Dietary nucleotides and human immune cells. II. Modulation of PBMC growth and cytokine secretion

Introduction 

The immune system is dependent on adequate nutrition for different vital processes such as energy production, protein synthesis, proliferation, and other specific metabolic pathways. Immune cells require several nutrients to increase their growth rate and function 1, 2. Dietary nucleic acids and nucleotides influence optimal growth and cell function 3, 4. In rodents, nucleotide-free diets decrease cellular and humoral immune responses 5, 6, 7 and resistance to bacterial and fungal pathogens 8, 9, 10, 11. Further, dietary nucleotides can modulate type 1 and type 2 T-helper (Th) cell responses by promoting a shift in the Th1/Th2 balance toward Th1-dominant immunity 12, 13, 14. In fish, dietary nucleotides seem to enhance non-specific immune responses [15] and vaccine efficacy [16]. In humans, dietary nucleotides increase immune responses to Haemophilia influenzae virus b and diphtheria [17], and polynucleotides significantly increase in vitro antibody production in response to T-cell–dependent antigen [9].

Nucleotides, in addition to being involved in well-known energy-dependent processes, exhibit functions as pluripotent signaling molecules that elicit direct effects on cell functions by purine and pyrimidine receptors but probably also other receptors to induce gene expression. Because dietary nucleotides may exert a wide range of effects depending on nucleotide concentration, structure, and possible receptor affinity, reflected by their target cells, it is necessary to evaluate the effect of various nucleotides on different cell types.

In this study we investigated DNA isolated from fish soft roe, RNA from bakers yeast, 2′deoxyadenosine 5′-monophosphate sodium (dAMP-Na+), 2′deoxycytidine 5′-monophosphate sodium (dCMP-Na+), 2′deoxyguanosine 5′monophosphate sodium (dGMP-Na+), 2′deoxyuridine 5′-monophosphate disodium (dUMP-Na+), and Thymidine 5′monophosphate sodium (TMP-Na+) to determine their capacity to stimulate immune cell growth and function and used influenza virus antigen as an immunostimulant in vitro.

Results 

Growth responses 

DNA-Na+ from fish soft roe, RNA from bakers yeast, or mononucleotides did not trigger growth of PBMCs from healthy subjects (n = 10, P > 0.05; Fig. 1). Influenza virus antigen (0.33 μg/mL) induced significant immune cell proliferation compared with the control casein (P = 0.0367). In contrast, 1.3 μg/mL of virus antigen produced a three-fold lower mean response (Fig. 1) and was not significantly different from the mean of the control group. Interestingly, when cultured with virus antigen, DNA-Na+, and RNA (Fig. 1), dAMP, dCMP, and dUMP (Fig. 2) significantly (P < 0.05) increased virus antigen-induced PBMC proliferation responses. Nucleotide concentrations of 50 μg/mL in particular, but also 1 mg/mL (except 1 mg/mL of dUMP and DNA), significantly triggered virus-induced PBMC growth. The highest proliferation responses, compared to virus antigen alone (1897 mean dpm) were observed in PBMC cultures in which RNA, dAMP, and dCMP were added with virus antigen (3970-, 4326-, 4086- mean dpm respectively, Fig. 1, Fig. 2). In contrast, dGMP and TMP (at all concentrations tested) showed no significant effect on influenza virus-antigen induced growth responses (Fig. 3).

View Large Image Download to PowerPoint

Fig. 1. DNA-Na+ from fish soft roe (50 μg/mL) and RNA from bakers yeast (50 μg/mL or 1 mg/mL) significantly increases Inf-induced immune cell growth. Inf (0.33 μg/mL) induced significant proliferation of peripheral blood mononuclear cells compared with control casein (P = 0.0367), Triplicate measurements of peripheral blood mononuclear cells from 10 healthy subjects were investigated. Results are presented as mean dpm ± standard error of the mean. Differences are significant at P < 0.05. dpm, determinations per minute; Inf, influenza virus antigen.

View Large Image Download to PowerPoint

Fig. 2. Nucleotides dAMP, dCMP (50 μg/mL, 1 mg/mL), and dUMP (50 μg/mL) significantly increase Inf-induced immune cell growth. Triplicate measurements of peripheral blood mononuclear cells from 10 healthy subjects were investigated. Growth responses are presented as mean dpm ± standard error of the mean. Differences are significant at P < 0.05. dAMP, deoxyadenosine monophosphate; dCMP, deoxycytidine monophosphate; dpm, determinations per minute; dUMP, deoxyuridine monophosphate; Inf, influenza virus antigen.

