Thioredoxin-2 but not Thioredoxin-1 is a Substrate of Thioredoxin Peroxidase-1 from Drosophila melanogaster Isolation and characterization of a second thioredoxin in D. melanogaster and evidence for distinct biological functions of Trx-1 and Trx-2

(DmTrx-2) from Drosophila melanogaster and a closely related thioredoxin from the mosquito Anopheles gambiae (AgTrx-1). In order to specify the functions of DmTrx-2 we studied its reaction with DmTrxR-1 as well as the important dithiol-disulfide exchange with GSSG, and its role in peroxide detoxification. For this objective we recombinantly expressed and characterized thioredoxin peroxidase-1 from D. melanogaster (DmTPx-1). DmTrx-2 but not DmTrx-1 was found to be a substrate for this enzyme. This result indicates a clear separation of cellular functions between thioredoxins from the same organism.


INTRODUCTION
Aerobic life conditions that are based on the metabolization of molecular oxygen are inseparably linked to the occurrence of reactive oxygen species (ROS). These compounds are cytotoxic both for the organism but even more so for invaders such as microorganisms and tumor cells. This oxidative challenge is counteracted by a shield of enzymatic and nonenzymatic defense systems including superoxide dismutase (2 O 2 . The main difference between GR and TrxR is an additional redox center located in the Cterminal region of high M r thioredoxin reductases as they are found in most eukaryotes (6).
Mammal TrxRs have an adjacent redox active cysteine-selenocysteine pair (7,8) whereas in the thioredoxin reductase of the malaria parasite Plasmodium falciparum this redox center is represented by two cysteines separated by a spacer of 4 amino acids (9). Thioredoxin reductase is also present in the fruit fly Drosophila melanogaster. These insect thioredoxin reductases (DmTrxR-1 and DmTrxR-2) are the prototype of a third class of high M r TrxRs: they possess two neighboring C-terminal cysteine residues that are essential for thioredoxin reduction (3).

Thioredoxin-dependent GSSG Reduction Assay (GHOST assay)
The GHOST assay was conducted as described previously (15,19). The mixture contained 340 nm. After thioredoxin reduction was complete, 1 mM GSSG was added and GSSG reduction was followed by further NADPH consumption. This allows to determine k 2 , the second order rate constant for the reaction Trx(SH) 2 + GSSG TrxS 2 + 2 GSH.

Antibody production and immunoblots
One milligram of recombinant DmTrx-2 and AgTrx-1, respectively, was injected as an antigen into rabbits (Bioscience, Göttingen, Germany). Insect cell extracts (10 µg and 50 µg total protein) as well as 500 ng purified DmTrx-1 and DmTrx-2 were applied to a 15% SDS-PAGE. After completion of the electrophoresis the proteins were blotted overnight onto a nitrocellulose membrane (20). After incubation with the anti(Trx) IgGs the second antibody (swine anti(rabbit IgG) antibody-horseradish peroxidase conjugate) was added. For staining, the nitrocellulose paper was exposed to a solution of 1-chloronaphthol and hydrogen peroxide in sodium citrate buffer.

