Dorsal root ganglia in Friedreich ataxia: satellite cell proliferation and inflammation

The Institutional Review Board of the Veterans Affairs Medical Center in Albany, NY, USA, has approved the research described in this paper. DRG of 15 FA patients and 12 normal controls were available for routine staining of paraffin-embedded tissue sections, immunohistochemistry, and immunofluorescence. Tissues of FA patients were collected under a formal donation program supported by Friedreich’s Ataxia Research Alliance. Control DRG were obtained during autopsies conducted at Veterans Affairs Medical Center and Albany Medical College; and from National Disease Research Interchange, Philadelphia, PA, USA. Detailed clinical and genetic information was available for all patients (8 male, 7 female). Age of onset ranged from 2 to 18 years (mean ± standard deviation [S.D.]: 10 ± 5), and age of death from 10 to 69 years (mean ± S.D.: 36 ± 18). All patients had homozygous GAA repeat expansions, ranging from 249 to 1200 for GAA1 and 566–1200 for GAA2 (means ± S.D.: GAA1, 734 ± 251; GAA2, 955 ± 204). Autopsy delays were 2–96 h. The age range of the controls (9 male, 3 female) was 48–68 years (mean ± S.D.: 60 ± 6). Autopsy delays in the control cases ranged from 1 to 48 h.

Paraffin sections of 6 μm thickness were processed to visualize selected proteins by immunohistochemistry and immunofluorescence. The overall approach was to visualize proteins in or around satellite cells with antibodies that were successfully used in animal experiments [4, 6, 7, 9, 17, 19, 22] or on human DRG [13, 15]. Table 1 provides detailed information on antibodies, sources, catalogue numbers (Cat. No.), and antigen retrieval methods. Details of immunohistochemistry and double-label immunofluorescence of DRG were described in previous publications [13‐16]. Briefly, for immunohistochemistry, paraffin sections were rehydrated and oxidized in hydrogen peroxide-containing methanol, processed through antigen retrieval (Table 1), blocked by 10 % normal horse serum in phosphate-buffered saline (PBS), and incubated overnight at 4 °C in antibodies (Table 1) diluted in PBS, also containing 1 % normal horse serum. The next step was incubation at room temperature for 2 h in biotinylated anti-mouse, rabbit, or goat IgG (Vector Laboratories, Burlingame, CA USA), depending on the nature of the primary antibody. After repeated washing steps, the sections were immersed in a dilute solution of horseradish peroxidase-labeled streptavidin for 1 h, followed by a chromogenic solution of diaminobenzidine-urea-hydrogen peroxide (Sigma-Aldrich, St. Louis, MO, USA). Slides were dehydrated and cover-slipped by standard techniques, viewed in a Zeiss Axiophot microscope, and photographed. For immunofluorescence, sections were rehydrated in the same manner though oxidation by hydrogen peroxide in methanol was omitted. Antigen retrieval followed the plan outlined in Table 1 for each protein-of-interest. The suppressor protein was 10 % donkey serum in PBS (vol/vol). After incubation with primary antibodies, the sections were exposed to Alexa Fluor 488-labeled donkey anti-mouse IgG (Jackson ImmunoResearch Laboratories, West Grove, PA, USA) or Cy3-labeled donkey anti-rabbit or goat IgG (Jackson ImmunoResearch Laboratories). Sections were covered by a 50 % solution of glycerol in PBS (vol/vol). For selected pairs of Alexa Fluor 488- or Cy3-labeled antigens, 4,6′-diamidino-2-phenylindole (DAPI, 1.3 μmol/l) was added to the glycerol-PBS mixture to reveal nuclei. Sections were viewed and processed in a Zeiss LSM 510 Meta-NLO laser scanning confocal microscope. Exciting wavelengths for Alexa Fluor 488 and Cy3 were 488 nm and 543 nm, respectively. Band pass filters were 500–530 nm for Alexa Fluor 488 and 565–615 nm for Cy3. DAPI fluorescence was generated by two-photon technology at an exciting wavelength of 740 nm, emitted by a titanium-sapphire laser, and recorded through a band pass filter of 435–485 nm.

We used several methods to validate the identity of proteins in tissue sections. Four proteins listed in Table 1 (ATP5B, connexin 43, EAAT1, and IBA1) were validated by tandem mass spectrometry including multiple reaction monitoring-initiated detection and sequencing (MIDAS), as described by Unwin et al. [28]. For this purpose, proteins in samples of frozen DRG (35–65 mg) were extracted by ultrasonication in a buffer found suitable for the assay of tissue frataxin [5, 16]. After centrifugation at 14,000 × g for 45 min, the supernatant was collected, and an aliquot was used for sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting by established procedures. The apparent molecular weights of the detected protein bands were calculated by reference to a series of prestained standard proteins and compared with information on mass in UniProt [27]. The protein bands on matching Coomassie Blue-stained gels were excised for MIDAS. One or more signature peptides were identified for each of the four aforementioned proteins by in-gel trypsin digestion, alkylation, and MIDAS at the Center for Functional Genomics at State University of New York at Albany, Rensselaer, NY, USA. The authenticity of staining with antibodies to frataxin, laminin, glutamine synthetase, mGluR2/3, Kir4.1, CD68, and cytosolic ferritin was supported by absorption of the antibodies with the corresponding purified antigenic protein, recombinant protein, or antigenic peptide (frataxin: recombinant human protein, courtesy of Dr. Grazia Isaya, Mayo Clinic, Rochester, MN, USA; laminin: purified laminin, AbD Serotec, Raleigh, NC, USA, Cat. No. 5620–0604; glutamine synthetase: full length recombinant protein, Abcam, Cambridge, MA, USA, Cat. No. ab98145; mGluR2/3: peptide GREVVDSTTSSL, US Biological, Salem, MA, USA, Cat. No. M3884-76C; Kir4.1: antigenic peptide CKLEESLREQAEKEGSALSVR, Alomone Labs, Jerusalem, Israel, Cat. No. APC-035; CD68: recombinant protein, Abcam, Cat. No. ab38260; cytosolic ferritin: purified human liver ferritin, EMD-Millipore, Billerica, MA, USA, Cat. No. 341482). To further characterize S100 reaction products, we used a monovalent polyclonal anti-S100α antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA, Cat. No. SC-7849-R) and absorption of the monoclonal anti-S100 antibody by an excess of recombinant S100α (Abnova, Taipei, Taiwan, Cat. No. H00006271-P01) or full-length S100β-peptide (Abcam, Cat. No. ab30380).