Cross-species activation of trypanosome S-adenosylmethionine decarboxylase by the regulatory subunit prozyme

Abstract

The protozoan parasite Trypanosoma cruzi is the causative agent of Chagas disease (American trypanosomiasis), a neglected disease of Central and South America. Polyamines are small organic cations that are required for cell growth and their biosynthesis has been the target of drug discovery efforts in both T. cruzi and the related T. brucei parasites. Here we show that, as previously demonstrated for T. brucei, S-adenosylmethionine decarboxylase (AdoMetDC) from T. cruzi forms a heterodimer with prozyme, an inactive homolog that arose by gene duplication of the canonical enzyme uniquely in the trypanosomatids. The T. cruzi AdoMetDC/prozyme heterodimer is 110-fold more active than homodimeric AdoMetDC. Unlike for T. brucei AdoMetDC, the activity of the T. cruzi heterodimer is further stimulated by putrescine to generate an enzyme with similar catalytic efficiency to the fully activated T. brucei enzyme. The effects of prozyme on T. cruzi AdoMetDC are mediated by an increase in k_cat, while the predominant effect of putrescine is to lower the K_m. Finally we demonstrate that the cross-species heterodimers of T. cruzi and T. brucei AdoMetDC and prozyme pairs are functional, and that putrescine is required for prozyme to fully activate the mixed species heterodimers. These data demonstrate that prozyme mediated activation of AdoMetDC is a common mechanism required to regulate AdoMetDC activity in the trypanosomatids.

Introduction

The trypanosomatid parasites, Trypanosoma cruzi and Trypanosoma brucei cause significant human and veterinary disease worldwide [1-3]. In Africa 50 million people are at risk for contracting African sleeping sickness caused by T. brucei. The disease is fatal if untreated. In South America the causative agent of Chagas disease,T. cruzi, has an incidence of 200,000 cases each year and over 13 million people are chronically infected with the parasite. Chagas disease is a major cause of cardiomyopathy in the region. Current therapeutics for both Chagas disease and African sleeping sickness are limited, underscoring the need for the development of new and improved anti-trypanosomal drugs and drug targets.

Polyamines are small organic cations that are required for cell growth and proliferation [4]. Polyamines are synthesized from L-ornithine and decarboxylated S-adenosylmethionine (dcAdoMet) (Scheme 1), and the biosynthetic enzymes ornithine decarboxylase (ODC) and S-adenosylmethionine decarboxylases (AdoMetDC) have been studied as chemotherapeutic targets. α-Difluoromethylornithine (DFMO), a suicide inhibitor of ODC, is used clinically for the treatment of African sleeping sickness [5]. In contrast, T. cruzi does not encode an ODC gene, and therefore DFMO is not effective against this parasite (Scheme 1) [6, 7]. Instead, the intracellular T. cruzi pathogen relies on putrescine uptake to obtain this polyamine precursor, which is facilitated by the abundance of putrescine found within host cells [8]. T. cruzi epimastigotes are also able to transport spermidine [9]. In contrast, T. brucei, an extracellular parasite, relies on de novo polyamine synthesis since there are limiting amounts of polyamines in the blood and lymph.

Scheme 1. The polyamine biosynthesis pathway in T. cruzi.

AdoMet (S-adenosylmethionine); dcAdoMet (decarboxylated AdoMet)Put (putrescine); Spd (spermidine); SpdSyn (Spd synthetase); γ-GC (gamma-glutamylcysteine); γ-GCS (γ-GC synthetase); GSH (glutathione); GS (GSH synthetase); GSH-Spd (glutathionyl-spermidine); TSH (trypanothione); TrypSyn (TSH synthetase); TS₂ (oxidized TSH); T(SH)₂ (reduced TSH); TrypRed (TSH reductase).

