Human Chitotriosidase Is an Endo-Processive Enzyme

Human chitotriosidase (HCHT) is involved in immune response to chitin-containing pathogens in humans. The enzyme is able to degrade chitooligosaccharides as well as crystalline chitin. The catalytic domain of HCHT is connected to the carbohydrate binding module (CBM) through a flexible hinge region. In humans, two active isoforms of HCHT are found–the full length enzyme and its truncated version lacking CBM and the hinge region. The active site architecture of HCHT is reminiscent to that of the reducing-end exo-acting processive chitinase ChiA from bacterium Serratia marcescens (SmChiA). However, the presence of flexible hinge region and occurrence of two active isoforms are reminiscent to that of non-processive endo-chitinase from S. marcescens, SmChiC. Although the studies on soluble chitin derivatives suggest the endo-character of HCHT, the mode of action of the enzyme on crystalline chitin is not known. Here, we made a thorough characterization of HCHT in terms of the mode of action, processivity, binding, and rate constants for the catalysis and dissociation using α-chitin as substrate. HCHT efficiently released the end-label from reducing-end labelled chitin and had also high probability (95%) of endo-mode initiation of processive run. These results qualify HCHT as an endo-processive enzyme. Processivity and the rate constant of dissociation of HCHT were found to be in-between those, characteristic to processive exo-enzymes, like SmChiA and randomly acting non-processive endo-enzymes, like SmChiC. Apart from increasing the affinity for chitin, CBM had no major effect on kinetic properties of HCHT.


Introduction
Chitin is an essential structural component of different fungi, nematodes, arthropodes, insects and crustaceans. Chitin is composed of linear β-1-4-linked N-acetylglucosamine (NAG) units. The chitin chains are packed into crystals with intervening amorphous regions. In nature chitin is degraded by chitinases, hydrolytic enzymes that are synthesized by different bacteria, fungi and higher eukaryotes. In humans, two chitinases are expressed, chitotriosidase (HCHT) and acidic mammalian chitinase (AMCase) [1,2]. While bacteria degrade chitin for energy supply, human chitinases are parts of innate immune system. Chitin is a good candidate for a a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 found in subsites -2/-1 [21]. Strong binding of dimers to product binding sites (+1 and +2 in case of SmChiA) has been shown for other processive GHs and such strong product binding has been shown to drive the processive movement of the enzymes [32][33][34][35]. Thus, the question of the directionality of HCHT remains still open.
Here we made a thorough characterization of HCHT in terms of the mode of action, processivity, binding, and rate constants for the catalysis and dissociation using α-chitin as substrate. Both HCHT isoforms 50 kDa full-length protein (HCHT50) and its 39 kDa truncated version without the C-terminal CBM (HCHT39) were included into the study. The kinetic properties of HCHT were found to be in-between those characteristic to processive exoenzymes, like SmChiA, and randomly acting non-processive endo-enzymes, like SmChiC.

