Reprogramming fibroblasts and peripheral blood cells from a C9ORF72 patient: A proof‐of‐principle study

Abstract As for the majority of neurodegenerative diseases, pathological mechanisms of amyotrophic lateral sclerosis (ALS) have been challenging to study due to the difficult access to alive patients' cells. Induced pluripotent stem cells (iPSCs) offer a useful in vitro system for modelling human diseases. iPSCs can be theoretically obtained by reprogramming any somatic tissue although fibroblasts (FB) remain the most used cells. However, reprogramming peripheral blood cells (PB) may offer significant advantages. In order to investigate whether the choice of starting cells may affect reprogramming and motor neuron (MNs) differentiation potential, we used both FB and PB from a same C9ORF72‐mutated ALS patient to obtain iPSCs and compared several hallmarks of the pathology. We found that both iPSCs and MNs derived from the two tissues showed identical properties and features and can therefore be used interchangeably, giving the opportunity to easily obtain iPSCs from a more manageable source of cells, such as PB.


| INTRODUC TI ON
Amyotrophic lateral sclerosis (ALS) is a devastating and fatal neurodegenerative disease. This disorder is characterized by a progressive motor neuron (MNs) loss in brain and spinal cord, causing paralysis, respiratory failure and death within 5 years from the diagnosis. 1 At molecular level, the DNA/RNA-binding protein TAR (trans-activation response element) DNA-binding protein (TDP-43) was described to be the major component of the pathological aggregates found in brains of ALS patients. 2 The presence of the GGGGCC (G 4 C 2 )-repeat expansion in the noncoding region of the open reading frame 72 (C9ORF72) gene on the chromosome 9 is the most frequent genetic cause of ALS and frontotemporal dementia (FTD). Usually, in healthy subjects, the number of G 4 C 2 -repeats is lower than 30 but conversely can be in the order of thousands in ALS/FTD patients. The C9ORF72 repeat expansion determines several pathological hallmarks, among which there is a toxic gain of function of RNA repeats, accompanied by foci formation. No effective cure for this disease is yet available, mainly due to the fact that pathological mechanisms are difficult to study owing to the impossibility to obtain affected cells from alive patients. In addition, post mortem brain tissues represent only the end stage of the disease limiting the comprehension of cellular and structural defects leading to the onset of neurodegeneration. Studies to elucidate ALS pathological mechanisms have also been limited by the lack of models able to fully mimic affected MNs until the advent of induced pluripotent stem cells (iPSCs). 3 This powerful in vitro model allows the generation of patient-specific cell lines that can be differentiated into MNs, affected in ALS. iPSCs can be obtained by reprogramming different cell types such as CD34 + cord blood cells, 4 keratinocytes 5 or peripheral blood (PB), 6,7 but fibroblasts (FB) remain the most widely used. [8][9][10][11] However, FB do not necessarily represent the best cell source to generate iPSCs, not even displaying the highest reprogramming efficacy and needing in vitro passages before reprogramming increasing the risk to accumulate genetic alteration. Additionally, if compared to PB, a biopsy punch is doubtless more invasive than a venipuncture. PB has been already shown to be an easily accessible source of patient tissue without the need to extensively maintain cells in culture prior to reprogramming experiments. 6 Furthermore, using PB as starting source may allow to achieve longitudinal studies of the same patient over many years and easily create a bio-bank of collected samples. In this work, we investigated and compared the key aspects of iPSCs and iPSC-derived MNs obtained from FB and PB of a C9ORF72-mutated patient.

| MATERIAL AND ME THODS
This study was approved by the ethics committee of IRCCS Istituto Auxologico Italiano.

