High salt intake does not exacerbate murine autoimmune thyroiditis

Abstract

Recent studies have shown that high salt (HS) intake exacerbates experimental autoimmune encephalomyelitis and have raised the possibility that a HS diet may comprise a risk factor for autoimmune diseases in general. In this report, we have examined whether a HS diet regimen could exacerbate murine autoimmune thyroiditis, including spontaneous autoimmune thyroiditis (SAT) in non-obese diabetic (NOD.H2^h4) mice, experimental autoimmune thyroiditis (EAT) in C57BL/6J mice challenged with thyroglobulin (Tg) and EAT in CBA/J mice challenged with the Tg peptide (2549–2560). The physiological impact of HS intake was confirmed by enhanced water consumption and suppressed aldosterone levels in all strains. However, the HS treatment failed to significantly affect the incidence and severity of SAT or EAT or Tg-specific immunoglobulin (Ig)G levels, relative to control mice maintained on a normal salt diet. In three experimental models, these data demonstrate that HS intake does not exacerbate autoimmune thyroiditis, indicating that a HS diet is not a risk factor for all autoimmune diseases.

Introduction

Recently, two independent studies have reported that a high salt diet (HSD) can exacerbate the severity of experimental autoimmune encephalomyelitis (EAE) in mice, compared to the level of severity developed in mice fed a normal diet [1,2]. These findings were fascinating as well as intriguing, in view of the fact that data generated by the World Health Organization (WHO) do not support a direct correlation between salt intake and prevalence of multiple sclerosis (MS) in humans [3,4]. Nevertheless, the reported dramatic exacerbation of EAE in mice on a HSD has highlighted the possibility that increased salt intake may aggravate clinical symptoms in individuals genetically predisposed to MS and, most importantly, has raised the question of whether HSD might represent a risk factor for the development of other autoimmune diseases.

In this report, we have examined potential HSD effects on the development of autoimmune thyroiditis in mice. The experimental design was made to monitor the incidence and severity of disease in three mouse models that have been used extensively to study pathogenetic mechanisms analogous to those operating in Hashimoto's thyroiditis (HT) in humans. In the first group we used non-obese diabetic (NOD).H2^h4 mice, which are known to develop spontaneous autoimmune thyroiditis (SAT) with age [5–7]. In the second and third groups, we asked whether mild forms of experimental autoimmune thyroiditis (EAT), induced after challenge with thyroid antigen in adjuvant, could be exacerbated by concomitant high salt (HS) intake. To this purpose, we challenged C57BL/6 mice with thyroglobulin (Tg) and CBA/J mice with the known pathogenic Tg peptide (2549–2560) containing thyroxine (T4) at amino acid position 2553 [(T4)2553] [8,9]. We also monitored Tg-specific serum immunoglobulin (Ig)G responses as an index of autoreactivity in all strains.

Materials and methods

Mice and diet

Female C57BL/6J and CBA/J mice were purchased from Jackson Laboratories and were used at 6–8 weeks of age. NOD.H2^h4 mice, kindly provided to us by Dr H. Braley-Mullen (University of Missouri, Columbia), were bred and maintained in the Animal Facility, Faculty of Medicine, Memorial University of Newfoundland. Following the protocol of Kleinewietfeld et al. [1], control mice received a regular sodium diet (0·49% NaCl, Teklad TD.96208) and deionized distilled water (autoclaved before distribution) ad libitum. The HSD groups received a combination of salt-enriched chow (4% NaCl Teklad TD.92034) and water supplemented with 1% NaCl. The commercial rodent chow had a natural ingredient base containing wheat, corn, soy and alfalfa; (w/w) 19% protein; 48% carbohydrates; 5·2% fat; 0·9% Ca; 0·8% K. Water intake was monitored weekly and is expressed as grams consumed per mouse per day.

