Reliable generation of induced pluripotent stem cells from human lymphoblastoid cell lines

Abstract

Patient-specific induced pluripotent stem cells (iPSCs) hold great promise for many applications, including disease modeling to elucidate mechanisms involved in disease pathogenesis, drug screening, and ultimately regenerative medicine therapies. A frequently used starting source of cells for reprogramming has been dermal fibroblasts isolated from skin biopsies. However, numerous repositories containing lymphoblastoid cell lines (LCLs) generated from a wide array of patients also exist in abundance. To date, this rich bioresource has been severely underused for iPSC generation. We first attempted to create iPSCs from LCLs using two existing methods but were unsuccessful. Here we report a new and more reliable method for LCL reprogramming using episomal plasmids expressing pluripotency factors and p53 shRNA in combination with small molecules. The LCL-derived iPSCs (LCL-iPSCs) exhibited identical characteristics to fibroblast-derived iPSCs (fib-iPSCs), wherein they retained their genotype, exhibited a normal pluripotency profile, and readily differentiated into all three germ-layer cell types. As expected, they also maintained rearrangement of the heavy chain immunoglobulin locus. Importantly, we also show efficient iPSC generation from LCLs of patients with spinal muscular atrophy and inflammatory bowel disease. These LCL-iPSCs retained the disease mutation and could differentiate into neurons, spinal motor neurons, and intestinal organoids, all of which were virtually indistinguishable from differentiated cells derived from fib-iPSCs. This method for reliably deriving iPSCs from patient LCLs paves the way for using invaluable worldwide LCL repositories to generate new human iPSC lines, thus providing an enormous bioresource for disease modeling, drug discovery, and regenerative medicine applications.

Keywords: B lymphocytes; Developmental biology; Endoderm; Induced pluripotent stem cells; Neural differentiation; Neuron; Pluripotent stem cells; Reprogramming.

Figure 1.

Generation and characterization of LCL-iPSCs. (A): Schematic depicting the episomal reprogramming process and timeline of iPSC generation from LCLs. (B): Bright-field images of the reprogrammed iPSC colonies from control and SMA LCLs show a high nuclear-to-cytoplasmic ratio and are alkaline phosphatase-positive and immunopositive for the surface antigens, SSEA4, TRA-1-60 and TRA-1-81 and nuclear pluripotency markers OCT3/4, SOX2, and NANOG. Scale bar = 75 μm. (C): Flow cytometry analysis of representative clones from the 49iCTR (49iCTR-n2) and 84iSMA (84iSMA-n4) lines showing >96% of cells were immunopositive for SSEA4 and OCT3/4. (D): All the LCL-iPSC lines maintained a normal G-band karyotype as shown from representative lines. (E): Gene chip- and bioinformatics-based PluriTest characterization of eight LCL- or Fib-iPSC lines. (F): Quantitative reverse transcription-polymerase chain reaction (PCR) analyses of POU5F1 (OCT4), SOX2, LIN28, L-MYC, and KLF4 expression in 49iCTR and 84iSMA LCL-iPSCs relative to H9 hESCs. (G): Three sets of PCR primers were used to detect immunoglobulin heavy locus rearrangements occurring in the LCLs and iPSC lines. A clonal positive control and negative H9 hESCs control were included. (H): Epstein Barr virus-related genes (EBNA-1, EBNA-2, BZLF-1, LMP-1, and OriP) were analyzed by PCR analysis of genomic DNA obtained from H9 ESCs, parental LCLs, and daughter iPSC lines. GAPDH was used a loading control. (I): PCR-restriction fragment-length polymorphism assay using Dde1 restriction enzyme digest illustrating maintenance of SMA genotype in SMA LCL-iPSCs as shown by undigested SMN, along with SMN1 and SMN2 in all four LCL-iPSC lines. GAPDH was used as a loading control. Abbreviations: ALK.PHOS., alkaline phosphatase; CDS, coding DNA sequence; CTR, control; Fib, fibroblast; hESC stem cells, human embryonic stem cells; hNPC, human neural progenitor cells; iPSC, induced pluripotent stem cell; LCL, lymphoblastoid cell line; Pla, plasmid; SMA, spinal muscular atrophy.

