Introduction
Systemic lupus erythematosus (SLE) is a multi-system autoimmune disease characterized by immune dysregulation that results in the production of antinuclear and other autoantibodies, as well as immune complex deposition in the kidneys and other organs. The disease course of SLE is heterogeneous and characterized by unpredictable flares and remissions. Thus, there is a pressing need to identify biomarkers that will facilitate better assessment of disease activity and organ involvement, and provide insight into the relationships between pathogenesis and clinical manifestations.
Recently, we and others have used gene expression microarrays to identify a group of type I IFN-inducible genes (IFIGs) that are significantly upregulated in peripheral blood cells from SLE patients [
1‐
4]. The expression of these IFIGs, often referred to as IFN signatures, was later found to be closely associated with increased disease activity, specific autoantibody profiles and significant organ damage in SLE patients [
5,
6]. In addition to carrying markers of the IFN signature, peripheral blood cells from SLE patients are also elevated in a variety of chemokines [
7]. Chemokines are a group of small molecules with the ability to recruit specific leucocytes to target tissue sites, thereby contributing to the organ damage seen in SLE. Other functions of chemokines include their ability to influence dendritic cell maturation, induction of B-cell and T-cell development, determination of peripheral cell localization, and involvement in T-helper-1 and T-helper-2 polarization [
8].
A number of studies have identified increased plasma concentrations of chemokines, including 'regulated upon activation normal T-cell expressed and secreted' (RANTES), monocyte chemotactic protein (MCP)-1, IL-8, IFN-inducible protein 10 (IP-10), and monokine induced by IFN-γ (MIG), in patients with active SLE [
9‐
12]. In addition, the
ex vivo production of chemokines by peripheral blood cells from SLE patients appears to be significantly higher than that of cells from normal control individuals, after stimulation by lipopolysaccharide or phytohaemagglutinin [
10], which suggests that the elevated expression of chemokines is involved in the immune dysregulation seen in this disorder.
Although the contributions made by chemokines in the pathogenesis of SLE have been studied extensively, the mechanisms that give rise to the increased chemokine responses in peripheral blood cells from SLE patients remain unclear. It has been reported that certain chemokine responses are strongly dependent upon IL-2 [
13]. Another study [
10] revealed that the plasma concentrations of IP-10 and MIG are significantly correlated with that of IL-18. A recent study [
9] found that several serum chemokines were significantly elevated in SLE patients with increased expression of IFIGs, implying that the production of certain chemokines may be regulated by the type I IFN pathway. It is also interesting that IFN-inducible chemokines are significantly elevated in active SLE patients, a fact that raises the possibility that they might serve as novel biomarkers for SLE disease activity, and which adds a new link between these two essential aspects of SLE pathogenesis. However, the associations between the IFN-inducible chemokines and the clinical features of SLE have not been fully studied. Moreover, the finding that IFN-inducible chemokines may serve as a biomarker in active SLE requires verification in a larger cohort of patients, as well as in patients from different races and backgrounds.
In the present study we measured the transcription levels of seven IFN-inducible chemokines, as well as those of five classical IFIGs, in peripheral blood cells drawn from 67 patients with SLE, 20 with rheumatoid arthritis (RA), and 23 healthy donors, and calculated a chemokine score and an IFN score for each participant. We found that the transcriptional levels of IFN-inducible chemokines in peripheral blood cells were closely associated with disease activity and organ damage in SLE, and may be useful in disease monitoring and prognostication.
Discussion
In the present study, we selected seven IFN-inducible chemokines (RANTES, MCP-1, CCL19, MIG, IP-10, CXCL11 and IL-8), and we investigated the associations between their combined expression level and specific clinical features of SLE. Of these seven chemokines, MCP-1, RANTES and CCL19 are members of the CC family, and preferably recruit monocytes, macrophages, T cells and dendritic cells. In contrast, MIG, IP-10, CXCL11 and IL-8 are from the CXCL family, the first three of which are chemoattractants of activated T cells, whereas IL-8 is chemotactic for neutrophils [
8]. All of these chemokines have been reported to have consensus sequences for IFN-responsive elements, including ISRE (IFN-stimulated responsive element), GAS (IFN-γ activation site) and IRF (interferon regulatory factor), within their gene promoter regions [
19‐
22]. Consequently, the expression levels of these chemokines can be regulated by the IFN pathway, making them IFN inducible. These chemokines have been studied extensively, and their contributions to SLE have been confirmed by several different investigative teams [
23‐
26].
Rather than focusing on individual chemokines, as most previous investigators have done, we investigated the expression of multiple chemokines and assessed the impact that overall chemokine expression has on SLE disease features. We measured the transcription levels of these chemokines in peripheral leucocytes and calculated a chemokine score by combining their expression levels. Given that there are various sources of serum chemokines, other than those produced by peripheral blood leucocytes, measurement of the mRNA levels of these chemokines in peripheral blood cells is possibly a direct indicator of the dysregulation of chemokine expression that exists in peripheral immune cells in patients with SLE. Moreover, the method is simple, inexpensive and has high throughput, making it a suitable approach to gaining an overview of the expression of multiple chemokines.
In the SLE patients included in the study, IFN score was significantly correlated with chemokine scores (Figure
1c), implying that expression levels of the IFN-inducible chemokines are associated with those of classical IFIGs in SLE. This result, however, was difficult to interpret because we did find elevated chemokine scores in some SLE patients with a low IFN score (IFN-low) and low chemokine scores in patients with a high IFN score (IFN-hi). In addition, we found that the overall chemokine score was significantly higher in SLE patients than in RA patients or healthy donors (Figure
1a), whereas IFN score was elevated in both of the disease groups compared to healthy donors. This result verifies previous reports that IFIGs are notably elevated in a subgroup of RA patients [
27] but fails to identify any increase in the expression of IFN-inducible chemokines in RA, indicating that an elevated chemokine score might be more specific for SLE than for RA.
