Do cytokines have any role in Wilson's disease?

Abstract

The aim of this study was to determine the serum cytokine levels in patients with Wilson's disease (WD) and correlate with phenotype, therapeutic status and laboratory data. In this cross-sectional study, the serum levels of cytokines were estimated in 34 patients (M : F, 23 : 11; drug-naive, 11) with WD (mean age: 13·8 ± 8·6 and 19·6 ± 9·03 years) and compared with 30 controls. The following serum cytokines were analysed using enzyme-linked immunosorbent assay: (i) tumour necrosis factor (TNF)-α, (ii) interferon (IFN)-γ, (iii) interleukin (IL)-2, (iv) IL-6 and (v) IL-4. Serum TNF-α (P < 0·001), IFN-γ (P = 0·005) and IL-6 (P < 0·001) were detectable in WD compared with controls. However, serum level elevation of IL-4 (P = 0·49) and IL-2 (P = 0·11), although detectable compared with controls, was statistically insignificant. The disease severity and therapeutic status did not affect the cytokines. Presence of anaemia, leucopenia, thrombocytopenia, pancytopenia and hepatic dysfunction did not influence cytokine levels. There was a significant negative correlation between IL-6 and ceruloplasmin (P = 0·04) and anti-inflammatory cytokines (IL-4) and copper level (P = 0·01). Serum cytokines, both proinflammatory and anti-inflammatory subtypes, were elevated significantly in patients with WD. Further studies would establish their role in its pathogenesis.

Introduction

Elevated serum cytokines have been documented in neurological disorders: cerebral ischaemia, epilepsy, brain trauma, multiple sclerosis, Creutzfeldt–Jakob disease, amyotrophic lateral sclerosis, Alzheimer's disease and Parkinson's disease. However, their exact role in the pathophysiology of neurological disorders is not clear [1]. Copper acts as a co-factor for a variety of enzymes such as cytochrome c oxidase by its ability to accept and donate electrons under physiological conditions [2]. It is also required by immune cells for stress-induced release of fibroblast growth factor 1 and interleukin (IL)-1α and cytokines such as transforming growth factor (TGF)-β, tumour necrosis factor (TNF)-α and IL-1β. Biologically active extracellular polypeptides, which participate in inflammation, are dependent upon copper and its presence in cell culture is essential for T cell proliferation induced by macrophages or macrophage-dependent cytokines [3–5]. Copper also plays a key role in multiple steps along the angiogenesis pathway [6,7].

Wilson's disease (WD) is an inherited disorder resulting from mutations in the ATP7B gene, deficiency of which results in positive copper balance, tissue copper accumulation and copper-induced oxidant damage [8,9]. Treatment of this disease requires lowering of tissue copper levels with lifelong use of de-coppering drugs (D-penicillamine, zinc acetate, etc.) [10]. A copper-induced rise in levels of cytokines may play an important role in the pathogenesis of WD. It is hypothesized that drugs which decrease tissue levels of copper decrease serum cytokine levels in WD, thereby decreasing inflammatory response and tissue damage [11]. Although it is believed that increased free copper levels in various tissues in these patients may lead to free radical-mediated damage, a previous study from this centre did not find a significant increase in serum malondialdehyde in patients with WD compared with controls [8]. Thus, the mechanism whereby copper deposition leads to tissue damage remains largely unanswered, and literature regarding the role of cytokines in WD is lacking.

The present study was thus designed to ascertain the role of various cytokines [IL-2, IL-4, IL-6, interferon (IFN)-γ, TNF-α] in patients with WD and correlate with clinical and laboratory data, and (b) to compare serum levels of cytokines in drug-naive patients and those on de-coppering therapy.

Methods

This cross-sectional study was carried out from October 2006 to January 2007 on 34 patients with confirmed WD (age: 19·65 ± 9·03 years; M : F, 23:11) attending the speciality clinic for WD at a tertiary care university teaching hospital in South India. The study had the approval of the Institutional Ethics Committee. The diagnosis of WD was established by clinical examination, the presence of Kayser–Fleischer (KF) ring by slit lamp evaluation, low serum ceruloplasmin and total copper levels, and increased 24 h urinary copper excretion. Data were recorded on a predesigned proforma and the patients were subdivided into three groups: (i) drug-naive patients, newly diagnosed patients or those who had stopped treatment for >6 months (n = 11); (ii) on zinc therapy (n = 10); and (iii) on both penicillamine and zinc (n = 13). The dosages of medications were: zinc sulphate 150–300 mg/day and penicillamine 250–750 mg/day. The severity of WD was assessed by the following scales: (i) neurological symptom score (NSS) [12]; (ii) Schwab and England score (S&E) [13]; and (iii) Chu staging [14].

