Samples of saccular intracranial aneurysms
A previously published sIA series [11,12,13, 17,18,19] of 36 sIA samples (16 unruptured and 20 ruptured) were studied. The sIA samples were resected after surgical clipping at the Department of Neurosurgery, Helsinki University Hospital (HUH), Helsinki, Finland. The samples were immediately snap-frozen in liquid nitrogen after harvesting and stored at − 80 °C. For immunohistochemical and immunofluorescence stainings, the frozen samples were embedded in Tissue-Tek (Sakura, Alphen aan den Rijn, the Netherlands) and cryosectioned at 4 µm. Clinical data were collected from the patients’ medical records and sIA dimensions were obtained from preoperative computed tomography angiography images. The HUH Ethics Committee approved this study.
Basic characteristics of the aneurysm walls
The sIAs were classified into categories A–D according to the characteristics of their walls: type A (9/36; 25%), type B (12/36; 33%), and type C (11/36; 31%) [11]. Only two samples representative of wall type D (2/36; 6%) were present and therefore they were excluded from further analyses. The following wall classification criteria were originally published by Frösen et al., 2004 [20]: wall type A displays a wall with intact endothelium and an organized layer of smooth muscle cells (SMCs), type B displays a thickened wall and a disorganized layer of SMCs, type C displays a hypocellular wall with either myointimal hyperplasia or organized thrombus, and type D displays a very thin hypocellular wall with organized thrombus. This sIA series has also been analysed earlier for inflammatory and lipid characteristics [11,12,13, 18, 19].
Immunohistochemistry and immunofluorescence stainings
For immunohistochemical stainings, the histological sections were fixed with ice-cold acetone for 3 min, and sequentially incubated with the EnVision Kit’s blocking reagent (Dako, Santa Clara, CA, USA) and 3% normal horse serum (NHS; Vector, Burlingame, CA, USA) at room temperature (RT) for 30 min. The primary antibodies against lymphatic vessels i.e. against the studied LEC-markers: podoplanin, Prox1, LYVE-1, and VEGFR-3 (Additional file 1: Table S1) were diluted to the buffer solution from the kit and incubated on the sections for 60 min at RT. The secondary detection was performed with the EnVision Kit’s horseradish peroxidase reagent (Dako) for the mouse primary antibodies and with the anti-goat secondary antibody diluted 1:200 for the goat primary antibodies according to the manufacturer’s protocol. For the detection of the positive signal, the sections were incubated in diaminobenzidine for 4 min at RT. Finally, the sections were background stained with Mayer’s haematoxylin (Sigma-Aldrich, St. Louis, MO, USA) or with Lillie’s Modification (Dako) and embedded in a mounting aqueous medium (Faramount, Dako). An irrelevant mouse monoclonal antibody (IgG1 or IgG2a, depending on the subclass of primary antibody; Serotec, Oxford, UK) served as a substitute for the primary antibody in negative controls in the podoplanin, LYVE-1, and VEGFR-3 stainings. In the Prox1 stainings, the primary antibody was omitted in the negative controls. Freshly frozen human tonsil tissue served as a positive control in both immunohistochemical and immunofluorescence stainings.
For immunofluorescence double stainings the frozen sIA sections were fixed and blocked as described above. The primary antibodies against podoplanin or LYVE-1 (Additional file 1: Table S1) were diluted to the buffer solution and incubated on sections for 60 min at RT. The secondary detection was performed by incubating the sections with Alexa fluor 488 (green) F(ab’)2 fragments of goat anti-mouse or rabbit anti-goat IgG antibodies (Thermo Fisher Scientific, Eugene, OR, USA), respectively, at RT for 20 min. Thereafter, the sections were re-blocked and incubated in primary antibodies against α-smooth muscle actin (αSMA) or CD68 (Additional file 1: Table S1). The secondary detection of CD68 was performed by incubating the sections with Alexa fluor 594 (red) F(ab’)2 fragments of rabbit anti-goat IgG antibody (Thermo Fischer Scientific). Finally, the sections were stained for nuclei with DAPI (Sigma-Aldrich) and embedded in fluorescence mounting medium (Faramount, Dako). The primary antibody was omitted in the negative controls.
Analysis of immunohistochemical and immunofluorescence stainings
The immunohistochemical and immunofluorescence stainings were scanned using 3DHISTECH Pannoramic 250 FLASH II digital slide scanner (Budapest, Hungary). Positive stainings for LYVE-1, podoplanin, VEGFR-3, and Prox1 and their histological locations were analysed semiquantitatively from all 36 stained sIAs.
The samples were scored semiquantitatively as 0–5 depending on the extent and location of the positively stained area (Fig. 1A) using the highest score for each sample. The presence and the number of positively stained ring-shaped structures were considered as cross-sections of lymphatic vessels and counted in each sample of LYVE-1, podoplanin, Prox1, and VEGFR-3 stainings.
The semiquantitative analysis of LYVE-1, podoplanin, Prox1, and VEGFR-3 and the presence and the number of positively stained ring-shaped structures, i.e. lymphatic vessels were compared to sIA rupture status, sIA wall degeneration (wall type), presence of thrombus, and markers for lipids, inflammation, and angiogenesis as defined in Additional file 1: Table S1 and in our previous studies [11,12,13, 18, 19]. The percentual proportion of the thrombus area in the sIA wall was defined by measuring the area of the thrombus in the sample section using the 3DHISTEC Slide viewer area tool and dividing it by the total sIA wall area (including the thrombus). The overlap of a LYVE-1-positive area with areas positive for podoplanin, Prox1, VEGFR-3, or angiogenesis (CD34 staining, [11]) on consecutive sections was defined as complete, partial, or absent, as demonstrated in Fig. 2. The presence and location of double-positive staining for podoplanin and αSMA, and for LYVE-1 and CD68 were determined in scanned immunofluorescence double stainings.
Statistics
Data analysis was performed using the IBM SPSS Statistics Software, version 27. For categorical variables, proportions were calculated, and Fisher’s exact test (F) was used. For continuous variables, Kruskal–Wallis (KW) multiple comparison test, Mann–Whitney-U (MWU) test, and Spearman (S) correlation test were used. P-values < 0.05 were considered statistically significant.