Antigens and immunizations

Tg was purified from frozen thyroid glands of female outbred CD1 mice (Charles River, Quebec, Canada) by gel filtration using a Sepharose CL-4B (Amersham Biosciences AB, Uppsala, Sweden) column, as described previously [7]. The Tg peptide (T4)2553 was synthesized by Biosynthesis Inc. (Lewisville, TX, USA) at >85% purity and was blocked with an acetyl group at the N terminus and with an amide group at the C terminus. C57BL/6 and CBA/J mice were challenged subcutaneously (s.c.) with 100 μg of Tg or 100 nmol (T4)2553 peptide, respectively, in 100 μl of complete Freund's adjuvant (CFA) emulsion (with Mycobacterium butyricum; Difco Laboratories, Detroit, MI, USA). After 3 weeks, C57BL/6 mice were boosted with 50 μg of Tg and CBA/J mice with 50 nmol of (T4)2553 peptide, both emulsified in Freund's incomplete adjuvant (IFA) (Difco). EAT induction was assessed 5 weeks after the first antigenic challenge. All experimental procedures were reviewed and approved by the Animal Care Committee at Memorial University of Newfoundland.

Serological assays

Serum aldosterone was assessed by radioimmunoassay (RIA) (Coat-a-count Aldosterone kit TKAL1; Siemens, Oakville, ON, Canada), according to the manufacturer's instructions. Serum IgG antibodies specific for Tg were detected by a two-step enzyme-linked immunosorbent assay (ELISA), as described previously [7]. Briefly, wells of polyvinylchloride plates were coated overnight with 10 μg/ml Tg in carbonate buffer, pH = 9·6 and blocked with 0·1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) for 1 h. Serum samples, diluted in PBS/Tween 20 containing 0·1% BSA, were added to wells for 1 h and specific antibody binding was assessed with an alkaline phosphatase-conjugated goat anti-mouse IgG antibody (Jackson Immunoresearch Laboratories, Inc., West Grove, PA, USA). After 20 min, conversion of the p-nitrophenyl phosphate substrate (Sigma, St Louis, MO, USA) into the p-nitrophenol product was determined by light absorption at 405 nm using a Vmax plate reader (Molecular Devices, Sunnyvale, CA, USA).

Histological assessment of thyroiditis

Fixation, embedding in methacrylate and sectioning of thyroid lobes was performed as described previously [10]. Histological sections were stained with haematoxylin and eosin, and the mononuclear cell infiltration index (II) was scored blindly as follows: 0, no infiltration; 1, interstitial accumulation between two or more follicles; 2, one or two foci of inflammatory cells larger than one follicle; 3, diffuse infiltration of 10–40% of total area; and 4, extensive infiltration of 40–80% of total area. Each mouse was assigned the maximum II score observed. Statistical comparison between groups was performed by the non-parametric Mann–Whitney U-test.

Results and discussion

Following placement of NOD.H2^h4, C57BL/6 and CBA/J mice on HSD (n = 10, per strain) or control diet (n = 10 per strain), water intake was monitored weekly over the 5-week observation period as an index of response to the dietary regimen [11]. Mice on the HSD showed a highly significant increase of water intake with average [mean ± standard error (s.e.)] daily consumption rates 15·5 ± 0·71, 9·1 ± 0·3 and 7·3 ± 0·2 g/mouse for NOD.H2^h4, C57BL/6 and CBA/J strains, respectively, while control values ranged between 3·2 and 4·3 g/day (Fig. 1a). At the end of the study period, the effectiveness of the HSD was assessed further by measurement of serum aldosterone levels to check for the HS-induced suppression reported by others [12,13]. A profound suppression of serum aldosterone was observed in the experimental versus the control groups in all strains: 13 versus 381 pg/ml in NOD.H2^h4, 13 versus 254 pg/ml in C57BL/6 and 21 versus 386 pg/ml in CBA/J mice (Fig. 1b).

Figure 1.

Effects of high salt diet (HSD) versus control diet (CD) on physiological and immunological parameters in mice developing autoimmune thyroiditis. (a) Water intake (g/day) in mice placed on HSD or CD for 5 weeks, expressed as mean ± standard error (s.e.) of values (n = 10 per group). Asterisks denote statistical significance assessed by a t-test (P-values for all groups <10⁻⁷). (b) Serum aldosterone levels measured by radioimmunoassay (RIA) at the end of the 5-week observation period. Five samples were run for each experimental and control group, with each sample being pooled from two randomly selected mice within a group. The data represent the mean ± s.e. values in each group and asterisks denote statistical significance assessed by a t-test (P < 0·036 for non-obese diabetic (NOD).H2^h4 mice; P < 10⁻⁴ for C57BL/6 and CBA/J). (c) Enzyme-linked immunosorbent assay (ELISA)-based thyroglobulin (Tg)-specific immunoglobulin (Ig)G responses from individual mouse sera, observed spontaneously (NOD.H2^h4 mice) or after antigenic challenge, as described in Materials and methods (C57BL/6 and CBA/J mice). Serum dilutions for the assay were 1:25 for NOD.H2^h4, 1:50 for C57BL/6 and 1:100 for CBA/J mice. Data represent the mean ± standard deviation (s.d.) optical density values of triplicate wells.