Thirty age- and sex-matched healthy volunteers served as control. Care was taken to ensure that the control population had no inflammatory or infectious disease which may alter serum levels of cytokines. The following serum cytokines were analysed: (i) TNF-α, (ii) IFN-γ, (iii) IL-2, (iv) IL-6 and (v) IL-4. The cytokines were analysed using a standard kit provided by BD-Pharmingen (San Diego, CA, USA). Using the standard protocol prescribed in the kit, the standards were run each time along with the assay samples. The test adopted here was sandwich enzyme-linked immunosorbent assay (ELISA), where cytokine from the given sample was captured on the solid surface of an ELISA plate. The test was considered ‘undetectable’ if it was <0·4 pg/ml. All the controls had undetectable values.

The serum levels and degree of rise of these cytokines were analysed in relation to: (i) clinical severity and disability (severe disease – any one: NSS > 25, S&E < 50% and Chu stage 3; non-severe disease – NSS < 25, S&E > 50% and Chu stage 1, 2); (ii) therapeutic status (drug-naive, on de-coppering therapy – zinc alone, zinc and penicillamine); (iii) age at onset (young onset =15 years of age; late onset >15 years of age); (iv) duration of illness and treatment (≤ 12 months; >12 months); (v) haematological parameters [anaemia–haemoglobin (Hb) <10 g%; leucopenia <4000/mm³, thrombocytopenia <1·5 lakhs/mm³, pancytopenia – any two of the above]; (vi) hepatic dysfunction (raised liver enzymes, low serum albumin, raised serum bilirubin); and (vii) serum copper/ceruloplasmin and urinary copper excretion. Patients were considered to have improved completely if they were asymptomatic at follow-up and as partially improved if they had shown improvement, but did not become symptom-free.

Statistical analysis was performed using spss version 13. For group comparisons, the cytokine levels were analysed as dichotomous variables and compared using the χ² test. Fisher's exact test was used when any number in the cross-tabulation was less than 5. A P value of <0·05 was taken as significant. The Mann–Whitney U-test was used for comparing non-normally distributed variables among patient groups.

Results

Demographic, clinical and laboratory data

The mean duration of illness at diagnosis and evaluation was 15·4 ± 22·6 months (range: 0–96 months) and 80·7 ± 79·5 months (range: 1–300 months) respectively. These patients were from 30 families and consanguinity was noted in 28 of 30 families (93·3%); 13 families (43·3%) had a positive family history suggestive of WD. Based on dominant clinical manifestations, patients could be grouped into the following forms: neuropsychiatric (n = 17), a combination of neuropsychiatric and hepatic (n = 9), pure hepatic (n = 6) and presymptomatic (n = 2). All subjects had KF ring on slit lamp examination. Other systemic observations were: pallor (n = 8), darkening of the skin (n = 4), hepatosplenomegaly (n = 3), ascites (n = 1) and joint pain (n = 1). The laboratory parameters in patients with WD are given in Table 1. Therapeutic status was: complete improvement in nine, partial improvement in six and status quo in six patients. Two asymptomatic patients were on zinc at the time of diagnosis and remained so until the current evaluation.

Table 1.

Laboratory parameters in patients with Wilson's disease (n = 34).

