SIGLEC1 (CD169): a marker of active neuroinflammation in the brain but not in the blood of multiple sclerosis patients

We aimed to evaluate SIGLEC1 (CD169) as a biomarker in multiple sclerosis (MS) and Neuromyelitis optica spectrum disorder (NMOSD) and to evaluate the presence of SIGLEC1+ myeloid cells in demyelinating diseases. We performed flow cytometry-based measurements of SIGLEC1 expression on monocytes in 86 MS patients, 41 NMOSD patients and 31 healthy controls. Additionally, we histologically evaluated the presence of SIGLEC1+ myeloid cells in acute and chronic MS brain lesions as well as other neurological diseases. We found elevated SIGLEC1 expression in 16/86 (18.6%) MS patients and 4/41 (9.8%) NMOSD patients. Almost all MS patients with high SIGLEC1 levels received exogenous interferon beta as an immunomodulatory treatment and only a small fraction of MS patients without interferon treatment had increased SIGLEC1 expression. In our cohort, SIGLEC1 expression on monocytes was—apart from those patients receiving interferon treatment—not significantly increased in patients with MS and NMOSD, nor were levels associated with more severe disease. SIGLEC1+ myeloid cells were abundantly present in active MS lesions as well as in a range of acute infectious and malignant diseases of the central nervous system, but not chronic MS lesions. The presence of SIGLEC1+ myeloid cells in brain lesions could be used to investigate the activity in an inflammatory CNS lesion.

www.nature.com/scientificreports/ SIGLEC1 (Sialic acid-binding immunoglobulin-type lectins-1, CD169) is a sialic acid binding cell-surface protein, exclusively expressed on monocytes and macrophages 2 . Its expression is upregulated upon contact with type I interferons and to a lesser degree, other activatory stimuli such as LPS 3,4 . As such, SIGLEC1 expression has been used as a surrogate marker for type I interferon activity in autoimmune diseases like Systemic lupus erythematosus 5 or primary Sjögren syndrome 6 as well as interferonopathies 7 and viral infections 8,9 . In the healthy brain, SIGLEC1 is only expressed by some perivascular and choroid plexus macrophages, but not by microglia 10 . In mice, mechanical insult to the brain led to an accumulation of SIGLEC1 + myeloid cells in the damaged area 10 . The cause of this accumulation, however, is unclear-either the contact of serum components or inflammatory signals with resident myeloid cells induces SIGLEC1 expression or blood-derived SIGLEC1 + myeloid cells infiltrate the CNS.
The role of type I interferons and SIGLEC1 in the pathophysiology of MS and NMOSD is not yet clear. In relapsing-remitting MS (RRMS), type I interferons are used as an immunomodulatory treatment that reduces the rate of relapses 11 . It does, however, not protect against clinical progression in the progressive forms of MS (primary progressive MS (PPMS) and secondary progressive MS (SPMS)) 11 . In a recent report, SIGLEC1 positive myeloid cells were found within MS lesions and specific ablation of SIGLEC1 expressing cells in the experimental autoimmune encephalitis (EAE) model of MS led to an amelioration of disease 12 . Additionally, plasmacytoid dendritic cells, the main type I interferon-producing cells, were found in increased frequencies in the cerebrospinal fluid of MS patients during a flare 13 . Two studies found SIGLEC1 expression to be increased on blood monocytes of MS patients, especially those with a progressive type of MS 12,14 . Taken together, there is conflicting evidence on the role of type I interferons and the interferon-induced expression of SIGLEC1 in MS as being protective or pathogenic. This is likely due to the heterogeneity of patients and disease stages and potentially also the confounding effect of interferon therapy.
In NMOSD, there is some evidence that type I interferons play a role in its pathogenesis: many patients with NMOSD have an overlap with additional, type I interferon-dependent diseases, such as SLE 15 and a recent report describes a series of patients with increased levels of endogenous or exogenous interferon α who went on to develop a seropositive NMOSD 16 . Interferon treatment does not prevent NMOSD relapses 17 and anecdotally even increases the disease activity 18 . Transcriptomic profiling of blood from NMOSD patients identified a "interferon high" signature in 16 of 38 (42.1%) of NMOSD patients 19 . This signature included the transcript for SIGLEC1 and was associated with higher disease severity. Additionally, we recently described the presence of low-density granulocytes in MS and NMOSD patients 20 , a subset of granulocytes that produce high levels of type I interferons in SLE 21 . Thus, we hypothesized that type I interferons could also play a role in the pathogenesis of NMOSD.
In this work we investigated expression of SIGLEC1 on monocytes of patients with MS and NMOSD and correlated the expression with clinical parameters. In addition, we analysed human brain tissue sections of active, inflammatory and chronic MS lesions as well as other neurological and systemic diseases.

