc content analysis.

After elimination of the metal ions, an AA 140/240 atomic absorption flame spectrophotometer was used to determine the zinc content of the WT LF and LF mutants (19, 32). All analyses of the samples were performed in triplicate, and a zinc standard curve was established by using a zinc standard solution designed for atomic absorption spectrophotometry. The amount of zinc in each protease preparation was determined after subtracting the background readings from an equivalent amount of solution from the parental cell line prepared in the same manner as the purified proteins.

Proteolytic activity.

LFPS-2 (Alpha Diagnostic) was used to measure the LF metalloproteolytic activity. It is an N-acetylated and C-p-nitroanilide (pNA), 14-mer peptide substrate (Ac-Gly-Tyr-bAla-Arg-Arg-Arg-Arg-Arg-Arg-Arg-Arg-Val-Leu-Arg-pNA), originating from LF substrate MEK-2. One milligram of lyophilized LFPS-2 was dissolved in 100 μl 1× phosphate-buffered saline (PBS) buffer to a final concentration of 10 μg/μl. WT LF or LF mutants at final concentrations of 0.1, 1, 10, and 50 μg/ml in 100 μl PBS buffer were added to a 96-well plate, followed by addition of 3 μl of 10 μg/μl LFPS-2 in each well. Triplicate wells were used for each concentration. After 3 h of incubation at 37°C, the proteolytic activity of proteins was monitored by recording the absorption at 405 nm.

Toxicity of anthrax toxin proteins to BALB/c mice.

A total of 25 6-week-old mice were randomly divided into five groups. To study the toxicities of LF mutants, the mice were injected intravenously with the mixture of 50 μg LF mutants mixed with 120 μg PA. Meanwhile, the mixture of 50 μg WT LF and 120 μg PA was administered as a positive control, and 120 μg PA alone was administered as the negative control. At the same time, a 1× 50% lethal dose (LD₅₀) of LT (12 μg PA and 5 μg WT LF) was applied to the mice to induce histopathology. The final volume injected in mice was 100 μl for all of the doses by using PBS buffer as the diluent. The animals were monitored postinjection for 10 days. After 10 days postinjection, the surviving mice were sacrificed. The spleens were collected immediately and fixed with 10% neutral buffered formalin. The paraffin-embedded tissue blocks were cut into 4-μm sections, deparaffinized, and stained with hematoxylin and eosin (H&E) according to the conventional protocol, and the histopathological analysis was performed under a light microscope.

Protein structure accession number.

The active channel of LF was reconstructed by using Pymol and has been deposited in the Protein Data Bank under PDB accession no. 1JKY.

RESULTS

Site-directed mutagenesis and cytotoxicities of the mutants.

For each site mutagenesis, five colonies were picked and sequenced to determine the accuracy of the mutation. As a result, we successfully got alanine replacement at His-277, His-280, and His-424 in domain II; His-309 in domain III; and His-588, His-645, His-654, His-669, His-745, and His-749 in domain IV. Except for the desired site mutagenesis, the other parts of the LF gene remained unchanged.

The bacterial lysate was then used to screen the cytotoxicity of these LF mutants. Nine of these mutants, when combined with PA, killed the cells and showed similar cytoxicities to the WT LF. Only one mutant, H669A (H→A), was nontoxic to the cells, suggesting that residue His-669 was important to the LF activities.

The activity of the purified LF H669A protein was quantitatively compared with that of the WT LF. The purity of the proteins was up to 90%, as determined by SDS-PAGE (Fig. 1). The cytotoxicity results showed that WT LF, when combined with PA, was highly active and could kill 50% of the cells when the concentration of WT LF was as low as 2.7 ng/ml. In contrast, 100% of the cells survived the attack of the H669A mutant with a concentration of up to 20 μg/ml. Thus, the cytotoxicity of the LF H669A mutant was at least 7,500 fold less than that of WT LF.

FIG. 1.