Parameters	Mean ± s.d. (range)	Abnormal (n %)
Haemoglobin (g %)	11·7 ± 1·79 (7·2–14·4)	Anaemia – 6 (17·7%)
TLC (mm3)	5532 ± 1772 (2400–9600)	Leucopenia – 5 (14·1)
Platelet count (mm3)	153 911 ± 81 663 (43 000–439 000)	Thrombocytopenia – 17 (50)
Pancytopenia		6 (17·7%)
Serum total copper normal (75–160 µg/dl)	87·8 ± 61·8 (9–260)	Decreased in 16 (47·1)
Serum ceruloplasmin normal (15–35 mg/dl)	12·2 ± 21·7 (1–124)	Decreased in 28 (82·4)
24 h urinary Cu normal (<70 µg/day)	402·9 ± 273·2 (25–994)	Increased in 26/28 (92·8)
RBS (mg/dl)	96·3 ± 27·8 (56–203)	Hyperglycaemia in 1 (5·9)
Blood urea (mg/dl)	23·2 ± 5·5 (13–35)	Normal in all
Serum creatinine (mg/dl)	0·83 ± 0·22 (0·5–1·4)	Normal in all
Serum calcium (mg/dl)	8·7 ± 0·77 (6·4–10·2)	Hypocalcaemia – 13 (38·2)
Total bilirubin (mg/dl)	1·02 ± 0·77 (0·2–3·6)	Elevated in 8 (23·5)
SGOT (IU)	48·7 ± 41·8 (18–182)	Elevated transaminases (>5 times): nil
SGPT (IU)	38·1 ± 29·2 (10–160)
S. Alkaline phosphatase (IU/l)	256·7 ± 214·6 (57–1020)	Elevated in 23 (67·6)
Total serum protein (g/dl)	6·7 ± 0·71 (4·4–8·4)	Decreased in 17 (50)
Serum albumin (g/dl)	3·6 ± 0·59 (1·5–4·6)	Decreased in 10 (29·4)
A/G ratio	–	Reversed in 9 (26·5)

A/G, albumin/globulin; RBS, random blood sugar; s.d., standard deviation; SGOT, serum glutamic oxaloacetic transaminase; SGPT, serum glutamate pyruvate transaminase; TLC, total leukocyte count.

The WD and cytokines

Serum cytokines in patients versus controls

Serum levels of TNF-α (P < 0·001), IFN-γ (P = 0·005), IL-6 (P < 0·001) were elevated in patients with WD, and this difference was statistically significant compared with controls. There was an insignificant difference in serum IL-4 (P = 0·49) and IL-2 (P = 0·11) levels compared with controls (Table 2).

Table 2.

Comparison of serum cytokines in patients and controls (n = 34).

Cytokines	Subjects	Detectable†	Non-detectable†	P-value
TNF-α	Patient (n = 33)	13	20	<0·001*
	Control	0	30
IFN-γ	Patient	8	26	0·005*
	Control	0	30
IL-6	Patient	11	23	0·001
	Control	0	30
IL-4	Patient	2	32	0·5
	Control	0	30
IL-2	Patient (n = 33)	4	29	0·1
	Control	0	30
Any cytokines	Patient	22	12	<0·001
	Control	0	30

^*P < 0·05: statistically significant.

^†Detectable if >0·4 pg/ml; undetectable if <0·4 pg/ml. IFN, interferon; IL, interleukin; TNF, tumour necrosis factor.

Serum cytokines in patients and clinical parameters

Patients with WD were subdivided into two groups: duration of illness at time of evaluation ≤12 months or >12 months. A statistically significant negative correlation (P = 0·037) was found between serum IL-4 levels and duration of illness, implying that as the duration of illness increased, the serum IL-4 levels decreased. The patients were divided into two subgroups, based on NSS, S&E and Chu scores: (i) severe disease (NSS > 25, S&E < 50%, Chu: 3); and (ii) less severe disease (NSS ≤ 25, S&E > 50%, Chu: 1, 2). There was no significant difference in any of the cytokines in the severe and less severe groups. A negative trend (P = 0·07) was seen when serum levels of IL-6 were correlated with age at onset of illness. There was no other correlation between the levels of individual cytokines or their combination with respect to whether or not the patient received treatment, or whether they received zinc alone or both zinc and penicillamine (Table 3).

Table 3.

Comparison of serum cytokines with clinical and laboratory parameters.