View Large Image Download to PowerPoint

Fig. 3. Nucleotides dGMP and TMP, when cocultured with Inf, inhibit virus-induced growth of peripheral blood mononuclear cells. Responses are presented as dpm mean ± standard error of the mean (n = 10, P < 0.05). dGMP, deoxyguanosine monophosphate; dpm, determinations per minute; Inf, influenza virus antigen; TMP, thymidine monophosphate.

DNA/RNA and mononucleotide effects on PBMC cytokine secretion 

Healthy (Table 1) unstimulated (control) and casein-Na+ (control) stimulated PBMCs produced TNF-α and IFN-γ in comparable amounts. Although these did not act like growth inducers, all nucleotide derivatives tested affected human PBMC cytokine secretion when compared with control and casein-Na+–stimulated cultures. All nucleotide derivatives (P < 0.05), but especially RNA and TMP, supported PBMC TNF-α secretion.

Table 1. DNA-, RNA-, and mononucleotide-induced secretion of TNF-α, IL-10, and INF-γ by peripheral blood mononuclear cells from 10 healthy subjects⁎

	Culture	TNF-α (pg/mL)	IL-10 (pg/mL)	INF-γ (pg/mL)
	Control	171±45	0	109±31
	DNA 50 μg	309±58	0	88±40
	DNA 1 mg	315±42	0	0
	DNA 5 mg	254±64	0	0
	RNA 50 μg	773±293	299±21	340±73
	RNA 1 mg	805±152	598±64	438±31
	RNA 5 mg	1172±98	653±32	348±21
	dAMP 50 μg	226±12	0	1242±455
	dAMP 1 mg	177±54	0	320±67
	dCMP 50 μg	345±32	153±55	153±22
	dCMP 1 mg	586±55	212±33	0
	dGMP 50 μg	141±34	92±34	190±62
	dGMP 1 mg	472±81	161±12	56±2
	dUMP 50 μg	245±18	0	240±54
	dUMP 1 mg	222±32	0	69±15
	TMP 50 μg	760±68	349±23	0
	TMP 50 μg	871±97	289±41	0
	Casein 1 mg	128±52	0	200±121

dAMP, deoxyadenosine monophosphate; dCMP, deoxycytidine monophosphate; dGMP, deoxyguanosine monophosphate; dUMP, deoxyuridine monophosphate; IFN-γ, interferon-γ; IL-10, interleukin-10; TMP, thymidine monophosphate; TNF-α, tumor necrosis factor-α

⁎ Values are mean ± standard error of the mean.

In addition, RNA, dCMP, dGMP, and TMP induced IL-10 secretion, a cytokine that was not secreted by control PBMCs or those stimulated by DNA, dAMP, or dUMP. Only RNA and dAMP (particularly 50 μg/mL) significantly (P < 0.05) triggered IFN-γ secretion above control levels.

DNA/RNA and mononucleotide effects on influenza virus antigen-induced PBMC cytokine secretion 

Influenza (Table 2) virus antigen (0.33 μg/mL) induced no significant mean PBMC cytokine secretion above levels of control and casein-stimulated PBMCs. Some individual virus antigen-stimulated PBMCs showed significant differences in TNF-α secretion from control PBMCs, but this significance disappeared after calculating mean PBMC responses. RNA and dAMP cocultured with virus antigen significantly (P < 0.05) increased PBMC secretions of IFN-γ, IL-10, and TNF-α. Interestingly, compared with control, DNA increased (P < 0.05) virus antigen-induced cell secretion of IFN-γ only, whereas dUMP significantly increased secretion of IL-10 only. IL-10 was not secreted by control cultures or PBMCs stimulated by virus antigen, dAMP, dUMP, or DNA (with or without virus antigen). Nucleotide dGMP completely inhibited virus-triggered cell secretion of IFN-γ (P < 0.05), whereas TMP did not change the virus-induced secretion pattern of the measured cytokines compared with secretion of TMP alone and virus antigen alone.

