Region-specific generation of cholinergic neurons from fetal hu #3

November 17, 2002

In a message dated 11/17/2002 11:44:43 PM Pacific Standard Time, FVJAMES

writes:

> To exclude the possibility of host cells picking up gfp DNA or GFP proteins

> leaked from damaged fetal hNSCs, we performed control transplantations

> using freeze-thawed hNSCs, which were pre-treated with the same priming

> procedure and rAAV-transduction as the experimental groups. We did not

> detect any green fluorescent (GFP+) cells in the host tissue in these

> control rats for at least ten days after surgery (data not shown). In

> animals transplanted with live fetal hNSCs, all nucleated GFP+ cells

> originated from the grafted hNSCs as verified by positive labeling with a

> monoclonal antibody specifically against human nuclei (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4e</A>), and

> negative labeling with a specific antibody against rat cytochrome P450 side

> chain cleavage enzyme (P450scc, <A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4f</A>). Therefore, our data indicate that

> the GFP+ cells were indeed of human origin, and that host cells were not

> labeled by GFP leakage from damaged hNSCs or by spontaneous fusion between

> human and rat cells<A

HREF= " http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/vaop/ncurre\

nt/full/#B28 " >28</A>, <A

HREF= " http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/vaop/ncurre\

nt/full/#B29 " >29</A>.

>

> Morphological examination of fetal hNSC grafts showed that GFP+ cells in

> the pyramidal cell layer of the CA1 region of the hippocampus acquired

> typical pyramidal cell characteristics (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4g</A>). Numerous spines on their

> processes indicated a functional maturation of these grafted hNSCs.

> Immunocytochemical analyses using the neuron-specific markers NeuN (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4h</A>

> ) and TuJ1 (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4iâ€“t</A>) revealed that most GFP+ cells acquire neuronal

> labeling 1 month after transplantation into cortex (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4iâ€“k</A>), hippocampus

> (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4lâ€“n</A>), medial septum (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4oâ€“q</A>) or spinal cord (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F4.html " >F\

ig. 4râ€“t</A>). Only a

> few scattered GFP+ cells were double-labeled with an astrocyte-specific

> markerâ€”glial fibrillary acidic protein (GFAP) (data not shown). No GFP+

> cells were immunoreactive to a monoclonal antibody against

> galactocerebroside (GalC, data not shown), indicating the absence of

> oligodendrocyte differentiation from grafted hNSCs. In addition, negative

> staining using an undifferentiated neural stem cell marker, nestin (data

> not shown), suggested that all grafted cells had differentiated by one

> month after transplantation.

>

> Unprimed hNSCs become astrocytes in adult rat CNS Although CAG-egfp

> transduction did not alter the in vitro differentiation patterns of hNSCs

> with or without priming, as previously reported<A

HREF= " http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/vaop/ncurre\

nt/full/#B26 " >26</A>, it is necessary to

> determine whether this is true for these cells grafted in vivo. To exclude

> the possibility that GFP expression in fetal hNSCs could be responsible for

> the neuronal differentiation observed in vivo, we grafted CAG-egfp-treated

> but not FHL-primed hNSCs into spinal cord, cortex, medial septum and

> hippocampus (2 104 cells and n = 4 for each region).

>

> One month after transplantation, stereological analyses showed that

> percentages of surviving hNSCs were 11.8 2.0% in cortex, 13.8 3.5% in

> medial septum, 16.8 3.2% in hippocampus and 4.8 0.6% in spinal cord.

> These survival rates were similar to those of their primed counterparts.

> Many of the surviving cells, 61 6.2% in cortex (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >F\

ig. 5a</A>), 60.1 4.7% in

> medial septum (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >F\

ig. 5b</A>), 58.7 3.8 in CA1 of hippocampus (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >F\

ig. 5c</A>), 48.9

> 2.3 in dentate gyrus (DG) of hippocampus and 56.9 5.6% in spinal cord (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >

> Fig. 5d</A>), still expressed nestin one month after transplantation. Neither

> neurons nor their GFP+ fibers were detected in the non-neurogenic areas we

> examined. However, some grafted hNSCs became TuJ1+ in the neurogenic

> dentate gyrus of the hippocampus (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >F\

ig. 5e</A>, 11.2 4.8%), which is consistent

> with previous observations<A

HREF= " http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/vaop/ncurre\

nt/full/#B17 " >17</A>. On the other hand, 38.6 8.7%, 39.9 6.0%,

> 40.7 3.7%, 39.6 2.9% and 42.7 5.1% of grafted cells became GFAP+ when

> they were grafted into cortex, medial septum, CA1, DG and spinal cord (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F5.html " >F\

ig.