One of the potential explanations for the discrepancy between the expression of IFIGs and IFN-inducible chemokines is the highly complicated regulation of chemokine expression that exists in various diseases. Stimuli other than type I IFNs, such as IL-18 or IL-2, as suggested by previous studies [
10,
13], may be playing a role in driving the expression of chemokines in SLE. Moreover, medication used by the patients at the time of blood donation may elicit different responses in the expression of chemokines or IFIGs. The use of multiple drugs (and probably different drugs) by patients in the two patient groups might also complicate data interpretation. Nevertheless, regardless of the precise mechanism, these data suggest that the chemokine score we present here, although closely linked to IFN score, is an independent index for research and has novel and specific clinical significance.
In the present study we found that chemokine scores were associated with disease activity, as assessed using the SLEDAI-2K score and C3 level, and with ongoing or cumulative organ damage, as assessed based on the presence of active LN or SDI score in SLE patients. An elevated chemokine score may thus be helpful to identify SLE patient with active and severe disease. The preliminary longitudinal data also show that these chemokine scores are responsive to treatment and may change in conjunction with disease activity and severity, suggesting that chemokine score might be used to monitor disease progression and guide therapy. One of the weaknesses of the SLEDAI-2K score is its insensitivity in detecting improvement or worsening in a manifestation, because this can only be recorded as absent or present. For example, although patient 3 (see Figure
3e) had a dramatic decrease in urinary protein level from 6.5 g/24 hours to 0.8 g/24 hours, the SLEDAI-2K score failed to capture the improvement because she was still scored as positive in the proteinuria category. Her chemokine score, however, exhibited a significant decrease in concordance with the clinical improvement. This result, although limited and preliminary, lent further support to the chemokine score as a new and valuable marker of SLE disease activity and severity. However, prospective longitudinal studies with a larger sample size and more visits are needed to assess the role of chemokine score as a reliable biomarker in SLE.
Our conclusion that increased overall production of IFN-inducible chemokines by peripheral blood cells is important in the pathogenesis of SLE is supported by a number of published papers. Chemokines may contribute to SLE by recruiting immune and inflammatory cells to target tissues and by altering the normal trafficking and localization of certain populations of immune cells in the body; hence, they may impair the normal function of such cells. In cutaneous lupus erythematosus, MIG and IP-10 have been found to be significantly upregulated in inflamed skin and to help in the recruitment of plasmacytoid dendritic cells (pDCs), the major producers of type I IFN, to the skin [
28]. This result could explain, at least in part, the observation that the number of pDCs is reduced in the peripheral blood of SLE patients [
29], and that pDCs are recruited into and enriched within inflamed target tissues [
30,
31]. Moreover, ectopic expression of CCL19 can retain dendritic cells in target tissue and prevent their normal homing and migration to lymph nodes [
32]. Previous investigators have reported that systemic over-expression of MCP-1 in mice can impair the homing and migration of monocytes to a localized MCP-1 gradient [
33], and that MCP-1 may inhibit the normal differentiation of monocytes, which is possibly one of the mechanisms that is involved in autoimmunity [
34]. In confirmation of these reports, our current data demonstrate that the overall production of IFN-inducible chemokines, as measured using a chemokine score, may serve as a useful indicator of the ongoing state of immune dysregulation in SLE.
In addition, in a small-scale study we also observed that the expression levels of those IFN-inducible chemokines were notably elevated in CD14
+ monocytes compared with T and B lymphocytes from SLE patients, indicating that monocytes might be more important contributors to the chemokine score than lymphocytes (data not shown). Therefore, the number as well as the activation state of the circulating monocytes might be a valuable clinical marker in SLE. In accordance with this assumption, it was recently reported [
35] that activated renal macrophages are markers of disease onset and remission in LN, adding the possibility that active circulating monocytes might also be useful in disease monitoring in SLE. However, additional studies are needed in this area.
The patients with anti-Sm or anti-RNP autoantibodies had higher chemokine scores than did SLE patients without these two autoantibodies. An association of chemokine score with anti-Sm and anti-RNP antibodies is, to our knowledge, reported here for the first time. The underlying pathophysiological mechanism for this remains unknown. One possible explanation, however, is that the expression of IFN-inducible chemokines is somewhat linked to the IFN signature. The association between the IFN signature and anti-RNP autoantibodies was reported in earlier studies [
5,
6,
36]. Although the mechanisms are unclear, it has been proposed that activation of pDCs by single-stranded or double-stranded RNA, through Toll-like receptors, might lead to the induction of type I IFN production and enhanced presentation of RNA-associated materials to autoreactive T and B cells. This, in turn, has the potential to cause upregulation of IFIGs and the appearance of anti-RNA-associated protein autoantibodies. Given that patients who are positive for anti-Sm or anti-RNP antibodies exhibit increased IFN scores, it is not surprising that such patients also have higher expressions of IFN-inducible chemokines and exhibit higher chemokine scores.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
QF, NS and CB designed the study. YG and JC collected clinical data and blood samples. HC participated in RNA extraction and cDNA preparation. QF and XC performed real-time PCR and conducted data analysis. QF, NS and XC wrote the manuscript. CB and NS supervised the study. All authors read and approved the final manuscript.