RESULTS
When we attempted to characterize D. melanogaster thioredoxin peroxidase-1 (DmTPx-1) (21,22) in detail, it became obvious that the known thioredoxin isoprotein DmTrx-1 ( A. gambiae that is putatively isofunctional to DmTrx-2 was also studied. Like thioredoxin-2 from Drosophila, AgTrx-1 was found to be a highly expressed protein in vivo. The concentrations are in the same order of magnitude as estimated for DmTrx-2. As shown in DmTrx-2 and AgTrx-1 as substrates of Drosophila thioredoxin reductase. Wild-type as well as HIS-tagged DmTrx-2 are equally good substrates of D. melanogaster thioredoxin reductase-1 ( Table I). The measured kinetic parameters, notably the k cat /K m ratio of 3 x 10 6 M -1 s -1 , are in the same order of magnitude as for the DmTrx-1/DmTrxR-1 system and for other homologous thioredoxin systems. Catalysis most likely depends on a dithioldisulfide exchange between the C-terminal redox center of TrxR and its substrate Trx (6). The rate k of this reaction is equal to k cat /K m if a dithiol-disulfide exchange is the rate limiting chemical reaction during catalysis; otherwise k is even greater, that is >3 x 10 6 M -1 s -1 .
AgTrx-1 was found to be an excellent heterologous substrate of DmTrxR-1, with the k cat /K m ratio being 1.2 x 10 6 M -1 s -1 .
Since the Drosophila genome does not encode a genuine glutathione reductase (3) the maintenance of redox homeostasis and antioxidative defense by the thioredoxin system is a key issue (Fig. 3). Thioredoxin reductase is likely to substitute for GR in an indirect way because the broad TrxR substrate spectrum does not include GSSG. Glutathione is most probably kept in reduced state by a nonenzymatic dithiol-disulfide exchange reaction with Trx(SH) 2 . Using the GHOST assay (see Experimental Procedures), k 2 was determined to be 170 M -1 s -1 (= 0.01 µM -1 min -1 ) at 25 °C. This is basically identical with the published value for DmTrx-1 as a GSSG-reducing protein (3).
Thioredoxin peroxidase-1 (DmTPx-1). While DmTrx-1 and DmTrx-2 reduce glutathione disulfide equally well this is not the case for the reduction of DmTPx-1. The occurrence of this enzyme had already been shown in different developmental stages of Drosophila (21). As described by Orr and collaborators (22), DmTPx-1 is a cytosolic protein and overexpression of the gene in Schneider cells results in higher survival rate under peroxide treatment.
DmTPx-1, a protein of 194 amino acids and a molecular mass of 21.7 kDa per subunit, shows sequence elements that are typical for type I TPx, notably in the environment of the two conserved cysteines Cys 47 and Cys 168 (Fig. 4).
The tBuOOH and cumeneOOH as peroxide compounds. Furthermore transspecies activities of the enzyme were examined using AgTrx-1, PfTrx-1 and hTrx(C73S) as reductants. DmTrx-2, in tagged or wild-type form, turned out to be the preferred reducing substrate. The apparent K m and V max values were found to be 9 µM and 14 U mg -1 , respectively, when the assays were conducted in the presence of 70 µM tBuOOH (Table II). The thioredoxin from A. gambiae (AgTrx-1) exhibits only slightly higher K m and lower V max values than DmTrx-2. It was surprising to find that DmTrx-1 was an even poorer substrate for DmTPx-1 than PfTrx-1.
With a K m of 52 µM and a k cat /K m ratio 60fold lower than for the system DmTrx-2/DmTPx-1, thioredoxin-1 is probably not of physiological relevance as a peroxidase-1 substrate.
Peroxides as substrates of DmTPx-1. DmTPx-1 is inactivated by peroxides, an effect that has been previously described for other peroxidases (26). Under the used coupled assay conditions with 1 U/ml DmTrxR-1, this phenomenon is visible as a constantly decreasing rate of NADPH consumption that can be reversed by addition of fresh peroxidase aliquots (Fig. 5).  (Table I).
This value is almost identical when comparing the two cytosolic thioredoxins, DmTrx-1 and DmTrx-2. We could demonstrate that DmTrx-2 is indeed a major protein in situ. As suggested by Western blot analysis it represents approximately 0.3 to 1% of the protein extracted from Schneider cells or whole insects (Fig. 1). Assuming a protein concentration of 100 mg/ml cytosol and 50% cytosol in the cell, the cytosolic Trx-2 concentration is estimated to be above (1 U/ml), the resulting GSSG flux v = k 2 [GSSG][Trx(SH) 2 ] would therefore be in the range of physiological demand, that is approx. 100 µM min -1 (29), and may be further adapted to oxidative challenge by increasing the thioredoxin concentration.