SDS-PAGE analysis of purified WT LF (lane 1), the H669A mutant (lane 2), and the H669R mutant (lane 3). Lane M, reference protein markers.

To further confirm that the His-669 residue is essential to LF toxicity, the histidine was further replaced with arginine by site-directed mutagenesis. Arginine has similar properties to histidine. The recombinant mutant protein was expressed and purified by the same protocol described previously. This mutant, H669R (H→R), also lost its toxicity to the cells. In this cytotoxicity assay, the final concentrations of purified WT LF or the H669A or H669R mutant ranged from 0.0003 μg/ml to 10 μg/ml. All of them were mixed with 1 μg/ml PA. The cytopathic effect of the toxins to the cells was observed, and the 50% effective concentration (EC₅₀), which represented the toxin activities, was calculated. As a result, the EC₅₀ of WT LF was 0.0027 μg/ml, while both the H669A and H669R mutants were completely nontoxic to cells, even at the highest concentration of 20 μg/ml, which was in agreement with the above results.

Binding of LF mutants to PA in solution.

LF binds to trypsin-nicked PA and forms PA-LF complex in solution, which can be detected by nondenaturing polyacrylamide gel electrophoresis. Here, we investigated whether the H669A and H669R mutants could bind to PA in solution. As shown in Fig. 2 A, a low-mobility band was observed not only from the WT LF but also from the H669A and H669R mutants. This implied that the H669A and H669R mutants, like WT LF, could bind to PA. The mutants with mutation at H669 didn't lose their activity to bind to PA.

FIG. 2.

Binding of WT LF or the mutants to PA. (A) Native gel electrophoresis detected (PA₆₃)_7/8-LF, which was formed by incubating PA₆₃ with WT LF or the LF mutants. (B) Competitive binding of LF mutants and WT LF to PA. Increasingly concentrations of the H669A mutant or H669R mutant were added to macrophages with a fixed concentration of LT (1 μg/ml LF and PA). The binding of LF mutants to PA inhibited the LT toxicities on RAW264.7 cells. The inhibitory effect was dose dependent.

The ability of the H669A and H669R mutants to compete with WT LF to bind to PA was also tested on a macrophage cell line. The concentrations of LT components (WT LF and PA) were fixed at 1 μg/ml, which was fully toxic to the cells, as assessed by a complete lysis of the treated macrophage cells. Whereas the H669A or H669R mutants were 2-fold serially diluted, ranging from 0.1 to 25.6 μg/ml, WT LT and these mutants at various concentrations were mixed, and the cytotoxicities of the mixture were evaluated. As shown in Fig. 2B, the cells lost the viability absolutely without the mutants, whereas, when the H669A or H669R mutants were added with LT, the viabilities of cells increased with the increased concentrations of mutant proteins. When the mutants were used at the concentration of 25.6 μg/ml, which was 25.6-fold greater than that of WT LF, the ratio of surviving cells reached 75%.

Therefore, the H669A or H669R mutant could effectively bind to PA oligomer. These results indicated that the toxicity losses of both LF H669A and H669R mutants were not attributed to the binding efficiency of LF mutants to PA oligomer.

Zinc content of mutants.

The activities of LF are also affected by Zn²⁺ stabilization of the protein. Thus, the zinc content of the H669A and H669R mutants was detected by an atomic absorption spectrophotometer, with WT LF used as a control. As shown in Fig. 3 A, the average zinc content of the H669A or H669R LF mutants was similar to that of the WT LF, which was 1 molar zinc per molar protein after being deposed with Chelex-100. It could be concluded that residue His-669 is not involved in the zinc stabilization of the LF. Furthermore, the crystal structure model was assayed, and the results revealed that residue His-669 is far from the Zn²⁺ and the distance between them is about 14.59 Å. In contrast, the distance between the zinc binding motif residues His-686 and His-690 and zinc is within 3 Å (Fig. 3B). Taken together, these findings demonstrated that the residue His-669 is not involved in the stabilization of the zinc of the LF.

FIG. 3.

