|
  
- 积分
- 0
- 威望
- 0
- 包包
- 483
|
作者：Sebastian Jessberger, Gregory D. Clemenson, Jr, Fred H. Gage作者单位：Laboratory of Genetics, The Salk Institute for Biological Studies, La Jolla, California, USA
/ C5 J' {& S. W" Y" |# ]: J
& L' Z: ?( }. f! @/ Y" x3 U' I, _0 C
; Y: F+ _, A# b5 H+ M! v2 ]. D- k 4 L3 }( b3 l. ~! w0 O) W4 [
\$ @% I& l6 r; I3 l1 j: q
' W7 N# c( n* ~" c
1 u3 B: q: _4 p/ |" S, y) {& } & L# ~& N+ J# c) s
' I2 d& X$ I, _4 [& @% Y
& ~9 J; G3 g# K/ `7 W; L , s6 J2 U D) s% x$ F
6 ^2 L1 [. M3 Z. V% H1 m 8 F5 ^8 @8 {7 s5 c w! X& k
【摘要】( S- @7 b# [& v, E" F9 w w& }& C
Multipotent neural stem cells (NSCs) can be isolated from various regions of the adult brain and propagated in vitro. Recent reports have suggested spontaneous fusion events among NSCs when grown as free-floating neurospheres that may affect the genetic composition of NSC cultures. We used adult NSCs expressing either red fluorescent protein (RFP) or green fluorescent protein (GFP) to analyze the fusion frequency of rat and mouse NSCs. Fluorescence-activated cell sorting (FACS) revealed that, under proliferating conditions, approximately 0.2% of rat and mouse NSCs coexpressed RFP and GFP irrespective of whether the cells were grown as neurospheres (mouse NSCs) or as attached monolayers (rat and mouse NSCs). Fused cells did not proliferate and could not be propagated, suggesting that aberrantly fused cells are not viable. Furthermore, we found that neither neurospheres nor monolayers grew clonally, because even very low-density cultures had spheres containing both GFP- and RFP-expressing cells and monolayer patches with GFP- and RFP-expressing cells in close proximity. The nonclonal growth between distinct NSC populations strongly suggests the use of careful and precise culture conditions, such as single-cell assays, to characterize potency and growth of NSCs in vitro.
* ^% _* o( A# x6 K
+ x& }. w" L# T: V3 V0 dDisclosure of potential conflicts of interest is found at the end of this article. / G/ I0 U* g8 i, P" C2 C. ?* m
【关键词】 Adult neural stem cell Fusion Neurosphere Monolayer Chimera Nonclonal growth
T0 F. M+ w; B D! a" T INTRODUCTION+ |6 C' K2 y- Q7 ^: E4 C6 a) |3 W; i
) q7 @& s, q; zThe isolation, expansion, and differentiation of adult NSCs in vitro have allowed the extensive molecular and biochemical characterization of the multipotent progenitors that persist in the adult brain. Furthermore, the in vitro expansion of adult NSCs opened the perspective of a potential therapeutic use of propagated NSCs in human disease. Since the first description of adult NSC culture conditions was published . This approach allowed us to screen for fusion events of large-cell numbers using fluorescence-activated cell sorting (FACS) without the need to depend on post hoc identification methods such as immunocytochemistry. We found that, irrespective of the culturing method, approximately 0.2% of mouse and rat NSCs coexpressed GFP and RFP, thus indicating cellular fusion. However, fused cells did not divide and could not be propagated, and died shortly after the fusion event. Surprisingly, we found that neither neurospheres nor monolayers grew within clonal boundaries, but formed heterogeneous spheres or monolayer areas of GFP- and RFP-expressing NSCs, even under low-density conditions. The data presented here show that, under standard culture conditions, NSC fusion is a rare event that does not crucially alter the cellular composition of adult NSC cultures. However, given the heterogeneous growth of NSC clones, precise and accurate culture conditions should be used when determining the behavior and potency of NSCs in vitro.- i' ?2 V. {; \0 S2 G
5 `/ P; Y( c% O1 UMATERIALS AND METHODS
1 q+ J1 `& M# Z' V2 ]4 }
5 @, s( N1 n2 u$ @Neural Stem Cell Isolation and Growth
3 D: B+ i+ r* [+ F1 n! e7 U3 C N: [% _
