CD analysis of nucleic acid structures was performed as described35 (link), EMSAs were performed as recommended36 (link),37 (link), and DMS protection assays followed a previously described protocol26 (link),36 (link). The RNase protection assays were performed following Ambion’s recommendations. Plasmids were constructed in a pCR8 TOPO vector (Invitrogen), and GGGGCC HRE inserts were generated using a self-templating PCR protocol38 (link). in vitro transcription reactions were performed with these plasmids and analyzed on sequencing gels. R-loop assays were adapted from previously described methods39 (link). The cDNA from B lymphocytes or RNA from human tissues with or without the C9orf72 HRE were generated from total RNA following manufacturer’s protocols and relative levels were then measured. NanoString RNA analysis followed standard protocols as previously described20 (link). The RNA pulldown with isotopically labeled HEK293T lysates and biotinylated RNA conjugated to streptavidin beads followed a previously described protocol40 (link), with an additional KCl gradient wash. Quantitative mass spectrometry was performed by employing a three-state SILAC analysis using a filter-aided sample preparation (FASP) method followed by analysis on an LTQ-Orbitrap Elite mass spectrometer30 (link),41 (link). Peptides were identified using the Mascot search algorithm. Western blotting was performed on RNA pulldown fractions according to the manufacturer’s recommendations for each antibody. IF staining of lymphocytes, HEK293T cells, fibroblasts, and iPS motor neurons followed a standard protocol described in detail in Methods. RNA FISH with IF on human motor cortex tissue was performed essentially as previously described20 (link).