NOD.H2^h4 mice, developed by Dr L. Wicker and colleagues for studies on diabetes [14], were chosen as a first animal model for our study because it is the only mouse strain known to develop SAT at an increasing incidence and severity with age. In our colony, the first signs of spontaneous thyroiditis appear in 12–14-week-old male or female mice [7], and at 55 weeks of age practically all mice exhibit SAT of advanced severity and seropositivity for Tg-specific IgG (unpublished observations). While thyroiditis incidence and severity may differ among various colonies, analogous age-dependent effects on SAT have been reported by others [5,6]. In this report, our aim was to test whether a HSD, initiated at the early stages of SAT, would influence the development of disease. As shown in Table 1, placement of 12-week-old NOD.H2^h4 mice on a HSD for 5 weeks did not significantly raise the incidence (6/10 versus 9/10) or severity (II = 1·3 versus 1·6) of SAT versus those of the controls. In addition, there was no significant increase in the number of mice spontaneously developing IgG antibodies to Tg (Fig. 1c).

Table 1.

Thyroiditis incidence and severity in non-obese diabetic (NOD).H2^h4, C57BL/6 and CBA/J mice.

Strain	NaCl†	Immunizing antigen	Thyroiditis							P-valueb
			Incidence	Severity (score)					Mean
				0	1	2	3	4
NOD.H2h4‡	+	-	6/10	4	1	4		1	1·3	0·63
	−	-	9/10	1	5	2	1	1	1·6
C57BL6/J§	+	Tg	9/10	1	7	1	1		1·2	0·39
	−	Tg	6/10	4	4	1	1		0·9
CBA/J¶	+	T4p2553	6/10	4	3	3			0·9	0·35
	−	T4p2553	4/10	6	3	1			0·5

^†Mice received normal chow and water ad libitum or chow containing 4% NaCl and water containing 1% NaCl ad libitum during a 5-week period, as described in the Mice and diet section.

^bStatistical significance was assessed by the non-parametric Mann–Whitney U-test.

^‡Each group consisted of five female and five male 12-week-old thyroiditis-prone NOD.H2^h4 mice.

^§Female C57BL/6J mice (n = 10 per group) were challenged subcutaneously (s.c.) with 100 μg thyroglobulin (Tg) in complete Freund's adjuvant (CFA) and the experimental group was placed simultaneously on a high salt diet (HSD). Three weeks later, all mice received 50 μg Tg in incomplete Freund's adjuvant (IFA). Experimental autoimmune thyroiditis (EAT) was determined 5 weeks after the initial antigenic challenge.

^¶Female CBA/J mice (n = 10 per group) were challenged s.c. with 100 nmol of T4p2553 peptide in complete Freund's adjuvant (CFA) and the experimental group was simultaneously placed on a HSD. Three weeks later, all mice received 50 nmol of the same peptide in IFA. EAT was determined 5 weeks after the initial antigenic challenge.

In a second series of experiments, we followed the experimental design of Wu et al. [2] and Kleinewietfeld et al. [1] by placing C57BL/6 mice on the same HS regimen (4% NaCl in chow, 1% NaCl in drinking water) that exacerbates EAE in this strain and a similar period of observation (5 weeks). An added attractive feature was that, in the field of EAT, this strain is well known to develop relatively low levels of thyroiditis after challenge with Tg in adjuvant [15], thus favouring investigation into possible enhancing effects of HS intake on pathology in this model. The results showed that a dietary regimen of HS, continued throughout the period of antigenic challenge, was associated with a tendency towards an increase in EAT incidence (9/10 versus 6/10) or severity (II = 1·2 versus 0·9) versus that of the controls, but these differences were not significant (P = 0·39) (Table 1). Moreover, strong, but not significantly different, Tg-specific IgG responses were detected in both the HS and control groups (Fig. 1c).