> 5f</A>), respectively, which is similar to previous reports<A

HREF= " http://www.nature.com/cgi-taf/DynaPage.taf?file=/neuro/journal/vaop/ncurre\

nt/full/#B20 " >20</A>. In summary, CAG-

> egfp-modified hNSCs without FHL priming remained either undifferentiated or

> acquired an astroglial phenotype when grafted into non-neurogenic areas of

> the adult rats. Therefore, both in vitro and in vivo data support the

> conclusion that the FHL priming procedure, but not CAG-egfp transduction,

> is responsible in part for the neuronal differentiation of fetal hNSCs.

>

> Generation of ChAT+ neurons is region-specific Subtypes of fetal

> hNSC-derived neurons in rat CNS were identified by single-channel or merged

> confocal xy images with a z-thickness of either 1 m or 0.3 m after

> immunofluorescence analyses with various specific antibodies. Many of the

> GFP+ neurons were double-labeled with the ChAT-specific antibody in medial

> septum (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F6.html " >F\

ig. 6aâ€“c</A>) and spinal cord (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F6.html " >F\

ig. 6dâ€“i</A>) of the host brain, which

> are regions containing many endogenous ChAT+ neurons. A few GFP+/ChAT+

> double-labeled neurons were found in prefrontal cortex, an area with a

> limited number of endogenous ChAT neurons. We did not find any GFP+/ChAT+

> neurons in the hippocampus, where there are no endogenous ChAT neurons. In

> the spinal cord, some transplanted ChAT+ neurons had a size and morphology

> that was indistinguishable from endogenous motoneurons (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F6.html " >F\

ig. 6gâ€“i</A>).

> Immunohistochemical analyses with antibodies that specifically recognize

> other neuronal subtypes also revealed region-specific patterns. For

> example, glutamate immunoreactivity was found in the majority of GFP+ cells

> grafted into cortex (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F6.html " >F\

ig. 6jâ€“l</A>), some cells in spinal cord and the dentate

> gyrus of hippocampus, a few in medial septum, and none in the CA1 region of

> hippocampus (data not shown). In contrast, the majority of grafted GFP+

> cells in the CA1 region (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F6.html " >F\

ig. 6mâ€“o</A>) as well as some cells in all other

> regions tested (data not shown) were double-labeled with a GABA antibody.

> No GFP+ cells were immunoreactive to TH.

>

> To quantify neuronal differentiation of grafted fetal hNSCs, we counted the

> number of GFP+ cell profiles and the number of each phenotype. The vast

> majority of surviving GFP+ cells differentiated into neurons, as determined

> by TuJ1 staining, including 94.9 1.8% in cortex, 95.4 0.9% in medial

> septum, 96.3 1.0% in CA1, 95.1 1.1% in DG and 94.8 1.9% in spinal cord.

> In contrast, only small percentages of surviving cells became GFAP+

> astrocytes: 4.4 0.9% in cortex, 3.9 1.1% in medial septum, 3.4 0.7% in

> CA1, 4.8 0.7% in DG and 4.4 0.8% in spinal cord. As no significant

> differences were observed between FHL-primed and SFHL-primed hNSCs (n = 5

> for each region in each group), ten animals for each region were pooled to

> obtain quantitative analyses of neuronal subtypes (ChAT, glutamate or GABA)

> with the means ( s.e.m.) (<A

HREF= " http://www.nature.com/neuro/journal/vaop/ncurrent/fig_tab/nn974_F7.html " >F\

ig. 7</A>). Specifically, 61.3 5.4% and 55.5 3.2%

> of the GFP+ cells become cholinergic neurons when grafted in medial septum

> and spinal cord, respectively. Fetal hNSC-derived glutamatergic neurons

> were mainly detected in prefrontal cortex (51.1 1.5%) and in a much lower

> percentage (13.9 1.7%) in spinal cord. Whereas 71.3 4.9% of GFP+ cells

> turned into GABAergic neurons in the CA1 region of hippocampus, smaller

> fractions (20â€“30%) of such neurons were observed in all the other areas

> that we transplanted.

>

November 17, 2002