- Email: [email protected]
Detection of serum soluble HLA-G levels in patients with acute ischemic stroke: A pilot study
Human Immunology xxx (xxxx) xxx–xxx
Contents lists available at ScienceDirect
Human Immunology journal homepage: www.elsevier.com/locate/humimm
Detection of serum soluble HLA-G levels in patients with acute ischemic stroke: A pilot study Enrico Fainardia, , Daria Bortolottib, Massimiliano Castellazzic, Ilaria Casettad, Tiziana Bellinid,e, Roberta Rizzob ⁎
a
Neuroradiology Unit, Department of Experimental and Clinical Biomedical Sciences “Mario Serio”, University of Florence, University Hospital Careggi, Florence, Italy Section of Microbiology and Medical Genetics, Department of Medical Sciences, University of Ferrara, Ferrara, Italy c Section of Neurological, Psychiatric and Psychological Sciences, Department of Biomedical and Specialist Surgical Sciences, University of Ferrara, Ferrara, Italy d Section of Medical Biochemistry, Molecular Biology and Genetics, Department of Biomedical and Specialist Surgical Sciences, University of Ferrara, Ferrara, Italy e University Center for Studies on Gender Medicine, University of Ferrara, Ferrara, Italy b
ARTICLE INFO
ABSTRACT
Keywords: Acute ischemic stroke HLA-G Serum levels Clinical outcome
The aim of this study was to investigate the potential role of soluble Human Leukocyte Antigen-G (sHLA-G) molecules as biomarkers predicting outcome in acute ischemic stroke (AIS). Serum levels of total sHLA-G (sHLAG1/HLA-G5) and its soluble isoforms sHLA-G1 and HLA-G5/G6 were measured by enzyme-linked immunosorbent assay (ELISA) in 92 AIS patients and healthy donors (HD). Incidence of hemorrhagic transformation (HT), size of final infarct volume (FIV) and clinical outcome at 3 months were recorded in AIS patients. Detectable serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 were present in a small proportion of AIS patients (26.1%, 17.4% and 16.3%, respectively) and HD (12.5%, 10.7% and 10.7%, respectively) and were more elevated in AIS patients without HT than in those with HT (p < 0.01; p < 0.05; p < 0.01, respectively). HT was less frequent (p < 0.01) in AIS patients with measurable serum concentrations of sHLA-G1/HLA-G5 and HLA-G5/G6. Serum levels of sHLA-G1/HLA-G5 and sHLA-G1 were inversely correlated to FIV (p < 0.02), whereas good outcome was more common (p < 0.01) in AIS patients with detectable serum concentrations of sHLA-G1/HLA-G5. Taken together, these findings suggest that total sHLA-G could exert a protective effect in a subset of AIS patients, irrespective of its soluble isoforms sHLA-G1 and HLA-G5/G6, and indicate that the prognostic value of serum levels of sHLA-G remains to be established.
1. Introduction Acute ischemic stroke (AIS) accounts for 85% of all cerebrovascular diseases and represents a major cause of death and permanent disability worldwide [1]. Reperfusion therapies, such as intravenous thrombolysis and endovascular treatment, have recently been introduced in clinical practice as standard of care because they have demonstrated to be effective in improving outcome of patients with AIS [2]. However, many patients do not achieve a good clinical outcome after successful recanalization due to the lack of reliable clinical and radiological criteria for selection and outcome prediction of AIS patients [3]. Therefore, as an intense inflammatory response is believed to be implicated in evolution of ischemic lesions, in the last year the potential prognostic value of measurable blood biomarkers has been largely explored [4]. Nevertheless, although several molecule have been tested, the
identification of blood biomarkers predicting outcome in AIS patients still remains inconclusive [4,5]. Human Leukocyte Antigen-G (HLA-G) are immunonologically functional non-classical HLA-I antigens, structurally related to classical class Ia HLA products (HLA-I), which display a limited polymorphism, a restricted tissue distribution and alternative splicing of their primary transcript encoding seven distinct isoforms which include four membrane bound (G1, G2, G3 and G4) and three soluble (G5, G6, G7) proteins [6]. Membrane-bound HLA-G1 and soluble HLA-G5/G6 (HLA-G5/G6) are the most expressed HLA-G isoforms [6] and are currently considered the most important functionally active soluble HLA-G isoforms [7]. HLA-G5/G6 molecules are actively secreted as soluble isoforms, whereas soluble HLA-G1 (sHLA-G1) proteins are proteolytically released after shedding from cell surface [6,7]. In this setting, HLA-G disulfide-linked dimers are presumed to be more biologically active than HLA-G monomers [8]. These soluble HLA-G
⁎ Corresponding author at: Struttura Organizzativa Dipartimentale di Neuroradiologia, Dipartimento di Scienze Biomediche, Sperimentali e Cliniche “Mario Serio”, Università degli Studi di Firenze, Ospedale Universitario Careggi, Largo Brambilla 3, 50134 Firenze, Italy. E-mail address: [email protected] (E. Fainardi).
https://doi.org/10.1016/j.humimm.2019.11.004 Received 18 July 2019; Received in revised form 31 October 2019; Accepted 6 November 2019 0198-8859/ © 2019 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Enrico Fainardi, et al., Human Immunology, https://doi.org/10.1016/j.humimm.2019.11.004
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
(sHLA-G) molecules are at present postulated to have tolerogenic properties in some autoimmune, neoplastic and viral neurological disorders [9]. In particular, the anti-inflammatory function of sHLA-G antigens was particularly documented in Multiple Sclerosis (MS), the prototypic brain autoimmune disease, in which elevated levels of intrathecally produced HLA-G5/G6 in dimeric form were associated with magnetic resonance imaging (MRI) inactive disease, suggesting a role of these antigens in the resolution of MS inflammatory response within the brain [10–12], probably under the influence of the individual genetic background represented by HLA-G polymorphisms [13]. However, no data are currently available on level of sHLA-G concentrations in AIS patients. For these reasons, in this study we aimed to investigate whether serum sHLA-G levels are detectable in AIS patients and their possible association with clinical outcome.
HLA-G held in Paris, 2003. In particular, the antibody MEM-G/9 reacts with native form of human HLA-G1 on the cell surface as well as with soluble HLA-G5 isoform in its beta2-microglobulin associated form. The antibody 5A6G7 was generated to a peptide corresponding to C-intron 4-encoded sequence. This antibody does not crossreact with the fulllength HLA-G1 isoform and thus allow to distinguish between secreted HLA-G5 and HLA-G6 isoforms from shedded HLA-G1. After ELISA measurements of serum values of sHLA-G1/HLA-G5 and HLA-G5/G6, the amount of serum sHLA-G1 was calculated by the mathematical the difference between sHLA-G1/HLA-G5 and HLA-G5/G6 concentrations. Serum sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 levels were expressed in ng/ml. The intra-assay coefficient of variations (CV), the inter-assay CV and the limit of sensitivity were 1.4%, 4% and 1 ng/ml for sHLA-G1/HLA-G5 determination, 2%, 3.5% and 1 ng/ml for HLAG5/G6 assay. In summary, total sHLA-G concentrations, given by the sum of sHLA-G1 and HLA-G5 amounts, was defined as sHLA-G1/HLAG5 levels, amounts of HLA-G5/G6 soluble isoforms was referred as HLA-G5/G6 levels and the mathematical difference between sHLA-G1/ HLA-G5 and HLA-G5/G6 concentrations was indicated as sHLA-G1 levels.