AdoMetDC catalyzes the decarboxylation of S-adenosylmethione to form dcAdoMet, which then serves as the amino propyl donor for the formation of spermidine from putrescine (Scheme 1). Both T. cruzi and T. brucei contain the gene for AdoMetDC, and inhibitors of this enzyme have been demonstrated to cure murine models of infection with either trypanosome species [10-12]. AdoMetDC is a pyruvoyl-dependent enzyme that derives the pyruvoyl cofactor from an auto processing reaction that cleaves the polypeptide chain into a larger α-chain, containing the pyruvate at the N-terminus, and a smaller β-chain [13]. The human AdoMetDC enzyme is a homodimer (α₂β₂) that is allosterically activated by putrescine [14, 15]. Previous work has shown that the recombinant AdoMetDC enzyme from T. cruzi forms a homodimer that is stimulated by putrescine, however it has a much lower catalytic efficiency than human or plant homologs [16-18]. We recently demonstrated that the T. brucei AdoMetDC is allosterically activated by a regulatory protein, prozyme [19]. Prozyme is an inactive paralog of AdoMetDC that arose by gene duplication and is found only in the trypanosomatids. T. brucei prozyme and AdoMetDC form a high affinity heterodimer that is the functional species of the enzyme in the parasite. Furthermore prozyme protein expression levels are regulated in T. brucei, providing a mechanism to control pathway flux that is unique to these parasites [20].

The functional role for prozyme as an activator has so far only been demonstrated for T. brucei AdoMetDC, though phylogenic analysis suggests that both T. cruzi and Leishmania AdoMetDC will also be activated by prozyme [19]. Here we show that T. cruzi AdoMetDC is also activated by prozyme. The AdoMetDC/prozyme heterodimer from T. cruzi forms with high affinity stimulating the activity of the enzyme; however, in contrast to the T. brucei enzyme, the activity of the T. cruzi heterodimer is putrescine-dependent. Finally we demonstrate that prozyme and AdoMetDC from T. brucei and T. cruzi are able to form functional cross-species heterodimers, demonstrating that the dimer interface and mechanism of activation are conserved between the species.

Materials and Methods

Cloning of T. cruzi prozyme and AdoMetDC

The expression construct (TcAdoMetDC) for T. cruzi AdoMetDC has been described previously [16]. The gene for T. cruzi prozyme was found by blast search of geneDB (http://www.genedb.org/, ID: Tc00.1047053509167.110) using T. brucei prozyme as the search sequence. T. cruzi prozyme was cloned and amplified by PCR from genomic DNA, and ligated into the pET 15b vector generating an N-terminal His₆-tagged construct. A double expression vector (DEV) that allows for coexpression of His-tagged AdoMetDC and untagged prozyme in E.coli was created. First, the T. cruzi prozyme gene was introduced into the pET22b with its stop codon to produce an untagged protein. Then, the ribosomal binding site (rbs), the T7 promoter and the His₆-AdoMetDC coding region was amplified by PCR from the TcAdoMetDC construct and ligated into this vector to create DEV. Primers used for cloning were as follows: Tc prozyme-Forward for pET 15b (5′-CGCCATATGTTGGAGAGCACCTGGGCAGCCG-3′, includes an Nde I site); Tc prozyme-Reverse for pET 15b (5′-CGCGGATCCTTATTCGGCGCGAATATAGCTGG -3′, includes a BamHI site); Tc DEV-Forward for pET 22b-DEV (5′-CGATAGTCGACTCGAAATTAATACGACTCACTATAGG-3′, includes a SalI site); Tc DEV-Reverse for pET 22b-DEV (5′-TATGCGGCCGCTTACTACTCTTCCACAGAATCTGTGG-3′, contains a NotI site).

Expression and purification of T. cruzi prozyme and AdoMetDC/prozyme complex

T. cruzi AdoMetDC or prozyme were expressed individually in Escherichia coli BL21/DE3 cells containing the respective construct, and purified in two steps by Ni²⁺-agarose (Qiagen) and anion exchange column chromatography (Amersham Mono Q 5/50 GL column) as described [16], except that the buffer was exchanged after elution from Ni⁺² agarose and Mono Q by a HiPrep 26/10 desalting column (Amersham) equilibrated in storage buffer (50 mM Hepes pH 8.0, 50 mM NaCl, 1 mM dithiothreitol). For co-purification, the T. cruzi DEV construct was expressed in E. coli, and the proteins were purified as above. Purified proteins were quantified using their respective extinction coefficients (at 280 nm): AdoMetDC, 61.3 mM^-1 cm^-1; prozyme, 36.9 mM^-1cm^-1; AdoMetDC/prozyme complex, 98.2 mM^-1 cm^-1. SDS-PAGE was used to assess purity. Cloning, expression and purification of His₆-tagged T. brucei AdoMetDC and prozyme has been described previously [19].

