Nitric oxide

Nitric oxide (NO) plays a central role in the metabolic coordination of immune networks. This is, on the one hand, due to its involvement with the l-arginine and l-kynurenine pathways (Thomas et al. Reference Thomas, Mohr and Stocker1994; Noel et al. Reference Noel, Raes, Hassanzadeh Ghassabeh, De Baetselier and Beschin2004) that exert pathogen control and elimination. NO and NO-related reactive derivatives, on the other hand, are generated and used by different immune cells, such as macrophages, granulocytes and natural killer cells for direct killing, one of the most important active defence mechanisms of the innate immune system. Therefore, a variety of NO derivatives possessing microbicidal capacity are generated, including nitrogen dioxide (NO₂), nitrous acid (HNO₂) and peroxynitrite (ONOO⁻), which is formed by the reaction of NO and O₂⁻. DNA seems to be the main target of the reactive nitrogen intermediates whereby mutation and inhibition of repair mechanisms and protein synthesis are mainly affected (Bogdan et al. Reference Bogdan, Rollinghoff and Diefenbach2000). Interestingly, some pathogens possess detoxifying agents such as peroxiredoxins in Mycobacterium tuberculosis and Salmonella typhimurium, which degrades ONOO⁻ to nitrite (Bogdan, Reference Bogdan2001).

l-arginine pathway

One way to generate NO is by l-arginine degradation via inducible nitric oxide synthase (iNOS) within the l-arginine pathway (Fig. 1), whereby the pathway consists of three reactive sequences. The fate of l-arginine depends on its availability and the activity of the involved enzymes. Alternatively, l-arginine can be converted into agmatine by arginine decarboxylase (ADC), or metabolised to l-ornithine and urea via arginase. The two enzymes, ornithine amino transferase (OAT) and ornithine decarboxylase (ODC), then further convert l-ornithine to proline and polyamines such as putrescine, spermidine and spermine.

Fig. 1. Cross-regulation between l-arginine and l-kynurenine pathways within a macrophage. l-arginine is substrate for three different reactions that result in the production of either l-citrulline and NO by inducible nitric oxide synthase (iNOS), urea and l-ornithine, or agmatine, whereas l-tryptophan degrades stepwise to NAD⁺. While IFN-γ stimulates expression of both key enzymes, iNOS and indoleamine 2,3-dioxygenase (IDO), picolinic acid stimulates iNOS but inhibits IDO expression. Agmatine and l-hydroxy arginine both exert negative feedback on iNOS and arginase (arg), respectively and NO is an important inhibitor of IDO. The broken arrows indicate negative regulation and the solid arrows positive regulation. Key: 1, arginine decarboxylase; 2, agmatinase; 3, ornithine amino decarboxylase; 4, ornithine decarboxylase.

l-arginine itself derives from l-citrulline, which is the direct precursor and produced in the small intestinal enterocytes from proline, glutamate, and glutamine (Wu, Reference Wu1998; de Jonge et al. Reference De Jonge, Kwikkers, Te Velde, Van Deventer, Nolte, Mebius, Ruijter, Lamers and Lamers2002). Both metabolites, l-citrulline and l-arginine, were depleted during inflammatory processes (Bansal and Ochoa, Reference Bansal and Ochoa2003). l-arginine supplementation shows beneficial effects in wound healing after surgical intervention and other tissue-destructing processes (Barbul et al. Reference Barbul, Lazarou, Efron, Wasserkrug and Efron1990; Yan et al. Reference Yan, Peng, Huang, Zhao, Li and Wang2007).

Apart from its involvement in tissue regeneration, l-arginine exerts additional immune regulatory functions. It seems to be, for instance, essential for an effective switch from pro- to pre-B cells in the bone marrow (de Jonge et al. Reference De Jonge, Kwikkers, Te Velde, Van Deventer, Nolte, Mebius, Ruijter, Lamers and Lamers2002), and depletion of the metabolite leads to a reduced B cell presence in the lymph nodes, spleen and other secondary lymphoid organs. Also, the T cell lineage depends on l-arginine to establish appropriate growth and signalling processes, including T cell proliferation and up-regulation of the T cell receptor CD3ζ chain (Rodriguez et al. Reference Rodriguez, Zea, Desalvo, Culotta, Zabaleta, Quiceno, Ochoa and Ochoa2003). The central position of l-arginine in the differential activation of macrophages in response to infection is well described (Noel et al. Reference Noel, Raes, Hassanzadeh Ghassabeh, De Baetselier and Beschin2004).