Figure 2.

Spontaneous and directed differentiation from LCL-iPSCs. (A): Spontaneous in vitro EB differentiation of all four LCL-iPSC lines illustrating iPSC (TDGF) and germ-layer (NCAM1, HAND1, MSX1, and AFP) specific gene expression. GAPDH was used as a loading control. (B): TaqMan human pluripotent stem cell scorecard table showing the trilineage potential of all four LCL-iPSCs. (C): Pairwise correlation coefficients, scatterplots, and expression pattern plots from selected genes in four gene groups (pluripotency, ectoderm, endoderm, and mesoderm) comparing EBs derived from all four LCL-iPSC lines. (D): Representative iPSCs directed to generate ectodermal (Sox2+/β₃-tubulin+ neuronal cells), endodermal (CDX2+/FABP2+ intestinal enterocytes), and mesodermal (CD73+/Collagen type 1+ chondrocytes) cell types from the 49iCTRn2 line. Scale bar = 75 μm. Abbreviations: CTR, control; EB, embryoid body; Ecto, ectoderm; Endo, endoderm; iPSC, induced pluripotent stem cell; LCL, lymphoblastoid cell line; Meso, mesoderm; Pluri., pluripotency; SMA, spinal muscular atrophy.

Figure 3.

LCL- and dermal fibroblast-derived iPSCs can similarly be directed to differentiate into disease-relevant cell types. (A): All LCL-iPSCs were capable of being directed to generate cells that are immunopositive for Nkx6.1, Islet1, SMI32, and ChAT, which represent different stages of MN development. The percentage of efficiency of iPSCs differentiated in to Nkx6.1 (63% ± 6%), Islet1 (44% ± 5%), SMI32 (56% ± 3%), and ChAT (51% ± 4%) immunopositive cells was relative to total Hoechst-positive cells in the culture. Scale bar = 75μm. (B): Quantitative reverse transcription-polymerase chain reaction expression and Western blot analysis showing maintenance of depleted SMN in SMA patient LCL-iPSC directed to i-MNs, at both the mRNA and protein level, as compared with control i-MNs. COX-IV was used as a loading control. (C): iPSCs derived from either LCLs or dermal fibroblasts generate i-MNs at different developmental stages and general neurons (β₃-tubulin) that are virtually indistinguishable from each other. Scale bar = 10 μm. The percentage of efficiency of iPSCs differentiated into cell types immunopositive for Islet1 (LCL: 32% ± 4%, fibroblast [fib]: 33% ± 3%), Nkx6.1 (LCL: 14% ± 2%; fib: 12% ± 3%), SMI32 (LCL: 56% ± 3%; fib-iPSC: 52% ± 3%), ChAT (LCL: 51% ± 4%; fib: 47% ± 5%), and β₃-tubulin (LCL: 74% ± 2%; fib: 72% ± 5%) was relative to total Hoechst-positive cells in the culture. (D): LCL- or fib-iPSCs generate intestinal organoids that were all immunopositive and had similar morphologies for the intestinal transcription factor, CDX2, and the intestinal subtypes including goblet cells (Muc2+), enteroendocrine cells (CGA+), enterocytes (FABP2+), and Paneth cells (Lysozyme+). Scale bar = 75 μm. Abbreviations: ChAT, choline acetyltransferase; COX-IV, Cytochrome c Oxidase Subunit IV; CTR, control; i-MNs, iPSC-derived motoneurons; iPSC, induced pluripotent stem cell; LCL, lymphoblastoid cell line; MN, motoneuron; SMA, spinal muscular atrophy.