- Email: [email protected]
Novel lacZ-based recombination vectors for mammalian cells
Gene, 137 (1993) 1633169 0 1993 Elsevier Science Publishers
GENE
B.V. All rights reserved.
163
0378-l 119/93/$06.00
07523
Novel ZacZ-based recombination (Genetic
recombination;
PCR; reporter
vectors for mammalian cells
genes; transgenic
Laura B.K. Herzing” and M. Stephen
animals;
embryonic
stem cells)
MeynaTb
Departments of aGenetics and bPediatrics. Yale University School of Medicine, New Haven, CT 06510, USA Received by W.P. Sisk: 31 December
1992; Revised/Accepted:
17 March/25
March
1993; Received at publishers:
20 August
1993
SUMMARY
We have constructed two sets of Escherichia coli &Z-based vectors for use in studies of general mitotic recombination, both in somatic mammalian cells grown in culture and in transgenic animals. The vectors use two mutant copies of the E. coli 1acZ gene as their recombination substrates and contain a neo gene for selection of stable transformants. In one vector, plrec, an SV40 promoter drives lacZ, while the other vector, pArec, utilizes a human non-muscle p-actin promoter for 1ucZ expression. Gene conversions, unequal sister chromatid exchanges and reciprocal exchanges between the two 1ucZ genes result in expression of P-galactosidase, which can be detected in situ by histochemical staining. These vectors yield rates and frequencies of mitotic intrachromosomal recombination in human and rodent cell lines which are similar to rates reported using conventional recombination vectors. Molecular analysis of recombinational events involving the lucZ-based vectors can be carried out on genomic DNA isolated from clonally expanded populations and individual LacZ+ cells and cell clusters can be analyzed using PCR amplification. These reporter gene-based vectors may facilitate the study of recombination in cells with limited proliferative capacities, allow for analysis of both products of an unequal sister chromatid exchange, and permit in situ detection of recombination in the tissues of transgenic animals.
INTRODUCTION
Mitotic genetic recombination is an integral part of many cellular processes (e.g., Austerberry et al., 1984). Recombination can influence the evolution of genes (Bajaj and Blundell, 1984) yet is an important component of many repair processes which maintain genomic fidelity (reviewed in Bollag et al., 1989). Recombination also plays a role in oncogenesis through chromosome rearrangements leading to homozygosis of recessive oncogene alleles (reviewed in Cavanee, 1989). Correspondence to: Dr. MS. Meyn, Department of Genetics, Yale School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. Tel. (l-203) 785-6546; Fax (l-203) 785-7227; e-mail: [email protected] Abbreviations: A-T, ataxia-telangiectasia; act& gene encoding human non-muscle p-actin; Ap, ampicillin; BGal, )3-galactosidase; bp, base pair(s); A, deletion; DMEM, Dulbecco’s modified Eagle’s medium; ES cells, embryonic stem cells; FACS, fluorescent-activated-cell-sorter;
Much of our current understanding of mammalian recombination has been obtained by studying recombination of dominant selectable marker genes (e.g., tk, hph, and neo) contained in vectors transfected into permanent cell lines (Bollag et al., 1989). Using these genes as recombination substrates allows for selection and clonal expansion of cells containing a reconstituted wt gene, one of the product DNAs resulting from a recombinational event. Such vectors have limitations; the requirement of selection to detect a recombinational event usually results in the loss of cells containing the other recombination FDG, fluorescein di-b-o-galactopyranoside; kb, kilobases or 1000 bp; Lx1 and Lx2, mutants of IacZ (5’ and 3’, respectively; see section a and Fig. 1); nro, gene encoding Tn5 aminocyclitol phosphotranserfase; nt, nucleotide(s); PBS, phosphate-buffered saline (0.009 M CaCI,/ 0.0027 M KC1/0.0012 M KH,P0,/0.0005 M MgCI,,6H,0/0.138 M NaC1/0.0081 M Na,HP0,,7H,O); PCR, polymerase chain reaction; SCE, sister chromatid exchange (see Fig. 2); SV40, simian virus 40; tk, thymidine kinase-encoding gene: wt, wild type; XGal, 5-bromo-4chloro-3-indolyl-P-o-galactopyranoside.
164
product,
so that results
obtained
when all products
analysis
are not as informative
in yeast, and selection
transient
as those
are recoverable,
To bypass culture
phenomena
and in transgenic
dergone
activity
and expand
the range
animals,
in mammalian
using
other
two
Unlike
other tacZ-based
1
2
t-, _ll
vector analysis
mammalian
8Y4~~~~(A)
3
animals
pArec vectors recombination
(Murti
systems,
A 5’ la& mutant
iF= Xhd
3’ iacz mutant
ALxl linker (1351 bp)
, .:: ‘. :.w
I
of recombinational
recombination
vec-
6. 0
2
t
4
6
and
can be used to study general mitotic in a variety of mammalian cell lines and
8kb
0
H
B
X
I
X
B PLraC
-I LX2 H
ALxl
*neo_ B
recombi-
et al., 1992), the pLrec
)
X
B
I
_I
LX1 H
X
neo
A Lx1
4
.
hPh
* B
contained
p(lacZ)Lrec
)
)
X
B p(Lxz)Lrec
LX.?
Fig. 1. Construction
A Lx2
of recombination
in the vector pCH
vectors.