Parameters	TNF-α		IFN-γ		IL-6		IL-4		IL-2
	D	ND	D	ND	D	ND	D	ND	D	ND
Severe disease	11	17	8	21	10	19	2	27	4	24
Not severe disease	2	3	0	5	1	4	0	5	0	5
P-value	1·00		0·4		1·00		1·00		1·00
Drug-naive	3	7	4	10	6	8	2	12	2	11
De-coppering	10	13	4	16	5	15	0	20	2	18
P-value	0·7		1·0		0·43		1·0		1·0
= 12 mon DOI	3	3	2	5	2	5	2	5	1	5
>12 mon DOI	10	17	6	21	9	18	0	27	3	24
P-value	1·00		0·69		0·46		0·16		1·00
Anaemia: present	2	4	6	22	8	20	2	26	4	23
Anaemia: absent	11	16	2	4	3	3	0	6	0	6
P-value	1·00		0·61		0·36		1·00		1·00
Leucopenia: present	2	5	6	21	9	18	1	26	3	23
Leucopenia: absent	11	15	2	5	2	5	1	6	1	6
P-value	0·67		1·00		1·00		0·37		1·00
Thrombocytopenia: present	5	10	5	11	5	11	2	14	2	2
Thrombocytopenia: absent	8	10	3	15	6	12	0	18	13	16
P-value	0·72		0·42		1·00		0·21		1·00
Pancytopenia: present	3	6	5	20	8	17	1	24	3	21
P-value	1·00		0·65		1·00		0·46		1·00
LF: abnormal	5	8	5	15	6	14	0	20	3	17
LF: normal	8	12	3	11	5	9	2	12	1	12
P-value	1·00		0·65		1·00		0·16		1·00
†Serum copper (P-value)	−0·046 (0·799)		−0·134 (0·450)		−0·238 (0·174)		−0·223 (0·206)		−0·044 (0·810)
†Serum ceruloplasmin (P-value)	−0·132 (0·465)		−0·205 (0·244)		−0·353 (0·041)**		−0·0212 (0·229)		−0·201 (0·262)
†24 h urinary copper (P-value)	0·109 (0·580)		0·168 (0·384)		0·239 (0·212)		0·0248 (0·194)		0·174 (0·465)

^**P < 0·05.

^†Serum levels of various cytokines were correlated with the biochemical parameters such as serum copper, serum ceruloplasmin and 24 h urinary copper using Spearman's correlation test. Mon, month; DOI, duration of illness; D, detected; ND, not detected; LF, liver function tests; Urin, urinary.

Serum cytokines in patients and laboratory parameters

There was a significant negative correlation between IL-6 levels and serum ceruloplasmin (P = 0·041) and between the levels of anti-inflammatory cytokines (IL-4) and serum copper levels (P = 0·014). There was a significant negative correlation between IL-4 and albumin (P = 0·037). There was no correlation between the cytokines and the presence of anaemia, leucopenia, thrombocytopenia or pancytopenia, raised bilirubin and aminotransferase.

Quantitative analysis of cytokines versus clinical and laboratory parameters

For correlating the serum levels of TNF-α patients were divided into two groups, based on the serum levels of TNF-α: (i) >25 pg/ml and (ii) ≤ 25 pg/ml. For correlating the serum levels of IFN-γ and IL-6, patients were divided into two groups based on their serum levels: (i) >50 pg/ml; (ii) ≤ 50 pg/ml or not detectable. These two groups were analysed with respect to disability and impairment scales, haematological parameters, duration of illness and treatment, whether or not the patient received treatment, age at onset of illness and decompensated hepatic disease. The degree of increase in levels of various cytokines did not bear any relationship to any of these parameters.

Discussion

For many years the central nervous system was considered to be ‘immune privileged’, but it is now known that in response to injury, infection or other diseases it generates inflammatory mediators, including proinflammatory cytokines, prostaglandins, free radicals and complements. WD occurs because of a mutation in gene ATP7B, which leads to defective biliary excretion of copper, resulting in positive copper balance, and copper toxicity from oxidant damage [15]. This study was designed to ascertain whether serum cytokines have any role in WD. The phenotype in this study was comparable to previous studies [16,17]. Consanguinity among parents in this study was higher (93·3%) than the reported incidence of 33% observed in general population in South India [18].