Enzymes
Protein expression and purification of the two isoforms of HCHT is described in Stockinger et al. [29]. In brief, proteins were produced in HEK293-6E cells. Two vectors were constructed designated pHCHT50 and pHCHT39 expressing HCHT including its native signal peptide and with and without the C-terminal chitin binding domain, respectively. The genes were synthesized (Genescript) as BamHI-XbaI fragments and ligated into pTT5V5H8Q (NRC Biotechnology Research Institute) resulting in a C-terminal His-tag (8xHis) on the recombinant proteins produced. The cloning steps were performed in Escherichia coli DH5α. pHCHT50 and pHCHT39 were transfected into HEK293-6E cells grown in F17 medium (Invitrogen) supplemented with Kolliphor P188 (Sigma) and L-glutamate (Sigma) to final concentrations of 0.1% and 4 mM, respectively. Cultivation of cells took place in 90 mL medium in disposable 500 mL flasks with gentle shaking (70 rpm, at 37˚C, with 5% CO 2 and 80% humidity). Transfection was performed with PEIpro (Polyplus) when the cell density in the cultures was 1.7 × 10 6 cells/mL. Tryptone N1 feeding medium (TekniScience) was added to a final concentration of 0.5% 48 h after transfection, and harvesting of the protein containing culture supernatant was performed 96 h after transfection. Harvesting took place at a cell density of 2.2 × 10 6 cells/mL. Recombinant HCHT was purified using a HisTrap HP column (GE Healthcare) according to the instructions given by the manufacturer.
Protein expression and purification of SmChiA and SmChiB is described in Brurberg et al. [36]. Protein expression and purification of SmChiC is described in Synstad et al. [37]. The purity of all enzymes was > 95% as judged by SDS-PAGE (Fig 1). Enzymes concentrations were determined by using the Bradford method from BioRad. ethanol and incubated in acetone for 1 h with 2 changes at room temperature, washed repeatedly with water and grinded in mortar. To ensure that all amino-groups were acetylated the purified chitin was treated with acetic anhydride. For that the chitin was washed three times with methanol and finally re-suspended in methanol to give the concentration of 20 mg/mL. 1 mL of acetic anhydride was added per 1 g of chitin and the mixture was incubated overnight at room temperature, with stirring. Next, O-deacetylation was carried out by adding 100 mM KOH in methanol and incubating for 4 h at room temperature, with stirring. After that the chitin was washed repeatedly with water and 50 mM sodium acetate, pH 6.1. Finally 0.01% NaN 3 was added and the chitin was stored at 4˚C.
Chitin nanowhiskers (CNWs) and 14 C labelled CNWs ( 14 C-CNWs) were prepared as described in Kuusk et al. [38]. A total of 35 g of crab chitin was treated with HCl, NaOH, ethanol and acetone as described above. The purified chitin was suspended in 3 M HCl and incubated at 100˚C for 90 min, with stirring. The mixture was diluted 3-fold with water, neutralized by slowly adding solid NaOH and finally buffered with 10 mM sodium acetate, pH 6.1. CNWs were washed several times with 10 mM sodium acetate, pH 6.1, and stored at 4˚C. The N-acetylation and O-deacetylation of CNWs were carried out essentially as described for crystalline α-chitin (see above). To prepare 14 C-CNWs the N-acetylation of CNWs was carried out with 5 mCi of [1-14 C]acetic anhydride as described in Kuusk et al. [38]. The specific radioactivity of 14 C-CNWs was 4.18 × 10 6 dpm/mg. Amorphous chitin was prepared by the acetylation of chitosan as described in Kurashin et al. [35]. Chitosan was suspended in water, an equal volume of 20% acetic acid was added, with stirring, and the mixture was diluted five times by adding methanol, with stirring. 1 mL of acetic anhydride was added per 1 g of chitin, with stirring, and the mixture was incubated overnight at room temperature, without stirring. Next, the mixture was diluted further five times with water. The acetic acid was neutralized and the O-deacetylation was carried out by adding NaOH to the final concentration of 50 mM, followed by incubating overnight at room temperature, with stirring. The amorphous chitin was repeatedly washed with water and 50 mM sodium acetate, pH 6.1. Finally 0.01% NaN 3 was added and the chitin was stored at 4˚C.
Reduced chitin was prepared of α-chitin by NaBH 4 treatment as described in Kurashin et al. [35]. The purified α-chitin was washed twice with 0.25 M NaHCO 3 /Na 2 CO 3 pH 10 and resuspended in the same buffer to give the chitin concentration 2 mg/mL. The mixture was heated to 80˚C, 5 M sodium borohydride in 0.1 M NaOH was added to give the final concentration of sodium borohydride of 25 mM, followed by 1 h of incubation. The same amount of 5 M sodium borohydride in 0.1 M NaOH was added for four more times with 1 h of incubation at 80˚C after each. To quench the reaction, equal volume of 0.5 M acetic acid was added and the mixture was incubated overnight at room temperature, with stirring. The reduced chitin was washed repeatedly with water and 50 mM sodium acetate, pH 6.1, 0.01% NaN 3 was added and the chitin was stored at 4˚C.
Reducing-end AA-labeled crystalline α-chitin (AA-α-chitin) was prepared as described in Kurashin et al. [35]. For reducing-end AA-labeling of CNWs the CNWs in 50 mM sodium acetate pH 6.1 were added to buffered methanol until the methanol concentration was 80%. Sodium cyanoborohydride and AA were added to the concentrations of 0.5 M and 50 mM, respectively. The reaction was carried out for 2 h at 80˚C. The AA-labelled CNWs were washed repeatedly with 47.86 mM NaOH through centrifugation (5 min at 4000×g) and re-suspension steps. Finally, 20% glacial acetic acid was added to adjust pH to 6.1, and AA-CNWs were stored at 4˚C.
Determining the directionality of hydrolysis AA-CNWs (1 mg/mL) were incubated with SmChiA (10 nM), SmChiB (10 nM), HCHT50 (10 nM), HCHT39 (10, 20 or 50 nM) or SmChiC (100 nM) in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/mL) at 37˚C, with stirring. After 5, 10, 20, 40 and 60 min aliquots were taken out, the reaction was stopped by adding NaOH to 0.2 M, and the chitin was pelleted by centrifugation (2 min at 10 4 ×g). The concentration of soluble AA-sugars was determined by measuring the fluorescence in the supernatant using excitation and emission wavelengths set to 330 nm and 425 nm, respectively. The relative fluorescence of 450 intensity units/μM measured for AA-labeled NAG was used for calibration [35]. The concentration of the reducing groups in the supernatant was measured using the MBTH method [39]. The degree of total degradation of AA-CNWs (in %) was calculated from the released reducing groups assuming NAG 2 as the sole hydrolysis product. This assumption is plausible since NAG 2 is the predominant product (> 90%) in hydrolysis of crystalline chitin by S. marcescens chitinases [40] as well as HCHT [29]. The degree of the released reducing-end-label (in %) was calculated from the released AA-sugars and the total amount of AA-label in AA-CNWs (9.3 ± 0.4 μmol/g).