)
(A) The wt IacZ gene
1IO (obtained from Dr. G. Nolan, Stanford
University) was modified by inserting an 8-bp Xhol linker (New England Biolabs) into either of two HpuI sites, yielding frame-shift mutant iacZ genes tsl and Lx2. The recombi~tion substrate vector pLrec was constructed by ligating in direct-repeat orientation the
AND DISCUSSION
(a) Recombination We constructed
substrate series design two series of vectors that
tacZ as a reporter the recombination
gene (Fig. 1). Each series consists of vector, a positive control vector con-
use E. coli
taining a wt 1acZ gene and two negative control vectors containing either one or the other mutant 1acZ gene. The is the trpS-1acZ fusion
gene contained in the plasmid pCHll0, in which the first 24 nt of 1acZ have been replaced by the first 33 nt of the E. co/i trpS gene. The fusion gene encodes a hybrid protein that has functional PGal activity in both bacterial
and mammalian cells (Nolan et al., 1988). We inserted 8-bp XhoI linkers into two HpaT sites within the 1acZ gene, creating 5’ (Lxl) and 3’ (Lx2) mutations in the first third of the gene. Both of these mutations abolish detectable BGal activity when transfected into /acZ- E. coli strains WMllOO or CSH26 (data not shown). The recombination vectors contain, in direct repeat orientation, the Lx2 gene, a neo gene, and a 5’ deletion of the Lx1 gene. The neo gene is driven by its own HSV tk promoter (Thomas et al., 1986), which allows for selection of stable integrants of the vector. The first vector series, plrec,
Lx2
3
to
as well as
to detect site-specific
1acZ gene used in these constructs
Lx1
recombina-
tissues.
RESULTS
5kb I
1
between immune gene sequences (Matsuoka et al., or recombination in the male germ line in
transgenic
4
;
laczv
nation 1991)
ene
p8 HllO
Xhol linker
cells which have un-
events can be carried out both from clonally expanded populations and from individual cells and cell clusters. tors, which were designed
site and rrg.
from
Xhol linker (727 bp)
5’facZ
for detection
cell lines which are similar
that molecular
i&&ion bps
mutant
event. We tested these vectors
and mouse
rates determined
of
both in
we constructed designed
and report rates of mitotic intrachromosomal
demonstrate
I 0
vectors
a recombinational
tion in human
to study-
which can be studied
sets of E. coli lacZ-based of wt-PGal
SY40
promoter/enhancer
lad
$
E.co/ilad gene
I
in vivo.
these limitations
recolnbination
‘...,;.
regimes do not permit for
assays, nor can they be readily adapted
ing recombination
A.
e.g., tetrad
was constructed
within
iLlindlII-BumHI-flanked LNZ-lacZ gene, the BumHI-Xhol neo gene fragment from pCMI-neo(polA) (Stratagene, La Jolla, CA. USA), and the ‘~k~~t-B~~zHl fragment of the Lul-lucZ gene (which results in a 5’ deletion of the iurZ gene at the linker insertion site) within the vector pCH1 IO, downstream from the SV40 promoter. The same three fragments are inserted downstream from the human non-muscle fi-actin promoter in the vector pH-B-APr-I-gpt (obtained from Dr. R.M. Liskay) to generate the vector pArec. (B) Positive control vectors for pLrec or pArec, respectively, replace the Lx2 gene with wt iac2. The Lx2 negative control replaces the 5’ deletion of Lxl with a 5’ Xhol deletion of f,.~2, thus eliminating complementary sequence. The Lx1 negative control consists of the Lx1 gene and an izph gene (BatnHI fragment from the vector pHyg, obtained from Bill Sugden, University of Wisconsin). All vectors contain pBR322 ori sequences and ctmp genes for function in E. co/i (not shown). Abbreviations: H. HindHI; B, BarnIll; X, XhoI. Methods: Restriction digests and ligations were conducted using standard methods (Sambrook et al., 1991). Constructs were electroporated into the lacZ_ E. coli strains CSH26 and WMllOO using a BioRad gene pulser following the protocol provided by the manufacturer. 2 ul plasmid DNA (to ng) were added to 40 ul (2 x IO&) electrocompetent bacteria in a 0.1 cm cuvette and electroporated using settings of 25 uF, 200 W, 2.4 kV. Bacteria were then resuspended in SOC media (2% w/v Bacto Tryptone (Difco)/O.S% (w/v) yeast extract/O.010 M NaC1/0.0025 M KCI/O.OlO M MgCl,/O.OlO M MgSO,/O.OZO M glucase), incubated for I h to allow expression and then plated onto LB agar plates containing 100 ltg Ap/ml. For BGaI assays, constructs were electroporated into the /ucZ- E. coli strain CSH26 and plated onto LB agar plates overlaid with 29’,, w/v Bluogal (BRL). After overnight incubation
at ?7’ C, blue
( LacZ’) cells were scored visually.
165
EcoR I
promoter
Xho I Xho I
-q----&-_ Lx2
Xh:l
neo
Xho I
pLrec (LaCiT) neo
A Lxi
I( lacZ+
A Lx1
neo Gene
Fig. 2. The pLrec and pArec recombination
ConversIon (Lacz+)
vectors.