| Sample collection and processing
Skin biopsy and PB were collected from a 49 years old female affected by ALS and carrying a C9ORF72 repeat expansion. Written informed consent was obtained from the patient. FB were obtained from skin biopsy as follow: the skin tissue fragment was transferred in a 60 mm culture dish and subcutaneous residuals removed. Few drops of FB medium (RPMI 1640 (Euroclone) supplemented with 10% foetal bovine serum (FBS, Sigma-Aldrich), 2 mmol/L l-glutamine, 2.5 μg/mL amphotericin B (Sigma-Aldrich), 100 U/mL penicillin and 100 μg/mL streptomycin) were added, and the sample was allowed to adhere to the bottom of the dish overnight. The following day, fresh medium was added and changed twice a week after attachment and outgrowth of cells from skin dissection. FB were maintained in medium and expanded at 37°C with 5% CO 2 . Reprogramming was performed by the fifth passage. Peripheral blood mononuclear cells (PBMCs) were isolated by layering diluted blood sample on a density gradient (1.077 g/mL) (Histopaque ® -1077, Sigma-Aldrich) followed by centrifugation. PBMCs layered in the plasma-density medium interface were washed, counted and cryopreserved in FBS/10% DMSO until reprogramming.
Briefly, 2 days before reprogramming, FB were seeded in a well of a 6-well plate in FB medium at 30%-60% confluence. After reprogramming, they were maintained in FB medium until day 7 and therefore plated onto Matrigel (Corning) coated dishes to allow colony attachment. For PB, 500 000 PBMCs were seeded in a 24-well plate and cultured with the addition of IL-3 (20 ng/mL), IL-6 (20 ng/mL), SCF (100 ng/mL) and FTL-3 ligand (100 ng/mL) (all from Gibco) in StemPro-34 medium (Thermo Fisher Scientific) for 4 days. The third day after transduction, cells were transferred on Matrigel-coated dishes and cultured in StemPro-34 without cytokines.
For both samples, the eighth day after reprogramming, medium was replaced with the iPSCs-specific Essential 8 medium (Thermo Fisher Scientific). Spent medium was daily replaced with fresh medium and the cultures monitored for the emergence of iPSC colonies. When colonies reached the appropriate size, six clones per sample were picked and transferred on new Matrigel-coated dishes for expansion. Colonies, passaged using an EDTA 0.5 mmol/L solution, were expanded in Essential 8 medium for at least six passages before being characterized and differentiated.

| Characterization of iPSCs clones
After at least six passages in culture, karyotyping and characterization were performed on clones from each starting material.

| Karyotyping
Colcemid solution (KaryoMAX™, Thermo Fisher Scientific) was added overnight to cultured cells in logarithmic phase. Chromosomes were stained with the fluorescent dye quinacrine mustard (Sigma-Aldrich). Q-Band stained chromosomes were analysed.

| Repeat primed PCR
The maintenance of repeat expansion was evaluated by repeat primed PCR as previously described. 12

| Differentiation of iPSCs into neural stem cells (NSCs)
iPSCs were cultured in Induction medium (Neurobasal and 2% Neural Induction Supplement [Thermo Fisher Scientific]) for 7 days. On day eight, cells were harvested by Accutase (Euroclone) and seeded on Matrigel-coated dishes in Expansion medium (50% Neurobasal, 50% DMEM/F12 and 4% Neural Induction Supplement) with addition of 10 μmol/L RHO-associated kinase (ROCK) inhibitor Y27632 (Selleckchem). Cells were allowed to grow and expanded for at least five passages, until they showed a homogeneous neural morphology.

| Differentiation of iPSCs into MNs
MNs were differentiated as previously described. 13 Briefly, iPSCs were grown in 100 mm dishes until confluence, harvested and RT-PCR was performed as previously described to identify expression of Chat and HB9 markers (see Appendix S1).

| TDP-43 localization
Immunofluorescence analysis was performed as previously de-    in SSC 0.1X, the dried blot was placed in the film cassette with the auto-radiographic film and exposed for at least two nights at −20°C before developing.

| Sodium arsenite treatment and stress granules evaluation
MNs were seeded on coverslips at a concentration of 10 × 10 5 cells/ well and on day 10 exposed to 0.5 mmol/L sodium arsenite (Merck).
After 1 hour of treatment, cells were fixed for immunocytochemistry. The presence of stress granules was revealed by immunofluorescence as previously described using a specific anti-TIAR-1 (1:300, Cell Signaling Technology) antibody.

| Quantification and statistical analyses
Bar charts represent the mean ± standard error mean (SEM).
Statistical differences were assessed by Fisher's exact test and t test.
A P value of <.05 was considered significant.