Measuring apparent processivity (P app )
Reduced chitin (1 mg/mL) was incubated with 10 nM HCHT50, HCHT39 or 1 nM SmChiC in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/mL) at 37˚C, with stirring. At defined times, aliquots were withdrawn and the reaction was stopped by adding NaOH to 0.2 M. The chitin was pelleted by centrifugation (5 min at 10 4 ×g) and the amount of soluble reducing groups (SRGs) in the supernatant was measured using the MBTH method. The amount of enzyme generated insoluble reducing groups (IRGs) was determined by fluorescence labeling of the enzyme treated reduced chitin with AA as described in Kurashin et al. [35]. Briefly, the chitin pellet was washed twice with water, once with 50 mM sodium acetate, pH 6.1, and twice with water. The chitin was re-suspended in 200 μL water and the AA labeling was carried out in 80% buffered methanol with 0.5 M sodium cyanoborohydride and 50 mM AA at 80˚C for 2 h. The AA-labeled chitin was washed three times with water and three times with 50 mM sodium acetate, pH 6.1. Finally, the chitin was re-suspended in 50 mM sodium acetate, pH 6.1 to the final concentration of 0.5 mg/mL and the fluorescence of the suspension was measured using excitation and emission wavelengths set to 330 nm and 425 nm, respectively. Relative fluorescence of 310 intensity units/μM determined for AA-labeled NAG in 0.5 mg/mL chitin suspension was used for calibration [35]. The value of P app was found as a slope of the linear regression line of the data plotted in coordinates of ([IRG] + [SRG])/[IRG] [35,41,42].

Measuring the probability of endo-mode initiation (P endo )
Before using, AA-α-chitin was treated with NaOH to remove nonspecific label. For that the AA-α-chitin was incubated in 0.2 M NaOH for 15 min at room temperature, followed by three washes with 50 mM sodium acetate, pH 6.1. Washed AA-α-chitin (1 mg/mL) was incubated with 10 nM HCHT50 or HCHT39 in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/mL) at 37˚C, with stirring. At defined times, the reaction was stopped by adding NaOH to 0.2 M and the chitin was pelleted by centrifugation (2 min at 10 4 ×g). The concentration of the reducing groups in the supernatant was measured using the MBTH method. The concentration of soluble AA-sugars was determined by measuring the fluorescence in the supernatant using excitation and emission wavelengths set to 330 nm and 425 nm, respectively. The relative fluorescence of 450 intensity units/μM measured for AA-labeled NAG was used for calibration [35]. The number of AA-sugars released from AA-α-chitin ([AA-sugars]) was taken equal to the number of exo-initiations from the reducing-end. The sum of the numbers of reducing-end exo-initiations and endo-initiations was taken equal to the number of IRGs ([IRG]) generated to the reduced α-chitin under exactly the same experiment conditions (see measuring P app above). The P endo was calculated according to P endo = ([IRG]-[AA-sugars])/[IRG] [35,41].

Measuring initial rates
Crystalline α-chitin, CNWs, or amorphous chitin (0.1-10 mg/mL) were incubated with 100 nM HCHT50 or HCHT39 in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/ mL) at 37˚C for 1 min, without stirring. The reaction was stopped by adding NaOH up to 0.2 M. For t = 0, NaOH was added before the enzyme. Chitin was sedimented by centrifugation (5 min at 10 4 ×g) and the concentration of the reducing groups in the supernatant was measured using the MBTH method. In the case of CNWs as substrate an additional amount of CNWs (to 3 mg/mL) were added to the NaOH stopped reactions before centrifugation. This was done in order to improve the sedimentation of CNWs during centrifugation [38].