The hippocampal NSCs (HCNs) isolated from the adult rat brain that we used in this study were described earlier , plus or minus 5 µg/ml heparin for mouse NSCs). Rat NSCs were grown on polyornithine (PORN) and laminin coated plates; mouse NSCs were plated on uncoated tissue culture plates. At approximately 80%¨C90% confluency (for monolayers) or after 5 days (for neurospheres), cells were trypsinized, and a 10th of the plate was propagated.! V8 N! B) Y# s! s8 A
" q' Z& `. r7 [Retroviral Transduction and Coculture of RFP- and GFP-Expressing NSCs
; W9 C: d' w. `: `4 l' ]& L' K; S+ z( B4 [
At passage six (mouse) or passage eight (rat), 1 x 106 NSCs were transduced with approximately 2 x 108 infectious particles of retroviruses expressing either GFP or RFP under the chicken ß-actin promoter . After 4 days under proliferating conditions, transduced NSCs were sorted using FACS to obtain pure populations of either GFP- or RFP-expressing NSCs./ }/ B4 t7 ~- o; ]# m( j% f4 o
) N$ D2 @0 Q" [% v( y
GFP- and RFP-expressing cells were plated at a 1:1 ratio at high-density (100 cells/µl) or low-density (5 cells/µl) conditions. FACS analyses were performed after 4 days of growth under proliferating conditions. To identify potential fusion events after the induction of neuronal differentiation, cells were plated for 1 day in medium containing FGF-2 (rat), or FGF-2, EGF, and heparin (mouse) and then transferred to medium containing 1 µM retinoic acid (RA) and 5 µM forskolin (FSK) for rat NSCs and 5 µM FSK plus 0.5% FBS for mouse NSCs.9 J q, `% M r9 g" ~
6 }1 j+ U& A8 ~' R* ]
Fluorescence-Activated Cell Sorting
* j: _4 `; `# S U
, }4 l: o0 n' F/ _For each condition (monolayer or neurosphere growth) and species (rat or mouse), we performed at least three independent FACS experiments of separate cell populations using a BD FACSvantage Diva system (BD Biosciences, San Diego, http://www.bdbiosciences.com). GFP was excited with a 488-nm water-cooled argon laser; RFP was excited with a 561-nm solid-state laser. The filters used were 530/30 nm for GFP and 630/22 nm for RFP. GFP- and RFP-coexpressing cells were collected in DMEM:F12 medium containing FGF-2 (rat) or FGF-2, EGF, plus or minus heparin (mouse). Cells were either immediately fixed and analyzed using a confocal microscope or pulsed with bromodeoxyuridine (BrdU).
" W8 j- n, W- c7 U) ?( r# D: F
1 {& ?3 M; f3 O' O/ T S& VBrdU Labeling and Immunocytochemistry
0 ~. d1 c/ D! t* s, L3 w6 o! x5 [4 z2 P" j3 ^; t5 h
To analyze the proliferative capacity of fused NCSs, we plated GFP-/RFP-coexpressing cells on 4-well glass chamber slides (coated with PORN-laminin for rat NSCs or uncoated for mouse NSCs). Either 4 or 24 hours after FACS, 5 µM BrdU (Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com) was added into the medium. One hour after the addition of BrdU, cells were fixed (4% paraformaldehyde in 0.1 M PO4 buffer for 20 minutes at 4¡ãC) and nuclei were counterstained with 4'-6-diamidino-2-phenylindole (DAPI). To detect BrdU incorporation, cells were treated with 1 M HCl (30 minutes at 37¡ãC) followed by 10 minutes in 0.1 M borate buffer, and then extensively washed with tris-buffered saline (TBS). Cells were then incubated for 24 hours with the primary antibody at 4¡ãC (rat anti-BrdU; (Harlan Sera-Lab, Loughborough, United Kingdom, http://www.harlaneurope.com; 1:500 in TBS containing 3% donkey serum and 0.01% Triton X-100). A Cy5-conjugated anti-rat IgG (Jackson ImmunoResearch, West Grove, PA, http://www.jacksonimmuno.com; 1:250) was used (2 hours at room temperature) as secondary antibody. BrdU labeling and the endogenous GFP and RFP signals for all other images were analyzed after fixation of the cells and nuclear counterstaining with DAPI using an inverted confocal microscope (Bio-Rad Radiance 2100; Bio-Rad, Hercules, CA, http://www.bio-rad.com). Only general contrast enhancements and color adjustments were carried out using Adobe Photoshop 7.0 for Mac OsX (Adobe Systems, San Jose, CA, http://www.adobe.com).