Cytokines are soluble proteins involved in the regulation of growth, development and activation of immune system cells and mediation of the inflammatory response [15]. There are reports of increased levels of cytokines in other neurodegenerative disorders, but their role in the pathophysiology is speculative. Perry suggested that microglial activation in the brain might be associated with the production of anti-inflammatory rather than proinflammatory cytokines [19]. Lucas et al. reported that the exact role of cytokines in Parkinson's disease remains controversial [1]. In our cohort, a significant increase in levels of TNF-α (P < 0·001), IFN-γ (P = 0·005) and IL-6 (P < 0·001) were detected in WD compared with controls. To the best of our knowledge, this is the only study in WD and these observations have not been documented. Elevation of serum levels of TNF-α, IFN-γ and IL-6 might suggest an ongoing inflammatory process. It is also likely that anti-inflammatory cytokines (IL-4) are increased in WD as a compensatory measure due to the natural defence mechanisms of the body, or might be due to the pleiotropic nature of the cytokines. It might therefore strengthen the hypothesis that inflammation might have an important role in WD. The underlying mechanism could be secondary to copper-induced injury.

In the present study, there was no significant association between the serum levels of various cytokines and severity of the disease. While this might suggest the absence of cause and effect relationship between cytokines and diseaseseverity, the results need to be interpreted with caution. We had evaluated these patients cross-sectionally, and several of them were on long-term treatment. It is likely that some patients, with an initially severe form of the disease, might have improved. Disease severity as measured by disability scales during the course of illness might not actually reflect the severity of underlying inflammatory process at the time of sampling and might simply reflect the consequence of an insult in the past, prior to initiation of therapy. The ongoing inflammation in this subset of patients is expected to be less severe. Therefore, an absence of correlation between disease severity and cytokines should not negate an association between them. Better-designed studies may resolve this issue.

A statistically significant negative correlation (P = 0·037) was noted between IL-4 and the duration of illness. However, the significance of this association is not certain, as serum IL-4 was not elevated significantly in cases compared with controls. There was no other correlation between serum cytokines and duration of illness and treatment. This might be expected, as neither of the above parameters reflects disease activity at the time of sampling. Patients with lesser duration of illness may have more active disease and vice versa. The significance of association of IL-4 and albumin is not definitive. Moreover, the serum IL-4 level was not elevated significantly in cases compared with controls.

There was no significant relation between the serum cytokines and therapeutic status, therefore the modifying role of the de-coppering agents on the cytokines could not be substantiated. Ziche et al. found that the administration of penicillamine decreased serum cytokine levels in experimental models of cancer [7]. Brem et al. reported penicillamine to be an effective inhibitor of intracerebral glioma growth, invasiveness and angiogenesis in experimental animals, acting by decreasing serum copper levels and thereby serum levels of copper-dependent cytokines [20]. We did not find any effect of zinc on the serum cytokine levels, although it is reported to increase serum levels of cytokines in literature. Our findings did not support the views of Brewer et al.[11]. Serial analysis of cytokines before and after initiating treatment may establish the relationship. Song et al. summarized that tetrathiomolybdate, a drug used in WD, attenuated bile duct ligation-induced cholestatic liver injury and fibrosis in mice, in part by inhibiting TNF-α and TGF-β1 secretion. The protective mechanism might be independent of oxidative stress [21].

Patients with low serum total copper had higher levels of anti-inflammatory cytokine IL-4 (P = 0·014), suggesting a compensatory increase in level of anti-inflammatory cytokine in patients with severe disease as reflected by low serum copper levels. Tapiero et al. had documented high free copper in inflammatory conditions [22]. The estimation of free copper levels could not be carried out due to lack of facility. A significant negative correlation between IL-6 and ceruloplasmin (P = 0·04) and anti-inflammatory cytokine (IL-4) and copper (P = 0·01) was noted. Thus, IL-6 levels were elevated significantly in patients with lower serum ceruloplasmin. These may reflect a compensatory increase in the level of cytokines in patients with very low ceruloplasmin and copper. Tapiero et al. proposed that IL-6 stimulates the synthesis of ceruloplasmin [22].

Limitations

The sample in this study was heterogeneous, with variable duration of illness and treatment and differential outcome. Another limitation could have been due to high consanguinity; there may be a restricted gene pool and the results could be affected by certain common polymorphisms in the genes encoding the cytokines.

To conclude, the serum levels of various cytokines were elevated in patients with WD. An increase in the level of cytokines in patients with serum low ceruloplasmin and copper was observed. These observations suggest a possible underlying role of cytokine-mediated inflammation in WD and warrant further studies to explore therapeutic potentials of anti-cytokine agents, particularly in severe and refractory forms of the disease.