Table 2. DNA, RNA, and mononucleotides, in culture with Infl (0.33, μg/mL), modulate secretion of TNF-α, INF-γ, and IL-10 by peripheral blood mononuclear cells from 10 healthy subjects⁎

	Cultures	TNF-α (pg/mL)	IL-10 (pg/mL)	INF-γ (pg/mL)
	Control	171±45	0	109±31
	Casein	128±52		200±121
	Infl	358±171	0	159±56
	DNA 50 μg + Infl	440±220	0	309±78
	DNA 1 mg + Infl	304±156	0	412±123
	DNA 5 mg + Infl	291±68	0	298±145
	RNA 50 μg + Infl	846±245	493±38	593±245
	RNA 1 mg + Infl	1038±420	658±254	553±156
	RNA 5 mg + Infl	1493±376	664±159	839±298
	dAMP 50 μg + Infl	393±125	130±45	592±222
	dAMP 1 mg + Infl	320±56	180±67	1534±520
	dCMP 50 μg + Infl	430±256	142±98	231±134
	dCMP 1 mg + Infl	985±199	230±124	226±24
	dGMP 50 μg + Infl	227±79	119±34	0
	dGMP 1 mg + Infl	412±165	146±54	0
	dUMP 50 μg + Infl	318±155	227±103	172±76
	dUMP 1 mg + Infl	397±267	165±76	178±45
	TMP 50 μg + Infl	675±398	231±34	173±76
	TMP 1 mg + Infl	1122±412	310±68	235±34

dAMP, deoxyadenosine monophosphate; dCMP, deoxycytidine monophosphate; dGMP, deoxyguanosine monophosphate; dUMP, deoxyuridine monophosphate; IFN-γ, interferon-γ; IL-10, interleukin-10; TMP, thymidine monophosphate; TNF-α, tumor necrosis factor-α

⁎ Values are mean ± standard error of the mean of triplicate measurements.

Discussion 

In this study we found that DNA from fish soft roe, RNA from bakers yeast, and mononucleotides exerts direct action on healthy human immune cells in vitro.

Using influenza virus antigen as a specific antigen, DNA, RNA, dAMP, dCMP, and dUMP increased antigen-induced PBMC proliferation responses. This corresponded with an increase in PBMC secretion of selected cytokines. RNA and dAMP acted as general cytokine inducers by increasing secretions of IFN-γ, TNF-α, and IL-10, whereas DNA increased PBMC secretion of IFN-γ only and dUMP stimulated secretion of IL-10 only. IL-10 was not secreted by relevant control cultures, by dAMP- or dUMP-stimulated PBMCs, or by any of the cultures in which DNA was added with or without virus antigen. Nucleotides dGMP and TMP inhibited influenza virus-induced PBMC growth and dGMP completely blocked virus-induced IFN-γ secretion. TMP did not change cytokine secretion in any culture.

Our results indicated that coculturing virus antigen and nucleotides probably influence and activate distinct cell fractions in PBMCs and involve antigen-presenting cells and T cells. It is tempting to speculate that specific nucleotides may act as immunomodulators or costimulatory molecules by exerting their effect, perhaps mostly, on specific antigen-activated cells among PBMCs.

Jyonouchi et al. [18] reported that yeast RNA enhances immunoglobulin (Ig) M and IgG production to T-cell–dependent stimuli but not to T-cell–independent stimuli. Decomposing RNA to pyrimidine bases decreased the enhancing action of RNA on Ig production. It has been proposed that polynucleotides modulate the process of antigen- and allergen-mediated Th-cell activation, thus enhancing IFN-γ secretion 12, 13, 14, 19, 20 In our investigation, the various DNA/RNA/nucleotide derivatives showed similar and different effects on immune cell growth and cytokine secretion, a behavior that also was observed when investigating nucleotide effects on intestinal cell lines [21]. This indicates that configuration and type of nucleotide bases influence the resulting effects.

Immunostimulatory and “non-immunostimulatory” CpG synthetic oligonucleotides showing adjuvant effects were described by McCluskie et al. [22]. Further, prepriming and antigen covaccination with immunostimulatory DNA sequences promoted a Th1 response to antigen [23]. Therefore, the supporting ability of nucleotide derivatives on growth and cytokine secretion in our virus-stimulated immune cell cultures raises the possibility that specific sequences of DNA/RNA or specific nucleotides may act as potent Th1-directing adjuvants and converters of Th2 immune responses into Th1 responses. This has to be investigated further by stimulating human allergic PBMCs and different PBMC cell fractions with specific allergen and DNA/RNA/nucleotides and assessing cell proliferation and cytokine profiles. Further, influenza virus antigen-induced PBMC responses (Th1 like) should be compared with PBMC responses after measles virus infection (Th2 like). Thus, nucleotide sequences with optimal adjuvant activity could be detected.

