Exosome microRNA signatures in patients with complex regional pain syndrome undergoing plasma exchange

Standard protocol approvals, registrations, and patient consents

This is a retrospective case series study of CRPS patients seen at the Drexel University College of Medicine pain clinic that met the Budapest consensus criteria for CRPS and received PE as treatment for their illness between September 2012 and June 2014. All participants agreed to provide blood samples for this study after giving informed consent as approved by the institutional review board.

Plasma exchange and patient evaluation

The patients were asked to provide blood samples before and after PE. Patients with autoimmune or immunodeficiency conditions were excluded. Patient records were reviewed from which data regarding demographics, CRPS signs and symptoms, duration of illness and response to PE were obtained. PE was performed in patients with refractory CRPS over 2 weeks at the Hahnemann University Hospital, Philadelphia. All patients had full cardiac and neuropsychological clearance before PE. Patients were asked to rate their overall pain before, during, and after the apheresis using a 11-point numerical rating scale (NRS) from 0 (no pain) to 10 (the worst pain imaginable).

Characterization of exosomes by nanoparticle tracking analysis, electron microscopy and western blot analysis

One milliliter of serum from CRPS patients was diluted in equal volume of PBS and centrifuged at 2000×g for 30 min at 4 °C. The supernatant was diluted to final volume of 24 ml in 1× PBS and centrifuged at 12,000×g for 45 min at 4 °C. The supernatant was filtered through a 0.22 µ filter and centrifuged at 110,000×g for 70 min at 4 °C. The exosome pellet obtained was washed in 25 ml 1× PBS without ions and centrifuged at 110,000×g for 70 min at 4 °C. The exosome pellet was resuspended in 100 µl of PBS for use in nanoparticle tracking analysis, electron microscopy and protein estimation.

Nanoparticle tracking analysis

Exosomes in PBS were analyzed for size and concentration using the NanoSight NS300 according to the manufacturer’s protocol (Malvern Instruments, MA, USA). Samples were diluted to ~ 10⁷–10⁹ particles/ml and continuously injected with a syringe pump and three videos (30 s each) were captured for particle analysis. Nanoparticle tracking analysis was performed using NTA 3.2 software.

Electron microscopy

Ten microliters of PBS resuspended exosomes were coated on Ni-formvar grids and incubated for 20 min at RT. The grids were washed on 50 µl drops of 0.1 M Sorensen’s phosphate buffer (pH 7.2) for 5 s each for a total of five times. The grids were blot dried perpendicularly on whatman #1 filter paper. Negative staining and embedding were performed by incubating the grids on 0.5% uranyl acetate (in a 0.2% methyl cellulose solution) for 10 min at 4 °C. The excess solution was blotted on a Whatman paper, air dried and imaged in a JEOL Transmission Electron Microscope (JEM 1230). Alternatively, the exosomes were immunolabelled for the exosome marker CD81 and crosslinked with 1% glutaraldehyde, and probed with 6 nm gold secondary antibody, followed by negative staining and embedding.

Western blotting

Protein concentration was estimated using a DC Protein assay (Bio-Rad Laboratories, CA, USA); 5 µg of exosomes isolated from CRPS serum were resolved on a reducing 12% SDS-PAGE, transferred to PVDF membrane and blocked in Odyssey Blocking buffer (927-50100, LI-COR Biosciences) for 2 h. The membrane was incubated in rabbit anti-CD63 antibody (ab68418, Abcam) overnight at 4 °C, washed thrice in TBST, 10 min each, and incubated in goat anti-rabbit 680RD IgG (925-68071, LI-COR Biosciences) for 45 min at RT, washed thrice in TBST for 10 min each, and imaged on an Odyssey Fc imaging system.

Exosome miRNA profiling

RNA was isolated from exosomes using a miRvana miRNA isolation kit (Life technologies) following the manufacturer’s protocol. Taqman low-density array microfluidic cards version A and B (Applied Biosystems, Foster City, CA) were used to profile miRNAs in 100 ng of total RNA as previously described [19]. miRNA species with CT values 35 and higher were treated as undetected. Fold change was calculated from raw CT values using the 2^−ΔΔCT method [20]. The mean CT values of the 10 miRNAs with the lowest standard deviations across all samples were used as the endogenous control in the calculation of ΔCT. Statistical significance of differences in ΔCT values was calculated by a two-tailed paired t-test for comparison of pre- and post-PE samples and by a two-tailed independent t-test for comparison of other experimental groups. A p-value threshold of 0.05 and a fold-change of 2 were used to select significant differentially expressed miRNAs between experimental groups.