Measuring the total binding of HCHT to crystalline α-chitin ([HCHT bound ])
Crystalline α-chitin (1 mg/mL) was incubated with 10 nM HCHT50 or HCHT39 in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/mL) at 37˚C, with stirring. At selected times, 400 μL aliquots were withdrawn and chitin was sedimented by centrifugation (1 min 10 4 ×g). The concentration of the free enzyme in the supernatant ([HCHT free ]) was assesseded by measuring the MU-NAG 2 hydrolyzing activity in the supernatant. For that 100 μL of the supernatant was added to 100 μL of 10 μM MU-NAG 2 and incubated for 2 min at 37˚C, without stirring. The reactions were stopped by adding NaOH to 10 mM. The concentration of MU ([MU]) was quantified by fluorescence with excitation and emission wavelengths set to 360 nm and 450 nm, respectively. Before fluorescence measurements the volumes of the reaction mixtures were brought to 1 mL with 0.1 M ammonium hydroxide.
[HCHT free ] was found from the released [MU] using the calibration curves made for the hydrolysis of MU-NAG 2 (5 μM) at different HCHT concentrations (0.5-10 nM). Concentration of HCHT bound to chitin ([HCHT bound ]) was found as a difference between the total concentration of HCHT and [HCHT free ].

Measuring the concentration of HCHT with active site occupied by the chitin ([HCHT bound-OA ])
[HCHT bound-OA ] was measured by following the inhibition of MU-NAG 2 hydrolyzing activity of HCHT by crystalline α-chitin [38]. α-chitin (1 mg/mL) was incubated with 10 nM HCHT50 or HCHT39 in 50 mM sodium acetate, pH 6.1, supplemented with BSA (0.1 mg/mL) at 37˚C, with stirring. At defined times 100 μL aliquots of reaction mixtures were pipetted to 100 μL of 10 μM MU-NAG 2 and incubated for 2 min at 37˚C, without stirring. The reactions were stopped by adding NaOH to 10 mM and the volumes of the reaction mixtures were brought to 1 mL with 0.1 M ammonium hydroxide. Chitin was pelleted by centrifugation (1 min 10 4 ×g) and the concentration of MU in the supernatant was quantified by fluorescence (see above). The concentration of HCHT with free active sites ([HCHT FA ]) was found from the rates of MU-NAG 2 hydrolysis in the presence of crystalline α-chitin using the calibration curves made without chitin. The possible inhibition of MU-NAG 2 hydrolyzing activity of HCHT by NAG 2 released from chitin was judged to be negligible. [HCHT bound-OA ] was found as a difference between the total concentration of HCHT and [HCHT FA ].
Hydrolysis of MU-NAG 2 and inhibition with NAG 2 MU-NAG 2 (1-50 μM) was incubated with 1 nM HCHT50 or HCHT39 in 50 mM sodium acetate pH 6.1 supplemented with BSA (0.1 mg/mL) at 37˚C, without stirring. At defined times the reactions were stopped by adding NaOH to 10 mM and the volumes were brought to 1 mL with 0.1 M ammonium hydroxide. The concentration of released MU was determined by fluorescence with excitation and emission wavelengths set to 360 nm and 450 nm, respectively. For zero time points NaOH was added before the enzyme. Inhibition of MU-NAG 2 hydrolysis by NAG 2 was studied at three different MU-NAG 2 concentrations (0.5, 5.0 and 50 μM) by varying the concentration of NAG 2 between 0.05-2 μM. NAG 2 inhibition of the hydrolysis of chitin was assessed using 14 C-CNWs as substrate. For that, 14 C-CNWs (1.0 or 5.0 mg/mL) were incubated with 1 nM HCHT50 or HCHT39 in 50 mM sodium acetate pH 6.1 supplemented with BSA (0.1 mg/mL) at 37˚C, without stirring. A defined amount of NAG 2 was added to the reactions. The reaction was stopped at selected times by adding NaOH to 0.2 M. Non-labeled CNWs (to 3 mg/mL) were added and chitin was separated by centrifugation (5 min 10 4 ×g) and the amount of radioactivity in the supernatant was quantified using a liquid scintillation counter. [38].