Unequal SCE (Lacz+)
Reciprocal lntrachromatid Exchange (LacZ+)
(A) Vector maps. (B) Recombination
mechanisms
by which a wt lacZ gene can be reconstituted
from the mutant lacZ genes contained in the recombination substrate vectors pLrec or pArec include unidirectional transfer of wt nt sequence from the ALxl gene to the Lx2 gene (gene conversion) and reciprocal exchange. Reciprocal exchange of sequence between the two mutant lacZ genes can occur either within the vector (reciprocal intrachromatid exchange) or between vectors located on sister chromatids (unequal SCE). In both cases of reciprocal
exchange,
the neo gene and the ALxl
gene are lost.
the plasmid pCHl10, which contains an SV40 promoter/enhancer element. In the second vector series, pArec, constructed using the plasmid pH+APr-1-gpt (Gunning et al., 1987), lacZ is driven by a human nonmuscle l3-actin promoter, which exhibits a wide range of tissue-type expression in transgenic mice (e.g., Galski et al., 1989). For pLrec and pArec those recombinational events that result in the reconstitution of a wt 1acZ gene and expression of functional PGal can be detected. These events include gene conversion, unequal sister chromatid exchange and reciprocal intrachromatid exchange (Fig. 2B). Both recombination vectors yield rare PGal+ cells (putative recombinants) when transfected into AlacZ- E. coli strains, and constructs containing wt 1acZ show ubiquitous expression in these bacteria (data not shown). (b) Detection of LacZ+ recombinants and determination of recombination rates In human fibroblasts transfected with vectors pLrec or pArec and stained with XGal, rare PGal-producing cells and cell clusters could be identified easily and clearly under 25 x magnification in cultures containing l-100 positive cells per lo7 (Fig. 3). The number of cells in each cluster were then verified by further examination under higher magnification.
Fig. 3. colony curred LacZ+
Staining and identification of LacZ’ fibroblasts. A sectored A-T containing LacZ+ cells. A putative recombinational event oclate in colony formation, resulting in a cluster of two turquoise cells. Methods: XGal staining: Fibroblasts in monolayer were
stained as described by Lim and Chae ( 1984). After aspiration of media, cells were fixed for 5 min at 4°C in a solution of 0.2% glutaraldehyde/2% formaldehyde in PBS. Cells were then washed once in PBS and covered with staining solution: 44 mM HEPES buffer/l5 mM NaCl/1.3 mM MgClJ3 mM K,ferrocyanide/3 mM K.ferricyanide/l.25 ug per ml XGal (Sigma). Cells were incubated for 4424 h at 37°C and turquoise cells scored visually.
Using cell lines determined to be single-copy for our vectors by Southern blot analysis, recombination frequencies and rates were measured for two human
166
A. w.4
-9
4174-
~1463-406-1140-34(!2$
I
(I&)
(713)
H
yP I
A
(6li3,
(4exgS)(S7)
I
I A+
(29)
I I
W-C
I
Lx2
(1726)
(4774)
B
yP
X
(5892)
d
facZ
B
YP
I
c
pLrec
A Lx1
1 C
B
gene conversion A Lx1
reciprocal exchange/ unequal SCE
IacZ
c.
B. left-hand right-hand IacZ gene IacZ gene A Lx1 Lx2
left-hand IacZ gene wild We
right-hand /acZgene A Lx1
Clone I 4.2
F3
LN L4
-4.17 -3.86
-3.03 -2.41
5L20-F4
Fig. 4. Molecular analysis of La&? recombinants. (A) Restriction maps for unrearranged plrec, gene conversion of pLrec and reciprocal exchange/un~qual SCE of ptrec. Restriction enzyme sites are shown above and homologous target sequences for the PCR primers are shown below each map. Map positions, in base pairs from the vector‘s unique Hind111 site, are indicated in parentheses. B, Burnt-II: H, ~~~dIl1; Hp. Hpnl: X, Xkol. (B) PCR analysis: 5L20 is an A-T fibroblast clone that contains a single unrearranged copy of the pLrec vector while 5L20-F4 is a FACS-sorted LacZ+ derivative of 5L20. PCR primer pair A-C amplifies a 1.01.kb fragment specific to the left-hand IacZ gene; primer pair B-C amplifies a 1.14-kb fragment and is specific for the right-hand neo-/acZ junction. The Lx2 gene PCR product of SL20 yields 905 and 10%bp fragments upon Hpul digestion and 740- and 273-bp fragments upon XhoI digestion; indicating that the amplified DNA contains a single restriction site for each enzyme. XhoI and Hpul digests are diagnostic of 5L20-F4 having undergone a gene conversion event involving the Ieft-hand fucZ gene (Lx?); the Xhol linker, represented by an X&I site, has been replaced by sequence donated by &xl, thus regenerating the HpaI site found in wt In& and giving three fragments (624, 273 and 108 bp) on NPLII digest of the left-hand gene. (C) Southern analysis: LNL4 is a Lesch-Nyhan syndrome human fibroblast line that contains a single unrearranged copy of the pLrec vector while clones 4.2 and F3 are FACS-sorted LacZ’ derivatives of LNLC Genomic DNAs from these clones were obtained and digested with either Hind111 and BamHI or HindHI, XhoI and BurnHI, then electrophoresed, blotted and probed, as described in Table I, using the 4.2.kb BarnHI-BurnHI fragment from pLrec that contains both nco and 1acZ sequence. The probe detects 4.2- and 3.8.kb HindIII-BumHI fragments containing 1ucZ sequence in the parent cell line (LNL4) and clone 4.2, but the 3.8-kb fragment is resistant to XizuI-digestion in clone 4.2, indicating that clone 4.2 is the product of a gene conversion event. DNA from clone F3 contains only a single 3.8.kb icrcZ fragment that is resistant to Xhof digestion, consistent with F3 containing a wt lacZ gene that is the product of a reciprocal exchange or unequal SCE. Methods: FDG staining of trypsinized fibroblasts was as described by Nolan et al. (1988) and cells were brought to a final concentration of 2 x IO’ cells/ml. Fluorescent cells were sorted utilizing a modified Beckton-Dickinson fluorescence-activated cell sorter. Sorted cells were plated and grown without G418 selection. Clones were isolated and analyzed for BGal activity using XGal. Genomic DNA was then obtained
167 TABLE
I
Spontaneous
intrachromosomal
recombination
Total number
Cell line”
of LacZ+
rates of lacZ genes in mammalian
cells containing
Total number
cells detectedb
the pLrec vector Rate of conversion
of cells screened
(conversions/cell (x 10-y Human cell lines GM847-LNL6 GM5849-5L20
Murine ES cell line CCE-4-BGal-9
“Each cell line contains
0.72
10
2.10 x lo6 3.76 x lo3
443
43.0 x lo3
14 72
5.36 x lo5
1.3
7.23 x lo5
3.1
a single copy of the pLrec vector integrated
with 0.25% trypsin (Gibco) dSpontaneous recombination
152 225
into their genomic
‘As determined by microscopic examination following XGal staining. “Determined by direct microscopic examination if < 10000 cells. For populations
analysis tables Methods: For 500-2000 cells fibroblast line,
0.70
7.60 x lo6
206 41
and then counting the resulting cell suspension rates were derived from the number of LacZ+
to LacZ’ generation)d
DNA (see Methods
below).