" \! }! }# r5 O& U( k7 i8 C: H) {2 _) q, R# G
RESULTS+ b8 f3 h7 Y B! u6 p
3 H2 z. M: t7 b' h
Spontaneous Fusion of Mouse and Rat Adult NSCs* b0 Z/ M* Q( @0 c% d# E7 o: ^
" p; m# G/ \9 d7 p" R
We isolated adult NSCs from whole mouse brain or the rat hippocampus as described earlier . At passage six (mouse) or passage eight (rat), we transduced NSCs with a retrovirus coding for either GFP or RFP under the chicken ß-actin promoter. After 4 days under proliferating conditions, cells were sorted for the expression of GFP or RFP. For subsequent fusion analyses, we used pure populations of either RFP- or GFP-expressing adult NSCs. RFP- and GFP-expressing cells were cocultured at two different densities (high density of approximately 100 cells/µl or low density of approximately 5 cells/µl in suspension) under proliferating conditions. After 4 days of coculturing, RFP- and GFP-expressing adult NSCs, we performed FACS to identify RFP-/GFP-coexpressing cells. Using FACS, we found that, independent of the initial plating density, approximately 0.2% (within a range from 0.1% to 0.5%) of NSCs coexpressed RFP and GFP, indicating that an earlier fusion event had occurred (Figs. 1, 2). Figure 1 shows examples of fused NSCs derived from rat (Fig. 1A¨C1C) or mouse (Fig. 1D, 1E) brain that were grown for 4 days under high-density conditions. Note the distinct localization of GFP and the monomeric RFP within the cells, illustrating the actual coexpression of the two transgenes in the same cell.; Q# p. v2 p% c) K2 j) A+ f. O
3 ?2 Q1 s4 [5 K2 o8 t/ @' e EFigure 1. Spontaneous fusion of mouse and rat adult neural stem cells in vitro. (A¨CC): Three examples of fused rat NSCs. (D, E): Examples from mouse NSCs that were grown as monolayers under proliferating conditions. Note the distinct intracellular localization of red fluorescent protein (RFP) and green fluorescent protein (GFP), clearly demonstrating the true coexpression of the two transgenes in the same cell. Fused cells commonly showed condensed DNA, indicating the abnormal state of the cells. The large images are merged depictions of the separate channels for RFP (red), GFP (green), and 4'-6-diamidino-2-phenylindole (blue; single channels are shown in the small images). Scale bar = 10 µm in (C, E) and 5 µm in (D).! @" h' I/ `7 `8 l
0 N7 X& l4 ~: K8 n* C8 U4 VFigure 2. Fusion frequency of adult NSCs in vitro. Fluorescence-activated cell sorting (FACS) analyses reveal low but consistent fusion frequency of adult neural stem cells isolated from rat hippocampus (A) and mouse whole brain (B, C). Shown are examples of FACS experiments from NSCs grown as monolayers (A, C) or neurospheres (B). GFP- and RFP-expressing cells are depicted in blue. Note that, even though the experiments were started with pure populations expressing either GFP or RFP, FACS revealed a nonfluorescent cell population that may result from silencing of the transgene or very low levels of GFP or RFP, respectively. Abbreviations: GFP, green fluorescent protein; RFP, red fluorescent protein.
+ Y# [2 h8 D: W- p e8 l2 J3 y4 l8 y1 ]: w
We also examined rat and mouse NSCs grown as monolayers for the occurrence of fusion events after the induction of neuronal differentiation with RA and FSK. When we analyzed NSC monolayers 4 days after the addition of RA and FSK, we did not find any neuronal cells that coexpressed RFP and GFP (data not shown), suggesting that cells that fused while cycling do not survive.