2. Patients and methods 2.1. Subjects Serum samples were prospectively collected in 92 AIS patients admitted within 4.5 h from onset receiving intravenous fibrinolysis with recombinant tissue-plasminogen activator (t-PA) according to the current recommended guidelines [2]. Inclusion criteria were: 1) presentation at the hospital within 4.5 h from symptom onset; 2) baseline and follow-up confirmation of hemispheric AIS by clinical assessment and brain non-contrast computerized tomography (NCCT) scans; 3) baseline and follow-up CT imaging carried out at onset and at 24 h after admission CT. Exclusion criteria included: 1) detection of intracerebral hemorrhage at admission brain NCCT; 2) evidence of lacunar or brain stem infarct 3) evidence of undetermined or posterior circulation stroke; 4) previous strokes with residual deficit; 5) pregnancy; 6) age < 18 years. Disease severity at onset was scored using the National Institutes of Health Stroke Scale (NIHSS) [14]. Clinical outcome was measured by the modified Rankin scale (mRS) at 3 months. mRS ≤ 2 and > 2 were defined as good and poor outcomes, respectively [15]. Hemorrhagic transformation (HT) was classified on NCCT at 24 h post ictus according to the European Cooperative Acute Stroke Study (ECASS)-II criteria [16] into four different categories: hemorrhagic infarction type 1 (HI-1), HI type 2 (HI-2), Parenchymal hemorrhage type 1 (PH-1), and PH type 2 (PH-2). Final infarct volume (FIV) was measured on follow-up NCCT at 24 h after symptom onset with a multi-slice planimetric method by summation of the hypodense areas, manually traced on each slice in which they were detectable, multiplied by slice thickness [17]. Fifty-six healthy donors (HD) served as controls. Clinical assessment, CT scans and blood sampling were performed during routine clinics, with written informed consent and local ethical board approval.
2.4. Statistics After checking data for normality by using the Kolmogorov-Smirnov test, continuous variables were compared using Mann-Whitney U test, whereas their correlations were assessed by linear regression or Spearman's rank correlation coefficient. Categorical variables were compared by means of Chi-square test. A value of p < 0.05 was assumed as statistically significant. 3. Results 3.1. Clinical and radiological findings Demographic, clinical and radiological data recorded in our AIS patients are listed in Table 1. In our study population, a good outcome (mRS ≤ 2) was achieved in 72/92 (78.3%) cases, whereas a poor outcome (mRS > 2) was present in 20/92 (21.7%) patients. HT was observed in 28/92 (30.4%) patients (13 HI and 15 PH) appearing 24 h after stroke onset. 3.2. Serum levels of total sHLA-G, HLA-G5/G6 and sHLA-G1 in AIS patients and controls Detectable serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 were more frequent in AIS patients than in HD, without any statistical difference between the two groups (Table 1). As reported in Fig. 1, serum concentrations of sHLA-G1/HLA-G5 were higher in AIS than in HD (p < 0.05), while serum titers of HLA-G5/G6 and sHLA-G1 were statistically equivalent between AIS and HD, even if they resulted greater in AIS. As reported in Table 2, serum levels of sHLA-G1/HLAG5, HLA-G5/G6 and sHLA-G1 were more elevated in AIS patients without HT than in those with HT (p < 0.01; p < 0.05; p < 0.01, respectively). Conversely no significant differences were found between AIS patients with good and poor outcome for serum levels of sHLA-G1/ HLA-G5, HLA-G5/G6 and sHLA-G1 (data not shown).
2.2. Handling of serum samples Serum samples were obtained from centrifugation of blood specimens withdrawn by puncture of an anterocubital vein, collected under sterile conditions at admission, stored in aliquots at −80 °C until assay and measured under exactly the same conditions. 2.3. ELISA assay for sHLA-G1 and HLA-G5/G6 isoforms As previously described [11], serum levels of total sHLA-G (sHLAG1/HLA-G5) and HLA-G5/G6 were measured by enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies MEM-G/9 and MoAb 5A6G7 (Exbio, Praha, Czech Republic) as capture antibodies, respectively, and anti-beta2 microglobulin MoAb-peroxidase (HRP) conjugated and biotinylated anti-HLA class I MoAb W6/32 (Exbio, Prague, Czech Republic) as detection antibody, respectively. As reported in Essen Workshop for sHLA-G quantification [18,19], these antibodies were validated during the third International Workshop on
3.3. Radiological and clinical characteristics in AIS patients with and without measurable serum levels of total sHLA-G, HLA-G5/G6 and sHLAG1 When AIS patients were stratified according to detectable levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 (Fig. 2), HT was more frequent (p < 0.01) in AIS patients without measurable serum concentrations of sHLA-G1/HLA-G5 (sHLA-G1/HLA-G5neg) and HLA-G5/ G6 (HLA-G5/G6neg) than in those with detectable serum levels of sHLA2
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
Table 1 Demographic, clinical and radiological characteristics, and measurable levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in patients with acute ischemic stroke (AIS) and healthy donors (HD).
Sex: Male/Female Age, years: mean ± SD NIHSS at entry: median, IQR, mean ± SD, range mRS at 3 months: median, IQR, mean ± SD, range Time between symptom onset and NCCT scan at admission (hours): median, IQR, mean ± SD, range Involved arterial territory at admission: n/total (%) Right MCA Left MCA Final infarct volume (ml) on NCCT at 7 day: median, IQR, mean ± SD, range Detectable levels (%) sHLA-G1/HLA-G5 HLA-G5/G6 sHLA-G1
AIS = 92
HD = 56
p value
50/40 65.4 ± 12.9 10, 6–14, 11.0 ± 6.2, 1–25 1, 1–2, 1.5 ± 1.1, 0–5 2, 1.3–3.0, 2.1 ± 1.0, 0.3–4.3
34/22 62.1 ± 9.9 / / /
p = 0.45a p = 0.09b
40/92 (43.5%) 52/92 (56.5%) 8, 0.8–37.5, 42.0 ± 78.3, 0–458
/ / /
24/92 (26.1%) 16/92 (17.4%) 15/92 (16.3%)
7/56 (12.5%) 6/56 (10.7%) 6/56 (10.7%)
p = 0.05b p = 0.27b p = 0.34b
SD = Standard deviation; NIHSS = National Institutes of Health Stroke Scale; IQR = Interquartile range; mRS = modified Rankin Scale: NCCT = non-contrast CT; MCA = middle cerebral artery. aChi-square; bMann-Whitney
3.4. Correlations between serum levels of total sHLA-G, HLA-G5/G6 and sHLA-G1 and AIS radiological and clinical features Serum concentrations of sHLA-G1/HLA-G5 and sHLA-G1 were inversely correlated to FIV (p < 0.02), without any definite associations between serum levels of HLA-G5/G6 and FIV. No relationships were found between serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 and mRS (Table 3). 4. Discussion 4.1. Interpretation of result To our knowledge, this is the first study investigating serum levels of sHLA-G molecules in AIS patients who currently lack convincing blood biomarkers predicting outcome [4,5]. These antigens seem to be implicated in the termination of autoimmunity in MS, acting as anti-inflammatory molecules. These hypothesis is supported by the evidence that high amounts of HLA-G5/G6 in dimeric form were associated with MRI inactive disease [10–12]. More important, serum concentrations of sHLA-G proteins were greater in patients with MS than in those with other non-inflammatory and inflammatory neurological disorders [20]. Based on these considerations, we tested the potential relevance of serum levels of sHLA-G antigens as prognostic factor in AIS patients. In this way, we examined the relationship of serum concentrations of sHLA-G molecules with both tissue outcome assessed by FIV and clinical outcome evaluated by mRS. In addition, we explored the possible association between serum values of sHLA-G and HT, representing one of the most dangerous complications of thrombolytic therapy and a factor affecting outcome [21]. Our results indicate that serum levels of total sHLA-G (sHLA-G1/HLA-G5) and its soluble isoforms sHLA-G1 and HLA-G5/G6 are detectable in a small proportion of AIS patients and HD. In addition, levels these antigens differed between these two groups only for serum sHLA-G/HLA-G5 titers, which were slightly
Fig. 1. Serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in 92 acute ischemic stroke (AIS) patients and 56 healthy donors (HD). Serum sHLA-G1/ HLA-G5, HLA-G5/G6 and sHLA-G1 were higher in AIS than in HD, with statistically significant values only for sHLA-G1/HLA-G5 (p < 0.05). Median, interquartile range (IQR), mean ± standard deviation (SD) and range were 0, 0–0.9, 6.9 ± 15.8 and 0–78.1 ng/ml for sHLA-G1/HLA-G5, 0, 0–0, 3.7 ± 9.6 and 0–50.6 ng/ml for HLA-G5/G6, 0, 0–3, 3.3 ± 9.6 and 0–76.7 ng/ml for sHLA-G1 in AIS patients. Median, IQR, mean ± SD and range were 0, 0–0.9, 1.0 ± 3.2 and 0–16.0 ng/ml for sHLA-G1/HLA-G5, 0–0, 1.0 ± 2.9, 0–12.0 ng/ ml for HLA-G5/G6, and 0, 0–0 0.3 ± 1.6 and 0–11.0 ng/ml for sHLA-G1 in HD.