> 10 000, total cell number
was determined
by treating
using a hemocytometer. cells detected and the total number
the cells
of cells screened
using the fluctuation
provided by Capizzi and Jameson (1973). each experiment, dishes were seeded with enough cells to provide 100-500 colonies/dish, grown until individual colonies contained each, then stained for LacZ expression with XGal and LacZ+ and LacZ- cells counted. Cell culture: GM847, a human Lesch-Nyhan and GM5849, an A-T fibroblast line of complementation group D, were obtained from the Human Genetic Mutant Cell Repository,
Camden, NJ. The murine ES line CCE was obtained from Dr. J. Eisenstadt, Yale University. The pLrec vector was electroporated into these three cell lines, stable transformants isolated and their genomic DNA analyzed in order to identify clones that contained single copies of the pLrec vector integrated into their genomic DNA (e.g., GM847-LNL6, GM5849-5L20 and CCE-4-BGal-9) Cells were maintained to dMEM supplemented with 15520% fetal calf serum. ES cells were grown using feeder layers or supplemented with 50% buffalo rat liver cell conditioned media. Cell transformation: Cells were grown exponentially and passaged 2-3 days prior to electroporation. Cells at 50&75% confluence were trypsinized. washed twice in cold in PBS to a final concentration of (l-2) x 10’ cells/ml. Vector DNA, linearized using Seal, was added to PBS w/o Ca’+ or Mg’+ and resuspended a final concentration of 10 pg/ml, 0.8 ml of the mixture placed into a 0.4-cm gap cuvette (BioRad) and placed on ice for 10 min. Electroporation was performed on a BioRad Gene Pulser (settings 7OOV, 25 pF.) The cells were incubated on ice for a further 10 min and then plated in 15% DMEM. After 48 h, G418 (Gibco) at 400 ug/ml was added for selection. After 12-21 days growth, drug-resistant clones were isolated and grown in the presence of antibiotic. Genomic DNA was prepared according to Sambrook et al. (1989) and Southern hybridization using Hybond-N nylon membranes was carried out as recommended by the manufacturer (Amersham). Blots were probed lacZ sequence, labeled with 32P using a random-priming kit (Boehringer-Mannheim).
SV40-transformed fibroblast lines and for a murine embryonic stem cell (ES) line (Table I). Rates for GM847, a Lesch-Nyhan syndrome line, averaged 0.71 x 10m5 events/cell generation, which are comparable to rates determined for other mammalian fibroblast lines using conventional recombination vectors containing dominant selectable marker genes as recombination substrates (Bollag et al., 1989). Similar rates were also observed for CCE, a pluripotent mouse embryonic stem cell line. Rates for GM5849, a cell line derived from an A-T patient, averaged 1.9 x 10m3 events/cell generation, > 200-fold higher than those seen in the Lesch-Nyhan line, confirming the hyper-recombinogenic A-T phenotype that was previously observed using a dominant-selectable marker-
for Southern analysis and PCR amplification, Digests bromide-stained 0.7% agarose gel while PCR products
with the 3.8-kb HindIII-BamHI
fragment
of pLrec
that contains
based recombination vector (Meyn, 1993). We screened five independently derived stable transformants of GM5849 containing the negative control vector p(Lxl)Lrec and three containing the negative control vector p(Lx2)Lrec for the presence of LacZ+ cells. We found no LacZ+ cells in over lo8 cells screened for each stable transformant and conclude that reversion frequencies of the mutant lacZ genes are < 10m8. On average, lo-15% of cells in a population of G418-resistant fibroblasts or murine embryonic stem cells stably transformed with the positive control vectors were found to be LacZ+ by XGal staining, with some clones showing no detectable expression and many with >50% of the population producing detectable PGal.
of genomic DNA for Southern analysis were electrophoresed overnight through an ethidium were electrophoresed through a 1.3% agarose gel. Standard PCR: 1 pg of genomic DNA from
each recombinant was amplified enzymatically in a 25 ~1 reaction mixture which included 15 ng each of amplimer pairs A-C or B-C/200 pm of each deoxynucleotide triphosphate/l.25 units of Tuq polymerase/l.5 x Tay buffer (Boehringer-Mannheim). PCR was performed in a programmable thermal cycler (Perkin-Elmer Model 9600) for 30 cycles, each consisting of 30 s of denaturation at 94”C, annealing at 64°C and polymerization at 72°C. Primers: A, 5’-AAT CCG CCG TTT GTT CCC AC; B, 5’-TGC GAA GTG GAC CTG GGA CC; C, 5’-ACG ATC GCG CTG CAC CAT TC.