Cell culture

HEK293 cells obtained from the American Type Culture Collection (ATCC) was maintained in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% fetal bovine serum at 37 °C in 5% CO₂. Human monocytic THP-1 cells (TIB-202, ATCC) cells were maintained in RPMI-1640 medium containing 10% fetal bovine serum (FBS).

Luciferase reporter assay

The 3′UTR luciferase reporter constructs for IL-6 (NM_000600.2) was purchased from GeneCopoeia. The miRNA mimics for human miR-338-5p (MC12825), anti-miR-338 (MH12825) and scrambled negative control (4464058) was purchased from Life technologies. HEK293 cells were co-transfected with mir-338-5p, anti-miR-338 or miRNA scrambled control and luciferase reporter plasmid containing the 3′UTR of human IL-6 using Lipofectamine LTX (Life Technologies, Carlsbad, CA) for 48 h. The Luc-Pair miR Luciferase assay kit (GeneCopoeia) was used to measure firefly and Renilla luciferase activity according to the manufacturer’s instruction. Firefly luciferase measurements normalized to Renilla luciferase was used as a transfection control. The data expressed as percentage of control is the average of three independent experiments.

miR-338-5p overexpression in THP-1 cells

Transfections were performed following the manufacturer’s protocol for RNAiMax transfection reagent using either miR-338-5p or negative control with the following modifications. For each well of a 6-well plate, 7.5 µl RNAiMax reagent was diluted in 150 µl of serum-free media, and 30 pmol of miR-338-5p or control mimic was diluted in 150 µl of serum-free media individually. The dilutions were combined and incubated at room temperature for 15 min. This transfection complex (300 µl) was added to 0.5 × 10⁶ cells/well in 1.7 ml serum containing media in 6-well plates and incubated for 6 h at 37 °C, after which the media was changed. After 24 h, cells were treated with 1 µg/ml lipopolysaccharide (LPS) in complete culture media for 6 h. Exosome deplete media was used in all experiments. Cells were collected by centrifugation at 135×g for 5 min at 4 °C and the conditioned media was stored at 4 °C. The cell pellet was washed with 1× PBS and resuspended in either RNA lysis buffer (mirVana kit; Life Technologies) containing 0.5 U/µl RNAsin Plus (Promega; Madison, WI) for RNA isolation or 1× radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitor cocktail (Thermo Scientific; Waltham, MA) for protein analysis.

cDNA synthesis and qPCR for mRNAs

mRNA was isolated using the miRVana kit (Life technologies). The Maxima cDNA synthesis kit (Thermo Scientific) was used to generate cDNA and 2 µl cDNA was used for Taqman based quantitative real time mRNA analysis containing 10 µl Taqman Fast Universal polymerase chain reaction (PCR) master mix (2×) no AmpErase UNG (Life Technologies), 1 µl Taqman primer–probe (20×), in up to 20 µl nuclease-free water. GAPDH was used as the normalizer and one-way ANOVA was used to perform statistical analysis. Assay IDs were as follows: Hs00985639_m1 [IL-6], 4325792 (GAPDH) (Applied Biosystems, Carlsbad, CA).

Enzyme-linked immunosorbent assay (ELISA) for IL-6 in cell culture media

Supernatants collected after miR-338-5p or control miRNA transfections in THP-1 cells were used to perform ELISA for secreted IL-6 using the human IL-6 quantikine ELISA kit (D6050) according to the manufacturer’s protocol (R&D Systems; Minneapolis, MN).

Cytokine measurement in patient plasma

For plasma isolation, blood was collected into EDTA-coated (purple top) vacutainers. The plasma was separated by centrifugation (3000×g for 15 min at 4 °C), split into 250 µl aliquots and stored at − 70 °C until assayed. The MILLIPLEX MAP Human High Sensitivity T Cell Panel, HSTCMAG-28SK (Millipore, Billerica, MA) was used to determine plasma levels of 14 cytokines. The minimum detectable concentration for IL-6 and TNF-α was 0.11 and 0.16 pg/ml respectively. All assays were performed in duplicate according to the manufacturers’ instructions. Assay results were determined on a Luminex-200 (Luminex, Austin, TX).