G1/HLA-G5 (sHLA-G1/HLA-G5pos) and HLA-G5/G6 (HLA-G5/G6pos). On the contrary, HT rates did not significantly differ between AIS patients with (HLA-G1pos) and without (HLA-G1neg) measurable serum titers of sHLA-G1. Favorable outcome was more represented (p < 0.01) in sHLA-G1/HLA-G5pos than in sHLA-G1/HLA-G5neg AIS patients (Fig. 3), without significant differences between HLA-G5/G6pos and HLA-G5/G6neg, as well as between HLA-G1pos and HLA-G1neg AIS patients (data not shown).
Table 2 Serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in patients with acute ischemic stroke (AIS) categorized according to hemorrhagic transformation (HT).
Serum sHLA-G1/HLA-G5 (ng/ml) median, IQR, mean ± SD, range Serum HLA-G5/G6 (ng/ml) median, IQR, mean ± SD, range Serum sHLA-G1 (ng/ml) median, IQR, mean ± SD, range
HT (n=28)
no HT (n=64)
p value (Mann- Whitney)
0, 0-0, 1.4 ± 7.2, 0-39.2
0, 0-11.2, 9.6 ± 18.1, 0-78.1
p < 0.01
0, 0-0, 1.0 ± 5.7, 0-31.3
0, 0-0, 4.9 ± 10.8, 0-50.6
p < 0.05
0, 0-0, 0.4 ± 1.6, 0-7.9
0, 0-5.2, 4.7 ± 11.4, 0-76.7
p < 0.01
IQR = interquartile range; SD = standard deviation. 3
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
Fig. 2. Distribution of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) categorized according to detectable serum levels of sHLA-G1/HLA-G5, HLA-G5 and sHLA-G1. HT was more frequent (p < 0.01) in sHLA-G1/HLA-G5neg (28/69; 40.6%) than in sHLA-G1/HLA-G5pos (2/23; 8.7%) AIS patients and in HLA-G5/G6neg (29/77; 37.7%) than in HLA-G5/G6pos (1/15; 6.7%) AIS patients, whereas there were no differences between sHLA-G1pos (25/78; 32.1%) sHLA-G1neg (5/14; 35.7%) AIS patients.
further confirmed by the inverse correlation between serum sHLA-G/ HLA-G5 and sHLA-G1 levels and FIV, suggesting that infarct size was smaller when serum concentrations of sHLA-G proteins were higher. Consistent with these findings was also the demonstration that a favourable outcome was more frequently achieved in AIS patients with detectable levels of sHLA-G1/HLA-G5. Overall, these results argue for an apparent protective role sHLA-G molecules in AIS patients, without a clear predominance of a particular isoform. However, as the detection of these molecules in serum was restricted to a subset of patients, the use of serum levels of sHLA-G antigens collected at admission as prognostic biomarker in AIS is difficult to be proposed at this time. Moreover, in absence of data on cellular expression, function and activity, the mechanisms leading to the release of these antigens in peripheral blood of AIS patients can only be hypothesized. Cellular sources of sHLA-G proteins are likely represented by monocytes and macrophages which are strongly activated in the acute phase of stroke and subsequently migrate into the brain across the blood-brain barrier (BBB) and accumulate within the infarct and peri-infart tissue [22]. Elevated serum levels of sHLA-G1 could be due to an increase in membrane shedding promoted by matrix metalloproteinase-2 (MMP-2) via proteolytic cleavage [23]. In fact, a MMP-2 overexpression occurred in AIS where this gelatinase probably contributes to BBB breakdown facilitating development of vasogenic edema in ischemic tissue and HT [24]. On the contrary, factors generating the secretion of HLA-G5/G6 antigens by peripheral blood monocytes [25] remain currently unknown, but they are probably produced as a part of the inflammatory reaction associated to AIS [22,24]. In any case, as the measurements of serum sHLA-G levels in AIS patients did not reveal any predominant isoform, the increase in serum concentrations of total sHLA-G and its soluble isoforms sHLA-G1 and HLA-G5/G6 observed in these patients
Fig. 3. Distribution of favourable outcome (mRS ≤ 2) in patients with acute ischemic stroke (AIS) categorized according to detectable serum levels of sHLAG1/HLA-G5, HLA-G5/G6 and sHLA-G1. A good outcome was more frequent (p < 0.01) in sHLA-G1/HLA-G5pos (21/23; 91.3%) than in sHLA-G1/HLAG5neg (54/76; 71.1) AIS patients.
higher in AIS patients than in HD. More interesting was the analysis of AIS patients showing that the presence of both measurable and elevated serum levels of sHLA-G1/HLA-G5 and its soluble isoforms sHLA-G1 and HLA-G5/G6 was associated with low rates of HT, with a greater extent for HLA-G5/G6. This beneficial effect of sHLA-G molecules in AIS was
Table 3 Correlations between serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 and final infarct volume (FIV) in patients with acute ischemic stroke (AIS).
Serum sHLA-G1/HLA-G5 (ng/ml) Serum HLA-G5/G6 (ng/ml) Serum sHLA-G1 (ng/ml)
FIV
p value (Spearman)
mRS
p value(Spearman)
r = -242299 r = -0.1012; r = -249854
p < 0.02 p = 0.1685 p < 0.02
r = -0.144 r = -0.0678 r = -0.1554
p = 0.0855 p = 0.2604 p = 0.0695
4
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
might merely represent an indirect manifestation of inflammatory immune activation taking place in AIS, rather than a response directed to a specific target. The low incidence of HT found in AIS patients with both detectable and high serum levels of total sHLA-G, sHLA-G1 and HLAG5/G6 suggests that sHLA-G antigens could limit HT acting as procoagulant molecules. The concurrent increase in the expression of HLAG and pro-coagulation factors induced by preimplantation factor (PIF) recently reported in JEG-3 cytotrophoblast cells [26] may support this speculation. On the other hand, the beneficial effect exerted by total sHLA-G and sHLA-G1 on tissue and clinical outcome may be mediated by anti-inflammatory properties of these molecules through the inhibition of CD4+ T cell activation [27] and the suppression of CD8+ cytotoxic T effector cells [28]. In fact, both CD4+ and CD8+ T cells are considered to have a central role in inflammatory response occurring in AIS [22,29].
Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper References [1] G.J. Hankey, Stroke. Lancet 389 (2017) 641–654. [2] W.J. Powers, A.A. Rabinstein, T. Ackerson, O.M. Adeoye, N.C. Bambakidis, K. Becker, et al., Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association, Stroke 49 (2018) (2018) e46–e110. [3] M. Goyal, Poor clinical outcome despite successful arterial recanalization. What went wrong? How can we do better? Neuroradiology 52 (2010) 341–343. [4] C.C. Esenwa, M.S. Elkind, Inflammatory risk factors, biomarkers and associated therapy in ischaemic stroke, Nat. Rev. Neurol. 12 (2016) 594–604. [5] G.J.L. Ng, A.M.L. Quek, C. Cheung, T.V. Arumugam, R.C.S. Seet, Stroke biomarkers in clinical practice: A critical appraisal, Neurochem. Int. 107 (2017) 11–22. [6] E.D. Carosella, N. Rouas-Freiss, D. Tronik-Le Roux, P. Moreau, J. LeMaoult, HLA-G: an immune checkpoint molecule, Adv. Immunol. 127 (2015) 33–144. [7] J. LeMaoult, M. Le Discorde, N. Rouas-Freiss, P. Moreau, C. Menier, J. McCluskey, et al., Biology and functions of human leukocyte antigen-G in health and sickness, Tissue Antigens 62 (2003) 273–284. [8] S. Zilberman, C. Schenowitz, S. Agaugué, F. Benoît, B. Riteau, R. Rouzier, et al., HLA-G1 and HLA-G5 active dimers are present in malignant cells and effusions: the influence of the tumor microenvironment, Eur. J. Immunol. 42 (2012) 1599–1608. [9] F. Morandi, R. Rizzo, E. Fainardi, N. Rouas-Freiss, V. Pistoia, Recent Advances in our understanding of HLA-G Biology: Lessons from a wide spectrum of human diseases, J Immunol. Res. 2016 (2016) 4326495. [10] E. Fainardi, R. Rizzo, L. Melchiorri, L. Vaghi, M. Castellazzi, A. Marzola, et al., Presence of detectable levels of soluble HLA-G molecules in CSF of relapsing-remitting multiple sclerosis: relationship with CSF soluble HLA-I and IL-10 concentrations and MRI findings, J. Neuroimmunol. 142 (2003) 149–158. [11] E. Fainardi, R. Rizzo, L. Melchiorri, M. Stignani, M. Castellazzi, M.L. Caniatti, et al., Soluble HLA-G molecules are released as HLA-G5 and not as soluble HLA-G1 isoforms in CSF of patients with relapsing-remitting multiple sclerosis, J Neuroimmunol 192 (2007) 219–225. [12] E. Fainardi, D. Bortolotti, S. Bolzani, M. Castellazzi, C. Tamborino, G. Roversi, et al., Cerebrospinal fluid amounts of HLA-G in dimeric form are strongly associated to patients with MRI inactive multiple sclerosis, Mult Scler 22 (2016) 245–249. [13] R. Rizzo, D. Bortolotti, N.B. Fredj, A. Rotola, F. Cura, M. Castellazzi, et al., Role of HLA-G 14bp deletion/insertion and +3142C > G polymorphisms in the production of sHLA-G molecules in relapsing-remitting multiple sclerosis, Hum Immunol 73 (2012) 1140–1146. [14] T. Brott, H.P. Adams, C.P. Olinger, J.R. Marler, W.G. Barsan, J. Biller, et al., Measurements of acute cerebral infarction: a clinical examination scale, Stroke 20 (1989) 864–870. [15] J.C. van Swieten, P.J. Koudstaal, M.C. Visser, H.J. Schouten, J. van Gijn, Interobserver agreement for the assessment of handicap in stroke patients, Stroke 19 (1988) 604–607. [16] W. Hacke, M. Kaste, C. Fieschi, R. von Kummer, A. Davalos, D. Meier, et al., Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second EuropeanAustralasian Acute Stroke Study Investigators, Lancet 352 (1998) 1245–1251. [17] T. Brott, J.R. Marler, C.P. Olinger, H.P. Adams, T. Tomsick, W.G. Barsan, et al., Measurements of acute cerebral infarction: lesion size by computed tomography, Stroke 20 (1989) 871–875. [18] V. Rebmann, J. LeMaoult, N. Rouas-Freiss, E.D. Carosella, H. Grosse-Wilde, Report of the Wet Workshop for quantification of soluble HLA-G in Essen, Hum. Immunol. 66 (2005) (2004) 853–863. [19] V. Rebmann, J. LeMaoult, N. Rouas-Freiss, E.D. Carosella, H. Grosse-Wilde, Quantification and identification of soluble HLA-G isoforms, Tissue Antigens 69 (2007) 143–149. [20] F. Morandi, C. Venturi, R. Rizzo, M. Castellazzi, E. Baldi, M.L. Caniatti, et al., Intrathecal soluble HLA-E correlates with disease activity in patients with multiple sclerosis and may cooperate with soluble HLA-G in the resolution of neuroinflammation, J. Neuroimmune Pharmacol. 8 (2013) 944–955. [21] G. Lu, Q. He, Y. Shen, F. Cao, Potential biomarkers for predicting hemorrhagic transformation of ischemic stroke, Int. J. Neurosci. 128 (2018) 79–89. [22] R. Macrez, C. Ali, O. Toutirais, B. Le Mauff, G. Defer, U. Dirnagl, et al., Stroke and the immune system: from pathophysiology to new therapeutic strategies, Lancet Neurol. 10 (2011) 471–478. [23] R. Rizzo, A. Trentini, D. Bortolotti, M.C. Manfrinato, A. Rotola, M. Castellazzi, L. Melchiorri, et al., Matrix metalloproteinase-2 (MMP-2) generates soluble HLA-G1 by cell surface proteolytic shedding, Mol. Cell Biochem. 381 (2013) 243–255. [24] J. Castillo, I. Rodríguez, Biochemical changes and inflammatory response as markers for brain ischaemia: molecular markers of diagnostic utility and prognosis in human clinical practice, Cerebrovasc Dis. 17 (Suppl 1) (2004) 7–18. [25] V. Rebmann, A. Busemann, M. Lindemann, H. Grosse-Wilde, Detection of HLA-G5 secreting cells, Hum. Immunol. 64 (2003) 1017–1024. [26] M.S. Hakam, J.M. Miranda-Sayago, S. Hayrabedyan, K. Todorova, P.S. Spencer, A. Jabeen, et al., Preimplantation factor (PIF) promotes HLA-G, -E, -F, -C Expression
4.2. Limitations of the study This study is affected by several limitations. First, the relatively small sample size could weaken the consistency of our data. In particular, AIS patients with no HT and favourable outcome were more represented than those with no HT and unfavourable outcome and, therefore, the comparison between these groups could result unbalanced. Second, information about HLA-G cellular expression, secretion and shedding and genetic polymorphisms are needed for better understanding and interpreting the association between serum levels of sHLA-G molecules and AIS prognostic parameters found in this study. Third, functional experiments are requested for establishing the actual influence of HLA-G antigens on pro-inflammatory cells involved in AIS. Fourth, the relationship between HLA-G molecules and coagulation cascade should be clarified. Finally, the knowledge of temporal pattern of serum concentrations of sHLA-G antigens in relation to clinical and radiological evolution of ischemic lesion could be useful to verify the prognostic value of sHLA-G serum measurements in AIS. Fifth, the recent demonstration that novel HLA-G isoforms detectable by 5A6G7 antibody exist [30,31] opens new avenues not only for the investigation of HLA-G5/G6 typical isoform, but also for the analysis other less studied isoforms that could play a role in immune-system control in different diseases, including AIS. 4.3. Conclusions In summary, this pilot study indicates that detectable and elevated serum levels of total sHLA-G are associated with a lower occurrence of HT and a more frequent tissue and clinical good outcomes in a subset of AIS patients, irrespective of serum concentrations of its isoforms (sHLAG1 and HLA-G5/G6). These findings suggest that sHLA-G antigens could be implicated as protective molecules in mechanisms regulating AIS inflammatory response. Future studies in a larger patient population are warranted to verify whether sHLA-G proteins can really exert anti-inflammatory functions and can represent a potential biomarkers predicting outcome in AIS. Author Contribution Study concept and design: RR, EF. Acquisition of data: DB, MC, IC, RR. Laboratory analysis: DB, RR. Statistical analysis: DB, IC, RR, EF. Analysis and interpretation of data: all authors Drafting of the manuscript: DB, TB, RR, EF Critical revision of the manuscript for important intellectual content: all authors. Study supervision: RR, EF. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. 5
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al. in JEG-3 choriocarcinoma cells and endogenous progesterone activity, Cell Physiol Biochem 43 (2017) 2277–2296. [27] N. Lila, N. Rouas-Freiss, J. Dausset, A. Carpentier, E.D. Carosella, Soluble HLA-G protein secreted by allo-specific CD4+ T cells suppresses the allo-proliferative response: a CD4+ T cell regulatory mechanism, Proc. Natl. Acad. Sci. USA 98 (2001) 12150–12155. [28] P. Contini, M. Ghio, A. Poggi, G. Filaci, F. Indiveri, S. Ferrone, et al., Soluble HLAA,-B,-C and -G molecules induce apoptosis in T and NK CD8+ cells and inhibit cytotoxic T cell activity through CD8 ligation, Eur. J. Immunol. 33 (2003) 125–134.
[29] A. Bonaventura, L. Liberale, A. Vecchié, M. Casula, F. Carbone, F. Dallegri, et al., Update on inflammatory biomarkers and treatments in ischemic stroke, Int. J. Mol. Sci. 17 (12) (2016). [30] D. Tronik-Le Roux, J. Renard, J. Vérine, V. Renault, E. Tubacher, J. LeMaoult, et al., Novel landscape of HLA-G isoforms expressed in clear cell renal cell carcinoma patients, Mol Oncol 11 (2017) 1561–1578. [31] A. Lin, X. Zhang, R.L. Zhang, J.G. Zhang, W.J. Zhou, W.H. Yan, Clinical significance of potential unidentified HLA-G isoforms without α1 domain but containing intron 4 in colorectal cancer patients, Front. Oncol. 8 (2018) 361.