l-arginine-mediated macrophage activation

The l-arginine pathway gives rise to two sub-forms of macrophages, i.e. classically (caMΦ) or alternatively activated macrophages (aaMΦ), whereby the differential activation process influences the resulting macrophage function. Th1 cytokines, particularly IFN-γ, induce NO and l-citrulline production from l-arginine in macrophages by up-regulating iNOS, a process that increases the effective clearance of intracellular pathogens by the subsequent secretion of pro-inflammatory cytokines such as IL-1 and IL-6 and the direct antimicrobial capacity of NO. Additionally the l-arginine pool is depleted from the generation of polyamines (putrescine, spermine and spermidine), via arginase and ODC (Fig. 1). Helminth parasites use these host-derived polyamines for growth, and Leishmania spp. and T. cruzi also seem to rely on the uptake of polyamines for intracellular processes of growth and replication in aaMΦ (Wanasen and Soong, Reference Wanasen and Soong2008).

In contrast to caMΦ, an alternative activation pathway of macrophages is initiated by a mixture of Th2 cytokines, including IL-4 and IL-13 (Gordon, Reference Gordon2003). NO production is typically decreased in aaMΦ due to a down-regulation of iNOS, and the cells can perform certain immunosuppressive actions, such as the secretion of anti-inflammatory cytokines or suppressing T cell proliferation, which prevent effective parasite elimination. The anti-inflammatory functions of aaMΦ, on the other hand, can limit the tissue damage induced by the cell-mediated defence mechanisms.

l-kynurenine pathway

The l-arginine pathway is strongly interconnected with the l-kynurenine pathway (Fig. 1), in which l-tryptophan is degraded stepwise to NAD⁺ (Sanni et al. Reference Sanni, Thomas, Tattam, Moore, Chaudhri, Stocker and Hunt1998; King and Thomas, Reference King and Thomas2007). Indoleamine 2,3-dioxygenase (IDO), the key enzyme, initiates the conversion of l-tryptophan to a set of neuroactive intermediates, including kynurenic acid, quinolinic acid, and 3-hydroxyanthranilic acid. IFN-γ is, among other pro-inflammatory mediators, the most important inducer of IDO synthesis but, at the same time, it stimulates the l-arginine pathway and NO generation. NO, the key component of the l-arginine pathway is, in turn, able to inactivate directly IDO by binding to the haeme complex of the enzyme (Thomas et al. Reference Thomas, Terentis, Cai, Takikawa, Levina, Lay, Freewan and Stocker2007); hence it acts as a direct inhibitor of the l-tryptophan degradation process. This cross-pathway regulation has been confirmed by higher activity of IDO in response to NOS inhibitors, but IDO-expression is suppressed in murine macrophages after NO-supplementation (Thomas et al. Reference Thomas, Mohr and Stocker1994; Alberati-Giani et al. Reference Alberati-Giani, Malherbe, Ricciardi-Castagnoli, Kohler, Denis-Donini and Cesura1997; Thomas and Stocker, Reference Thomas and Stocker1999). Picolinic acid, an intermediate of the l-kynurenine pathway, substantially augments the IFN-γ mediated activation of iNOS (Melillo et al. Reference Melillo, Cox, Biragyn, Sheffler and Varesio1994), but suppresses IDO activity (Alberati-Giani et al. Reference Alberati-Giani, Malherbe, Ricciardi-Castagnoli, Kohler, Denis-Donini and Cesura1997).

Activating the l-kynurenine pathway and subsequent IDO-mediated l-tryptophan depletion can, on one hand, cause growth inhibition in certain microorganisms that depend on exogenous tryptophan, such as Toxoplasma gondii, Chlamydia pneumonia and certain bacteria, including Mycobacterium spp. (Mellor and Munn, Reference Mellor and Munn2004). On the other hand, a low tryptophan level can also impair T cell proliferation and hence induce immune suppression (Munn et al. Reference Munn, Shafizadeh, Attwood, Bondarev, Pashine and Mellor1999). A possible cell cycle arrest as a response to tryptophan depletion has been proposed as the main mechanism for the inhibiting effect on T cell proliferation and the concomitant increased sensitivity to apoptosis.