Sedimentation Equilibrium by Analytical Ultracentrifugation

The molecular weight of the T. cruzi AdoMetDC/prozyme complex was determined by equilibrium sedimentation analysis using a Beckman XLI analytical ultracentrifuge equipped with an AN60 Ti rotor. Samples in buffer (50 mM HEPES pH 8.0, 50 mM NaCl, and 1 mM β-mercaptoethanol) were loaded at 4, 6 and 8 μM (0.1-0.12 ml total volume) into a six-sector equilibrium centerpiece and equilibrated for data collection at 14,000 and 20,000 rpm. After equilibrium was reached (approximately 24 hours), absorption data were collected at 280 nm through sapphire windows using a radial step size of 0.001 cm. Base-line absorbance readings for each cell were acquired by over speed at 42,000 rpm.

Data sets were analyzed using equations describing a single ideal species model (eq. 1), or a monomer-heterodimer model (eq. 2), as previously described [21], or for the AdoMetDC/prozyme complex globally fitted to a single ideal species model using the Beckman XL-A/XL-I Data Analysis Software version 6.0. Both analyses gave similar results.

$$\mathsf{A}={\mathsf{c}}^{\ast }{\epsilon }^{\ast }{\mathsf{d}}^{\ast }\exp [({\mathsf{M}}^{\ast }{(1-{{\mathsf{v}}_{\mathsf{bar}}}^{\ast }\rho )}^{\ast }{\omega }^{2\ast }({\mathsf{r}}^2-{\mathsf{r}}_{\mathrm{o}}^2)∕2^{\ast }{\mathsf{R}}^{\ast }\mathsf{T}]+\delta$$ Eq. 1

$$\begin{array}{cc}\hfill & \mathsf{A}={\mathsf{c}}^{\ast }{{\epsilon }_{\mathsf{A}}}^{\ast }{\mathsf{d}}^{\ast }\exp [({{\mathsf{M}}_{\mathsf{A}}}^{\ast }{(1-{{\mathsf{v}}_{\mathsf{barA}}}^{\ast }\rho )}^{\ast }{\omega }^{2\ast }({\mathsf{r}}^2-{\mathsf{r}}_{\mathrm{o}}^2)∕2^{\ast }{\mathsf{R}}^{\ast }\mathsf{T}]\hfill \\ \hfill & +{\mathsf{c}}^{\ast }{{\epsilon }_{\mathsf{B}}}^{\ast }{\mathsf{d}}^{\ast }\exp [({{\mathsf{M}}_{\mathsf{B}}}^{\ast }{(1-{{\mathsf{v}}_{\mathsf{barB}}}^{\ast }\rho )}^{\ast }{\omega }^{2\ast }({\mathsf{r}}^2-{\mathsf{r}}_{\mathrm{o}}^2)∕2^{\ast }{\mathsf{R}}^{\ast }\mathsf{T}]\hfill \\ \hfill & +{\mathsf{c}}^{2\ast }{({\epsilon }_{\mathsf{A}}+{\epsilon }_{\mathsf{B}})}^{\ast }{\mathsf{d}}^{\ast }{{\mathsf{K}}_{a,\mathrm{AB}}}^{\ast }\exp [({({{\mathsf{M}}_{\mathsf{A}}}^{\ast }(1-{{\mathsf{v}}_{\mathsf{barA}}}^{\ast }\rho )+\left({{\mathsf{M}}_{\mathsf{B}}}^{\ast }(1-{{\mathsf{v}}_{\mathsf{barB}}}^{\ast }\rho )\right)}^{\ast }{\omega }^{2\ast }({\mathsf{r}}^2-{\mathsf{r}}_{\mathrm{o}}^2)∕2^{\ast }{\mathsf{R}}^{\ast }\mathsf{T}]+\delta \hfill \end{array}$$

Where: A = Absorbance at 280 nm at radial position r; c = concentration at the meniscus in abs units; ε= extinction coefficient (TcADoMetDC monomer 61310 M^-1cm^-1; Tcprozyme monomer 36900 M^-1cm^-1; heterodimer 98210 M^-1cm^-1; caclulated by ExPASy ProtParam tool: http://www.expasy.ch/tools/protparam.html); d= pathlength (1.2 cm); M = Molar mass (g/mol)(M_a = TcAdoMetDC monomer 44,000 dal; M_b = Tcprozyme monomer 37,000 dal.); V_bar = partial specific volume (0.73 mL/g TcADoMetDC; 0.732 mL/g Tcprozyme; 0.731 mL/g Tc heterodimer); ρ = buffer density (1.005 g/ml); ω (angular velocity) = (2*π*rpm)/60; r = radius; r_o=reference radius; R = gas constant (8.314 × 10⁷ ergK^-1mol^-1); T = temperature (277 °Kelvin); δ = offset (experimentally determined from the over speed); K_a = association constant (M). V_bar and ρ were calculated using the program SEDNTERP [22].