168
A
B # of cells amplified
0 0
# of cells amplified 23312M
1211643
kb ALxl
1.14)
bP 864
_/
-
Lx2 1.01)
624
273
108
Fig. 5. PCR ampfification of LacZ’ cell clusters. (A) One to six cells were picked from the same LacZ” sector of a colony of 5L20 fibroblasts and amplified using the two-step PCR method described below. As shown, neither of the blank lanes contains l.Ol- and 1.14-kb fragments characteristic of ~LICZgene amplification, while two of three reactions containing a single LacZ’ cell produced the characteristic fragments, as did the rest of the reactions. (B) HpnI digest of DNAs obtained from PCR amplification of samples of one to three ceils taken from the same LacZ+ sector. The resultant bands (864,624,273, and 108 bp) are diagnostic for a gene conversion event. Abbreviations: C, control pLrec plasmid; M, size marker. 0, 1, 2, 3, 4 and 6 indicate the number of ceils contained in each PCR reaction mixture. Methods: Two-step PCR: Individual LacZ+ clusters were harvested using a glass micropipette, transferred into PCR tubes containing 10 ~1 standard PCR and 0.5 mg/ml protease K, and incubated at 37°C for 2 h. 2.5 ~1 were transferred to a reaction tube containing for 20 min at 94°C. Remaining reagents were then added to a finai volume of 25 d, and PCR performed for as for standard PCR (Fig. 4 legend). Aliquots of PCR products were run directly on an ethidium-stained 1.3% with HpaI; HpaI buffer and enzyme
were added
directly
to the PCR reaction
(c) FDG staining and FACS sorting of individual ZacZ+ recombinants XGal staining is toxic to mammalian cells and does not allow for clonal expansion of recombinant LacZ+ cells after staining. We therefore have made use of fluorescein di-P-o-galactopyranoside (FDG), a non-toxic galactose analog which is cleaved by E. coli PGal to yield fluorescein. This lack of toxicity has allowed us to use FDG staining to identify, isolate and clonally expand LacZ+ colonies from mixed populations of La&+ and LacZ- colonies (data not shown). FDG staining also allows for isolation of LacZ+ cells by cell sorting. E.g., we used a fluorescent activated cell sorter (FACS) to collect the 0.3% most fluorescent cells from a population of FDG-stained A-T fibroblasts containing a single integrant of pLrec (5L20), initially seeded at low density to lessen the recovery of jackpots of previously recombined LacZ+ cells which may have been present in the original stock. Sorted cells were plated and clones picked and expanded for analysis. Over 50% of the colonies arising from the sorted population were LacZ+ by histochemical staining. Although
The numbers cells and cell
buffer supplemented with 0.1% NP40 20 ~1 of 1 x PCR buffer and incubated 50 cycles, 1 min/cycie at temperatures agarose gel, or first digested overnight
mixture.
not required for isolation of LacZ’ A-T fibroblasts, two consecutive sorts were necessary for efficient recovery of IacZ’ recombinant Lesch-Nyhan fibroblasts due to low spontaneous recombination rates. High levels of expression in FACS-sorted lines, as indicated by > 75% of cells showing intense XGal staining, persisted after passage and expansion. All FACS-sorted LacZ+ clones analyzed proved to be re~ombinants (data not shown). (d) PCR amplification and molecular analysis of individual recombinant clusters The two mutant IacZ genes contained in the recombination vectors differ from each other and the wt lacZ gene in having distinct 5’ and 3’ polymorphisms which allow for molecular analysis of the recombination products contained in a LacZ’ cell by Southern blotting or PCR amplification followed by restriction enzyme digestion (e.g., Fig. 4A and B). Using the primers described in Fig. 4, we have successfully amplified both of the _ I-kb 1ac.Z target sequences of plrec, starting from suspensions that contained fewer than seven LacZ+ cells (data not shown). We have suc-
169 cessfully
analyzed
attached
cells as well (Fig. 5A), sug-
gesting that it will be possible lacZ gene fragments
directly
to PCR-amplify from clusters
diagnostic
of LacZ+
cells
found in tissues of transgenic animals. Once amplified, the 1acZ gene fragments can then be digested to help determine
the nature
to BGal production
of the recombinant
events that led
(e.g., Fig. 5B). Although
ACKNOWLEDGMENTS
PCR ampli-
fication can be carried out on single LacZ+ cells, we have restricted our analysis to cells obtained from LacZ+ clus-
We wish to thank R. Carbone
E. Brainerd
for technical
of the Yale Comprehensive
assistance,
Cancer
Center
for FACS sorting, G. Ruano for advice on PCR amplification and Dr. R.M. Liskay for invaluable advice and discussion.
This work
was supported
from the National
Institutes
American
Society (MG-15 ).
Cancer
in part by grants
of Health (ES05327)
and the
ters of two or more cells. In that way, we avoid analyzing cells that contain a situation
both sister chromatids
that may complicate
(e) Conclusions (I ) We have constructed
of unequal
molecular
SCEs,
analysis.
two sets of E. coli &Z-based
vectors designed for detection of wt BGal activity in mammalian cells which have undergone a recombinational event. These are the first &Z-based vectors to be described that can be used in studies of general genetic recombination in a variety of mammalian cell lines in vitro and tissue types in vivo. (2) These vectors yield rates and frequencies of mitotic intrachromosomal recombination in human and rodent cell lines which are similar to rates reported using other vector systems. (3) Molecular analysis of recombinational events involving the lacZ-based vectors can be carried out on genomic DNA isolated from clonally expanded populations. It also may be possible to analyze both products of gene conversions and unequal sister chromatid exchanges, events that do not involve production of nonreplicating episomes, by means of FDG staining to identify colonies that contain both LacZ’ and LacZ- cells followed by isolation and clonal expansion of both cell types from such sectored colonies and analysis of their genomic DNA. Unlike dominant-selectable markerbased vectors, recombinant cells containing pLrec or pArec do not have to be clonally expanded prior to analysis, as individual LacZ+ cells and small clusters of cells can be studied using PCR amplification. (4) Advantages and potential applications of these vectors in studying mammalian mitotic recombination include: rapid determination of recombination frequencies in cell lines as well as cells with limited proliferative capacity (e.g., primary fibroblasts); detection and analysis of individual recombinant cells; detection of recombination in non-dividing cells; and in situ detection of recombination in the tissues of transgenic animals.