Comparison of zinc contents among the WT LF and its mutants (A) and the positions of residue His-669 and Zn²⁺ (B). Residues His-686 and His-690 formed a zinc binding motif, HEXXH. Their distance to Zn²⁺ is within 3 Å. Residue His-669 is far from Zn²⁺, implying that it is impossible for this residue to be involved in Zn²⁺ stabilization.

Proteolytic activity.

The cleavage activity of H669A and H669R mutants to LFPS-2 was measured by monitoring the release of pNA. No proteolytic activities were detected for the WT LF or mutants when the concentration of 0.1 μg/ml was applied to the substrate. However, WT LF showed greatly promoted proteolytic activities at the concentration of 10 μg/ml or 50 μg/ml. The OD₄₀₅ values in these two concentrations were 0.284 ± 0.018 and 0.301 ± 0.025, respectively. However, both H669A and H669R mutants did not present proteolytic activity at the concentrations of 10 μg/ml because the OD₄₀₅ values for the H669A mutant (0.061 ± 0.001) or H669R mutant (0.061 ± 0.001) were as low as those of the background values (0.061 ± 0.001) of the substrate (negative control). When the concentrations of these two mutants increased to 50 μg/ml, the OD₄₀₅ values of the substrate were only slightly improved, which were 0.078 ± 0.001 for the H669A mutant and 0.081 ± 0.002 for the H669R mutant, respectively (P > 0.05) (Fig. 4). These results demonstrated that the proteolytic activity of both H669A and H669R mutants was greatly reduced, in comparison with that of the WT LF. Combining these pieces of evidence with the above findings, it was concluded that the loss of the proteolytic activity contributed to the nontoxicity of the H669A mutant and H669R mutant.

FIG. 4.

Proteolytic activity of LF or the mutants to LF protease substrate 2. pNA was released by LF cleavage, and the reaction was monitored by recording the change in absorption at a wavelength of 405 nm (OD₄₀₅).

Toxicities of H669A and H669R mutants in vivo.

The activities of the LF mutants were further examined in mice. Fifty micrograms each of the purified WT LF, H669A mutant, or H669R mutant, in combination with 120 μg PA, was injected intravenously. This amount was equal to 10× the LD₅₀ of LT. The mice injected with LT began to die at 12 h after injection, and all died within 24 h. However, the mice injected with equal quantities of either the H669A mutant or H669R mutant in combination with PA did not present any clinical symptoms and survived for 10 days until the end of observation period. In the group challenged with 1× the LD₅₀ of LT, three mice survived. After 10 days of observation, the surviving mice were sacrificed, and their spleens were divided immediately. The spleens were used for histological examinations. The spleen looked normal by the naked eye, without any sign of splenomegaly in the group of mice injected with the mutants. The histopathological observation demonstrated that the spleen structure was seriously damaged in the group of mice injected with WT LT, as shown by the disappearance of red and white pulp and a massive loss of the splenocytes (Fig. 5 C). In contrast, the spleen showed normal histology with clear structure of red and white pulp and normal cell density for the mice challenged with LF mutants (Fig. 5B). This in vivo experiment confirmed that LF H669A and H669R mutants were detoxified.

FIG. 5.

Histopathological observation of mouse spleen at ×100 magnification. (A) Spleen from mice without any toxin injection. (B) Spleen from mice injected with LF mutants and PA shows a normal histology with a clear structure of red and white pulp and normal cell density. (C) Spleen of mice injected with 1× LD₅₀ of LT was seriously damaged, as shown by the disappearance of red and white pulp and a great loss of splenocytes.

