Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 24 (4), 484-490

Restorative Effects of Human Neural Stem Cell Grafts on the Primate Spinal Cord


Restorative Effects of Human Neural Stem Cell Grafts on the Primate Spinal Cord

Ephron S Rosenzweig et al. Nat Med.


We grafted human spinal cord-derived neural progenitor cells (NPCs) into sites of cervical spinal cord injury in rhesus monkeys (Macaca mulatta). Under three-drug immunosuppression, grafts survived at least 9 months postinjury and expressed both neuronal and glial markers. Monkey axons regenerated into grafts and formed synapses. Hundreds of thousands of human axons extended out from grafts through monkey white matter and synapsed in distal gray matter. Grafts gradually matured over 9 months and improved forelimb function beginning several months after grafting. These findings in a 'preclinical trial' support translation of NPC graft therapy to humans with the objective of reconstituting both a neuronal and glial milieu in the site of spinal cord injury.

Conflict of interest statement


The authors declare that they have no competing financial interests in this work.


Figure 1
Figure 1. Graft Concept, Procedure, Survival, And Differentiation
(A) Schematic of neural stem cell grafting and functional relay forma tion across a contusion injury site (brown). Host axons (blue) grow into the graft (1), graft axons (green) grow into the host spinal cord (2) to complete the relay circuit (3). (B) Surface of spinal cord, with dura retracted laterally and lesion cavity visible (arrow). (C) After grafting, the slightly opaque suspension of hNPCs in the fibrin matrix is visible in the lesion site (arrow). (D) Human spinal cord neural progenitor cell graft (GFP+) is well-integrated into a rhesus monkey C7 hemisection lesion site. Horizontal section, rostral to the left. (E–F) formula image and merged formula image image in 5-month graft. (G) formula image (early neuronal marker) and (H) formula image (early glial marker) were present in all grafts. (I) formula image (mature astrocyte marker) was only present in 5- and 9-month grafts (5 month graft shown). Vimentin and GFAP insets shown with and without formula image. (J) formula image (astrocyte and ependymal cell marker) was present in all grafts (arrows); at 5 and 9 months, 80% of these cells were astrocytes (GFAP+). (K) formula image (oligodendrocyte precursor and immature oligodendrocyte marker) was present in all grafts (arrow) (L) formula image (dividing cell marker) observed at low (1.5%) but detectable levels (arrow) at all time points. Scale bars: D, 1 mm; E–L, 10 μm. Panels E–F consist of five 0.5 μm confocal optical planes, Panels G, J, K, and L consist of two 0.5 μm confocal optical planes, and Panels H and I are single 0.5 μm confocal optical planes.
Figure 2
Figure 2. Changes in Graft Density and Cell Size
(A) Graft cell density significantly declined over time (P = 0.007, N=5 animals, Matlab corrcoef). (B) Mean cell size increased (P = 0.03, N=5 animals, Matlab corrcoef).
Figure 3
Figure 3. Axon Emergence from Grafts
(A) Large numbers of human, GFP-expressing axons emerge from lesion/graft site and grow caudally in linear arrays. Many axons travel at host white/gray matter interface (arrowheads). Horizontal section, subject 5. (B) Numerous linear axons present in host white matter 6mm caudal to the lesion site. (C) Large numbers of axons extend into spinal cord segment T1, 7mm caudal to the lesion. (D–F) Coronal sections at T1 demonstrating high density of human NSC-derived axons in (D) lateral white matter, (E) superficial laminae of the dorsal horn, and (F) lateral motor neuron column, densely surrounding ventral motor neurons (inset, G). (G) Single confocal plane shows many human axon terminals on spinal motor neurons (GFP+ and synaptophysin+; arrows). (H–J) Electron micrographs from T1 ventral gray matter confirm formation of synapses (arrowheads) between GFP+ human axons and host dendrites. Asymmetric synaptic morphology and circular, dense-core vesicles suggest an excitatory synapse. Panel I is a higher magnification of Panel H; Panel J is a second example. Scale bars: A, 250 μm; B, 20 μm; C, 1 mm; D,E,F, 50 μm; G, 10 μm; H, 250 nm; J, 100 nm.
Fig. 4
Fig. 4. Host Axons Regenerate into Human NPC Grafts
(A) Corticospinal axons ( formula image) regenerate short distances (up to 500 μm) into human NPC grafts. Dashed line indicates host/graft interface, revealed by formula image labeling. (B) Sample axon 300 μm within graft, with complex branching pattern and bouton-like swellings. (C–D) Serotonergic ( formula image) axons regenerate into human NPC grafts in all subjects with good rostral host-graft integration. (E–F) Host formula image axons extensively regenerate into NPC grafts. Insets: Graft-derived GFP+ axons do not label for NF200, either within the graft or in host white matter (shown). Scale bars: A,B,D,F, 50 μm; C, 200 μm; E, 1 mm.

Comment in

Similar articles

See all similar articles

Cited by 39 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources