Bacterial Interference of Arginase

Helicobacter pylori is a Gram-negative micro-aerophillic bacteria that selectively colonize the human stomach. It causes chronic gastritis, peptic ulcer, gastric carcinoma, and lymphoma, leading to its classification as a class I carcinogen [11]. H. pylori arginase is encoded by the gene rocF, which is constitutively expressed. Although both the wild-type and rocF mutant stimulate similar levels of iNOS mRNA, a significantly greater amount of NO production is elicited by the mutant strain in RAW macrophages at a physiologically relevant arginine concentration. This agrees with the hampered survival of the rocF mutant strain, thus indicating the role of the pathogen-induced arginase to quench arginine from iNOS [12], [13]. In a previous study, the Helicobacter arginase was found to protect the bacteria from acid stress, and an arginase-deficient strain showed attenuated colonization in the mouse model [5]. In addition, Helicobacter arginase also impairs host T cell function by reducing CD3ζ chain expression, and this phenotype might play a very important role during Helicobacter infection [14]. These unique functions imply that the Helicobacter arginase has evolved to allow the bacteria to effectively compete with its host in the mucous layer. The decrease in host NO production by Helicobacter arginase may have two important outcomes. The first one certainly is to avoid host nitrosative stress, and the second one might be to reduce the NO-mediated damage of the gastric mucosa. If the second possibility is true, then arginase must be one of the very essential factors for the long-term survival and proliferation of this pathogen in the gastric niche.

In addition, H. pylori upregulates the host arginase II in RAW macrophages and in mouse and human gastritis tissues and induces apoptosis. The product of the arginase pathway, ornithine, is acted upon by ornithine decarboxylase (ODC) to generate ornithine. H. pylori–mediated apoptosis was blocked in the presence of the host arginase inhibitor N^ώ-hydroxy-nor-L-arginine (nor-NOHA), and also in the presence of an ODC inhibitor, α-difluoromethylornithine (DFMO). iNOS inhibition had no effect on H. pylori–mediated apoptosis. Arginase II induction was not observed in mice infected with the nonpathogenic E. coli DH5α strain [15]. This clearly demonstrates that the induction of host arginase is directly correlated with the virulence of the organism. However, there are no reports of arginase induction in hosts with pathogenic E. coli, indicating the specific phenotype varies from pathogen to pathogen. Further, E. coli lacks endogenous arginase. Ornithine is converted to the polyamine spermidine by spermidine synthase (Spd Syn). The next step of polyamine production consists of the production of another polyamine spermine by action of the spermine synthase (Spm Syn) enzyme (Figure 1). H. pylori can utilize this spermine to restrain immune response in the activated macrophages by inhibiting pro-inflammatory gene expression. Spermine could also prevent the antimicrobial effects of NO by inhibiting iNOS translation in the macrophages infected by H. pylori [16]. Future work should provide insights into whether the nor-NOHA-mediated host arginase inhibition can also decrease H. pylori survival in human gastric tissues.

Another intracellular pathogen, Salmonella enterica serovar Typhimurium, also utilizes the host arginase for its own survival inside mouse macrophages. Salmonella is capable of causing clinical signs that range from self-limiting diarrhea to severe fibrinopurulent necrotizing enteritis and life-threatening systemic disease in various hosts. In the spleen of Salmonella Typhimurium–infected mice, a clear increase in the arginase II protein level and activity is observed. Inhibition of arginase via nor-NOHA treatment leads to increased NO production and decreased bacterial burden in the secondary lymphoid organs of the infected mice. Arginase induction is dependent on LPS or any other surface pathogen-associated molecular pattern of Salmonella [17]. The factor that is of importance here is the side effects of inhibition of the host urea cycle. Further, the specific response of the human pathogen Salmonella enterica serovar Typhimurium with respect to arginase induction and modulation of the host arginine metabolism also demands careful consideration.

The successful replication and survival of M. tuberculosis, the causative agent of tuberculosis, inside host macrophages depends on the intelligent strategies employed by the bacteria. In the case of Mycobacterium bovis (BCG) infection, it has been demonstrated that host urea production slowly increases with the infection time. After 24 hours and 72 hours of infection, there was a significant induction in J774.1 macrophage arginase activity. However, the replication of intracellular BCG increased when arginase activity was inhibited. Hence, it is indicative that the enhanced growth of BCG might be due to increased availability of the intracellular arginine pool to the bacteria in the arginase-blocked condition [18].