Results
HCHT is efficient in releasing the reducing-end label from crystalline chitin According to its active site architecture, HCHT is reminiscent to SmChiA, a processive enzyme moving towards the non-reducing end of chitin chain [43,44]. However, the experimental evidence for the directionality in the degradation of chitin polymer is lacking. As demonstrated with cellulases, the directionality of hydrolysis can be assessed upon the hydrolysis of reducing-end labeled polymeric substrates [45,46]. When plotted in coordinates released end-label (%) versus total degradation (%), the hydrolysis by reducing-end exo-acting enzymes results in convex curves whereas non-reducing-end exo-active enzymes generate concave progress curves [45,46]. Here, we followed the hydrolysis of reducing-end anthranilic acid (AA) labeled CNWs (AA-CNWs) by full-length HCHT and its 39 kDa isoform lacking the CBM. The S. marcescens chitinases with well characterized mode of action were also included. Most effective in releasing AA-label from AA-CNWs was HCHT followed by SmChiA (Fig 2A). Convex progress curves with high initial slopes are indicative of reducing-endo exo-mode initiations with low processivity [45]. Hydrolysis by both isoforms of HCHT resulted in overlapping convex progress curves indicating that CBM has no role in determining the mode of action of the enzyme. Experiments made at different concentrations of HCHT39 resulted in overlapping progress curves (Fig 2B) indicating that the mode of action is independent of enzyme concentration. In accord with its opposite directionality, SmChiB was much less efficient in releasing the reducing-end label from AA-CNWs compared to SmChiA. The progress curve of endochitinase SmChiC was similar to that of SmChiB. The use of endo-mode initiation obviously complicates the interpretation of progress curves [45]. Totally random mode of initiation with low processivity is expected to result in the equal extent of the release of end-label and total degradation [45]. However, this implies uniform distribution of the end-label, which is not possible to achieve in practice with crystalline substrates. This is because the chain ends buried inside the chitin crystal are not accessible for labeling. As a result, the relative amount of AA label in outer layers of chitin crystal is higher than that in inner layers [46].

HCHT preferentially uses endo-mode initiation in hydrolysis of α-chitin
It has been shown previously that HCHT acts as an endo enzyme on soluble substrates such as chitosans [27][28][29]. However, the soluble nature of chitosan may promote the use of the endomode of action of an enzyme [27]. Therefore, we tested the possible use of endo-mode initiation on insoluble α-chitin. For this we prepared two different substrates: one with a fluorescence label at the reducing ends of chitin chains-reducing-end AA-labeled α-chitin (AA-αchitin), and the other with the reducing ends reduced with NaBH 4 to corresponding alditolsreduced α-chitin. These two substrates were used in parallel hydrolysis experiments with HCHT in identical reaction conditions. We measured the release of AA-sugars from AA-αchitin and compared it with the number of insoluble reducing groups (IRGs) generated in the hydrolysis of reduced α-chitin (Fig 3B). The release of soluble reducing groups (SRGs) from both substrates was identical (Fig 3A), indicating that the general activity of the enzyme was not affected by the nature of the reducing end of α-chitin. The number of released AA-sugars from AA-α-chitin represents the number of initiations from the reducing end. The number of IRGs generated to reduced α-chitin represents the sum of the endo-mode initiations and the initiations from the reducing end. Therefore, the probability of endo mode initiation (relative to that of reducing-end exo) can be found as P endo = ([IRG]-[AA-sugars]) / [IRG] [35,41]. Both HCHT variants had P endo values around 0.95, indicating the predominant use of endomode initiation on crystalline α-chitin (Table 1). HCHT shows limited processivity in hydrolysis of α-chitin Apparent processivity of an enzyme (P app ) is the experimentally measured value of processivity on a real polymer. P app is defined as the number of catalytic events divided by the number of the initiations of processive runs. It has been shown that, for the GHs employing reducingend-exo and/or endo-mode initiation, P app can be found from the hydrolysis of a reduced Human Chitotriosidase Is an Endo-Processive Enzyme polymeric substrate under single-hit conditions (i.e. in the conditions that minimize the chance of the same chain being hit twice) [41,42]. The number of catalytic events is represented by the sum of soluble reducing ends (SRGs) and insoluble reducing ends generated to reduced polymeric substrate (IRGs). For an enzyme employing reducing-end exo-and/or endo-mode initiation, the number of the initiations of processive runs equals to the number of IRGs generated on reduced chitin. Thus, P app of HCHT can be calculated using the equation P app = (IRG + SRG) / IRG. Here we followed the hydrolysis of reduced α-chitin by both variants of HCHT. SmChiC was studied in parallel. The most efficient in generating IRGs was SmChiC followed by HCHT50 and HCHT39 (Fig 3B). The same trends were observed also in the release of SRGs (Fig 3A, note that the concentration of HCHT was 10-fold higher than that of SmChiC). The value of P app can be found as the slope of the linear regression line of the data plotted in coordinates of (IRG + SRG) versus IRG (Fig 3C). With all enzymes a slight deviation from linearity was observed at higher IRG concentrations, indicating a deviation from singlehit conditions at longer hydrolysis times. Therefore, only the data points within the linear regions of curves were used in calculating the P app values. Both HCHT isoforms displayed limited processivity, which was still about twofold higher than that measured for SmChiC ( Table 1).