Steady-state kinetic analysis

Steady state kinetic analysis was performed by trapping liberated ¹⁴CO₂ on a filter paper soaked in saturated barium hydroxide as described [16, 18, 19]. Reactions were performed over a range of enzyme (1 - 4 μM homodimer or 25-400 nM heterodimer) and substrate (1-[¹⁴C]-AdoMet (25 μM) plus unlabeled AdoMet (0 - 975 μM)) concentrations with or without saturating putrescine (5 mM) at 37 °C in buffer (100 mM HEPES pH 8.0, 50 mM NaCl, and 1 mM DTT). Reactions were allowed to proceed for various times (2.5 - 20 min) before quenching with 6 M HCl. Substrate titration data were fitted to the Michaelis-Menten equation (Eq. 3) to determine the steady-state kinetic parameters using GraphPad Prism version 5.00 for Windows (GraphPad Software, San Diego, CA, USA, www.graphpad.com). Samples were assayed in triplicate.

$${\mathsf{v}}_{\mathrm{o}}={\mathsf{V}}_{\max }\left[\mathsf{S}\right]∕({\mathsf{K}}_{\mathrm{m}}+\left[\mathsf{S}\right])$$ Eq. 3

To analyze the activation of AdoMetDC by prozyme from the other trypanosome species, the proteins were incubated together with an excess of prozyme (1:10 AdoMetDC: prozyme) during the AdoMetDC assay. Substrate titration was performed as described above, and the data was analyzed by Michaelis-Menten kinetics (Graph pad Prism).

Results

T. cruzi AdoMetDC and prozyme form a tight heterodimer

We have previously shown that T. brucei AdoMetDC and prozyme form a heterodimer, and here we examined the recombinant T. cruzi proteins to determine if they also formed a complex (Figure 1 A). Recombinant T. cruzi AdoMetDC undergoes auto-processing to generate the larger α (32 kdal) and smaller β (10 kdal) subunits (Figure 1A, lane A), while the T. cruzi prozyme was not processed and purified as a single polypeptide chain (Figure 1A, lane P) that had no catalytic activity. Recombinant His₆-tagged AdoMetDC and untagged prozyme were co-expressed from a single plasmid, and then co-purified. The subunits remained associated over three chromatography steps: Ni⁺²-agarose, anion exchange and size exclusion (Figure 1A, lane C), similar to the T. brucei proteins.

Figure 1. T. cruzi AdoMetDC forms a tight heterodimer with prozyme.

A. Purification of recombinant proteins from T. cruzi: His₆-tagged AdoMetDC (A), His₆-tagged prozyme (P) and the co-purified complex of His₆-tagged AdoMetDC and untagged prozyme (C). Proteins were purified by Nickel, anion exchange and size exclusion chromatography. B. Sedimentation equilibrium analysis of analytical ultracentrifugation of T. cruzi AdoMetDC/prozyme complex (8μM). Data were fitted to a single species model (left) and a heterologous association model (right).

To further analyze the AdoMetDC/prozyme complex, we used sedimentation equilibrium on the Analytical Ultracentrifuge to determine the molecular weight and dissociation constant of the complex (Figure 1B). The data were fitted to a single ideal species model (Equation 1), which predicted a species mass of 71.8 ± 4 kDa (Figure 1B, left panel). Since the predicted size of the heterodimer is 80.6 kDa, this indicates that the sample composition contains a mixture of dimers and monomers (predicted sizes of 44 kDa for AdoMetDC or 36.6 kDa for prozyme). To determine the affinity of prozyme and AdoMetDC in the heterodimer, the data were also fitted to a heterologous association model (Equation 2), and the K_d for heterodimerization was found to be < 0.5 μM (Figure 1B, right panel). The T. cruzi AdoMetDC homodimer was previously found to have a dissociation constant of 15 - 30 μM [17], thus the heterodimer forms with an affinity that is at least 25 - 50 fold higher than the homodimer, indicating that the heterodimer will be the physiologically relevant enzyme form.