SIGLEC1 expression is increased in a subset of patients with multiple sclerosis.
We investigated the SIGLEC1 expression on CD14 + monocytes in the peripheral blood of MS and NMOSD patients as well as in healthy controls. In most of the samples, monocyte SIGLEC1 expression was barely above the negative (fluorescence minus one) control (Fig. 1a,b). We defined a physiologic range of SIGLEC1 expression as two standard deviations from the mean SIGLEC1 expression in healthy controls (normal MFI range 0-564). Accordingly, 16/86 (18.6%) MS patients, 4/41 (9.8%) NMOSD patients and 1/31 (3.2%) healthy controls had increased SIGLEC1 levels. Almost all individuals with available longitudinal samples remained in their respective SIGLEC1 high or low category over a follow-up period of up to 1965 days. (Fig. 1c).
Treatment with type 1 interferon explains almost all increased. Next, we aimed to identify parameters that were associated with increased SIGLEC1 in MS patients. One obvious explanation would be exogenous type 1 interferon as treatment. 11/16 MS patients (68.6%) with increased SIGLEC1 expression levels received interferon beta treatment at the time of measurement (Fig. 1d). While a previous report found increased SIGLEC1 levels to be more prevalent in MS patients with a progressive form of MS, we found only 2/20 patients primary or secondary progressive MS with increased SIGLEC1 levels that were not explained by interferon treatment (Fig. 1e). SIGLEC1 expression on monocytes did not correlate with the Expanded Disability Status Scale (EDSS) (Fig. 1f), nor had it a temporal association with relapses. In the NMOSD patients, one of the four patients with increased SIGLEC1 expression had the additional diagnosis of a mixed connective tissue disease (MCTD), a disease that is also associated with increased type 1 interferon activity 22 .

SIGLEC1 expression on brain-infiltrating myeloid cells.
To investigate the presence and specificity of SIGLEC1+ myeloid cells in inflammatory MS lesion, we analysed brain tissue sections from 4 patients with MS lesions that had clinical, radiological and histological signs of activity (gadolinium uptake on MRI imaging, presence of myelin-laden macrophages), 5 patients with secondary-progressive MS (SPMS) who died of acute nonneurologic causes and with histologically classified chronic lesions and 6 patients who died of cardiovascular or multi-organ failure. In all control samples, SIGLEC1 positivity was limited to cells in the leptomeninges as well as perivascular macrophages, as previously reported 10 . In all four samples from patients with active MS lesions as well as in one patient with the Marburg variant of MS who died of the disease 14 days after onset, we found a dense infiltrate of CD68+HLA-DR+ myeloid cells that stained predominantly positive for SIGLEC1 (Fig. 2a). In contrast, in five samples from patients with secondary-progressive MS (SPMS) with histologically classified chronic lesions, we found CD68+HLA-DR+ infiltrates of varying density, but almost no SIGLEC1 expression www.nature.com/scientificreports/ ( Fig. 2a). This indicates that SIGLEC1 expression on myeloid cells could be used to distinguish active inflammatory lesions from chronic lesions.
To corroborate this hypothesis, we studied the SIGLEC1 expression in the brain tissue of 8 patients with other inflammatory neurological diseases. SIGLEC1+ myeloid infiltrates were present in patients with glioblastoma (n = 2), herpes simplex encephalitis (n = 2) and cerebral infarction (n = 2), as well as toxoplasmosis (n = 1), cerebral abscess (n = 2) and post-transplant lymphoproliferative disorders (n = 1, PTLD, Fig. 2b). No increase in SIGLEC1+ cells was noted in the brain of a patient who died of amyotrophic lateral sclerosis (ALS).
In summary, SIGLEC1+ is expressed on brain-infiltrating myeloid cells in a broad range of active inflammatory lesions, but not in chronic MS lesions.