The intermediates of the l-kynurenine pathway and their role within immune regulation have been described previously (Grohmann et al. Reference Grohmann, Fallarino and Puccetti2003; Moffett and Namboodiri, Reference Moffett and Namboodiri2003).

The eicosanoic cascade

Arachidonic acid is an n-6 polyunsaturated fatty acid (n-6 PUFA) that gives rise to an entire cascade of patent pro-inflammatory intra-cellular signalling molecules also called eicosanoids (Wong et al. Reference Wong, Chan, Lee, Jiang, Skrzypczak and Choy2000) (Fig. 2A). Several different sub-structures and functions have been identified, including prostaglandin E₂, which induces vasodilatation, fever and pain and increases the production of the pro-inflammatory cytokine IL-6 (Bagga et al. Reference Bagga, Wang, Farias-Eisner, Glaspy and Reddy2003; Calder, Reference Calder2008, Reference Calder2009). Also, the 4-series leukotrienes exert pro-inflammatory functions including chemotactic mediation of leukocytes, induction of release of lysosomal enzymes and reactive oxygen species by granulocytes, and increased production of TNF, IL-1 and IL-6 (Calder, Reference Calder2009).

Fig. 2. (A) Link between the choline pathway and arachidonic acid cascade. The choline-containing intermediates can be metabolised to phosphorylcholine, which represents an important precursor for n-6 PUFA arachidonic acid and the subsequent production of pro-inflammatory mediators. (B) Alternative cascade of mediators derived from n-3 PUFAs eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), which have certain anti-inflammatory functions and which can be competitively built into leukocytes. Adapted from Zeisel et al. (Reference Zeisel, Mar, Howe and Holden2003); Li and Vance (Reference Li and Vance2008) and Calder (Reference Calder2008). Key: CDP-Cho, CDP-choline; Cho, choline; GPC, glycerophosphocholine; HE, hydroxyeicosatetraenoic acid; Hp, hydroperoxyeicosatetraenoic acid; LT, leukotriene; PC, phosphatidylcholine; PG, prostaglandin; PhCho, phosphocholine; PL, phospholipase; SM, sphingomyelin; TX, thromboxane.

Eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), two other important inflammation-related fatty acids, are n-3 PUFA-precursors for a panel of eicosanoic mediators (Fig. 2B), which are less potent compared to their n-6 PUFA-derived analogues. LTB₅, for instance, which is EPA-derived, is significantly less effective in inducing neutrophil chemotaxis than LTB₄, which derives from n-6 PUFA arachidonic acid. The recently discovered E- and D-series resolvins, which are EPA- and DHA-derived respectively, have several anti-inflammatory functions, including inhibition of migration and chemotaxis of leucocytes (Serhan et al. Reference Serhan, Clish, Brannon, Colgan, Chiang and Gronert2000, Reference Serhan, Hong, Gronert, Colgan, Devchand, Mirick and Moussignac2002, Reference Shi, Meininger, Haynes, Hatakeyama and Wu2004; Seki et al. Reference Seki, Tani and Arita2009).

Some n-6 PUFA eicosanoids can act in an anti-inflammatory manner (Levy et al. Reference Levy, Clish, Schmidt, Gronert and Serhan2001; Vachier et al. Reference Vachier, Chanez, Bonnans, Godard, Bousquet and Chavis2002) and vice versa, PGE₂ and PGE₃ showed similarly effective inhibition of TNF-α and IL-1β (Dooper et al. Reference Dooper, Wassink, M'rabet and Graus2002; Miles et al. Reference Miles, Allen and Calder2002). Orally administering EPA and DHA in fish oil has shown beneficial effects in many chronic inflammatory conditions such as colitis (Calder, Reference Calder2008), and the overall counter-inflammatory effect of a higher n-3/n-6 PUFA-ratio has been proven in various other nutritional studies (Eritsland et al. Reference Eritsland, Arnesen, Gronseth, Fjeld and Abdelnoor1996; Troseid et al. Reference Troseid, Arnesen, Hjerkinn and Seljeflot2009). The interaction at the mediator level is not clear yet, but human (Caughey et al. Reference Caughey, Mantzioris, Gibson, Cleland and James1996; Belluzzi et al. Reference Belluzzi, Boschi, Brignola, Munarini, Cariani and Miglio2000; Cleland and James, Reference Cleland and James2000; Cleland et al. Reference Cleland, Caughey, James and Proudman2006; Troseid et al. Reference Troseid, Arnesen, Hjerkinn and Seljeflot2009) and animal studies (Serhan et al. Reference Serhan, Clish, Brannon, Colgan, Chiang and Gronert2000; Arita et al. Reference Arita, Yoshida, Hong, Tjonahen, Glickman, Petasis, Blumberg and Serhan2005; Hudert et al. Reference Hudert, Weylandt, Lu, Wang, Hong, Dignass, Serhan and Kang2006) have revealed that the beneficial immunosuppressive effects of such DHA and EPA supplements resulted mainly as a response to the competitive incorporation into inflammatory active cells, which minimises inflammation.