6
Contents lists available at ScienceDirect
Human Immunology journal homepage: www.elsevier.com/locate/humimm
Detection of serum soluble HLA-G levels in patients with acute ischemic stroke: A pilot study Enrico Fainardia, , Daria Bortolottib, Massimiliano Castellazzic, Ilaria Casettad, Tiziana Bellinid,e, Roberta Rizzob ⁎
a
Neuroradiology Unit, Department of Experimental and Clinical Biomedical Sciences “Mario Serio”, University of Florence, University Hospital Careggi, Florence, Italy Section of Microbiology and Medical Genetics, Department of Medical Sciences, University of Ferrara, Ferrara, Italy c Section of Neurological, Psychiatric and Psychological Sciences, Department of Biomedical and Specialist Surgical Sciences, University of Ferrara, Ferrara, Italy d Section of Medical Biochemistry, Molecular Biology and Genetics, Department of Biomedical and Specialist Surgical Sciences, University of Ferrara, Ferrara, Italy e University Center for Studies on Gender Medicine, University of Ferrara, Ferrara, Italy b
ARTICLE INFO
ABSTRACT
Keywords: Acute ischemic stroke HLA-G Serum levels Clinical outcome
The aim of this study was to investigate the potential role of soluble Human Leukocyte Antigen-G (sHLA-G) molecules as biomarkers predicting outcome in acute ischemic stroke (AIS). Serum levels of total sHLA-G (sHLAG1/HLA-G5) and its soluble isoforms sHLA-G1 and HLA-G5/G6 were measured by enzyme-linked immunosorbent assay (ELISA) in 92 AIS patients and healthy donors (HD). Incidence of hemorrhagic transformation (HT), size of final infarct volume (FIV) and clinical outcome at 3 months were recorded in AIS patients. Detectable serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 were present in a small proportion of AIS patients (26.1%, 17.4% and 16.3%, respectively) and HD (12.5%, 10.7% and 10.7%, respectively) and were more elevated in AIS patients without HT than in those with HT (p < 0.01; p < 0.05; p < 0.01, respectively). HT was less frequent (p < 0.01) in AIS patients with measurable serum concentrations of sHLA-G1/HLA-G5 and HLA-G5/G6. Serum levels of sHLA-G1/HLA-G5 and sHLA-G1 were inversely correlated to FIV (p < 0.02), whereas good outcome was more common (p < 0.01) in AIS patients with detectable serum concentrations of sHLA-G1/HLA-G5. Taken together, these findings suggest that total sHLA-G could exert a protective effect in a subset of AIS patients, irrespective of its soluble isoforms sHLA-G1 and HLA-G5/G6, and indicate that the prognostic value of serum levels of sHLA-G remains to be established.
1. Introduction Acute ischemic stroke (AIS) accounts for 85% of all cerebrovascular diseases and represents a major cause of death and permanent disability worldwide [1]. Reperfusion therapies, such as intravenous thrombolysis and endovascular treatment, have recently been introduced in clinical practice as standard of care because they have demonstrated to be effective in improving outcome of patients with AIS [2]. However, many patients do not achieve a good clinical outcome after successful recanalization due to the lack of reliable clinical and radiological criteria for selection and outcome prediction of AIS patients [3]. Therefore, as an intense inflammatory response is believed to be implicated in evolution of ischemic lesions, in the last year the potential prognostic value of measurable blood biomarkers has been largely explored [4]. Nevertheless, although several molecule have been tested, the
identification of blood biomarkers predicting outcome in AIS patients still remains inconclusive [4,5]. Human Leukocyte Antigen-G (HLA-G) are immunonologically functional non-classical HLA-I antigens, structurally related to classical class Ia HLA products (HLA-I), which display a limited polymorphism, a restricted tissue distribution and alternative splicing of their primary transcript encoding seven distinct isoforms which include four membrane bound (G1, G2, G3 and G4) and three soluble (G5, G6, G7) proteins [6]. Membrane-bound HLA-G1 and soluble HLA-G5/G6 (HLA-G5/G6) are the most expressed HLA-G isoforms [6] and are currently considered the most important functionally active soluble HLA-G isoforms [7]. HLA-G5/G6 molecules are actively secreted as soluble isoforms, whereas soluble HLA-G1 (sHLA-G1) proteins are proteolytically released after shedding from cell surface [6,7]. In this setting, HLA-G disulfide-linked dimers are presumed to be more biologically active than HLA-G monomers [8]. These soluble HLA-G
⁎ Corresponding author at: Struttura Organizzativa Dipartimentale di Neuroradiologia, Dipartimento di Scienze Biomediche, Sperimentali e Cliniche “Mario Serio”, Università degli Studi di Firenze, Ospedale Universitario Careggi, Largo Brambilla 3, 50134 Firenze, Italy. E-mail address: [email protected] (E. Fainardi).
https://doi.org/10.1016/j.humimm.2019.11.004 Received 18 July 2019; Received in revised form 31 October 2019; Accepted 6 November 2019 0198-8859/ © 2019 American Society for Histocompatibility and Immunogenetics. Published by Elsevier Inc. All rights reserved.
Please cite this article as: Enrico Fainardi, et al., Human Immunology, https://doi.org/10.1016/j.humimm.2019.11.004
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
(sHLA-G) molecules are at present postulated to have tolerogenic properties in some autoimmune, neoplastic and viral neurological disorders [9]. In particular, the anti-inflammatory function of sHLA-G antigens was particularly documented in Multiple Sclerosis (MS), the prototypic brain autoimmune disease, in which elevated levels of intrathecally produced HLA-G5/G6 in dimeric form were associated with magnetic resonance imaging (MRI) inactive disease, suggesting a role of these antigens in the resolution of MS inflammatory response within the brain [10–12], probably under the influence of the individual genetic background represented by HLA-G polymorphisms [13]. However, no data are currently available on level of sHLA-G concentrations in AIS patients. For these reasons, in this study we aimed to investigate whether serum sHLA-G levels are detectable in AIS patients and their possible association with clinical outcome.
HLA-G held in Paris, 2003. In particular, the antibody MEM-G/9 reacts with native form of human HLA-G1 on the cell surface as well as with soluble HLA-G5 isoform in its beta2-microglobulin associated form. The antibody 5A6G7 was generated to a peptide corresponding to C-intron 4-encoded sequence. This antibody does not crossreact with the fulllength HLA-G1 isoform and thus allow to distinguish between secreted HLA-G5 and HLA-G6 isoforms from shedded HLA-G1. After ELISA measurements of serum values of sHLA-G1/HLA-G5 and HLA-G5/G6, the amount of serum sHLA-G1 was calculated by the mathematical the difference between sHLA-G1/HLA-G5 and HLA-G5/G6 concentrations. Serum sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 levels were expressed in ng/ml. The intra-assay coefficient of variations (CV), the inter-assay CV and the limit of sensitivity were 1.4%, 4% and 1 ng/ml for sHLA-G1/HLA-G5 determination, 2%, 3.5% and 1 ng/ml for HLAG5/G6 assay. In summary, total sHLA-G concentrations, given by the sum of sHLA-G1 and HLA-G5 amounts, was defined as sHLA-G1/HLAG5 levels, amounts of HLA-G5/G6 soluble isoforms was referred as HLA-G5/G6 levels and the mathematical difference between sHLA-G1/ HLA-G5 and HLA-G5/G6 concentrations was indicated as sHLA-G1 levels.
2. Patients and methods 2.1. Subjects Serum samples were prospectively collected in 92 AIS patients admitted within 4.5 h from onset receiving intravenous fibrinolysis with recombinant tissue-plasminogen activator (t-PA) according to the current recommended guidelines [2]. Inclusion criteria were: 1) presentation at the hospital within 4.5 h from symptom onset; 2) baseline and follow-up confirmation of hemispheric AIS by clinical assessment and brain non-contrast computerized tomography (NCCT) scans; 3) baseline and follow-up CT imaging carried out at onset and at 24 h after admission CT. Exclusion criteria included: 1) detection of intracerebral hemorrhage at admission brain NCCT; 2) evidence of lacunar or brain stem infarct 3) evidence of undetermined or posterior circulation stroke; 4) previous strokes with residual deficit; 5) pregnancy; 6) age < 18 years. Disease severity at onset was scored using the National Institutes of Health Stroke Scale (NIHSS) [14]. Clinical outcome was measured by the modified Rankin scale (mRS) at 3 months. mRS ≤ 2 and > 2 were defined as good and poor outcomes, respectively [15]. Hemorrhagic transformation (HT) was classified on NCCT at 24 h post ictus according to the European Cooperative Acute Stroke Study (ECASS)-II criteria [16] into four different categories: hemorrhagic infarction type 1 (HI-1), HI type 2 (HI-2), Parenchymal hemorrhage type 1 (PH-1), and PH type 2 (PH-2). Final infarct volume (FIV) was measured on follow-up NCCT at 24 h after symptom onset with a multi-slice planimetric method by summation of the hypodense areas, manually traced on each slice in which they were detectable, multiplied by slice thickness [17]. Fifty-six healthy donors (HD) served as controls. Clinical assessment, CT scans and blood sampling were performed during routine clinics, with written informed consent and local ethical board approval.