Catalytic efficiency of the T. cruzi AdoMetDC is stimulated by prozyme

To characterize the catalytic efficiency of the T. cruzi AdoMetDC homodimer in comparison to the AdoMetDC/prozyme heterodimer, we measured the activity of the recombinant proteins over a range of substrate concentrations (Figure 2 A and B). As previously reported [16], the T. cruzi AdoMetDC enzyme was stimulated by putrescine and we observed a 17-fold increase in catalytic efficiency. However, in comparison to recombinant human AdoMetDC (k_cat = 2.6 s^-1 and k_cat/K_m= 44,000 M^-1s^-1 in the presence of 5 mM putrescine [16]), the putrescine activated enzyme has a slower catalytic rate (k_cat = 0.02 s^-1) and lower catalytic efficiency (k_cat/K_m= 150 M^-1s^-1). In contrast, the catalytic efficiency of the T. cruzi AdoMetDC/prozyme heterodimer is 110-fold greater than that of the basal T. cruzi AdoMetDC, and the addition of putrescine to the assay further stimulated the heterodimer by 5-fold (Figure 2b and Table 1). The effect of putrescine addition on the heterodimer is seen as a decrease in K_m (∼ 4 fold decrease), where the stimulation by prozyme is observed as an increase in k_cat (40-120 fold). In the presence of putrescine the T. cruzi AdoMetDC/prozyme heterodimer has a catalytic efficiency that is 550-fold greater than the unstimulated AdoMetDC homodimer, and has a similar activity to that observed for the T. brucei heterodimer. We have previously reported that T. brucei AdoMetDC was stimulated by prozyme by 1200-fold on k_cat [19]. Putrescine stimulates the catalytic efficiency of T. brucei AdoMetDC by 10-fold, however in contrast to the result for the T. cruzi heterodimer, the T. brucei heterodimer is fully active in the absence of added putrescine [19].

Figure 2. Steady-state kinetic analysis of recombinant T. cruzi AdoMetDC species.

A. Steady-state kinetics of T. cruzi AdoMetDC homodimer without (circles) and with (squares) 5 mM putrescine. B. Steady-state kinetics of T. cruzi AdoMetDC/prozyme heterodimer without (circles) and with (squares) 5 mM putrescine. Error bars represent n=3 replicates for each experimental point.

Table 1.

Kinetic comparison of T. cruzi AdoMetDC and AdoMetDC/prozyme heterodimer. AdoMetDC and prozyme proteins were coexpressed in E. coli followed by copurification as described in Materials and Methods. Error is the standard error of the fit

	No putrescine			5 mM putrescine			Fold change by putrescine
	kcat (s-1)	Km (μM)	kcat/Km (M-1s-1)	kcat (s-1)	Km (μM)	kcat/Km (M-1s-1)	kcat	Km	kcat/Km
A. T. cruzi AdoMetDC
TcAdoMetDC	0.005 ± 0.001	540 ± 350	9.3 ± 6.6	0.02 ± 0.003	130 ± 63	150 ± 76	4.0	0.24	17
TcAdoMetDC/Tcprozyme	0.60 ± 0.09	580 ± 200	1030 ± 400	0.84 ± 0.06	170 ± 42	5100±1900	1.4	0.29	4.9
Fold change by prozyme	120	1.1	110	41	1.3	32
B. T. brucei AdoMetDC*
TbAdoMetDC	0.0013	0.38	3.4	0.0082	0.24	35	6.3	0.63	10
TbAdoMetDC/Tbprozyme	1.4	0.11	12,000	1.7	0.17	11,000	1.2	1.5	0.92
Fold change by prozyme	1100	0.3	3500	210	0.71	310

^*Data taken from [19]