Amino acids

Most of the amino acids described previously relate to innate and adaptive immunity, which both depend on the availability of a constant pool of amino acids to synthesis the micro- and macromolecules necessary for an appropriate immune defence, such as immunoglobulins, acute phase proteins, reactive nitrogen species, histamine, major histocompatibility complex (MHC) and T cell receptors (Li et al. Reference Li, Yin, Li, Kim and Wu2007).

While some amino acids represent direct or indirect energy substrates for immune cells, such as glutamine for cells of the immune system in general (Wu et al. Reference Wu, Field and Marliss1991), or alanine, which is converted into glucose in the liver and acts as an important substrate for leucocytes (Newsholme and Newsholme, Reference Newsholme and Newsholme1989), others have regulatory functions within immune-related pathways. Lysine, for instance, acts as a direct antagonist to arginine, as both amino acids compete for the same transporter for cellular uptake, and excessive lysine minimises viral potency during a Herpes simplex infection, which might be due to restricted availability of polyamines for growth and replication of the pathogen (Griffith et al. Reference Griffith, Norins and Kagan1978). Moreover, phenylalanine and taurine can exert an indirect function by modifying the l-arginine pathway. The former seems to play an important role in synthesising an essential NOS-cofactor, tetrahydrobiopterin (Shi et al. Reference Shi, Meininger, Haynes, Hatakeyama and Wu2004; Li et al. Reference Li, Yin, Li, Kim and Wu2007), whereas the latter inhibits the expression of iNOS in macrophages (Wu and Meininger, Reference Wu and Meininger2002). Additionally, taurine can act as an effective antioxidant (Fang et al. Reference Fang, Yang and Wu2002). Its chlorinated form (i.e. taurine chloramine), has potent microbicidal (Schuller-Levis and Park, Reference Schuller-Levis and Park2004) and anti-inflammatory capacity, suppressing pro-inflammatory cytokine and prostaglandin E₂ expression (Weiss et al. Reference Weiss, Klein, Slivka and Wei1982; Chorazy et al. Reference Chorazy, Kontny, Marcinkiewicz and Maslinski2002).

The sulphur-containing proteogenic amino acids take a special position with regard to immune modulation, as they are essential for the synthesis of important immune regulatory molecules, such as glutathione (GSH) and the acute phase proteins which contain a high proportion of cysteine and methionine (Grimble and Grimble, Reference Grimble and Grimble1998; Li et al. Reference Li, Yin, Li, Kim and Wu2007). Whereas the main role of the acute phase proteins is the control of tissue inflammation, GSH acts as scavenger of reactive oxygen intermediates. Peterson and colleagues showed that it influences Th subgroup differentiation, whereby depleted GSH leads to suppression of Th1 cytokines, a subsequent shift towards a Th2 response and, moreover, to antibody production in non-responding mice (Peterson et al. Reference Peterson, Herzenberg, Vasquez and Waltenbaugh1998). Sulphur-containing amino acids are in higher demand during states of disease or injury, due to the production of acute phase proteins, and a competitive situation arises over the fate of cysteine. Studies on sulphur amino acid intake in rats by Hunter and colleagues showed that low abundance leads to relatively higher protein production, which might in turn affect the physiological control of oxidative damage by the depleted GSH levels (Hunter and Grimble, Reference Hunter and Grimble1994, Reference Hunter and Grimble1997).