In a third experimental model, we challenged the EAT-susceptible strain CBA/J [15] with the 12mer T cell epitope (T4)2553 of Tg [8,9] at doses that are mildly pathogenic. Mice in the experimental group were placed on a HSD during the 5-week induction phase of EAT. The HS regimen was again associated with a tendency towards an increase in the incidence (6/10 versus 4/10) or severity (II = 0·9 versus 0·5) of disease, but these differences were not statistically significant (P = 0·35) (Table 1). The (T4)2553 peptide elicited a Tg-specific IgG response of similar relative strength in both groups which was, as expected, lower than that obtained in C57BL/6 mice immunized with intact Tg (Fig. 1c).

The focus of the current work was to examine the thesis that there is a tangible link between a HS diet and autoimmune disease other than EAE, using mouse models of autoimmune thyroiditis. The data provided demonstrate strongly that dietary salt does not promote either antigen-induced or spontaneous thyroid autoimmune disease. The report that C57BL/6 mice on HS intake develop enhanced EAE after challenge with an encephalitogenic myelin oligodendrocyte glycoprotein (MOG) peptide in CFA [1,2] is in apparent contrast with our finding, that delivery of a thyroiditogenic molecule (Tg) to the same strain via the same route and under the same HS conditions does not lead to enhanced EAT. To the extent that the studies of Wu et al. [2] and Kleinewietfeld et al. [1] have assigned a pivotal role to pathogenic T helper type 17 (Th17) cells for the exacerbation of EAE, our data suggest that: either (i) Th17 cells are not as critical in EAT/SAT development as in EAE or (ii) additional pleiotropic physiological, hormonal and metabolic factors, triggered by a HSD, mediate EAE exacerbation in a unique manner that is not necessarily duplicated in EAT/SAT.

EAT and SAT are well known to be mediated by Th1 cells [16], but a role for Th17 cells in pathogenesis remains possible. IL-17^−/− NOD.H2^h4 mice have been reported to exhibit reduced iodide-accelerated SAT and diminished Tg-specific titres compared to wild-type (WT) controls, while intrathyroidal cytokine mRNA analysis has suggested the presence of both Th1 and Th17 cells in thyroidal lymphocyte infiltrates of WT NOD.H2^h4 mice [17]. In humans, an increased frequency of Th17 cells has been reported in the peripheral blood of patients with HT [18–20] or autoimmune thyroid disease (AITD) (HT and Graves' disease) [21,22] and it has been suggested recently that T cell-derived leptin may contribute to the increased frequency of Th17 cells in HT [23]. Associations of single nucleotide polymorphisms in AITD have been described for the IL-23R gene in Japanese populations [24] and for the IL-17A and IL-17F genes in Chinese populations [25]. However, as IL-17 is produced by several cell types, including CD8⁺ cells and γδ T cells [26–28], association data from both mouse and human studies may not necessarily reflect a cause-and-effect relationship between Th17 cells and thyroid pathology. Lastly, it is noteworthy that even in the pivotal EAE studies [1,2], exacerbating effects of a HSD on disease were demonstrable in WT mice during the first 20 days of the dietary regimen. When monitoring continued until day 25 in one of the studies, no significant differences between the HSD and control groups were found [2]. In the work presented here, while no significant effects of the HSD on EAT were observed using a well-established 5-week disease induction protocol, we cannot exclude the possibility that high salt intake might have had transient, non-sustained effects on EAT that were not apparent after 5 weeks.

Regardless of the mechanism involved, our data demonstrate clearly that excess salt in the diet should not be expected to be an environmental risk factor that broadly influences induction of autoimmune disease. Enhanced salt intake is accompanied by fluid, ionic and hormonal changes, including suppression of the renin–angiotensin system [29,30], which could differentially influence autoimmune responses depending on the anatomical site and metabolic status of the target organ. The planning of clinical trials to investigate low-salt diets for treating or preventing autoimmune disease should proceed with caution, taking into account relevant data from diverse animal model studies.

Disclosures

The authors declare that they have no conflicts of interest.