Binding of HCHT to chitin-differentiation between different populations of bound enzyme
Binding of enzymes to insoluble polysaccharides is often assessed by measuring the concentration of enzyme free in solution. For that, the polysaccharide bound enzyme is separated from that free in solution by centrifugation or filtration. The concentration of total bound enzyme can be found as a difference between total enzyme concentration and that free in solution. However, more detailed information about the binding is available from the experiments, where the hydrolysis of a low molecular weight reporter molecule is performed in the presence of polymeric substrate of interest [38,[47][48][49][50]. Only the enzyme molecules with active site free from polymer can hydrolyze the reporter molecule and this forms the basis for quantification of the population of enzymes with free active site. The concentration of bound enzyme with active site occupied by the polymer ([E] bound-OA ) can now be found as a difference between the total concentration of the enzyme and that with free active site. Furthermore, when the concentration of total bound enzyme ([E] bound ) is measured in parallel, the concentration of bound enzyme with free active site ([E] bound-FA ) can be found as a difference between [E] bound and [E] bound-OA . Here we assessed the kinetics of the binding of HCHT (10 nM) to α-chitin (1 mg/mL). MU-NAG 2 was used as a reporter molecule in measuring the active site mediated binding. For the full-length HCHT, the concentrations of all populations of bound enzyme were at their plateau value after the first 5 min of incubation with α-chitin (Fig 3D). The binding of HCHT39 was somewhat slower than that of HCHT50, especially at the level of the active site mediated binding (Fig 3D). It is noteworthy that both isoforms had a significant population of bound enzyme with free active site (Fig 3D). However, we note as a caveat here that there is a possibility that HCHT is bound to chitin through the substrate binding region of active site cleft so that subsites -2 to +1 are available for the hydrolysis of MU-NAG 2 . This binding mode reveals as [HCHT] bound-FA and not as [E] bound-OA in our experiment.

Dissociation rate constant of HCHT
Like P app , the dissociation rate constant (k off ) of GHs employing reducing-end-exo and/or endo-mode initiation can be found from the hydrolysis of a reduced polymeric substrate under single-hit conditions [41,42]. It has been shown that, when calculated as k IRG = v IRG / [E] bound-OA , the rate constant of the formation of IRGs (k IRG ) represents k off [47]. In this study, v IRG is the rate of the generation of IRGs (v IRG ) upon the hydrolysis of reduced chitin and [E] bound-OA stands for the concentration of bound HCHT with active site occupied by chitin (see above). SmChiC was studied in parallel with both HCHT isoforms. With all enzymes studied, v IRG decreased with hydrolysis time as evidenced by non-linear time curves of IRG formation ( Fig 3B). This non-linearity is common in hydrolysis of recalcitrant polysaccharides and is apparently caused by changes in substrate with hydrolysis time [51]. Therefore, we estimated the k off value using the rate of IRG formation and [E] bound-OA after the first 5 min of hydrolysis (Fig 3B & 3D). Both HCHT isoforms displayed similar k off values (Table 1) with half-lives around 1.7 s. Unfortunately there is no suitable reporter molecule for measuring the active site mediated binding of SmChiC. Using the total enzyme concentration instead of [E] bound-OA provides a minimum estimate for the k off value of SmChiC, which was still about an order of magnitude higher than the k off measured for HCHT (Table 1).

Catalytic constant and intrinsic processivity of HCHT
An important kinetic parameter of polymer-active enzymes is the enzymes intrinsic processivity (P Intr ) [41,42]. While P app is a processivity value measured under given experiment conditions, P Intr represents the processivity potential of an enzyme. P Intr is governed by catalytic constant (k cat ) and k off , and for the processive enzymes P Intr can be approximated by P Intr % k cat / k off [41,42,52] The estimates for k cat can be found from the measurements of initial rates at saturating substrate concentrations as k cat = V max / [E] total . Here we measured the release of SRGs by HCHT after 1 min of hydrolysis on 3 different substratesα-chitin, CNWs, and amorphous chitin. With all enzymes and substrates, the initial rates measured at different substrate concentrations followed the Michaelis-Menten saturation kinetics (Fig 4). Resulting V max /[E] total and K m values are listed in  (Table 1). The P Intr values of both isoforms were similar to each other within error limits. Although the P Intr values were somewhat higher than P app values measured on reduced α-chitin, the difference was far less prominent than that usually found for processive GHs [35,41].