Nucleotides

The role of nucleotides in the immune system has been assessed from different angles. For instance, nutritional supplementation showed relevant positive effects on mammalian health. Although de novo synthesis of nucleotides and the salvage pathway seem to produce sufficient quantities of nucleotides for basic physiological needs, certain pathological states, such as infection, surgery or growth processes, induce a higher demand which can be covered by externally provided nucleotides. Generally, nucleotide-enhanced diets show beneficial effects in various states of challenged immunity. Kulkarni and colleagues found that RNA and uracil supplementation increased T cell proliferation, and RNA in the diet has been directly linked to relatively higher levels of IL-2. Further studies showed the immunosuppressive consequences of nucleotide-free diets, resulting in increased susceptibility of cardio allografts in mice, and also indicated a role for nucleotides in T cell differentiation (Kulkarni et al. Reference Kulkarni, Rudolph and Van Buren1994). The main reason for impaired T cell differentiation and proliferation induced by depletion of purines and pyrimidines may be cell cycle arrest, impairing cell transit from G to S phase, a process which is mediated by IL-2 and IL-3 (Bender et al. Reference Bender, Van Epps and Stewart1986; Rudolph et al. Reference Rudolph, Kulkarni, Fanslow, Pizzini, Kumar and Van Buren1990; Jyonouchi, Reference Jyonouchi1994). Studies on infection-rodent models confirmed the positive influence of nucleotides on pathogen defence. When mice on a nucleotide-free diet were compared to mice fed with RNA and uracil supplements, the latter showed a markedly improved defence to Candida albicans (Fanslow et al. Reference Fanslow, Kulkarni, Van Buren and Rudolph1988) and Staphylococcus aureus (Kulkarni et al. Reference Kulkarni, Fanslow, Drath, Rudolph and Van Buren1986a, Reference Kulkarni, Fanslow, Rudolph and Van Burenb). Similarly, Carver and colleagues assessed spleen cells from mice after they were fed on a nucleotide-free diet and a five-nucleotide formula and found a generally lower activity of macrophages and natural killer cells in the tissue of the nucleotide-deprived animals (Carver et al. Reference Carver, Cox and Barness1990). Navarro and colleagues performed a more detailed assessment of supplemented nucleotides, whereby Balb/c mice were fed on single-nucleotide diets. Whereas AMP, GMP and UMP had a positive effect on the IgG response, IMP and CMP did not introduce any measurable changes (Navarro et al. Reference Navarro, Ruiz-Bravo, Jimenez-Valera and Gil1996).

The humoral branch of the immune system seems largely unaffected by a restriction of purines and pyrimidines, with the exception of T cell-dependent antigen presentation which resulted in relatively increased production of antibodies in vitro (Jyonouchi, Reference Jyonouchi1994).

Immunological relevance of identified infection biomarkers

A total of eight parasite-rodent models have been assessed so far via ¹H NMR-based metabolic profiling, including trematode infections, i.e. Schistosoma mansoni-mouse (Wang et al. Reference Wang, Holmes, Nicholson, Cloarec, Chollet, Tanner, Singer and Utzinger2004; Li, Reference Li, Holmes, Saric, Keiser, Dirnhofer, Utzinger and Wang2009), S. japonicum-hamster (Wang et al. Reference Wang, Utzinger, Xiao, Xue, Nicholson, Tanner, Singer and Holmes2006), Echinostoma caproni-mouse (Saric et al. Reference Saric, Li, Wang, Keiser, Bundy, Holmes and Utzinger2008b, Reference Saric, Li, Wang, Keiser, Veselkov, Dirnhofer, Yap, Nicholson, Holmes and Utzinger2009), and Fasciola hepatica-rat (unpublished data), nematode worms Trichinella spiralis-mouse (Martin et al. Reference Martin, Verdu, Wang, Dumas, Yap, Cloarec, Bergonzelli, Corthesy-Theulaz, Kochhar, Holmes, Lindon, Collins and Nicholson2006) and Necator americanus-hamster (Wang et al. Reference Wang, Xiao, Xue, Singer, Utzinger and Holmes2009), and two protozoan models, namely Plasmodium berghei-mouse (Li et al. Reference Li, Wang, Saric, Nicholson, Dirnhofer, Singer, Tanner, Wittlin, Holmes and Utzinger2008), and Trypanosoma brucei brucei-mouse (Wang et al. Reference Wang, Utzinger, Saric, Li, Burckhardt, Dirnhofer, Nicholson, Singer, Brun and Holmes2008).