In an interesting report, the exact mechanism of arginase induction upon mycobacteria infection was documented. In primary mouse macrophages, the BCG infection–mediated increase in the arginase I protein was shown to be MyD88 and TLR2 dependent and independent of the T helper type 2–associated STAT6 pathway. Deletion of arginase I leads to an enhanced NO response with BCG infection. Supporting the BCG data, an increased arginase I mRNA expression was seen in the lungs of M. tuberculosis–infected mice as well. A lower M. tuberculosis load was observed in the Arg-1-deficient mice. In the same report it was observed that the liver granuloma from BCG-infected mice produced greater bacterial nitrotyrosine when host arginase was knocked out [19]. The high NO in the Arg-1-deficient mice in response to the M. tuberculosis infection in conjunction with superoxide leads to the formation of the highly toxic peroxynitrite. Bacterial nitrotyrosine is formed when peroxynitrite nitrosylates tyrosine residues in bacteria. Although in the J774 macrophage cell line arginase inhibition promoted BCG growth, in the mouse model an opposite phenotype was observed. Arginase I–deficient mice were more efficient at clearing both M. tuberculosis and BCG by suppressing NO production from infected macrophages. Recently, in another report it was observed that the supernatant of BCG-infected wild-type macrophages induces arginase I in MyD88^−/− macrophages in a STAT3-dependent manner. This induction was attributed to IL-6/IL-10 secretion by the BCG-infected macrophages that induce arginase in neighboring macrophages by autocrine/paracrine IL-6/IL-10 activation of STAT3 [20]. It can be inferred from this that mycobacteria condition uninfected neighboring cells for low NO production by inducing arginase I.

Modulation of Arginase by Parasites

Trypanosoma cruzi is an obligate intracellular protozoan parasite that grows abundantly in the heart and other organs of patients with acute Chagas disease. Cruzipain (Cz), a major T. cruzi antigen, was found to increase urea production in splenic macrophages [21]. In a subsequent study, cardiocytes were cultured with Cz and a significant increase in host arginase II was observed. This enhanced arginase in turn promotes survival of the mouse cardiomyocyte [22]. During in vivo infection of T. cruzi in BALB/c mice, both arginase I and arginase II were induced in heart tissues. On the one hand, this induction could provide polyamine for the parasite's growth, and on the other might also downregulate the detrimental effects triggered by iNOS in the heart during infection [23]. The reduced apoptosis rate of the cardiomyocyte should ultimately generate an adequate environment for the parasite's growth and dissemination from the host heart. Hence, in T. cruzi infection, anti-apoptotic activity of arginase and its downstream enzymes are utilized by the pathogen for survival. We can speculate that NOHA-mediated arginase inhibition might increase this apoptosis and consequently decrease T. cruzi survival. This is in contrast to H. pylori infection, where there is a host arginase II–dependent polyamine-mediated increase in apoptosis, which might be explained by the differential cell and tissue types wherein polyamine-mediated apoptosis is taking place.

Trypanosoma brucei causes trypanosomiasis in both humans and animals in Africa. Both innate and adaptive immune responses of the host are involved in resistance. Moreover, arginase is also induced in T. brucei infection [24]. In the susceptible mouse strain BALB/c, arginase I and II mRNA and arginase activity were higher compared to the resistant mouse strain C57BL/6. Both of these phenotypes were directly induced by T. brucei infection. In accordance with this phenotype, NO production was significantly higher in the C57BL/6 mice and must be the cause for their resistance. Hence, the macrophage arginase inhibits NO-dependent trypanosome killing. NO generation and parasite survival was restored to the same level in both the susceptible and resistant strains on arginase inhibition. Thus, here arginase serves as a host marker for susceptibility to trypanosome infection [25]. Polyamines are further essential for trypanosome development, as they help in the synthesis of DNA and trypanothione. In protozoans such as Trypanosoma, polyamines like spermidine make trypanothione, which is an antioxidant and is required for parasitic proliferation [26], [27]. NOHA treatment decreased parasitic burden in the macrophage population [28]. This was achieved by an increase in arginine availability for iNOS, and it has been observed that supplementation of arginine restores NO-dependent parasite killing. At this point, the use of the T. brucei–mediated arginase pathway has been identified, but how the pathogen elicits arginase remains ambiguous.

Curiously, T. brucei does not use the arginase pathway for its spread like T. cruzi. This contrast might be explained by the fact that in the case of T. cruzi infection, apoptosis of the infected heart tissue cells is very much required for the pathogen to invade other organs, which is not the case for T. brucei, as it can spread via infected macrophages in the peritoneum cavity. Further investigation is required to determine the effect of arginase at the systemic level during experimental trypanosomiasis.

The protozoan Leishmania is an intracellular parasite of mammalian macrophages. To reside successfully in the very cells responsible for its clearance, Leishmania requires a bag full of immune evasion tricks. This involves avoidance of phagolysosomal fusion and prevention of activation of acquired immune mechanisms. Macrophages can control Leishmania infection when a T helper Type 1 response is mounted and pro-inflammatory cytokines like IFNγ and TNF-α are released. This leads to the induction of iNOS and NO production, which is the major Leishmania killer molecule in the murine system. Arginine, the substrate of NO production, is modulated by both Type 1 and Type 2 responses in a manner such that the Type 1 response increases IFNγ-induced iNOS-mediated conversion of arginine to NO, whereas the Type 2 response promotes arginase induction. [29].