Thioredoxin peroxidase and its reducing substrate DmTrx-2. The two cytosolic thioredoxins do not exhibit immunochemical cross-reactivity (Fig. 1). Another important difference is their behavior as electron donors for thioredoxin peroxidase (DmTPx-1). To study this crucial peroxide reduction pathway in more detail we cloned and characterized the enzyme. TPx-1 was found to be a well conserved homodecameric protein. For the structurally known enzyme of human erythrocytes it has been discussed that the decameric state may occur only in an oxidative environment (25). For DmTPx-1, however, we found that this aggregate is also the major molecular form in the presence of 5 mM dithiothreitol. TPx is not only involved in direct detoxification of reactive peroxides. In yeast it was found that at least one TPx-1 orthologue is essential for launching off the cell´s response to oxidative stress (30), a process which is based on up and down regulation of numerous gene activities. DmTPx-1 belongs to a group of proteins in Drosophila that are differentially expressed in the developmental stages of Drosophila (21). This is consistent with observations on other organisms where TPx is involved in redox signaling pathways (16,17) and repression of apoptosis (31).
Our kinetic data show that only one of the two cytosolic thioredoxins in D. melanogaster, namely DmTrx-2, serves as a substrate of DmTPx-1. DmTrx-1 is unlikely to be a substrate of physiological importance because its K m of 52 µM is too high and its k cat of 10 4 M -1 s -1 (Table II) is extremely low. In contrast, the kinetic parameters for DmTrx-2 (K m = 9 µM; k cat /K m = 6 x 10 5 M -1 s -1 ) are most favorable for serving as a TPx-1 substrate in situ. Thus Trx-2 has a function which is not shared by the isoprotein Trx-1 probably operating in the same cell compartment. To our knowledge this is the first clear example of functional differentiation of thioredoxin isoproteins. It will be interesting to see if Trx-2 can substitute for Trx-1 as a morphogenetic factor since Trx-1 is the deadhead-preventing protein of D.
melanogaster larvae (18,23). Obviously Trx-1 and Trx-2 share a number of functions. In their oxidized form TrxS 2 , both isoproteins are oxidating substrates of thioredoxin reductase, and in their reduced forms they are equally efficient in GSSG reduction and thus in redox buffering.
It should be noted that the functions of TPx-1 delineated above are influenced by the concentration and the redox-state of Trx-2 but not of Trx-1. Thus Trx-2 is expected to have additional indirect effects which are not shared by Trx-1.
The thioredoxin system of disease transmitting insects. The experimental results on the thioredoxin system of D. melanogaster probably apply also for disease-transmitting insects.
As a case in point, Anopheles gambiae thioredoxin-1 was found to be closely related to DmTrx-2 both in structure and function. The A. gambiae protein serves as a substrate of DmTrxR-1 and of DmTPx-1, and it reduces GSSG. Thus the redox metabolism on the basis of thioredoxins and thioredoxin reductases -instead of GR which is not present in Drosophila nor in Anopheles -offers a new target for the combat against insects and insect borne parasitic diseases like malaria (32,33). The causative agent of tropical malaria, Plasmodium falciparum, is particularly sensitive to oxidative stress. Severe disturbance of the redox buffering capacity in the human host erythrocyte is caused, for instance, by glucose-6phosphate dehydrogenase deficiency or inhibition of erythrocytic GR. These conditions in the host compromize the development of parasitic blood stages (34,35). Not only in the human host, but also in the insect vector, oxidative compounds like peroxynitrite are produced and launched against the parasite (36,37). In analogy to GR inhibitors as antiparasitic drugs in man (38), TrxR inhibitors would be promising agents against the insect stages of the parasite but also as insecticides. This perspective is supported by the fact that there are great structural and functional differences among the thioredoxin-reducing centers of the TrxRs from insects, malarial parasites and mammals (15). Only the mammalian enzymes contain a selenocysteine as a functional group so that the difference between disulfide and selenyl sulfide biochemistry can be used for the design of specific enzyme inhibitors. Ideally, such an inhibitor would inactivate both parasite TrxR and insect TrxR and thus serve as a prototype of parasiticide/insecticide that is taken up by the insect when feeding on blood of treated patients.