Although the metabolic signature of each infection is parasite-specific, particular groups of metabolites and macromolecules are a recurring theme across the models, including biomarkers of energy metabolism (e.g. creatine and tricarboxylic acid cycle intermediates), gut microbial co-metabolites (hippurate, p-cresol glucuronide, trimethylamine, etc.), osmolytes (e.g. betaine, myo-inositol, taurine), components of lipid metabolism (choline, phosphocholine, glycerophosphocholine and lipoproteins) and signals from N- and O-acetyl glycoprotein fragments, which are reflective of inflammatory processes.

Choline, phosphocholine (PC) and glycerophosphocholine (GPC) are frequently observed in plasma and tissues by ¹H NMR spectroscopy, and altered relative concentrations can be indicative of a change in lipid degradation processes. The intestinal fluke E. caproni, for instance, induced a depletion of GPC and choline, with a subsequent increase of lipoprotein fractions in plasma, as a consequence of an increased degradation of biological membranes (Saric et al. Reference Saric, Li, Wang, Keiser, Bundy, Holmes and Utzinger2008b). On the other hand, muscular hypertrophy has been described as a consequence of a T. spiralis-infection where choline and phosphocholine have been found at relatively increased levels, whereas glycerophosphocholine (GPS) and the lipoprotein fractions in plasma were depleted.

Choline can be generated de novo via the CDP-pathway or taken up with the diet. After intestinal translocation, choline can either be oxidised to betaine in an irreversible reaction and act as a methyl-donor or osmolite in liver and kidney (Zeisel et al. Reference Zeisel, Mar, Howe and Holden2003) (Fig. 2A), or it can function as a second messenger in the central nervous system via acetylcholine. Alternatively, choline phosphorylates to GPC or PC, the two major storage forms in the cytosol. PC then further converts to CDP-choline which forms, together with diaglycerol, phosphatidylcholine, the main phosholipid component of mammalian cell-membranes (Li and Vance, Reference Li and Vance2008). Apart from its membrane-related functions of general maintenance and signalling, phosphatidylcholine represents a major source for the generation of arachidonic acid, which is one of three immunologically relevant main fatty acids described in depth so far (Fig. 2AB).

Further disruptions amongst the choline-containing intermediates are found in the T. b. brucei-mouse model. This protozoan parasite breaks down and scavenges host lipids for integration into the parasite coat. Increased free choline and a relative decrease in phosphocholine and lipoprotein fractions in plasma have been described (Wang et al. Reference Wang, Utzinger, Saric, Li, Burckhardt, Dirnhofer, Nicholson, Singer, Brun and Holmes2008). In the same T. b. brucei model, acetylated plasma glycoproteins increased, which has subsequently been shown to be a common response to several parasite infections, including T. spiralis (Martin et al. Reference Martin, Verdu, Wang, Dumas, Yap, Cloarec, Bergonzelli, Corthesy-Theulaz, Kochhar, Holmes, Lindon, Collins and Nicholson2006) and F. hepatica (unpublished data). Fig. 3A and B demonstrate the relatively lower levels of the N and O-acetyl glycoprotein fragments in the ¹H NMR plasma spectra obtained from an uninfected control rat (A) compared to an age- and gender-matched rat 22 days after infection with the liver fluke F. hepatic (B). Fig. 4 further illustrates how infection increases the relative levels of acetylated glycoproteins, whereby the colour scale indicates the strength of the correlation between the metabolites and the classes (i.e. infected and uninfected), reaching from no correlation (blue) to high correlation (red). The orthogonal projection to latent structure discriminant analysis (O-PLS-DA) applied, is the multivariate method of choice for biomarker recovery. By relating a descriptor matrix X (e.g. spectral information) to a response matrix Y (e.g., class information) class-related separation is maximised which allows improved interpretation of the class-differentiating variables (Cloarec et al. Reference Cloarec, Dumas, Trygg, Craig, Barton, Lindon, Nicholson and Holmes2005).

Fig. 3. Aliphatic region of a ¹H NMR plasma spectrum obtained from a non-infected control female Wistar rat (A) and a gender and age-matched rat, infected with F. hepatica. Twenty-two days post-infection, (B) showing the relative differences in the N and O-acetyl glycoprotein resonance intensities. Key: 3HB, D-3-hydroxybutyrate; NAG1, N-acetyl glycoprotein fragment 1; NAG2, N-acetyl glycoprotein fragment 2; OAG, O-acetyl glycoprotein fragment.