DISCUSSION

Several studies have shown that histidine residues play an important role in the catalytic activities of enzymes (7, 9, 12). In this paper, we explored the histidine residues which may be important for LF activities. Earlier research determined that three histidine residues, His-35, His-42, and His-229, in LF domain I are important for LF binding to PA (2, 22); His-686 and His-690 in LF domain IV are crucial for LF Zn²⁺ stabilization (18). However, the functions of the remaining 10 histidine residues in LF have not been investigated yet. These histidine residues were replaced by alanine separately by site-directed mutagenesis. Of these 10 mutants, only mutation at residue His-669 caused LF to lose the ability to lyse the cells absolutely. A further conservative mutation was conducted at this residue (H→R). This mutation led to similar results, with a total loss of cytotoxicity of LF. The in vivo experiment demonstrated that these two mutants were totally detoxified and had no toxicity to the mice when the dose equal to 10× the LD₅₀ of LT was applied. We concluded here that residue His-669 in LF domain IV is a conservative amino acid of LF that plays an important role in LF activity. The other nine histidine residues are dispensable for normal function of LF.

Further experiments were performed to illustrate the mechanism of His-669 impact on LF activity. First, we demonstrated that the H669A mutant and H669R mutant kept the binding activity to the PA oligomer. When sensitive cells were treated with the mixture of the mutant and WT LF, the toxicities of the lethal toxin were inhibited in a dose-dependent way. The competitive reduction of the LT toxicity may be caused by two mechanisms. (i) The mutants compete with WT LF to bind to the limited binding site of PA oligomer. Less WT LF can be transported to the cell cytosol. (ii) The mutants interact with PA differently from WT LF, such that translocation is affected. Theoretically, the activity of LF could also be affected by the Zn²⁺ stabilization. However, the results showed that the zinc content of the H669A mutant and the H669R mutant was similar to that of the native LF. Thus, we concluded that His-669 was involved in neither LF binding to PA nor the Zn²⁺ stabilization.

Finally, the substrate proteolytic activities of both H669A and H669R mutants were explored. We found that these mutants displayed very low proteolytic activity, suggesting that residue His-669 is essential for LF catalytic activity. The crystal structure of LF with the N-terminal sequence of MEK2 is a well-demonstrated example of uncleaved substrate bound to LF (Fig. 6) (28). This structure was reconstructed to explain the potential role of residue His-669 in LF. As shown in Fig. 6, LF domains II, III, and IV come together and form a catalytic channel. The N-terminal tail of its substrate, MEK2, inserts into this active channel. The MEK2 scissile bond is distant from domain IV (the catalytic center) (36), suggesting that it's a precleavage complex. Furthermore, we imitated the structure of the mutants by using software and found that the mutant structure was similar to the wild-type structure, except for the amino acid changes in residue 669 (figure not shown). The residue His-669 is located on the outside β sheets of domain IV, with its side-chain imidazolyl extending to the groove. Because of the distance to the substrate, His-669 can directly contact the substrate and thus participates in the LF catalytic process by proton transportation or other activities. Taken altogether, we interpreted that residue His-669 on the surface of the LF catalytic groove helps the docking of the LF MEKK substrates (34, 35) and forms the precleavage complex. Then the substrate moves close to domain IV, which is prepared for the catalytic cleavage activity. Therefore, it was concluded that the function of His-669 is to catalyze the cleavage of its mitogen-activated kinase kinase substrates.

FIG. 6.

The active channel of LF was reconstructed by using Pymol. LF domains II to IV are blue. The side chain of His-669 (purple) is extended to the center of the active channel. The cleavage site of MEK2 (red) by LF is between P10 and A11 (green).

The structure between His-745 and His-749 (HSTDH) is similar to that of the zinc binding motif HEXXH; a previous report assumed that this site may correlate with the zinc binding for LF (13). Our data argues against this hypothesis because H745A and H749A mutants did not display reduced toxicity to the cells. In addition, according to the crystal structure of LF, the closest distance from this fragment to Zn²⁺ is 13.6 Å, which is too far to bind to the zinc. These findings all suggest that residues His-745 to His-749 do not form a zinc binding motif.

In summary, the evidence in this study demonstrated that His-669 is of great importance to LF function by participating in the proteolysis of its substrates. This finding will be of significance in elucidating the mechanisms of LF toxicity, developing antitoxin reagents, and utilizing the detoxified LF in other areas.