2.4. Statistics After checking data for normality by using the Kolmogorov-Smirnov test, continuous variables were compared using Mann-Whitney U test, whereas their correlations were assessed by linear regression or Spearman's rank correlation coefficient. Categorical variables were compared by means of Chi-square test. A value of p < 0.05 was assumed as statistically significant. 3. Results 3.1. Clinical and radiological findings Demographic, clinical and radiological data recorded in our AIS patients are listed in Table 1. In our study population, a good outcome (mRS ≤ 2) was achieved in 72/92 (78.3%) cases, whereas a poor outcome (mRS > 2) was present in 20/92 (21.7%) patients. HT was observed in 28/92 (30.4%) patients (13 HI and 15 PH) appearing 24 h after stroke onset. 3.2. Serum levels of total sHLA-G, HLA-G5/G6 and sHLA-G1 in AIS patients and controls Detectable serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 were more frequent in AIS patients than in HD, without any statistical difference between the two groups (Table 1). As reported in Fig. 1, serum concentrations of sHLA-G1/HLA-G5 were higher in AIS than in HD (p < 0.05), while serum titers of HLA-G5/G6 and sHLA-G1 were statistically equivalent between AIS and HD, even if they resulted greater in AIS. As reported in Table 2, serum levels of sHLA-G1/HLAG5, HLA-G5/G6 and sHLA-G1 were more elevated in AIS patients without HT than in those with HT (p < 0.01; p < 0.05; p < 0.01, respectively). Conversely no significant differences were found between AIS patients with good and poor outcome for serum levels of sHLA-G1/ HLA-G5, HLA-G5/G6 and sHLA-G1 (data not shown).
2.2. Handling of serum samples Serum samples were obtained from centrifugation of blood specimens withdrawn by puncture of an anterocubital vein, collected under sterile conditions at admission, stored in aliquots at −80 °C until assay and measured under exactly the same conditions. 2.3. ELISA assay for sHLA-G1 and HLA-G5/G6 isoforms As previously described [11], serum levels of total sHLA-G (sHLAG1/HLA-G5) and HLA-G5/G6 were measured by enzyme-linked immunosorbent assay (ELISA) using monoclonal antibodies MEM-G/9 and MoAb 5A6G7 (Exbio, Praha, Czech Republic) as capture antibodies, respectively, and anti-beta2 microglobulin MoAb-peroxidase (HRP) conjugated and biotinylated anti-HLA class I MoAb W6/32 (Exbio, Prague, Czech Republic) as detection antibody, respectively. As reported in Essen Workshop for sHLA-G quantification [18,19], these antibodies were validated during the third International Workshop on
3.3. Radiological and clinical characteristics in AIS patients with and without measurable serum levels of total sHLA-G, HLA-G5/G6 and sHLAG1 When AIS patients were stratified according to detectable levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 (Fig. 2), HT was more frequent (p < 0.01) in AIS patients without measurable serum concentrations of sHLA-G1/HLA-G5 (sHLA-G1/HLA-G5neg) and HLA-G5/ G6 (HLA-G5/G6neg) than in those with detectable serum levels of sHLA2
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
Table 1 Demographic, clinical and radiological characteristics, and measurable levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in patients with acute ischemic stroke (AIS) and healthy donors (HD).
Sex: Male/Female Age, years: mean ± SD NIHSS at entry: median, IQR, mean ± SD, range mRS at 3 months: median, IQR, mean ± SD, range Time between symptom onset and NCCT scan at admission (hours): median, IQR, mean ± SD, range Involved arterial territory at admission: n/total (%) Right MCA Left MCA Final infarct volume (ml) on NCCT at 7 day: median, IQR, mean ± SD, range Detectable levels (%) sHLA-G1/HLA-G5 HLA-G5/G6 sHLA-G1
AIS = 92
HD = 56
p value
50/40 65.4 ± 12.9 10, 6–14, 11.0 ± 6.2, 1–25 1, 1–2, 1.5 ± 1.1, 0–5 2, 1.3–3.0, 2.1 ± 1.0, 0.3–4.3
34/22 62.1 ± 9.9 / / /
p = 0.45a p = 0.09b
40/92 (43.5%) 52/92 (56.5%) 8, 0.8–37.5, 42.0 ± 78.3, 0–458
/ / /
24/92 (26.1%) 16/92 (17.4%) 15/92 (16.3%)
7/56 (12.5%) 6/56 (10.7%) 6/56 (10.7%)
p = 0.05b p = 0.27b p = 0.34b
SD = Standard deviation; NIHSS = National Institutes of Health Stroke Scale; IQR = Interquartile range; mRS = modified Rankin Scale: NCCT = non-contrast CT; MCA = middle cerebral artery. aChi-square; bMann-Whitney
3.4. Correlations between serum levels of total sHLA-G, HLA-G5/G6 and sHLA-G1 and AIS radiological and clinical features Serum concentrations of sHLA-G1/HLA-G5 and sHLA-G1 were inversely correlated to FIV (p < 0.02), without any definite associations between serum levels of HLA-G5/G6 and FIV. No relationships were found between serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 and mRS (Table 3). 4. Discussion 4.1. Interpretation of result To our knowledge, this is the first study investigating serum levels of sHLA-G molecules in AIS patients who currently lack convincing blood biomarkers predicting outcome [4,5]. These antigens seem to be implicated in the termination of autoimmunity in MS, acting as anti-inflammatory molecules. These hypothesis is supported by the evidence that high amounts of HLA-G5/G6 in dimeric form were associated with MRI inactive disease [10–12]. More important, serum concentrations of sHLA-G proteins were greater in patients with MS than in those with other non-inflammatory and inflammatory neurological disorders [20]. Based on these considerations, we tested the potential relevance of serum levels of sHLA-G antigens as prognostic factor in AIS patients. In this way, we examined the relationship of serum concentrations of sHLA-G molecules with both tissue outcome assessed by FIV and clinical outcome evaluated by mRS. In addition, we explored the possible association between serum values of sHLA-G and HT, representing one of the most dangerous complications of thrombolytic therapy and a factor affecting outcome [21]. Our results indicate that serum levels of total sHLA-G (sHLA-G1/HLA-G5) and its soluble isoforms sHLA-G1 and HLA-G5/G6 are detectable in a small proportion of AIS patients and HD. In addition, levels these antigens differed between these two groups only for serum sHLA-G/HLA-G5 titers, which were slightly
Fig. 1. Serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in 92 acute ischemic stroke (AIS) patients and 56 healthy donors (HD). Serum sHLA-G1/ HLA-G5, HLA-G5/G6 and sHLA-G1 were higher in AIS than in HD, with statistically significant values only for sHLA-G1/HLA-G5 (p < 0.05). Median, interquartile range (IQR), mean ± standard deviation (SD) and range were 0, 0–0.9, 6.9 ± 15.8 and 0–78.1 ng/ml for sHLA-G1/HLA-G5, 0, 0–0, 3.7 ± 9.6 and 0–50.6 ng/ml for HLA-G5/G6, 0, 0–3, 3.3 ± 9.6 and 0–76.7 ng/ml for sHLA-G1 in AIS patients. Median, IQR, mean ± SD and range were 0, 0–0.9, 1.0 ± 3.2 and 0–16.0 ng/ml for sHLA-G1/HLA-G5, 0–0, 1.0 ± 2.9, 0–12.0 ng/ ml for HLA-G5/G6, and 0, 0–0 0.3 ± 1.6 and 0–11.0 ng/ml for sHLA-G1 in HD.