: W7 H: b! e& r* y7 ~8 Z( T
" z4 ~4 B. f Z/ {# {2 mFused NSCs Do Not Divide and Cannot Be Propagated
% c/ ]) ?/ Z& B+ N1 N
$ }( t. @" g0 Q/ s7 _( W" FAfter FACS, we attempted to propagate GFP-/RFP-expressing cells, but were unable to grow fused cells for more than 24 hours. At that time, many pyknotic cells appeared in the dish, suggesting the death of fused cells. Whereas approximately 30% of control cells incorporated the thymidine analogue BrdU 4 and 24 hours after FACS (not shown), fused cells did not pick up BrdU, indicating that none of the GFP-/RFP-expressing cells were in S-phase at any point after FACS. Because of the lack of continuous growth of fused cells, we were not able to perform chromosomal spreads to characterize the chromosomal content of fused NSCs.5 l8 V# p0 L, U7 t) W
" d5 _9 j) }/ r g& X0 tNSC-Derived Neurospheres and Monolayers Grow Nonclonally
% [2 @2 y4 S% U% H
: d* X9 |3 m6 H eWhile performing the experiments to analyze the fusion frequency of NSCs, we noticed the heterogeneous composition of GFP- and RFP-expressing cells in single neurospheres (mouse NSCs) and monolayer cultures (mouse and rat NSCs; Fig. 3) under coculture conditions of NSCs expressing RFP and GFP. After 4 days of coculture, there were only very few spheres left that expressed GFP or RFP alone (Fig. 3B). This observation was true even when we cocultured RFP- and GFP-expressing NSCs at a low density of 5 cells/µl medium. As early as 24 hours after plating at low density, sphere-initiating cell clusters contained a mixed population of GFP- and RFP-expressing cells (Fig. 3C). Four days after plating the cells at low density, most of the spheres contained GFP- and RFP-expressing cells (Fig. 3E). As seen with the NSCs growing as spheres, we also found that monolayer cultures showed mixed growth of GFP- and RFP-expressing cells irrespective of the initial plating density (Fig. 3A ).. b/ n( `" f* T9 x, _
# V8 [4 `/ p& VFigure 3. Adult neural stem cells grow irrespective of clonal boundaries. Shown are examples of mouse NSCs grown as monolayers (A, D) or neurospheres (B, C, E) under proliferating conditions and either plated at either high (A, B) or low density (C, D, E). Note the patched pattern of red fluorescent protein (RFP)- and green fluorescent protein (GFP)-expressing cells in the vast majority of monolayer areas and neurospheres. As early as 24 hours after plating mouse NSCs at low density, the first chimeric cell aggregates could be observed (C). Four days after plating NSCs at low density, monolayers had already grown in mixed areas of RFP- and GFP-expressing cells, and only very few neurospheres consisted of cells expressing either RFP or GFP alone (E). Insets in (A¨CE) show bright-field images of the areas analyzed. Scale bar = 100 µm (B, E).
4 q# `) C2 G+ c9 } ]
( ?& L) B- N3 ?: SDISCUSSION4 {+ u' M" l4 i& k) X
- z( L6 W. n8 i6 l" M+ h+ V$ {, v
We report two major findings: (a) adult NSCs fuse at a low frequency without dramatically changing the genetic composition of adult NSC cultures, because the fused cells die soon after fusion; and (b) neurospheres and monolayer cultures of NSCs do not respect clonal boundaries and grow as chimerae.) s# Y" \& |1 j
" J& Q$ i/ F9 p/ Z" sThe fact that NSCs have the potential to fuse in vitro has several important implications. Cell fusion may change the genetic composition of cultures and, thus, might lead¡ªover a prolonged culture period¡ªto altered cellular composition of NSC cultures. The potential genetic plasticity of NSC cultures resulting from cellular fusion could, therefore, dramatically change the behavior of NSCs over time. Even though NSCs grown in cultures might change over time found a relatively high degree of aneuploidy in astrocyte monolayers derived from the subventricular zone. These cells have been previously shown to possess NSC characteristics in vitro. However, aneuploidy does not necessarily mean that cells have fused, because aneuploidy may also occur if chromosomes are not properly segregated during mitosis.