Discussion
Here, we report that, after accounting for interferon treatment, patients with multiple sclerosis (MS) and neuromyelitis optica spectrum disorder (NMOSD) did not have increased expression of SIGLEC1 on monocytes in the peripheral blood.
We identified SIGLEC1+ on CD68+ HLA-DR+ myeloid cells in active inflammatory MS lesions and a range of other inflammatory, infectious or malignant brain lesions. SIGLEC1+ expression was low on myeloid cells in chronic MS lesions of SPMS patients, indicating that SIGLEC1+ myeloid cells could serve as a marker of inflammatory activity. These findings are in line with a previous report that found SIGLEC1+ MHC-II+ myeloid cells to be present in mice after retinal transplantation, but not in the healthy retina or retinal degeneration 23 . Future studies will be needed to investigate, whether the increase in SIGLEC1+ cells in inflammatory brain lesions is due to the infiltration of SIGLEC1+ cells from the peripheral blood or an upregulation of SIGLEC1 in tissueresident microglia, as well as define the inflammatory signals leading to such an upregulation. We propose that SIGLEC1+ can serve as a marker to differentiate active from inactive MS lesions, which will require additional validation for use in routine histopathological diagnostics.
Our results on SIGLEC1 expression on blood monocytes are in contrast to two previous reports which identified increased levels of SIGLEC1 in cohorts of 44 MS patients 12,14 . The study by Malhotra et al. 14 focussed on   12 again identified small (approximately twofold) increases in SIGLEC1 expression on monocytes in 57 MS patients and found that these increases were independent of progressive disease and also of interferon treatment, an observation that is at odds with our own findings and biological plausibility. This disparity might be due to different reasons: it could be that the authors mostly observed small differences between individuals that would have all been classified as "SIGLEC1 low expressors" in our study, as the approximately twofold-differences they describe are significantly smaller compared to the five-tenfold increases in SIGLEC1 MFI in "SIGLEC1 high expressors" in our study. Different studies investigated the SIGLEC1 expression using different antibody clones which could also explain some of the differences; in this study, we used the same clone as previous studies that investigated SIGLEC1 in rheumatologic diseases 5,6 .
Additionally, our data are at odds with a previous report that found a type 1 interferon signature in NMOSD patients that was conducted with patients from the same patient cohort 19 . While 42% of NMOSD cohorts in the study by Agasing et al. showed an "high type 1 interferon" signature, we found < 10% of NMOSD patients had increased type 1 interferon activity as measured by SIGLEC1 expression; a result that is neither significantly different from healthy controls, nor likely of clinical consequence. As the results of whole blood bulk transcriptomic data are often subject to unforeseen external influences and changes in the cellular composition of the blood, they should usually be validated by protein level data.
Even though our data are not in agreement with multiple, previously published results, the validity of our assay is supported by the internal plausibility control of interferon treated MS patients. However, one significant weakness of our own study is the relative clinical quiescence of the patients enrolled; most did not have a relapse within the last six months before the measurement and most did not have a change in EDSS in longitudinal sampling. Additionally, most patients were under immunosuppressive treatment. Future studies in MS and NMOSD patients during a clinical relapse could thus still discover the presence of SIGLEC1 expressing monocytes in the peripheral blood.
In summary, our data indicate that type I interferons or SIGLEC1 expressing cells in the peripheral blood do not play a major role in the pathogenesis of most patients with stable NMOSD or MS, however SIGLEC1 + myeloid cells in the brain are present in active inflammatory MS lesions as well as in other inflammatory neurological diseases of the CNS.

Methods
Cohort. We analysed frozen peripheral blood mononuclear cells (PBMCs) from a total of 86 MS patients, 41 NMOSD patients and 31 healthy controls (HCs) that were included in observational cohort studies for MS and NMOSD at the NeuroCure Clinical ResearchCenter, Charité-Universitätsmedizin Berlin. PBMCs from all patients with sufficient available material were included in the study with no prior selection according to disease activity, treatment, age or sex. All MS patients fulfilled the 2017 revised McDonald criteria 24 , while the NMOSD patients fulfilled the 2015 Wingerchuck international consensus diagnostic criteria 25 . Further patient characteristics are provided in Table 1.
Histology of human brain tissue. We investigated archived cryo-and formalin preserved biopsy and autopsy tissue from patients who had been diagnosed in the Department of Neuropathology, Charité-Universitätsmedizin Berlin with inflammatory demyelination of the central nervous system (CNS) consistent with multiple sclerosis or other inflammatory, malignant or infectious diseases of the CNS (other neurological disease, OND). Tissue that was categorized as active MS lesions stemmed from patients with clinical and MRI evidence of active disease with gadolinium contrast agent uptake and/or perifocal edema in T2-weighted sequences. These tissue samples were all derived from diagnostic biopsies that were performed to diagnose an unclear tumefactive lesion that was diagnosed as inflammatory demyelinating lesion. After follow-up, these patients were then clinically diagnosed with MS. Histologically these lesions showed all active demyelination with myelin laden macrophages. In contrast, the patients with lesions categorized as chronic MS lesions were patients diagnosed with SPMS that had died of a non-neurologic cause and where only chronically active or inactive lesions with lack of myelin-laden macrophages were analysed 27 . Patients who died of cardiovascular cases or multi-organ failure served as controls. All patients with multiple sclerosis fulfilled clinical diagnosis criteria according to the 2017 revised McDonald criteria 24 and none of the patients were treated with interferon.
Immunohistology of human brain tissue. All  for Mac). A normal range for SIGLEC1 expression in FACS expression was calculated as two standard deviations from the mean SIGLEC1 expression in healthy controls. When multiple measurements of the same individual were available, the mean of the measurements was calculated and used for comparative analysis. The Chi-squared test or Kruskal-Wallis test with Dunn's correction was used to compare between the different groups for binary and continuous variables respectively.
Ethics approval and consent to participate. The analysis of blood was approved by the ethics committee of the Charité-Universitätsmedizin Berlin (MS: EA1/163/12; NMOSD: EA1/041/14) and informed consent was obtained from all participants in the study. The study of brain histology was approved approved by the ethics committee of the Charité-Universitätsmedizin (EA1/078/16). This study was performed in accordance with all relevant guidelines and regulations, such as the Declaration of Helsinki and the Good Clinical Practice guidelines of the International Council for Harmonization.

Data availability
All datasets used and analysed during the current study are available from the corresponding authors on request.