| No differences were observed between FBand PB-derived iPSCs regarding their reprogramming potential and stemness
We found no differences in the reprogramming efficiency comparing the two different cell sources. Both cell types gave rise to more than 12 emerging colonies after reprogramming and 6 clones for each source were picked and showed the same expansion rate.
iPSCs clones from both tissues formed compact colonies, presenting cells with high nucleus/cytoplasm ratio and distinct colony border. Colonies were able to be maintained over 50 passages and transgene vectors were lost within the first five passages, as shown in Figure 1C. iPSCs were subjected to a standard cytogenetic procedure to ascertain their normal karyotype ( Figure 1A), and the maintenance of the C9ORF72 expansion was established by repeat primed PCR ( Figure 1B).
To ensure their stemness, expression of specific intracellular markers was evaluated by RT-PCR (Sox 2, Oct 3/4 and Nanog) and by immunofluorescence (SSEA-4, TRA-1-60 and alkaline phosphatase). Expression of early transcription factors, absent or weakly detectable in starting tissues, confirmed staminality of our iPSCs ( Figure 1D). Additionally, both samples showed positivity for the expression of stemness proteins ( Figure 1E). We conducted pluripotency assays of the clones, assessing their ability to spontaneously differentiate into cells of the three germ layers. iPSCs formed spherical embryoid bodies (EBs) when cultured in suspension for 7 days in non-adherent dishes. Once allowed to adhere on Matrigel-coated dishes, EBs from both clones spontaneously differentiated into endodermal (α-fetoprotein positive), ectodermal (β III tubulin positive) and mesodermal (positive for Desmin) cells ( Figure 1F). Furthermore, if cultured in a specific neural induction medium, iPSCs changed their typical shape, becoming 100% positive for the neural stem/ progenitor cell marker nestin ( Figure 1G) either way.
Given these data, during this preliminary characterization phase, no differences were observed between clones in terms of stemness properties, pluripotency and ability to differentiate into neural cells.

| C9ORF72 mutation-specific hallmarks were present in the same way in both iPSCs and iPSCderived MNs
We subsequently investigated some specific hallmarks of the C9ORF72 mutation. In particular, it is known that the presence of the

| A prevalent nuclear localization of TDP-43 protein is shown in both iPSC-derived MNs
The presence of cytoplasmic aggregates of the TDP-43 protein is another common neuropathological hallmark, even if FB and iPSCs On the other hand, the total amount of TDP-43, evaluated by Western blot (WB), was comparable between FB-and PB-derived iPSCs and between iPSC-derived MNs of both tissues. Higher levels of the protein were observed in MNs vs iPSCs in both models (1.5 and 1.6-fold increase for FB and PB, respectively) ( Figure 3D).

| iPSC-derived MNs produce stress granules following acute oxidative stress treatment
ALS-related proteins are also involved in stress granules dynamics, a highly adaptive mechanism built up in response to stressful environmental insults. To mimic acute stress conditions in vitro, we treated iPSC-derived MNs at day 10 of differentiation with 0.5 mmol/L sodium arsenite for 1 hour and analysed cell responses. Following acute exposure to sodium arsenite, both iPSC-derived MNs were able to produce TIAR positive stress granules in the 100% of the cells as adaptive mechanism ( Figure 3E, with magnification panels).
No differences were found in terms of number of stress granules formed by both samples (average number of 2.5 granules/cell for FB vs 2.2 granules/cell for PB, 80 cells/sample analysed).