High splenic arginase I expression has been documented in the hamster model of progressive visceral leishmaniasis caused by Leishmania donovani [30]. Further, in subsequent studies, it was shown that even L. major infection leads to arginase I induction in macrophages, and that host arginase I induction supports Leishmania growth. Host arginase I is induced in both the resistant C57BL/6 and susceptible BALB/c mouse strains. However, in C57BL/6 mice, it is induced only during foot pad swelling, and in BALB/c it parallels time of infection. Specific inhibition of host arginase I by nor-NOHA treatment decreases the parasite load and delays lesion development in susceptible BALB/c mice. On the other hand, in resistant C57BL/6 mice ornithine supplementation increases the susceptibility of infection, clearly suggesting that in cutaneous leishmaniasis the host arginase pathway is hijacked by the parasite for polyamine acquisition [31]. L. major increases host arginase I for enhanced polyamine production, which acts as a growth factor. Further, spermidine and spermine inhibit the pro-inflammatory cytokine response of the host, and spermine even modulates immune function by inhibiting the LPS-TLR4 pathway. The inhibition of host arginase I activity has the therapeutic effect of reduced pathology and controlled L. major replication by decreasing polyamine synthesis [32].

Further, it has also been reported that the inhibition of L. major–encoded arginase controls parasite growth. This growth inhibition is caused by decreasing ornithine availability for polyamine synthesis and not by decreasing the host NO response. It was also observed that increasing parasite arginase activity by IL-4 induction further promotes Leishmania growth [33]. It interesting to note here that simultaneous inhibition of both the L. major–encoded arginase and the host arginase I did not result in an increase in the nitrite levels or in the Type 1 immune response of the host [32], [33]. Polyamine biosynthesis in Leishmania occurs by the arginase-ODC pathway only. Consequently, Δarg L. major are auxotrophic for polyamines. However, L. major lacking arginase retains infectivity in the susceptible BALB/c mouse strain. This clearly indicates that arginase-deficient Leishmania can survive in mice by salvaging the polyamines synthesized by the host. However, the pathology in the Δarg L. major infection emerged less rapidly than in the wild-type infection. [34].

Strikingly, the active role played by Leishmania arginase in diverting arginine away from the iNOS pathway was demonstrated by the increased host NO response to Δarg Leishmania mexicana infection in mice. The Δarg L. mexicana infection led to an enhanced Type 1–associated IFNγ response as well. This led to a significant growth attenuation of Δarg L. mexicana in mice [35]. This difference in survival between L. major and L. mexicana arginase knockouts can be attributed to the varied roles played by parasite-encoded arginase in the pathogenesis of different Leishmania species. There is also the possibility of the existence of an alternate arginase of L. major. Further, L. major might have an enhanced capacity to acquire host polyamines to support its growth in mice.

Host NO response is essential also in the control of Toxoplasma gondii infection [36], [37]. To circumvent this NO-mediated killing, T. gondii induces host arginase I protein expression in a STAT6-independent manner within 1 hour of infection. In accordance with the induction, mice lacking arginase I showed a survival advantage over control mice during experimental toxoplasmosis. Mice lacking arginase I did not lose weight and did not show any sign of toxoplasmosis, unlike the control mice [19]. Although the chief mechanism for host survival in T. gondii infection is attributed to a decrease in NO production, there might be additional functions of arginase in Toxoplasma infection.

Arginase is further involved in the pathogenesis of Schistosoma mansoni. It regulates the granulomatous pathology of schistosomiasis in vivo. In infected lung tissues, arginase activity is induced by scistosome eggs. Most of the arginase activity was derived from the host arginase I isoforms as observed by both RNA and protein data [38]. Mice carrying S. mansoni infection further showed a heightened arginase I expression in resident peritoneal macrophages. In the same study, a 10-fold higher level of circulating ornithine-derived polyamines were found in infected mice when compared to the control group [39]. This is of pathological significance as parasitic helminthes are known to depend on their host for uptake and interconversion of polyamines. The effect of arginase inhibition has not been tested on S. mansoni growth. However, an inhibition of the ODC enzyme by DFMO administered in drinking water after 5 weeks post-infection increases granuloma size and hepatic fibrosis in mice. This is explained by the enhanced bioavailability of L-ornithine in the absence of ODC towards ornithine amino transferase (OAT) for proline synthesis. Hence, it can be indirectly concluded from here that Schistosoma-mediated arginase I induction helps the parasite by increasing the available proline for increased collagen deposition [38]. However, direct proof can be obtained only after the development of a specific OAT inhibitor. In Schistosoma-infected livers, proteomic study shows that the overall abundance of arginase I protein is equal to that observed in uninfected livers. This data is consistent with various previous findings wherein arginase I is increased in the granuloma but not in the parenchyma [40].