G1/HLA-G5 (sHLA-G1/HLA-G5pos) and HLA-G5/G6 (HLA-G5/G6pos). On the contrary, HT rates did not significantly differ between AIS patients with (HLA-G1pos) and without (HLA-G1neg) measurable serum titers of sHLA-G1. Favorable outcome was more represented (p < 0.01) in sHLA-G1/HLA-G5pos than in sHLA-G1/HLA-G5neg AIS patients (Fig. 3), without significant differences between HLA-G5/G6pos and HLA-G5/G6neg, as well as between HLA-G1pos and HLA-G1neg AIS patients (data not shown).
Table 2 Serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 in patients with acute ischemic stroke (AIS) categorized according to hemorrhagic transformation (HT).
Serum sHLA-G1/HLA-G5 (ng/ml) median, IQR, mean ± SD, range Serum HLA-G5/G6 (ng/ml) median, IQR, mean ± SD, range Serum sHLA-G1 (ng/ml) median, IQR, mean ± SD, range
HT (n=28)
no HT (n=64)
p value (Mann- Whitney)
0, 0-0, 1.4 ± 7.2, 0-39.2
0, 0-11.2, 9.6 ± 18.1, 0-78.1
p < 0.01
0, 0-0, 1.0 ± 5.7, 0-31.3
0, 0-0, 4.9 ± 10.8, 0-50.6
p < 0.05
0, 0-0, 0.4 ± 1.6, 0-7.9
0, 0-5.2, 4.7 ± 11.4, 0-76.7
p < 0.01
IQR = interquartile range; SD = standard deviation. 3
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
Fig. 2. Distribution of hemorrhagic transformation (HT) in patients with acute ischemic stroke (AIS) categorized according to detectable serum levels of sHLA-G1/HLA-G5, HLA-G5 and sHLA-G1. HT was more frequent (p < 0.01) in sHLA-G1/HLA-G5neg (28/69; 40.6%) than in sHLA-G1/HLA-G5pos (2/23; 8.7%) AIS patients and in HLA-G5/G6neg (29/77; 37.7%) than in HLA-G5/G6pos (1/15; 6.7%) AIS patients, whereas there were no differences between sHLA-G1pos (25/78; 32.1%) sHLA-G1neg (5/14; 35.7%) AIS patients.
further confirmed by the inverse correlation between serum sHLA-G/ HLA-G5 and sHLA-G1 levels and FIV, suggesting that infarct size was smaller when serum concentrations of sHLA-G proteins were higher. Consistent with these findings was also the demonstration that a favourable outcome was more frequently achieved in AIS patients with detectable levels of sHLA-G1/HLA-G5. Overall, these results argue for an apparent protective role sHLA-G molecules in AIS patients, without a clear predominance of a particular isoform. However, as the detection of these molecules in serum was restricted to a subset of patients, the use of serum levels of sHLA-G antigens collected at admission as prognostic biomarker in AIS is difficult to be proposed at this time. Moreover, in absence of data on cellular expression, function and activity, the mechanisms leading to the release of these antigens in peripheral blood of AIS patients can only be hypothesized. Cellular sources of sHLA-G proteins are likely represented by monocytes and macrophages which are strongly activated in the acute phase of stroke and subsequently migrate into the brain across the blood-brain barrier (BBB) and accumulate within the infarct and peri-infart tissue [22]. Elevated serum levels of sHLA-G1 could be due to an increase in membrane shedding promoted by matrix metalloproteinase-2 (MMP-2) via proteolytic cleavage [23]. In fact, a MMP-2 overexpression occurred in AIS where this gelatinase probably contributes to BBB breakdown facilitating development of vasogenic edema in ischemic tissue and HT [24]. On the contrary, factors generating the secretion of HLA-G5/G6 antigens by peripheral blood monocytes [25] remain currently unknown, but they are probably produced as a part of the inflammatory reaction associated to AIS [22,24]. In any case, as the measurements of serum sHLA-G levels in AIS patients did not reveal any predominant isoform, the increase in serum concentrations of total sHLA-G and its soluble isoforms sHLA-G1 and HLA-G5/G6 observed in these patients
Fig. 3. Distribution of favourable outcome (mRS ≤ 2) in patients with acute ischemic stroke (AIS) categorized according to detectable serum levels of sHLAG1/HLA-G5, HLA-G5/G6 and sHLA-G1. A good outcome was more frequent (p < 0.01) in sHLA-G1/HLA-G5pos (21/23; 91.3%) than in sHLA-G1/HLAG5neg (54/76; 71.1) AIS patients.
higher in AIS patients than in HD. More interesting was the analysis of AIS patients showing that the presence of both measurable and elevated serum levels of sHLA-G1/HLA-G5 and its soluble isoforms sHLA-G1 and HLA-G5/G6 was associated with low rates of HT, with a greater extent for HLA-G5/G6. This beneficial effect of sHLA-G molecules in AIS was
Table 3 Correlations between serum levels of sHLA-G1/HLA-G5, HLA-G5/G6 and sHLA-G1 and final infarct volume (FIV) in patients with acute ischemic stroke (AIS).
Serum sHLA-G1/HLA-G5 (ng/ml) Serum HLA-G5/G6 (ng/ml) Serum sHLA-G1 (ng/ml)
FIV
p value (Spearman)
mRS
p value(Spearman)
r = -242299 r = -0.1012; r = -249854
p < 0.02 p = 0.1685 p < 0.02
r = -0.144 r = -0.0678 r = -0.1554
p = 0.0855 p = 0.2604 p = 0.0695
4
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al.
might merely represent an indirect manifestation of inflammatory immune activation taking place in AIS, rather than a response directed to a specific target. The low incidence of HT found in AIS patients with both detectable and high serum levels of total sHLA-G, sHLA-G1 and HLAG5/G6 suggests that sHLA-G antigens could limit HT acting as procoagulant molecules. The concurrent increase in the expression of HLAG and pro-coagulation factors induced by preimplantation factor (PIF) recently reported in JEG-3 cytotrophoblast cells [26] may support this speculation. On the other hand, the beneficial effect exerted by total sHLA-G and sHLA-G1 on tissue and clinical outcome may be mediated by anti-inflammatory properties of these molecules through the inhibition of CD4+ T cell activation [27] and the suppression of CD8+ cytotoxic T effector cells [28]. In fact, both CD4+ and CD8+ T cells are considered to have a central role in inflammatory response occurring in AIS [22,29].
Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper References [1] G.J. Hankey, Stroke. Lancet 389 (2017) 641–654. [2] W.J. Powers, A.A. Rabinstein, T. Ackerson, O.M. Adeoye, N.C. Bambakidis, K. Becker, et al., Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association, Stroke 49 (2018) (2018) e46–e110. [3] M. Goyal, Poor clinical outcome despite successful arterial recanalization. What went wrong? How can we do better? Neuroradiology 52 (2010) 341–343. [4] C.C. Esenwa, M.S. Elkind, Inflammatory risk factors, biomarkers and associated therapy in ischaemic stroke, Nat. Rev. Neurol. 12 (2016) 594–604. [5] G.J.L. Ng, A.M.L. Quek, C. Cheung, T.V. Arumugam, R.C.S. Seet, Stroke biomarkers in clinical practice: A critical appraisal, Neurochem. Int. 107 (2017) 11–22. [6] E.D. Carosella, N. Rouas-Freiss, D. Tronik-Le Roux, P. Moreau, J. LeMaoult, HLA-G: an immune checkpoint molecule, Adv. Immunol. 127 (2015) 33–144. [7] J. LeMaoult, M. Le Discorde, N. Rouas-Freiss, P. Moreau, C. Menier, J. McCluskey, et al., Biology and functions of human leukocyte antigen-G in health and sickness, Tissue Antigens 62 (2003) 273–284. [8] S. Zilberman, C. Schenowitz, S. Agaugué, F. Benoît, B. Riteau, R. Rouzier, et al., HLA-G1 and HLA-G5 active dimers are present in malignant cells and effusions: the influence of the tumor microenvironment, Eur. J. Immunol. 42 (2012) 1599–1608. [9] F. Morandi, R. Rizzo, E. Fainardi, N. Rouas-Freiss, V. Pistoia, Recent Advances in our understanding of HLA-G Biology: Lessons from a wide spectrum of human diseases, J Immunol. Res. 2016 (2016) 4326495. [10] E. Fainardi, R. Rizzo, L. Melchiorri, L. Vaghi, M. Castellazzi, A. Marzola, et al., Presence of detectable levels of soluble HLA-G molecules in CSF of relapsing-remitting multiple sclerosis: relationship with CSF soluble HLA-I and IL-10 concentrations and MRI findings, J. Neuroimmunol. 142 (2003) 149–158. [11] E. Fainardi, R. Rizzo, L. Melchiorri, M. Stignani, M. Castellazzi, M.L. Caniatti, et al., Soluble HLA-G molecules are released as HLA-G5 and not as soluble HLA-G1 isoforms in CSF of patients with relapsing-remitting multiple sclerosis, J Neuroimmunol 192 (2007) 219–225. [12] E. Fainardi, D. Bortolotti, S. Bolzani, M. Castellazzi, C. Tamborino, G. Roversi, et al., Cerebrospinal fluid amounts of HLA-G in dimeric form are strongly associated to patients with MRI inactive multiple sclerosis, Mult Scler 22 (2016) 245–249. [13] R. Rizzo, D. Bortolotti, N.B. Fredj, A. Rotola, F. Cura, M. Castellazzi, et al., Role of HLA-G 14bp deletion/insertion and +3142C > G polymorphisms in the production of sHLA-G molecules in relapsing-remitting multiple sclerosis, Hum Immunol 73 (2012) 1140–1146. [14] T. Brott, H.P. Adams, C.P. Olinger, J.R. Marler, W.G. Barsan, J. Biller, et al., Measurements of acute cerebral infarction: a clinical examination scale, Stroke 20 (1989) 864–870. [15] J.C. van Swieten, P.J. Koudstaal, M.C. Visser, H.J. Schouten, J. van Gijn, Interobserver agreement for the assessment of handicap in stroke patients, Stroke 19 (1988) 604–607. [16] W. Hacke, M. Kaste, C. Fieschi, R. von Kummer, A. Davalos, D. Meier, et al., Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second EuropeanAustralasian Acute Stroke Study Investigators, Lancet 352 (1998) 1245–1251. [17] T. Brott, J.R. Marler, C.P. Olinger, H.P. Adams, T. Tomsick, W.G. Barsan, et al., Measurements of acute cerebral infarction: lesion size by computed tomography, Stroke 20 (1989) 871–875. [18] V. Rebmann, J. LeMaoult, N. Rouas-Freiss, E.D. Carosella, H. Grosse-Wilde, Report of the Wet Workshop for quantification of soluble HLA-G in Essen, Hum. Immunol. 66 (2005) (2004) 853–863. [19] V. Rebmann, J. LeMaoult, N. Rouas-Freiss, E.D. Carosella, H. Grosse-Wilde, Quantification and identification of soluble HLA-G isoforms, Tissue Antigens 69 (2007) 143–149. [20] F. Morandi, C. Venturi, R. Rizzo, M. Castellazzi, E. Baldi, M.L. Caniatti, et al., Intrathecal soluble HLA-E correlates with disease activity in patients with multiple sclerosis and may cooperate with soluble HLA-G in the resolution of neuroinflammation, J. Neuroimmune Pharmacol. 8 (2013) 944–955. [21] G. Lu, Q. He, Y. Shen, F. Cao, Potential biomarkers for predicting hemorrhagic transformation of ischemic stroke, Int. J. Neurosci. 128 (2018) 79–89. [22] R. Macrez, C. Ali, O. Toutirais, B. Le Mauff, G. Defer, U. Dirnagl, et al., Stroke and the immune system: from pathophysiology to new therapeutic strategies, Lancet Neurol. 10 (2011) 471–478. [23] R. Rizzo, A. Trentini, D. Bortolotti, M.C. Manfrinato, A. Rotola, M. Castellazzi, L. Melchiorri, et al., Matrix metalloproteinase-2 (MMP-2) generates soluble HLA-G1 by cell surface proteolytic shedding, Mol. Cell Biochem. 381 (2013) 243–255. [24] J. Castillo, I. Rodríguez, Biochemical changes and inflammatory response as markers for brain ischaemia: molecular markers of diagnostic utility and prognosis in human clinical practice, Cerebrovasc Dis. 17 (Suppl 1) (2004) 7–18. [25] V. Rebmann, A. Busemann, M. Lindemann, H. Grosse-Wilde, Detection of HLA-G5 secreting cells, Hum. Immunol. 64 (2003) 1017–1024. [26] M.S. Hakam, J.M. Miranda-Sayago, S. Hayrabedyan, K. Todorova, P.S. Spencer, A. Jabeen, et al., Preimplantation factor (PIF) promotes HLA-G, -E, -F, -C Expression
4.2. Limitations of the study This study is affected by several limitations. First, the relatively small sample size could weaken the consistency of our data. In particular, AIS patients with no HT and favourable outcome were more represented than those with no HT and unfavourable outcome and, therefore, the comparison between these groups could result unbalanced. Second, information about HLA-G cellular expression, secretion and shedding and genetic polymorphisms are needed for better understanding and interpreting the association between serum levels of sHLA-G molecules and AIS prognostic parameters found in this study. Third, functional experiments are requested for establishing the actual influence of HLA-G antigens on pro-inflammatory cells involved in AIS. Fourth, the relationship between HLA-G molecules and coagulation cascade should be clarified. Finally, the knowledge of temporal pattern of serum concentrations of sHLA-G antigens in relation to clinical and radiological evolution of ischemic lesion could be useful to verify the prognostic value of sHLA-G serum measurements in AIS. Fifth, the recent demonstration that novel HLA-G isoforms detectable by 5A6G7 antibody exist [30,31] opens new avenues not only for the investigation of HLA-G5/G6 typical isoform, but also for the analysis other less studied isoforms that could play a role in immune-system control in different diseases, including AIS. 4.3. Conclusions In summary, this pilot study indicates that detectable and elevated serum levels of total sHLA-G are associated with a lower occurrence of HT and a more frequent tissue and clinical good outcomes in a subset of AIS patients, irrespective of serum concentrations of its isoforms (sHLAG1 and HLA-G5/G6). These findings suggest that sHLA-G antigens could be implicated as protective molecules in mechanisms regulating AIS inflammatory response. Future studies in a larger patient population are warranted to verify whether sHLA-G proteins can really exert anti-inflammatory functions and can represent a potential biomarkers predicting outcome in AIS. Author Contribution Study concept and design: RR, EF. Acquisition of data: DB, MC, IC, RR. Laboratory analysis: DB, RR. Statistical analysis: DB, IC, RR, EF. Analysis and interpretation of data: all authors Drafting of the manuscript: DB, TB, RR, EF Critical revision of the manuscript for important intellectual content: all authors. Study supervision: RR, EF. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. 5
Human Immunology xxx (xxxx) xxx–xxx
E. Fainardi, et al. in JEG-3 choriocarcinoma cells and endogenous progesterone activity, Cell Physiol Biochem 43 (2017) 2277–2296. [27] N. Lila, N. Rouas-Freiss, J. Dausset, A. Carpentier, E.D. Carosella, Soluble HLA-G protein secreted by allo-specific CD4+ T cells suppresses the allo-proliferative response: a CD4+ T cell regulatory mechanism, Proc. Natl. Acad. Sci. USA 98 (2001) 12150–12155. [28] P. Contini, M. Ghio, A. Poggi, G. Filaci, F. Indiveri, S. Ferrone, et al., Soluble HLAA,-B,-C and -G molecules induce apoptosis in T and NK CD8+ cells and inhibit cytotoxic T cell activity through CD8 ligation, Eur. J. Immunol. 33 (2003) 125–134.
[29] A. Bonaventura, L. Liberale, A. Vecchié, M. Casula, F. Carbone, F. Dallegri, et al., Update on inflammatory biomarkers and treatments in ischemic stroke, Int. J. Mol. Sci. 17 (12) (2016). [30] D. Tronik-Le Roux, J. Renard, J. Vérine, V. Renault, E. Tubacher, J. LeMaoult, et al., Novel landscape of HLA-G isoforms expressed in clear cell renal cell carcinoma patients, Mol Oncol 11 (2017) 1561–1578. [31] A. Lin, X. Zhang, R.L. Zhang, J.G. Zhang, W.J. Zhou, W.H. Yan, Clinical significance of potential unidentified HLA-G isoforms without α1 domain but containing intron 4 in colorectal cancer patients, Front. Oncol. 8 (2018) 361.
6