6 p& Z0 H( s6 |! } S- x% v; {2 m! Q" Q5 S( E
Nevertheless, there is strong evidence that cell fusion is a rather frequent event in other systems of progenitor/stem cell cultures such as embryonic stem cell (ES) cultures. Such fusion may account for the observed "transdifferentiation" (i.e., the differentiation from somatic progenitor/stem cells from a certain tissue into a different tissue cell type) . Unfortunately, the karyotype of fused cells identified by RFP and GFP coexpression could not be determined in our current study because fused cells did not continue to divide, which is a prerequisite for chromosomal spreads during metaphase. However, the clear coexpression of the GFP and RFP transgenes in mouse and rat NSCs using the two most commonly used culture techniques reliably demonstrated that fusion of NSCs under standard culture conditions is rare, promptly leading to the elimination of fused cells. Therefore, the behavior of NSC cultures under standard culture conditions is not or is only marginally influenced by the occurrence of NSC fusion.& F* z' A) e0 ], E
! e6 J7 I1 u, z) [% ?
While characterizing the fusion potential of NSCs, we were surprised to find that a large fraction of neurospheres consisted of a mixed population of RFP- and GFP-expressing cells. This finding suggests the chimeric growth of neurospheres rather than the clonal expansion of single sphere-forming cells. The apparent nonclonal growth of NSCs in vitro has several important implications. The neurosphere assay is commonly used to assess stem cell properties such as self-renewal and differentiation potential raised concern about the cellular composition of a neurosphere and suggested that the number of true stem cells in a single sphere is very low. Our data extend these concerns and call for caution when using the neurosphere assay as a quantitative measure to assess stem cell properties. If neurospheres are composed of cells that originated from different clones, it would be difficult to use the size of a neurosphere as a measure of proliferative capacity unless the experiment was started using single cells in a mini-well. In fact, the growth of neurospheres may be strongly influenced by the capability of free-floating cells to form aggregates, thus increasing the size of forming spheres.3 U* t3 n. t. [5 U0 x' k( f
& B( {8 i7 ?3 ?# |7 t. h: UCONCLUSION5 I( g/ a; E, ?/ v: p% }3 U
/ p2 I7 _( U! O4 Q$ |In summary, we showed herein that fusion of NSCs is a rare event when these cells are cultured in vitro. Fused cells do not survive and, thus, do not significantly change the genetic composition of NSC cultures. However, we provide evidence that growing neurospheres do not respect clonal boundaries; the vast majority of neurospheres consisted of a mixed cell population of GFP- and RFP-expressing cells. Therefore, attention should be paid when interpreting the results of growth, potency, and proliferative activity of NSC cultures unless the experiments were started with single cells in one mini-well.
2 P/ e+ H' w( o- u6 ~$ h' l; m7 r) j: A6 C$ v7 Y
DISCLOSURE OF POTENTIAL CONFLICTS OF INTEREST/ }2 q% @) g7 U; ?! V
. H+ B# {7 `2 x5 ?6 M% b' ZThe authors indicate no potential conflicts of interest.- z) ^% w1 R7 \ J1 v2 v
. ^/ Z" w1 \; A. [8 B. }& mACKNOWLEDGMENTS9 r) _8 _/ |( T7 ?0 {2 z
: e* {1 j- n2 Z0 j) o
We thank Jasodhara Ray, Stefan Aigner, Alysson Muotri, and Carol Marchetto for helpful comments, as well as Mary Lynn Gage for editing the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (Je297/1-1), the American Epilepsy Society (S.J.), and grants from the National Institute of Aging, the National Institute of Neurological Disorders an Strokes, and the Lookout Fund (F.H.G.).
% x( k) D4 z1 W 【参考文献】
+ Y; E9 T, x0 H, l7 B4 n# H2 I; W
) v6 G. V8 X. T# ?