Product inhibition of HCHT
Strong binding of disaccharide product to product binding subsites is required for processivity [34,38,53]. Therefore, processive GHs are more sensitive to product inhibition than their nonprocessive counterparts. Here we studied the NAG 2 inhibition of HCHT in the hydrolysis of low-molecular weight model substrate, MU-NAG 2 , as well as crystalline chitin ( 14 C-CNW). Inconsistent with Michaelis-Menten saturation kinetics, the release of MU from MU-NAG 2 showed distinct substrate inhibition at higher MU-NAG 2 concentrations (Fig 5A). The phenomenon of substrate inhibition is often reported with retaining GHs and can be accounted for by transglycosylation to substrate [25,54,55]. Breakdown of the Michaelis-Menten kinetics did not permit to assess the effects of inhibitor to V max and K m . Inhibition of MU-NAG 2 hydrolysis by NAG 2 was studied at substrate concentrations below (0.5 μM, and 5 μM) and above (50 μM) the optimal substrate concentration (around 15-20 μM). The NAG 2 inhibition appeared weaker in the case of 0.5 μM substrate concentration and was nearly equal in the case  Human Chitotriosidase Is an Endo-Processive Enzyme of 5 μM and 50 μM MU-NAG 2 concentrations. This pattern is reminiscent to uncompetitive inhibition, a mechanism of inhibition where the inhibitor binds to the enzyme-substrate complex but not to the free enzyme. Furthermore, the NAG 2 inhibition of HCHT was different from the "conventional" hyperbolic decrease of rate with increasing inhibitor concentration ( Fig 5B). The enzyme appeared to be slightly activated at low NAG 2 concentrations followed by inhibition at higher concentrations. Such an apparent activation by inhibitor is often observed with glucose tolerant β-glucosidases. Mechanistic interpretations of the phenomenon include transglycosylation reactions to inhibitor but also the competition of inhibitor with the non-productive binding of substrate [56][57][58]. A more detailed analysis of the inhibition mechanisms of HCHT was beyond the scope of present study. We also tested the possible inhibition of HCHT by NAG and glucose using 5 μM MU-NAG 2 as substrate. These monosaccharides were not inhibitory to HCHT at highest concentration (100 mM) tested (data not shown). Using 14 C-CNWs as substrate, the NAG 2 inhibition of HCHT was weak with about 80% activity retained in the presence of 2 mM NAG 2 , the highest concentration tested (Fig 5C). With 14 C-CNW substrate, the NAG 2 inhibition seemed independent on the concentration of substrate as judged by similar strength and pattern of inhibition observed at 1 mg/mL and 5 mg/mL 14 C-CNW concentrations (data not shown). With both substrates MU-NAG 2 and 14 C-CNWs the strength and the pattern of NAG 2 inhibition of two HCHT isoforms was identical within error limits.

