Human embryonic stem cell-derived GABA neurons correct locomotion deficits in quinolinic acid-lesioned mice

Abstract

Degeneration of medium spiny GABA neurons in the basal ganglia underlies motor dysfunction in Huntington's disease (HD), which presently lacks effective therapy. In this study, we have successfully directed human embryonic stem cells (hESCs) to enriched populations of DARPP32-expressing forebrain GABA neurons. Transplantation of these human forebrain GABA neurons and their progenitors, but not spinal GABA cells, into the striatum of quinolinic acid-lesioned mice results in generation of large populations of DARPP32(+) GABA neurons, which project to the substantia nigra as well as receiving glutamatergic and dopaminergic inputs, corresponding to correction of motor deficits. This finding raises hopes for cell therapy for HD.

Figure 1. Specification of LGE-like progenitors from hESCs

(A) A schematic procedure for differentiating LGE-like progenitors from hESCs. (B) Pax6-expressing dorsal progenitors decrease while Nkx2.1-expressing ventral progenitors increase in response to increasing concentrations of SHH. (C-E) Quantification of Pax6-, Nkx2.1- (C), Meis2- (D) expressing cell population among total cells and Gsx2 mRNA expression (E) in response to SHH (H9 and H1 cell lines, day 26). (F) Representatives of Meis2-, Mash1-, Nkx2.1-, and Pax6-expressing cells in cultures with medium concentration of SHH (200 ng/ml) or purmorphamine (0.65 μM). (G) Quantification of Meis2, Otx2, Mash1-expressing cells among total cells in the presence of SHH (200 ng/ml) or purmorphamine (0.65 μM). Data are presented as mean±s.e.m.. (*P<0.05. one-way ANOVA test) Blue, Hoechst-stained nuclei. Scale bars: 50 μm.

Figure 2. Differentiation and characterization of GABA neurons

(A) Schematic procedure of GABA neuron differentiation. (B) At day 47, GABA-expressing neurons are present in all cultures but DARPP32-expressing GABA neurons are present only in cultures that were treated with 200 ng/ml SHH but not in higher concentrations of SHH or in both RA and SHH. (C, D) Quantification of GABA and DARPP32 cell populations among ßIII-tubulin+ neurons under the three culture conditions (200 ng/ml SHH, 500 ng/ml SHH, and RA+200 ng/ml SHH). (E) GABA neurons generated under 200 ng/ml of SHH exhibit spiny morphology and express DARPP32, GAD, Meis2. (F) Quantitative analysis of DARPP32 and GAD positive cells among GABA positive cells. (G) HPLC measurement of GABA release from cultures treated with or without high concentration of potassium. (H) Whole cell patch clamping of neurons differentiated for 10 weeks show action potentials and spontaneous inward synaptic currents which are almost completely eliminated by bicuculine. P values (*P<0.05, ** P<0.01) in c, d, f and g were determined using the one-way ANOVA test. Data are shown as means±s.e.m.. Scale bars: 50 m.

Figure 3. Generation of differential GABAergic neuronal types from grafted human neural progenitors

(A) Grafted human forebrain (FBN) and spinal (SPN) GABA neural progenitors survive for 4 months, as revealed by human nuclei (hN) staining, and do not disrupt striatal structures, as indicated by Hoechst staining. (B) Both FBN and SPN generate GABA neurons but the GABA neurons in the FBN group are positive for Meis2. (C) GABA neurons in the FBN but not SPN grafts express DARPP32. (D) The GABA neurons in the FBN group are larger and more branched, and their dendrites, labeled by MAP2, possess more human-specific synaptophysin+ buttons (arrowheads) than those in the SPN group. (E) Stereological analysis estimates a similar number of total grafted cells and GABA neurons in the FBN and SPN grafts but the majority of GABA neurons in the FBN grafts are positive for DARPP32. Data are means±s.e.m.. (**P<0.01. n=10, one-way ANOVA test) (F) The cell body area of FBN GABA neurons is larger than that of SPN GABA neurons. Data are means±s.e.m.. (**P< 0.01. n=20, unpaired t test). Scale bars: 25 μm.

Figure 4. Projection and connections by grafted human GABA neurons

(A) In the striatum, GABA neurons of the FBN but not the SPN grafts possess dense TH+ terminals on cell bodies. GABA neurons of both FBN and SPN grafts have vGLUT1 positive terminals but GABA neurons in the FBN grafts have more (upper row). Most of the TH+ and vGlu+ fibers are negative for human-specific STEM121 (low row). (B) In anterior midbrain (STN/SN), human neuronal fibers, indicated by human specific synaptophysin, are present and surround TH+ dopamine neurons in the brain grafted with FBN but not SPN cells. The DARPP32+fibers in the FBN-grafted nigra are also positive for substance P but not enkephalin (lower row). Scale bars: 25 μm.

Figure 5. Transplantation of human GABA MSN corrects locomotive deficits

(A) Schematic presentation of temporal course of transplantation and behavioral analysis. (B) Rota-rod tests show increased latency in animals transplanted with FBN but not SPN. (C) Open field tests indicate increased center ratio, crossing times, and distances in animals grafted with FBN but not SPN. (D) Treadscan analysis reveals increased stride length, as well as decreased maximum longitudinal deviation, minimum lateral deviation and the foot base of their rear limbs in animals receiving FBN but SPN grafts. (B) and (D) were analyzed using the repeat measures of general linear model (GLM) analysis, and open field behaviors (E) were analyzed by one-way ANOVA. (*P<0.05, n=10).