Prozyme activates AdoMetDC homolog from the other trypanosome species

The prozyme proteins from T. cruzi and T. brucei share 48% sequence homology while the AdoMetDC proteins are 68% homologous. This high degree of conservation prompted us to examine the ability of prozyme from one species to activate the AdoMetDC from the other species. Each recombinant protein was expressed with a His₆-tag, and the proteins were assayed in a ratio of 1:10 AdoMetDC to prozyme. The specific activity of the T. brucei heterodimer was found to be lower when the two subunits are mixed rather than co-purified, probably due to the aggregation of prozyme when expressed alone [19]. The activity of the T. cruzi heterodimer generated by mixing was likewise slightly lower than for the co-purified heterodimer (Tables 1 and 2). Both sets of cross-species heterodimers (TbAdoMetDC/Tcprozyme and TcAdoMetDC/Tbprozyme) were evaluated by subunit mixing experiments. In the absence of putrescine, T. cruzi prozyme stimulated T. brucei AdoMetDC by 60-fold, while the T. brucei prozyme was able to increase the activity of T. cruzi AdoMetDC by 70-fold (Figure 3A, C and Table 2). This level of activation was significantly less than observed for the same species heterodimers. However when putrescine was included, both cross-species heterodimers were further activated yielding mixed species heterodimers with similar catalytic efficiency to the same species heterodimers (Figure 3B, C and Table 2). Thus, the orthologous prozyme can function with the AdoMetDC from another trypanosome species, but only in the presence of putrescine.

Table 2.

Kinetic comparison of T. cruzi and T. brucei AdoMetDC/prozyme heterodimers with cross-species heterodimers. AdoMetDC and prozyme proteins were expressed and purified individually prior to mixing for the kinetic analysis. Error is the standard error of the fit

	No putrescine			5 mM putrescine
	kcat (s-1)	Km (μM)	kcat/Km (M-1s-1)	kcat (s-1)	Km (μM)	kcat/Km (M-1s-1)
Tb AdoMetDC/Tb Prozyme	0.86 ± 0.04	130 ± 20	6600 ± 1100	1.3 ± 0.09	230 ± 46	5600 ± 1200
Tb AdoMetDC/Tc Prozyme	0.09 ± 0.01	450 ± 130	200 ± 62	0.21 ± 0.02	52 ± 17	4000 ± 1400
Tc AdoMetDC/Tc Prozyme	0.53 ± 0.04	200 ± 38	2700 ± 540	0.69 ± 0.05	160 ± 35	4300 ± 990
Tc AdoMetDC/Tb Prozyme	0.03 ± 0.003	440 ± 14	680 ± 210	0.57 ± 0.03	100 ± 20	5700 ± 1200

Figure 3. Steady-state kinetic analysis of cross-species prozyme activation of AdoMetDC orthologs.

A. Steady-state kinetics of AdoMetDC and prozyme combinations (1:10 AdoMetDC to prozyme) in the absence of putrescine. B. Steady-state kinetics of AdoMetDC and prozyme combinations in the presence of 5 mM putrescine. T. brucei AdoMetDC (black circles) was mixed with T. brucei prozyme (closed circles) and with T. cruzi prozyme (open circles). T. cruzi AdoMetDC (grey squares) was mixed with T. cruzi prozyme (closed squares) and with T. brucei prozyme (open squares). C. Comparison of the catalytic efficiency of AdoMetDC and prozyme mixtures in the absence (black) and presence (grey) of 5 mM putrescine. Error bars represent n=2 replicates for each experimental point.

Discussion

The requirement for AdoMetDC to be activated by heterodimer formation with the inactive homolog, prozyme, is unique to the trypanosomatid species. Here we have demonstrated that the T. cruzi prozyme can form a high affinity heterodimer with AdoMetDC and increase the catalytic efficiency, similar to what we have previously reported for the T. brucei heterodimer [19]. We have shown that the T. cruzi AdoMetDC/prozyme heterodimer is further activated by putrescine, and have demonstrated that the prozyme from one species can increase the catalytic efficiency of AdoMetDC from another species in the presence of putrescine. This confirms that prozyme activation of trypanosomatid AdoMetDC is a general mechanism in these parasites.

As previously described for T. brucei AdoMetDC [19], T. cruzi AdoMetDC forms a high activity heterodimer with T. cruzi prozyme. In the absence of putrescine, the T. cruzi AdoMetDC/prozyme heterodimer is less efficient than the T. brucei heterodimer. However, in the presence of putrecine, the efficiency of the T. cruzi heterodimer increases to a similar level as observed for the T. brucei heterodimer. Thus, the prozyme is required for activation of both trypanosomal AdoMetDCs, while putrescine is important only for the T. cruzi heterodimer. The activity of the T. cruzi heterodimer is increased 5-fold in the presence of putrescine. Both k_cat and K_m of T. cruzi AdoMetDC homodimer are stimulated by putrescine [18]; in contrast the effect of putrescine on the heterodimer is confined to a reduction in K_m, while prozyme acts by increasing k_ca. Structural studies of the human enzyme show that the dimer interface and putrescine binding sites are distant from the active site [14, 23]. This indicates that both prozyme and putrescine will use allosteric mechanisms of activation; however the different molecules regulate activity by causing changes in different catalytic parameters.