REFERENCES Austerberry, C.F., Allis, C.D. and Yao, M.C.: Specific DNA rearrangements in synchronously developing nuclei of Tetrahymena. Proc. Natl. Acad. Sci. USA 81 (1984) 7383-7387. Bajaj, M. and Blundell, T.: Evolution and the tertiary structure tein. Annu. Rev. Biophys. Bioeng. 13 ( 1984) 453--492.
of pro-
Bollag, R., Waldman, A.J. and Liskay, R.M.: Homologous recombination in mammalian cells. Annu. Rev. Genet. 23 (1989) 1999225. Capizzi, R.L. and Jameson, J.W.: A table for the estimation of the spontaneous mutation rate of cells in culture. Mutation Res. 17 (1973) 1477148. Cavanee,
W.K.: Current
knowledge
of heritable
tumors.
Birth Defects
25 (1989) 585-599. H., Sullivan, M., Willingham, M., Khew-Voon, C., Galski. Gottesman, M., Pastan, I. and Merlino, G.: Expression of a human multidrug transgenic
resistance cDNA (MDRl) in the mice: resistance to daunomycin-induced
bone marrow of leukopenia. Mol.
Cell. Biol. 9 (1989) 4357-4363. Gunning, P., Leavitt, J., Muscat, G., Ng, S.Y. and Kedes, L.: A human B-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA 84 (1987) 4831-4835. Lim, K. and Chae, C-B.: A simple assay for DNA transfection by incubation of the cells in culture dishes with substrates for betagalactosidase. Matsuoka, Muller, somatic
Biotechniques
7 ( 1989) 5766579.
M., Nagawa, F., Okazaki, K., Kingsbury, L., Yoshida, K. U., Larue, D.T., Winer. J.A. and Sakano, H.: Detection of DNA recombination in the transgenic mouse brain. Science
254(1991)81-86. Meyn, M.S.: Elevated rates of spontaneous intrachromosomal recombination in ataxia telangiectasia. Science 260 ( 1993) 132771330. Murti, J.R., Bubulis, gene conversion
M. and Schimenti, J.C.: High-frequency in transgenic mice. Mol. Cell. Biol.
2545-2552. Nolan, G.P., Fiering, Fluorescence-activated on beta-o-galactosidase
germ line 12 (1992)
S., Nicolas, J.-F. and Herzenberg, L.: cell analysis and sorting of viable cells based activity after tranduction of Escherischia
coli lacZ. Proc. Natl. Acad. Sci. USA 85 (1988) 2603-2607. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Laboratory Manual. Cold Spring Harbor Laboratory
Cloning. A Press, Cold
Spring Harbor, NY, 1989. Thomas, K., Folger, K. and Capecchi, M.: High frequency targeting of genes to specific sites in the mammalian genome. Cell 44 (1987) 419-428.
GENE
B.V. All rights reserved.
163
0378-l 119/93/$06.00
07523
Novel ZacZ-based recombination (Genetic
recombination;
PCR; reporter
vectors for mammalian cells
genes; transgenic
Laura B.K. Herzing” and M. Stephen
animals;
embryonic
stem cells)
MeynaTb
Departments of aGenetics and bPediatrics. Yale University School of Medicine, New Haven, CT 06510, USA Received by W.P. Sisk: 31 December
1992; Revised/Accepted:
17 March/25
March
1993; Received at publishers:
20 August
1993
SUMMARY
We have constructed two sets of Escherichia coli &Z-based vectors for use in studies of general mitotic recombination, both in somatic mammalian cells grown in culture and in transgenic animals. The vectors use two mutant copies of the E. coli 1acZ gene as their recombination substrates and contain a neo gene for selection of stable transformants. In one vector, plrec, an SV40 promoter drives lacZ, while the other vector, pArec, utilizes a human non-muscle p-actin promoter for 1ucZ expression. Gene conversions, unequal sister chromatid exchanges and reciprocal exchanges between the two 1ucZ genes result in expression of P-galactosidase, which can be detected in situ by histochemical staining. These vectors yield rates and frequencies of mitotic intrachromosomal recombination in human and rodent cell lines which are similar to rates reported using conventional recombination vectors. Molecular analysis of recombinational events involving the lucZ-based vectors can be carried out on genomic DNA isolated from clonally expanded populations and individual LacZ+ cells and cell clusters can be analyzed using PCR amplification. These reporter gene-based vectors may facilitate the study of recombination in cells with limited proliferative capacities, allow for analysis of both products of an unequal sister chromatid exchange, and permit in situ detection of recombination in the tissues of transgenic animals.