Discussion
Human chitinases, HCHT and AMCase are expressed by different cells of immune system and their native substrates are chitin molecules of pathogen origin. This is α-chitin, a chitin with antiparallel orientation of polymer chains in the chitin crystal [59]. β-chitin with parallel orientation of polymer chains has been found only in a few marine organisms and has thus little probability to come into contact with human immune system [60]. However, to date detailed biochemical characterization of HCHT has been made using β-chitin or soluble chitin derivative, chitosan, as substrates [26][27][28][29][30]. Therefore, we made a thorough characterization of HCHT in terms of the mode of action, processivity, binding, and rate constants for the catalysis and dissociation using α-chitin as substrate. Since HCHT has been demonstrated to be active in two isoforms, 50 kDa full-length protein and its 39 kDa truncated version without the C-terminal CBM [19], both were included into the study. The performance of HCHT is discussed in comparison with well characterized chitinases of bacterium S. marcescens [61].
According to its active site architecture HCHT, is reminiscent to SmChiA, a processive enzyme moving towards the non-reducing end of chitin chain [43,44]. However, the location of CBM of HCHT is reminiscent to that of SmChiB, a processive enzyme moving towards the reducing end of chitin chain [44]. When moving towards the non-reducing end, HCHT needs to "drag" its CBM during processive movement. A recent structural study has revealed that the length of the linker peptide is sufficient to enable the location of CBM on both sides of catalytic domain [16] suggesting the possibility to move towards the non-reducing end without "dragging" CBM. The progress curves of the hydrolysis of AA-α-chitin measured here support the reducing-end exo-mode initiation with low processivity (Fig 2). At the same time, comparison of the release of reducing-end label from AA-α-chitin and IRGs generated to reduced α-chitin reveals predominant (95%) use of endo-mode initiation. The corresponding figure reported for SmChiA is 76% [35]. In parallel with the higher probability of endo-mode initiation, the processivity of HCHT was about 6 fold lower than that of SmChiA (36.5) measured using the same experimental approach [35]. In comparison with SmChiA (0.057 1/s for SmChiA [35]), HCHT had also higher off-rate constant (around 0.4 1/s, Table 1) as judged by the rate of generation of IRGs to reduced α-chitin. Combining the rate constants for catalysis and dissociation provides with the estimates of the P intr values, a parameter that represents the processivity potential of an enzyme [41,42]. Notably HCHT had similar P app and P intr values, a phenomenon characteristic to endo-enzymes [41]. Still, HCHT showed significantly higher processivity and lower off-rate constant than the typical endo-chitinase, SmChiC (Table 1). Collectively these properties place HCHT in-between processive exo-enzymes and randomly acting non-processive endo-enzymes. The endo-processive character of HCHT may provide explanation to its efficiency in complete degradation of crystalline β-chitin without the aid of synergistic enzyme components like endo-chitinases [29].
Regarding the role of CBM, we note that apart from the K m values the two HCHT isoforms had similar kinetic properties. CBM had no effect on the mode of action as revealed by overlapping progress curves (Fig 2) and similar probability of endo-mode initiation ( Table 1). Both isoforms had also similar V max /E 0 values on different chitin substrates ( Table 2) and similar off-rates as well as intrinsic processivities ( Table 1). The strength and the pattern of NAG 2 inhibition of both isoforms was also similar (Fig 5). Although the difference was not prominent, we note that the truncated isoform had somewhat higher apparent processivity on reduced α-chitin (Table 1). Similar trends have been observed in hydrolysis of β-chitin where the processivity (calculated as the NAG 2 /NAG product ratio) values of 11.4 and 7.6 were found for HCHT39 and HCHT50, respectively [29]. The hinge region connecting CBM with catalytic domain contains 31 amino acid residues [16]. This translates into the length of about 10 nm considering the maximum length of the polypeptide chain of approximately 0.34 nm per amino acid residue. Since NAG 2 is the product of processive hydrolysis and provided with the length of NAG 2 unit of 1 nm, the average length of processive run of HCHT50 is 5.6 nm ( Table 1). The processive run of this length can be achieved just by stretching the hinge region without the need to "drag" CBM. In this context, it is tempting to speculate that the somewhat lower apparent processivity of full-length isoform compared to the truncated version may reflect the constraints posed by the stretching of the hinge region.
The most prominent difference between the two HCHT isoforms is seen in the lower affinity of truncated version for chitin substrates ( Table 2). Since both isolated CBM [24,62] and catalytic domain [19] can bind to chitin, the higher affinity of full length enzyme apparently reflects the synergistic interaction between domains in binding. Regarding binding, it is worth noting that about 10% of both isoforms was bound to α-chitin with the active site free for the hydrolysis of MU-NAG 2 (Fig 3D). Whereas for the full length enzyme the E bound-FA can readily be interpreted as the population of enzyme bound through CBM only, the presence of E bound-FA is less obvious in the case of the CBM-less isoform. The E bound-FA of HCHT39 may represent the population with chitin chain bound to the subsites -6 to -3. Although not identified in HCHT, the presence of surface binding sites cannot be excluded.
It has been proposed that the character of immune response to chitin depends on the particle size (reviewed in [11][12][13]). This points to a relationship between chitin concentration and HCHT activity, since the surface area available for the binding of HCHT increases with decreasing particle size. The K m values of HCHT on α-chitin and CNWs (Table 2) were significantly higher than those measured for SmChiA in the same experiment system [35]. The low affinity of HCHT to chitin supports the hypothesis that the aim of chitin hydrolysis by HCHT is in the regulation of immune system response rather than in efficient degradation of chitin, since the regulation at substrate level is effective only at substrate concentrations well below K m . Also, the lysosomal processing of HCHT to remove its CBM further increases the K m of the enzyme and in this way may contribute to the sensitivity of the regulation depending on chitin concentration.

Author Contributions
Conceptualization: MS PV SK.

Formal analysis: SK PV.
Funding acquisition: MS PV.