The requirement for both prozyme and putrescine to fully activate the T. cruzi AdoMetDC/prozyme heterodimer shows that T. cruzi is able to regulate decarboxylation of AdoMet on two levels. This additional level of control by putrescine may possibly allow the parasite to monitor the outside environment. T. cruzi is an intracellular parasite, whereas the related T. brucei is an extracellular pathogen. While most cell types have abundant polyamine pools, the blood, where the T. brucei parasites live, is a poor source of polyamines. T. cruzi relies on putrescine uptake for the synthesis of spermidine and can also import spermidine much more efficiently than T. brucei [24]. The species difference in putrescine regulation of AdoMetDC/prozyme may be related to the differences in polyamine biosynthesis and environmental niche between these trypanosomes.

The structural basis for the trypanosome species differences in AdoMetDC putrescine activation are unclear as no X-ray structure of the trypanosome AdoMetDC’s has been reported. Site-directed mutagenesis of the T. cruzi AdoMetDC residues predicted to be within the putrescine binding-site based on the human AdoMetDC structure, demonstrated that the putrescine binding-site is not well conserved between the parasite and human enzymes [16, 17]. This finding makes it difficult to predict where the putrescine binding-site will be located in the parasite AdoMetDCs. Despite this several residues have been shown to have a possible role in putrescine binding to T. cruzi AdoMetDC (D174, S111 and F285) and all are conserved in T. brucei AdoMetDC. Thus if trypanosome species differences in the putrescine binding-site are present they have not yet been identified.

The ability of the prozyme to form functional cross-species heterodimers with AdoMetDC from the other trypanosome species indicates that the amino acid residues important for the energetics of dimerization and for the activation mechanism are conserved between the two species. Interestingly putrescine is required for full activation of the cross-species heterodimers, suggesting that the T. cruzi component of these heterodimers may drive this requirement. Functional cross-species dimers have also been reported to form between T. brucei and mammalian ODC [25]. T. brucei and mouse ODC share ∼60% sequence identity, slightly higher than the 50% sequence identity shared between T. brucei and T. cruzi prozyme.

We have recently discovered that prozyme is a key factor in the regulation of polyamine and trypanothione biosynthesis in T. brucei [20]. In unchallenged cells, prozyme is limiting, and not all of the cellular AdoMetDC is activated. Genetic or chemical depletion of AdoMetDC activity results in increased levels of prozyme protein. These data suggest that T. cruzi parasites may also use prozyme to regulate polyamine synthesis and pathway flux. T. cruzi has a requirement for putrescine uptake suggesting prozyme levels could be regulated in response to the available extracellular putrescine in order to maintain balance in the pathway. In support, a 10-fold increase in putrescine was observed after genetic knockdown of AdoMetDC in T. brucei, and this effect remains a possible trigger for the increased prozyme expression observed under these conditions [20].

In conclusion, we have shown that the prozyme from T. cruzi and T. brucei act similarly to activate AdoMetDC. Further functional cross-species heterodimers between prozyme from one species can be formed with AdoMetDC from the other species. Together, these data demonstrate that prozyme evolved as a general mechanism to regulate AdoMetDC activity in the trypanosomatids. The T. cruzi heterodimer can be further stimulated by putrescine, which may be a consequence of the lack of ODC and subsequent need for putrescine import from the invaded host cell environment of these parasites. The trypanosomatid AdoMetDC/prozyme heterodimer interaction is unique to the trypanosomatids and offers a species-selective mechanism to target AdoMetDC in the treatment of trypanosomiasis.

Abbreviations

AdoMetDC

S-adenosylmethionine decarboxylase

AdoMetDC/prozyme

the purified complex between AdoMetDC/prozyme

ODC

ornithine decarboxylase

DFMO

α-Difluromethylornithine

AdoMet

S-adenosylmethionine

dcAdoMet

decarboxylated AdoMet