INTRODUCTION
Mitotic genetic recombination is an integral part of many cellular processes (e.g., Austerberry et al., 1984). Recombination can influence the evolution of genes (Bajaj and Blundell, 1984) yet is an important component of many repair processes which maintain genomic fidelity (reviewed in Bollag et al., 1989). Recombination also plays a role in oncogenesis through chromosome rearrangements leading to homozygosis of recessive oncogene alleles (reviewed in Cavanee, 1989). Correspondence to: Dr. MS. Meyn, Department of Genetics, Yale School of Medicine, 333 Cedar Street, New Haven, CT 06510, USA. Tel. (l-203) 785-6546; Fax (l-203) 785-7227; e-mail: [email protected] Abbreviations: A-T, ataxia-telangiectasia; act& gene encoding human non-muscle p-actin; Ap, ampicillin; BGal, )3-galactosidase; bp, base pair(s); A, deletion; DMEM, Dulbecco’s modified Eagle’s medium; ES cells, embryonic stem cells; FACS, fluorescent-activated-cell-sorter;
Much of our current understanding of mammalian recombination has been obtained by studying recombination of dominant selectable marker genes (e.g., tk, hph, and neo) contained in vectors transfected into permanent cell lines (Bollag et al., 1989). Using these genes as recombination substrates allows for selection and clonal expansion of cells containing a reconstituted wt gene, one of the product DNAs resulting from a recombinational event. Such vectors have limitations; the requirement of selection to detect a recombinational event usually results in the loss of cells containing the other recombination FDG, fluorescein di-b-o-galactopyranoside; kb, kilobases or 1000 bp; Lx1 and Lx2, mutants of IacZ (5’ and 3’, respectively; see section a and Fig. 1); nro, gene encoding Tn5 aminocyclitol phosphotranserfase; nt, nucleotide(s); PBS, phosphate-buffered saline (0.009 M CaCI,/ 0.0027 M KC1/0.0012 M KH,P0,/0.0005 M MgCI,,6H,0/0.138 M NaC1/0.0081 M Na,HP0,,7H,O); PCR, polymerase chain reaction; SCE, sister chromatid exchange (see Fig. 2); SV40, simian virus 40; tk, thymidine kinase-encoding gene: wt, wild type; XGal, 5-bromo-4chloro-3-indolyl-P-o-galactopyranoside.
164
product,
so that results
obtained
when all products
analysis
are not as informative
in yeast, and selection
transient
as those
are recoverable,
To bypass culture
phenomena
and in transgenic
dergone
activity
and expand
the range
animals,
in mammalian
using
other
two
Unlike
other tacZ-based
1
2
t-, _ll
vector analysis
mammalian
8Y4~~~~(A)
3
animals
pArec vectors recombination
(Murti
systems,
A 5’ la& mutant
iF= Xhd
3’ iacz mutant
ALxl linker (1351 bp)
, .:: ‘. :.w
I
of recombinational
recombination
vec-
6. 0
2
t
4
6
and
can be used to study general mitotic in a variety of mammalian cell lines and
8kb
0
H
B
X
I
X
B PLraC
-I LX2 H
ALxl
*neo_ B
recombi-
et al., 1992), the pLrec
)
X
B
I
_I
LX1 H
X
neo
A Lx1
4
.
hPh
* B
contained
p(lacZ)Lrec
)
)
X
B p(Lxz)Lrec
LX.?
Fig. 1. Construction
A Lx2
of recombination
in the vector pCH
vectors.
)
(A) The wt IacZ gene
1IO (obtained from Dr. G. Nolan, Stanford
University) was modified by inserting an 8-bp Xhol linker (New England Biolabs) into either of two HpuI sites, yielding frame-shift mutant iacZ genes tsl and Lx2. The recombi~tion substrate vector pLrec was constructed by ligating in direct-repeat orientation the
AND DISCUSSION
(a) Recombination We constructed
substrate series design two series of vectors that
tacZ as a reporter the recombination
gene (Fig. 1). Each series consists of vector, a positive control vector con-
use E. coli
taining a wt 1acZ gene and two negative control vectors containing either one or the other mutant 1acZ gene. The is the trpS-1acZ fusion
gene contained in the plasmid pCHll0, in which the first 24 nt of 1acZ have been replaced by the first 33 nt of the E. co/i trpS gene. The fusion gene encodes a hybrid protein that has functional PGal activity in both bacterial
and mammalian cells (Nolan et al., 1988). We inserted 8-bp XhoI linkers into two HpaT sites within the 1acZ gene, creating 5’ (Lxl) and 3’ (Lx2) mutations in the first third of the gene. Both of these mutations abolish detectable BGal activity when transfected into /acZ- E. coli strains WMllOO or CSH26 (data not shown). The recombination vectors contain, in direct repeat orientation, the Lx2 gene, a neo gene, and a 5’ deletion of the Lx1 gene. The neo gene is driven by its own HSV tk promoter (Thomas et al., 1986), which allows for selection of stable integrants of the vector. The first vector series, plrec,
Lx2
3
to
as well as
to detect site-specific
1acZ gene used in these constructs
Lx1
recombina-
tissues.
RESULTS
5kb I
1
between immune gene sequences (Matsuoka et al., or recombination in the male germ line in
transgenic
4
;
laczv
nation 1991)
ene
p8 HllO
Xhol linker
cells which have un-
events can be carried out both from clonally expanded populations and from individual cells and cell clusters. tors, which were designed
site and rrg.
from
Xhol linker (727 bp)
5’facZ
for detection
cell lines which are similar
that molecular
i&&ion bps
mutant
event. We tested these vectors
and mouse
rates determined
of
both in
we constructed designed
and report rates of mitotic intrachromosomal
demonstrate
I 0
vectors
a recombinational
tion in human
to study-
which can be studied
sets of E. coli lacZ-based of wt-PGal
SY40
promoter/enhancer
lad
$
E.co/ilad gene
I
in vivo.
these limitations
recolnbination
‘...,;.
regimes do not permit for
assays, nor can they be readily adapted
ing recombination
A.