Fig. 4. O-PLS-DA analysis between F. hepatica-infected rats and uninfected control animals at Day 22 post-infection (n=10), when differences between glycoprotein levels were found to be most significant. Upwards-pointing peaks illustrate positive correlation with the infection, and vice versa. The elevated levels in the infected animal reflect the higher production of acute phase proteins as a response to inflammation. Key: NAG1, N-acetyl glycoprotein fragment 1; NAG2, N-acetyl glycoprotein fragment 2; OAG, O-acetyl glycoprotein fragment.

Acetyl glycoproteins include acute phase proteins (e.g. α1 acid glycoprotein, haptoglobin, trasferrin, α-1 antitrypsin) whose plasma levels can change substantially with tissue damage, and which are secreted by hepatocytes to exert their mostly anti-inflammatory properties (Goldsby et al. Reference Goldsby, Kindt and Osborne2001). The ¹H NMR structure of the glycoprotein components has been described and assessed in detail in human and rat plasma by Bell and colleagues (Bell et al. Reference Bell, Brown, Nicholson and Sadler1987a) and Grootveld and colleagues (Grootveld et al. Reference Grootveld, Claxson, Chander, Haycock, Blake and Hawkes1993). α1 acid glycoprotein has been found to have the largest contribution to the main observed glycoprotein-related signals: three broad singlets at 2·04, 2·08, and 2·14 in the ¹H NMR spectra of rat plasma.

A recent co-assessment of ¹H NMR spectral data and the relative cytokine levels in a Plasmodium berghei infection in a murine host showed a direct correlation between IFN-γ and several plasma metabolites. Positive correlation was found with lactate and creatine applying analysis, whereas α and β glucose were found to be anti-correlated with relative concentrations of IFN-γ (Fig. 5).

Fig. 5. O-PLS correlation plot of relative IFN-γ levels and individually matched ¹H NMR spectral plasma data obtained from mice pre-infection, and animals infected with P. berghei, at days 2 and 4 post infection (at each time-point n=5). The colour scale indicates the significance of the correlation, i.e., red peaks illustrate highest correlation between cytokine and metabolite levels. Upwards-pointing peaks show positive correlation, whereas downwards-pointing peaks indicate anti-correlation.

Metabolic coverage of ¹H NMR spectroscopy

A large variety of physiological compartments has been profiled by ¹H NMR. Urine has been most extensively characterized across species (Bollard et al. Reference Bollard, Stanley, Lindon, Nicholson and Holmes2005), disease conditions (Lenz et al. Reference Lenz, Bright, Knight, Westwood, Davies, Major and Wilson2005; Salek et al. Reference Salek, Maguire, Bentley, Rubtsov, Hough, Cheeseman, Nunez, Sweatman, Haselden, Cox, Connor and Griffin2007; Williams et al. Reference Williams, Cox, Walker, North, Patel, Marshall, Jewell, Ghosh, Thomas, Teare, Jakobovits, Zeki, Welsh, Taylor-Robinson and Orchard2009), nutritional effects (Wang et al. Reference Wang, Tang, Nicholson, Hylands, Sampson and Holmes2005b, Reference Wang, Lawler, Larson, Ramadan, Kochhar, Holmes and Nicholson2007), and geographical exposure (Holmes et al. Reference Holmes, Loo, Stamler, Bictash, Yap, Chan, Ebbels, De Iorio, Brown, Veselkov, Daviglus, Kesteloot, Ueshima, Zhao, Nicholson and Elliott2008), information which is strongly reflected in the metabolic composition of the urine. The majority of metabolites found in urine are micromolecules and reflect energy metabolism (e.g. creatine, succinate, lactate, citrate, etc.), gastrointestinal function (e.g. hippurate, methylamines, and p-cresol glucuronide), amino and ketoacids, such as taurine, 2-ketoisocaproate, and detoxification products of hepatic amino acid degradation (i.e. urea) and nucleotide decomposition (i.e. uric acid).