D# o! D3 f# n: x5 LReynolds BA, Weiss S. Generation of neurons and astrocytes from isolated cells of the adult mammalian central nervous system
1 c' h7 J) i7 l/ ?4 S4 d
/ m: |, I' I/ i8 LKim M, Morshead CM. Distinct populations of forebrain neural stem and progenitor cells can be isolated using side-population analysis. J Neurosci 2003;23:10703¨C10709.
! S7 E+ X1 T6 j3 p. p0 r2 H5 v# C# S
# |2 }) Q% A& B$ I6 XPalmer TD, Ray J, Gage FH. FGF-2-responsive neuronal progenitors reside in proliferative and quiescent regions of the adult rodent brain. Mol Cell Neurosci 1995;6:474¨C486.
, ]8 [! d6 X, O9 _" r! h8 X$ O+ d! t$ H1 Z8 p. Q4 G
Ray J, Gage FH. Differential properties of adult rat and mouse brain-derived neural stem/progenitor cells. Mol Cell Neurosci 2006;31:560¨C573.
' f3 m5 q& {/ \9 `7 M$ e/ B: d# N- t& w) g) f. s
Ray J, Peterson DA, Schinstine M et al. Proliferation, differentiation, and long-term culture of primary hippocampal neurons. Proc Natl Acad Sci U S A 1993;90:3602¨C3606.
' ^, `$ [4 O% a! s9 l" k
' ]3 I( w" e2 U! M0 }5 {; {/ LGage FH, Ray J, Fisher LJ. Isolation, characterization, and use of stem cells from the CNS. Annu Rev Neurosci 1995;18:159¨C192.
" @7 ]: F o: a+ D) s: N$ p: }
$ W# ~- M' m% q$ kZhao C, Teng EM, Summers RG Jr et al. Distinct morphological stages of dentate granule neuron maturation in the adult mouse hippocampus. J Neurosci 2006;26:3¨C11.
) o5 t' S5 m5 l& Z6 `9 L- M3 Z0 t! F% g- y2 N; x, ~6 o
Palmer TD, Takahashi J, Gage FH. The adult rat hippocampus contains premordial neural stem cells. Mol Cell Neurosci 1997;8:389¨C404.- a$ u2 J8 r+ X3 }& k
# r+ m7 N+ N3 \/ G! S
Zietlow R, Pekarik V, Armstrong RJ et al. The survival of neural precursor cell grafts is influenced by in vitro expansion. J Anat 2005;207:227¨C240.
" f$ ^( j+ F7 V
) q# d/ H. P. E, N2 }0 m2 nChen KA, Laywell ED, Marshall G et al. Fusion of neural stem cells in culture. Exp Neurol 2006;198:129¨C135.
9 r4 v7 S, V+ O* s0 c u3 k/ Q6 p ^+ H& r) C
Wurmser AE, Nakashima K, Summers RG et al. Cell fusion-independent differentiation of neural stem cells to the endothelial lineage. Nature 2004;430:350¨C356.
; T$ X- C5 Y. m% v, K) T7 a
& R1 K* n! K+ \/ r5 MWurmser AE, Gage FH. Stem cells: Cell fusion causes confusion. Nature 2002;416:485¨C487.
$ T7 Z3 L# h% h8 [& c E, E0 h0 k$ [% y1 u0 u
Seaberg RM, van der Kooy D. Adult rodent neurogenic regions: The ventricular subependyma contains neural stem cells, but the dentate gyrus contains restricted progenitors. J Neurosci 2002;22:1784¨C1793.
/ n$ z) o2 c# }: N% @$ z" u* M+ W
+ S6 M1 I6 K- h( \) JBull ND, Bartlett PF. The adult mouse hippocampal progenitor is neurogenic but not a stem cell. J Neurosci 2005;25:10815¨C10821.
5 j0 i" x1 J& a$ L9 F/ k) I" _) o/ _- h' ^0 |" O6 h4 t
Reynolds BA, Rietze RL. Neural stem cells and neurospheres¨Cre-evaluating the relationship. Nat Methods 2005;2:333¨C336. |
|
发表于 2009-3-5 00:56
发表于 2009-3-24 08:06
发表于 2009-3-25 14:01