Statistical analysis

Data are presented as mean ± the standard error of the mean from three or more independent experiments. Student t-test was used for determining the statistical significance. Treatment effects were analyzed with a one-way analysis of variance (ANOVA). Pairwise comparisons between means were tested using the post hoc Dunnet method. Error probabilities of p < 0.05 were considered statistically significant. Paired t-test with Bonferroni correction for multiple assays was used for analysis of the cytokine panel. Error probabilities of p < 0.00357 were considered statistically significant.

Isolation and characterization of exosomes

Exosomes were isolated from the serum obtained from six CRPS patients, grouped retrospectively as three responders and three non-responders both before and after PE (Fig. 1, Table 1). The exosomes displayed a size below 100 nm under a transmission electron microscope (TEM) (Fig. 2a, b). Exosomes are characterized by the presence or enrichment of endosome-derived membrane proteins such as tetraspanins (CD9.CD63, CD81), flotillins, Alix and Tsg101. The presence of such characteristic exosomal proteins was also assessed using immunogold labeling technique for CD81 by TEM (Fig. 2c) and western blotting for CD63 (Fig. 2d). Nanoparticle tracking analysis of serum exosomes indicated particles with a mean diameter of 85.7 ± 0.9 nm and a concentration of 3.94 × 10¹¹ ± 1.6 × 10¹⁰ particles/ml (Fig. 2e). The metrics of exosome isolation, purity and characterization was submitted on EV track [21] and the EV-metric score is 44%.

ID	Age	Sex	Duration of CRPS (years)	Pre PE NRS	Post PE NRS	HT (cm)	WT (kg)	BMI	Response	Comorbidities
P1	54	F	4	7	3	170	100.9	34.9	Good	Hypothyroidism, B12 deficiency, Iron deficiency, GERD, Seizure,
P2	40	M	11	8	4	195	126	33.1	Good	No past medical history other than CRPS
P3	57	M	> 6	8	2	165	120	44.1	Good	Asthma, DM, OSA, Hypercholestrolemia, Mitochondrial disease
P4	57	F	> 6	8	7	163	51.63	19.4	Poor	Fibromyalgia, Anemia
P5	28	F	7	7	7	160	81.1	31.7	Poor	Grave’s disease, Polycystic ovary syndrome, Migraine, Von Willebrand disease
P6	26	F	> 6	8	7	163	52.27	19.7	Poor	Stiff man syndrome, Seizure disorder

Pain assessment was conducted on CRPS patients before and 2 weeks following PE, using a numerical rating scale (NRS). CRPS patients with a difference in the NRS pain scores (pre PE-post PE) greater than three were considered as responders, while those less than three were classified as poor responders

Differentially expressed miRNAs between responders and poor responders of PE

Exosomal miRNAs isolated from six CRPS patients, before and after PE, were used to profile a total of 754 miRNAs. The mean CT values of the 10 miRNAs with the lowest standard deviations across all samples were used as the endogenous control in the calculation of ΔCT excluding miRNAs with a CT value of > 35 in any sample. Statistical significance of differences in ΔCT values was calculated by a two-tailed paired t-test for comparison of pre- and post-PE samples and by a two-tailed unpaired t-test for comparison of other experimental groups. A p-value threshold of 0.05 and a fold-change of 2 were used to select significantly differentially expressed miRNAs between experimental groups (Table 2, Fig. 3, Additional file 1: Table S1). The comparison between two groups is depicted as log2 fold change values. A positive number in the fold change for nb/rb, it means nb is higher than rb. A negative fold change number for nb/rb comparison means nb is lower. To investigate differential miRNA expression between responders and non-responders prior to PE, we analyzed the miRNAs from the pretreatment groups (column 2 of Table 2, Fig. 3). Of the nine miRNAs that were significantly different between responders and non-responders prior to treatment, two were upregulated and seven downregulated in non-responders relative to responders. This suggests that the exosomal molecular signature differed even prior to PE and could be explored further for biomarker potential in predicting treatment response. To determine which miRNAs changed in responders after treatment, miRNA profiles before and after treatment were compared (Table 2, column 3). Four miRNAs changed significantly in responders after treatment with two miRNAs upregulated and two downregulated (Fig. 3). Two miRNAs were altered after PE in non-responders; however, these were different from the miRNA changes observed in responders. To determine differences in miRNA alterations by PE between responders and non-responders, we compared the changes in miRNA levels of each patient. Results from this meta-analysis (Table 2, column 5) shows nine miRNAs that have different degrees of change in responders and poor responders. For example, hsa-let-7a is downregulated in both responders and non-responders, but the degree of change in responders is 4.5-fold greater than in non-responders.

Relative expression of exosomal miRNAs in six CRPS patients pre and post-PE were determined using a Taqman low density array. The significance was determined using two-tailed paired t-test for comparison of pre- and post-plasma exchange samples and by a two-tailed independent t-test for comparison of other experimental groups. A p-value threshold of 0.05 and a fold-change of 2 were used to select significantly differentially expressed miRNAs between experimental groups

Log2fc: Log 2 fold change, orange red: downregulation, green: upregulation, Rb: responders before PE, Nb: non-responders before PE, Ra: responders after PE, Na: non-responders after PE, Na-Nb/Ra-Rb (miRNA changes in non-responders over responders as a result of PE)

In order to ascertain the potential function of these miRNAs, we employed a combination of bioinformatic algorithms Targetscan and miRDB that predict miRNA targets based on the ability of the miRNA sequence to undergo specific base-pairing within the putative 3′UTR in the mRNA target [22, 23]. Among the pro-inflammatory targets, IL-6 was among the top candidates for both miR-19a/b and miR-338-5p [22, 24]. The regulatory role of miR-19 on IL-6 has been demonstrated and shown to be indirectly mediated by the binding of miR-19a and miR-19b to Toll-like receptor 2 [25]. The role of miR-338-5p in inflammation or CRPS however has not yet been explored. This led us to investigate the biological role of miR-338-5p in regulating inflammatory responses.

Confirmation of miR-338-5p binding to the 3′UTR of predicted targets

The predicted interaction of miR-338-5p and IL-6 mRNA (Fig. 4a) was studied using a luciferase reporter assay. The 3′UTR of human IL-6 harboring miR-338-5p binding sites cloned downstream of the luciferase open-reading frame was used to assess miRNA-mRNA binding. HEK293 cells were transiently transfected with plasmids encoding the reporter 3′UTR construct and either miR-338-5p mimic, anti-miR-338 or a miRNA control. Firefly luciferase measurements were normalized to Renilla luciferase as a transfection control. A significant reduction was observed in luciferase activity 48 h after transfection with miR-338-5p, (p < 0.05) confirming the binding of miR-338-5p to the 3′UTR of IL-6 mRNA (Fig. 4b). Anti-miRs are complementary to the target miRNAs and used to study loss of function effects of specific miRNAs. Transfection with anti-miR-338 individually or in combination with control miRNA did not alter relative luminescence but when combined with miR-338-5p, significantly reversed the inhibition indicating downregulation of miRNA activity.

Transcriptional and translational regulation of IL-6 mRNA by miR-338-5p

We next assessed the ability of miR-338-5p to modulate levels of the pro-inflammatory IL-6 mRNA in vitro. IL-6 mRNA and protein levels are usually very low in normal/unstimulated immune cells. Hence, to assess the observable effects of miR-338-5p on IL-6 mRNA, we stimulated human monocytic leukemia (THP-1) cells with bacterial lipopolysaccharide (LPS). THP-1 cells were transfected with either miR-338-5p mimic or control miRNA for 18 h and then stimulated with LPS (1 µg/ml) for 6 h, following which endogenous levels of target mRNAs were measured by qPCR. There was a significant reduction in IL-6 mRNA upon miR-338-5p transfection compared to control miRNA transfection, suggesting that miR-338-5p regulates IL-6 via mRNA degradation (Fig. 5a). Conditioned media from THP-1 cells transfected with miR-338-5p or control and stimulated with LPS were analyzed by ELISA for IL-6 protein levels. The overexpression of miR-338-5p significantly reduced protein levels of IL-6 secreted into the media (Fig. 5b). Thus, over expression of miR-338-5p resulted in downregulation of both mRNA and protein levels of IL-6.

Plasma levels of cytokines in CRPS patients

Persistent inflammation is a hallmark of CRPS and an upregulation of proinflammatory cytokines has been demonstrated by multiple approaches in various biofluids including plasma, serum, blister fluid and cerebrospinal fluid [26]. We investigated a panel of 14 cytokines (Additional file 2: Table S2) of which only IL-6 was significantly different following PE. As both miR-19 and miR-338-5p can target IL-6, we investigated IL-6 in a plasma sample repository available from 18 patients. A significant reduction was observed in the plasma levels of IL-6 following PE (n = 18), while TNFα levels did not change (Fig. 6).