Intense assessment of plasma via ¹H NMR spectroscopy has shown, in contrast to urine, a higher intra-individual stability due to the physiological importance of maintaining blood homeostasis. Plasma metabolic profiles contain, in contrast to urine, signals from a variety of macromolecular structures that are transported in the blood, such as glycoproteins and mobile lipid fractions. The components of lipid metabolism that provide a major contribution to the ¹H NMR composition of plasma include further choline, PC and GPC but amino acid and ketoacid resonances are also visible (Bell et al. Reference Bell, Brown, Nicholson and Sadler1987a, Reference Bell, Sadler, Macleod, Turner and La Villeb; Foxall et al. Reference Foxall, Spraul, Farrant, Lindon, Neild and Nicholson1993; Nicholson et al. Reference Nicholson, Foxall, Spraul, Farrant and Lindon1995).

Faecal extracts have only recently been assessed (Saric et al. Reference Saric, Wang, Li, Coen, Utzinger, Marchesi, Keiser, Veselkov, Lindon, Nicholson and Holmes2008a, Reference Saric, Li, Wang, Keiser, Bundy, Holmes and Utzingerb), and showed the highest degree of intra- and inter-individual variation. This may, at least partially, be explained by the complex composition of the gut microbiota, which contributes substantially to the faecal mass. The major components derived from gut microbial co-metabolism are bile acids and the fermentation products of complex non-digestible carbohydrates, such as cellulose and starches. The short-chain fatty acids butyrate, acetate, and propionate are typically found in the ¹H NMR spectra of faecal water extracts (Cummings, Reference Cummings1981; Guarner and Malagelada, Reference Guarner and Malagelada2003; Wong et al. Reference Wong, de Souza, Kendall, Emam and Jenkins2006).

The metabolic profiling of tissue compartments has become a standard method to assess systemic changes. It is either performed by magic angle spinning (MAS), a non-destructive semi-solid method, or conventional ¹H NMR on tissue extracts, comparing liver (Bollard et al. Reference Bollard, Garrod, Holmes, Lindon, Humpfer, Spraul and Nicholson2000, Reference Bollard, Contel, Ebbels, Smith, Beckonert, Cantor, Lehman-Mckeeman, Holmes, Lindon, Nicholson and Keun2009), spleen (Saric et al. Reference Saric, Li, Wang, Keiser, Veselkov, Dirnhofer, Yap, Nicholson, Holmes and Utzinger2009), kidney (Garrod et al. Reference Garrod, Humpfer, Spraul, Connor, Polley, Connelly, Lindon, Nicholson and Holmes1999), brain (Tsang et al. Reference Tsang, Griffin, Haselden, Fish and Holmes2005), heart (Bollard et al. Reference Bollard, Murray, Clarke, Nicholson and Griffin2003) and different intestinal compartments (i.e. ileum, jejunum and colon) (Wang et al. Reference Wang, Tang, Holmes, Lindon, Turini, Sprenger, Bergonzelli, Fay, Kochhar and Nicholson2005a; Marchesi et al. Reference Marchesi, Holmes, Khan, Kochhar, Scanlan, Shanahan, Wilson and Wang2007). The main groups of metabolites which can be visualised by ¹H NMR amongst the majority of the compartments are amino acids and components of the lipid metabolism, including choline-derivatives, nucleotides, muscle degradation products (e.g. creatine, creatinine) and sugars (e.g. glucose). Brain metabolite composition is distinguished from other tissues by the presence of γ-aminobutyric acid (GABA) and N-acetyl aspartate (Holmes et al. Reference Holmes, Tsang and Tabrizi2006; Tsang et al. Reference Tsang, Huang, Holmes and Bahn2006, Reference Tsang, Haselden and Holmes2009), whereas the kidney contains the largest variety of osmolytically active molecules, including sorbitol, taurine, myo-inositol, GPC and betaine (Garrod et al. Reference Garrod, Humpfer, Spraul, Connor, Polley, Connelly, Lindon, Nicholson and Holmes1999; Waters et al. Reference Waters, Garrod, Farrant, Haselden, Connor, Connelly, Lindon, Holmes and Nicholson2000). Table 1 shows metabolites which are commonly identified in rodent tissues and biofluids using ¹H NMR spectroscopy, and which have shown cross-regulatory interaction with the immune system.

Table 1 Rodent metabolites commonly recovered by ¹H NMR spectroscopy with established connection to immune regulatory networks

The chemical moieties represent the proton groups, whereby each group corresponds to a ¹H NMR signal at a determined chemical shift region. The direct electron environment of each proton group induces different splitting patterns or multiplicities: s, singlet; d, doublet; dd, doublet of doublets; dt, doublet of triplets; t, triplet; m, multiplet; q, quadruplet.