| D ISCUSS I ON
This is a proof-of-principle study aimed to compare FB-and PBderived iPSCs from the same individual.
In particular, our aim was to investigate the key aspects of iPSCs and iPSC-derived MNs obtained from both FB and PB of a C9ORF72mutated patient to establish their comparability. Differentiation of human neural progenitor cells from iPSCs has been already demonstrated to occur similarly regardless their somatic tissue origin. 14 However, a comparison of iPSC-derived MNs from two tissues of the same C9ORF72-mutated ALS patient has never been performed.
The C9ORF72 mutation was chosen in this study as the most common genetic cause of ALS and FTD, 15,16 two fatal neurodegenerative diseases. In addition, patient cells carrying this mutation exhibit well known specific pathogenic molecular hallmarks that can be explored, such as RNA foci and repeat expansion length. In our work, initial F I G U R E 2 Comparison of C9ORF72mutated iPSC-derived motor neurons. A, iPSCs from FB and PB were induced to generate embryoid bodies (EBs). After this first step, EBs were dissociated and differentiated into motor neurons (MNs). MNs were stained with the specific nuclear HB9 antigen (red) and co-stained with the microtubule-specific marker β III tubulin (green). The positivity for the pan-axonal neurofilament marker SMI312 was also assessed (red). Nuclei The presence of the G 4 C 2 -repeat expansion in the noncoding region of the C9ORF72 gene leads to a C9ORF72-reduced expression, suggesting haploinsufficiency as one of the potential disease-causing mechanisms. 17 Additional specific pathological hallmark of C9ORF72 patients is represented by the presence of nuclear and cytoplasmic RNA foci formed by sense and antisense G 4 C 2 -repeat transcripts 12,18 as well as accumulation of five different dipeptide repeat proteins (DPRs) generated by mean of a repeat-associated non-AUG (RAN) translation, mostly in the neuron cytoplasm. 19 Previous reports have already demonstrated the presence of both RNA foci and DPRs in C9ORF72-mutated iPSC-derived MNs. 20,21 In this study, we quantified the number of cells showing foci and also the number of foci per cell in our iPSCs and iPSC-derived MNs without finding statistically significant differences among all samples.
The G 4 C 2 -repeat expansion was confirmed by both primed PCR and SB analyses in our FB-and PB-derived iPSCs. In this study, not surprisingly, the two tissues displayed different sizes of the repeat expansion. Indeed, intra-individual dissimilarity of number of repeats between tissues has already been described. 22 In addition, after reprogramming, we observed a contraction in the size of the expansion in both samples. It is conceivable that this contraction, probably caused by reprogramming, is due to a selection of clones with a less extended repeat length and therefore higher stability. Furthermore,  25,26 while no aggregation was observed in two other studies. 13,27 No subcellular localization was observed for TDP-43 protein in iPSCs 20 and iPSC-derived MNs 21 from C9ORF72-mutated patients while nuclear depletion and cytoplasmic accumulation of the protein were reported by Zhang et al. 28 The total amount of the TDP-43 protein evaluated by WB was comparable between FB-and PB-derived iPSCs and between iPSC-derived MNs of both tissues, with a slight increase in MNs. This can be probably due to a higher metabolism and protein translation of the differentiated cells if compared to iPSCs.
No differences were found in the response to acute stress provoked by sodium arsenite treatment. Both iPSC-derived MNs were able to produce TIAR positive stress granules as adaptive mechanism to stressful environmental insults.

| CON CLUS ION
iPSCs and iPSC-derived MNs represent powerful patient-specific models to investigate pathological mechanisms in diseases such as ALS and FTD. In this study, for the first time, we demonstrated that reprogrammed cells from PB are fully comparable with reprogrammed FB, underling that the cell source does not affect iPSC peculiarities, differentiation potential and specific pathological features. For some diseases, such as rare or childhood pathologies, different types of tissue may be available from different patients and we have successfully demonstrated that the resulting iPSCs can be compared anyway. Our findings represent new evidence, leading to develop promising tools for future translational and therapeutic applications.

CO N FLI C T O F I NTE R E S T
The authors confirm that there are no conflicts of interest. F I G U R E 3 Comparison of specific pathological hallmarks in FB-and PB-derived cells. A, The percentage of cells with foci and the number of foci per cell were evaluated in both iPSCs and iPSC-derived MNs by FISH technique. Foci are indicated by head arrows. Scale bar, 5 µm. B, Southern blot analysis on DNA obtained from PB and FB (lanes 1 and 4) and PB-and FB-derived iPSCs (lanes 2 and 5). Sample from a healthy donor (negative control) is shown in lane 3. C, Motor neurons derived from FB-and PB-iPSCs were analysed to observe localization and aggregation of TDP-43 protein (green) by immunofluorescence. Cells were co-stained with the pan-axonal neurofilament marker SMI312 (red). Scale bar, 20 µm. Magnification of cells displaying a cytoplasmic mislocalization. D, Western blot analysis of the total amount of TDP-43 protein in iPSCs (lanes 1 and 3) and iPSC-derived MNs (lanes 2 and 4). Densitometry analysis in lower panel: for both FB and PB-derived MNs, the expression was calculated using as reference the TDP-43 protein level of their corresponding iPSCs. E, Representative images of differentiated MNs following acute treatment with sodium arsenite (0.5 mmol/L, 1 h). After treatment, MNs were immunostained for the stress granule marker, TIAR1 (green) and the pan-axonal neurofilament marker SMI312 (red). Scale bar, 20 µm. Stress granules are indicated by head arrows in the magnification panels. MNs images are representative of two independent experiments

AUTH O R CO NTR I B UTI O N S
DB performed the research, analysed the data and wrote the paper. PB designed the research study, performed part of the experiments, analysed the data, wrote the paper and supervised the all study. All authors read and approved the final manuscript.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data that support the findings of this study are available from the corresponding author upon reasonable request.