In our investigations, RNA from bakers yeast and dAMP acted as a general inducer of TNF-α, IFN-γ, and IL-10, whereas DNA from fish soft roe seemed to be linked to regulation of IFN-γ secretion in particular. The RNA effects observed resemble the wide cytokine-inducing ability of CpG-DNA described by Magnusson et al. [24]. Nucleotides dGMP and TMP had no apparent growth effect, but TMP supported high TNF-α secretion and dGMP cocultured with virus antigen inhibited IFN-γ secretion. Several researchers have reported that immunostimulatory DNA suppresses IgE synthesis by promoting IgG, IFN-γ, IFN-α, IFN-β, IL-12, and IL-18 but also TNF-α secretion 25, 26, 27. TNF-α has an unequivocal beneficial function in activating host defences because it can mediate resistance to infectious diseases by controlling intracellular pathogen multiplication and limiting the duration and extent of inflammatory processes [28]. However, overproduction of TNF-α can result in tissue damage and is followed by the emergence of IL-10. In septic shock, IL-10 decreases TNF-α levels [29]. This probably makes DNA from fish soft roe a better candidate for a vaccine adjuvant than RNA from bakers yeast or dAMP to avoid the generally high RNA cytokine-inducing ability. In any case, it should be possible to identify DNA/RNA/nucleotide sequences that activate beneficial cytokines differently to avoid secretion of “harmful” cytokines.

The mechanism of action of immunostimulatory DNA is unknown. Nucleotide receptors (which respond to purines and pyrimidines) have a very widespread tissue distribution 20, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32. Natural nucleosides permeate mammalian cells by means of several carrier-mediated pathways [33] and soluble oligonucleotides, regardless of stimulating potential, bind to a B-cell surface receptor [34]. Recent information about the Toll-like receptor (TLR) family [35] has shown that TLR-9 [36] recognizes immunostimulatory CpG motifs, whereas TLR-3 recognizes viral double-stranded RNA [37]. Single-stranded RNA has been identified as the natural ligand for mouse TLR-7 and human TLR-8 38, 39. Gelman et al. [40] found that stimulation with CpG-containing DNA increased the survival of a purified population of activated CD4+ T cells in vitro by increasing the viable cell number without changing proliferation. Emerging data have indicated that targeting TLR-3, TLR-4, and TLR-9 pathways can enhance responses to certain viral infections [41]. Most tissues express at least one TLR, whereas spleen and peripheral blood leukocytes and especially phagocytes express a great variety of TLRs. Engagement of TLR induces activation of nuclear factor-κB and microtubule-associated protein kinase signaling pathways and cytokine production. However, striking differences have been reported in gene expression profile activation of TLRs by their specific ligands that produce differences in cytokine and chemokine expressions [42]. This may explain the qualitative and quantitative differences observed when measuring growth and cytokine secretion by the different DNA/RNA/nucleotide derivatives in our investigation. Although this has not been demonstrated, nucleotides and nucleotide fragments (depending on structure, concentration, and affinity) may act as ligands or costimulatory molecules for some or restricted TLRs.

DNA from fish soft roe and RNA from bakers yeast and deoxy mononucleotides possess immunomodulatory properties that might be used in vaccine adjuvancy and in allergy protection.

Conclusion 

The nucleotides used in this study did not, by themselves, induce any proliferation of PBMCs derived from healthy individuals. The nucleotides affected PBMC cytokine secretion.

Cultured with a specific antigen, RNA, DNA, dAMP, dCMP, and dUMP increased antigen-induced PBMC proliferation responses and PBMC secretion of selected cytokines. DNA increased specific antigen-induced PBMC secretion of IFN-γ only. dUMP increased antigen-induced PBMC secretion of IL-10 only. Nucleotides dGMP and TMP inhibited antigen-induced PBMC growth and dGMP completely inhibited antigen-induced PBMC IFN-γ secretion. TMP did not change antigen-induced PBMC secretion of the measured cytokines.