Schistosoma S. mansoni expresses its own endogenous arginase as well. Although there are structural differences between the host and Schistosoma arginase, they are both functionally similar. Arginase activity increases in case of Schistosoma invasion of the host skin, and this increase is directly attributable to the Schistosoma endogenous arginase. It is probable that Schistosoma attempts to control toxic NO production by the resident innate immune cells in the skin through arginase. However, the exact role of the parasite arginase in Schistosoma pathogenesis has to be further elucidated [41].

Survival Advantage Conferred by Arginase to Other Pathogens

The opportunistic fungal pathogen Candida albicans is part of the normal microflora but can cause systemic infection in immune-compromised individuals when it reaches the bloodstream. In order to escape from macrophages after being ingested, Candida employs a very fascinating strategy of inducing its own intracellular arginase (Car1p) and urea amidolyase to achieve hyphal switching. Once inside the macrophages, Candida rapidly upregulates its arginine biosynthetic genes. Arginine is further metabolized to ornithine and urea by arginase. The resulting urea is degraded to CO₂ and NH₃ by urea amidolyase (Dur1, 2p). CO₂ further activates adenyl cyclase and the cAMP-dependent protein kinase A pathway, thereby activating Efg1p, which triggers the yeast-to-hypha switch of Candida inside macrophages, enabling its release. In addition, Candida induces two other endogenous arginases that are secreted out. These extracellular arginases may provide a survival benefit to Candida by reducing nitrosative stress via quenching the iNOS substrate arginine [42].

It is further reported that in hepatitis C virus (HCV) infection, arginase I mRNA and protein expression is elevated. siRNA-mediated inhibition of arginase I leads to the inability of HCV to stimulate hepatocellular growth. Arginase inhibition also increased NO-mediated cell death. Hence, arginase I plays a very significant role in HCV-mediated hepatocellular growth and survival [43].

Spatial Localization of Arginases

Mammalian arginase I has a cytoplasmic localization, whereas mammalian arginase II is present in the mitochondrial matrix [2]. These two arginases can access the cytoplasmic arginine pool and modulate iNOS function by means of substrate quenching. However, this brings us to an intriguing question about how pathogenic arginases that are intracellular and not secreted outside the cytosol can get access to the host arginine pool. It is well documented that Schistosoma arginase is localized to the head of the organism and is not secreted upon infection [41], nor is the H. pylori arginase, which is again intracellular [44], or the Leishmania arginase, which contains a peroxisomal targeting signal (PTS-1) that directs it to the glycosome, an organelle unique to Leishmania that again is not secreted out [45]. The answer would probably lie in the unique ability of the pathogens to recruit the host arginine transporters to their vacuoles to access the cytoplasmic pool and also to utilize their own endogenous arginine uptake systems. It is hypothesized for Leishmania that it recruits host mCAT2B transporters to its parasitophorous vacuole [29]. It has been further reported that the L. donovani promastigote uses its own arginine permease, LdAAP3, to transport arginine across its membrane [46]. Mycobacteria infection also upregulates host arginine transport and utilizes this host-derived amino acid for its own benefit instead of synthesizing its own [18]. Helicobacter is known to possess its own arginine transport protein, RocE, to uptake arginine present in the extracellular milieu [44]. Further, the recruitment of the host arginine transporter mCAT1 to Salmonella–containing vacuoles and involvement of ArgT, the arginine permease of Salmonella in arginine uptake from host cytosol, has also been observed (unpublished data, P. Das, A. Lahiri, Ay. Lahiri, D. Chakravortty, et al.). Pathogens employ this clever strategy of channeling the host arginine pool to the intracellular pathogen-containing vacuole, and then uptake by its own arginine transporters inside the cytosol of the pathogen. Thus, pathogen-encoded arginases can modulate iNOS activity irrespective of their spatial localization by modulating the cellular distribution of arginine.

Taken together, it is evident that several intracellular pathogens, such as H. pylori, Salmonella Typhimurium, and M. tuberculosis, survive nitrosative stress by inducing the counteractive enzyme of iNOS, arginase, inside the host macrophages. Arginase promotes Toxoplasma and HCV infection as well by providing protection from host-induced NO stress. However, it is interesting to note that in the case of intracellular survival of parasites like Trypanosoma, Leishmania, and Schistosoma, arginase offers a survival advantage mainly by a polyamine-dependent and NO-independent mechanism. Arginase comes into play even in the establishment of fungal infections. Candida escape from host macrophages is mediated by arginase induction. Thus, arginase induction is clearly a very widespread and essential response in pathogenic infections.