e.g., tetrad
was constructed
within
iLlindlII-BumHI-flanked LNZ-lacZ gene, the BumHI-Xhol neo gene fragment from pCMI-neo(polA) (Stratagene, La Jolla, CA. USA), and the ‘~k~~t-B~~zHl fragment of the Lul-lucZ gene (which results in a 5’ deletion of the iurZ gene at the linker insertion site) within the vector pCH1 IO, downstream from the SV40 promoter. The same three fragments are inserted downstream from the human non-muscle fi-actin promoter in the vector pH-B-APr-I-gpt (obtained from Dr. R.M. Liskay) to generate the vector pArec. (B) Positive control vectors for pLrec or pArec, respectively, replace the Lx2 gene with wt iac2. The Lx2 negative control replaces the 5’ deletion of Lxl with a 5’ Xhol deletion of f,.~2, thus eliminating complementary sequence. The Lx1 negative control consists of the Lx1 gene and an izph gene (BatnHI fragment from the vector pHyg, obtained from Bill Sugden, University of Wisconsin). All vectors contain pBR322 ori sequences and ctmp genes for function in E. co/i (not shown). Abbreviations: H. HindHI; B, BarnIll; X, XhoI. Methods: Restriction digests and ligations were conducted using standard methods (Sambrook et al., 1991). Constructs were electroporated into the lacZ_ E. coli strains CSH26 and WMllOO using a BioRad gene pulser following the protocol provided by the manufacturer. 2 ul plasmid DNA (to ng) were added to 40 ul (2 x IO&) electrocompetent bacteria in a 0.1 cm cuvette and electroporated using settings of 25 uF, 200 W, 2.4 kV. Bacteria were then resuspended in SOC media (2% w/v Bacto Tryptone (Difco)/O.S% (w/v) yeast extract/O.010 M NaC1/0.0025 M KCI/O.OlO M MgCl,/O.OlO M MgSO,/O.OZO M glucase), incubated for I h to allow expression and then plated onto LB agar plates containing 100 ltg Ap/ml. For BGaI assays, constructs were electroporated into the /ucZ- E. coli strain CSH26 and plated onto LB agar plates overlaid with 29’,, w/v Bluogal (BRL). After overnight incubation
at ?7’ C, blue
( LacZ’) cells were scored visually.
165
EcoR I
promoter
Xho I Xho I
-q----&-_ Lx2
Xh:l
neo
Xho I
pLrec (LaCiT) neo
A Lxi
I( lacZ+
A Lx1
neo Gene
Fig. 2. The pLrec and pArec recombination
ConversIon (Lacz+)
vectors.
Unequal SCE (Lacz+)
Reciprocal lntrachromatid Exchange (LacZ+)
(A) Vector maps. (B) Recombination
mechanisms
by which a wt lacZ gene can be reconstituted
from the mutant lacZ genes contained in the recombination substrate vectors pLrec or pArec include unidirectional transfer of wt nt sequence from the ALxl gene to the Lx2 gene (gene conversion) and reciprocal exchange. Reciprocal exchange of sequence between the two mutant lacZ genes can occur either within the vector (reciprocal intrachromatid exchange) or between vectors located on sister chromatids (unequal SCE). In both cases of reciprocal
exchange,
the neo gene and the ALxl
gene are lost.
the plasmid pCHl10, which contains an SV40 promoter/enhancer element. In the second vector series, pArec, constructed using the plasmid pH+APr-1-gpt (Gunning et al., 1987), lacZ is driven by a human nonmuscle l3-actin promoter, which exhibits a wide range of tissue-type expression in transgenic mice (e.g., Galski et al., 1989). For pLrec and pArec those recombinational events that result in the reconstitution of a wt 1acZ gene and expression of functional PGal can be detected. These events include gene conversion, unequal sister chromatid exchange and reciprocal intrachromatid exchange (Fig. 2B). Both recombination vectors yield rare PGal+ cells (putative recombinants) when transfected into AlacZ- E. coli strains, and constructs containing wt 1acZ show ubiquitous expression in these bacteria (data not shown). (b) Detection of LacZ+ recombinants and determination of recombination rates In human fibroblasts transfected with vectors pLrec or pArec and stained with XGal, rare PGal-producing cells and cell clusters could be identified easily and clearly under 25 x magnification in cultures containing l-100 positive cells per lo7 (Fig. 3). The number of cells in each cluster were then verified by further examination under higher magnification.
Fig. 3. colony curred LacZ+
Staining and identification of LacZ’ fibroblasts. A sectored A-T containing LacZ+ cells. A putative recombinational event oclate in colony formation, resulting in a cluster of two turquoise cells. Methods: XGal staining: Fibroblasts in monolayer were
stained as described by Lim and Chae ( 1984). After aspiration of media, cells were fixed for 5 min at 4°C in a solution of 0.2% glutaraldehyde/2% formaldehyde in PBS. Cells were then washed once in PBS and covered with staining solution: 44 mM HEPES buffer/l5 mM NaCl/1.3 mM MgClJ3 mM K,ferrocyanide/3 mM K.ferricyanide/l.25 ug per ml XGal (Sigma). Cells were incubated for 4424 h at 37°C and turquoise cells scored visually.
Using cell lines determined to be single-copy for our vectors by Southern blot analysis, recombination frequencies and rates were measured for two human
166
A. w.4
-9
4174-
~1463-406-1140-34(!2$
I
(I&)
(713)
H
yP I
A
(6li3,
(4exgS)(S7)
I
I A+
(29)
I I
W-C
I
Lx2
(1726)
(4774)
B
yP
X
(5892)
d
facZ
B
YP
I
c
pLrec
A Lx1
1 C
B
gene conversion A Lx1
reciprocal exchange/ unequal SCE
IacZ
c.
B. left-hand right-hand IacZ gene IacZ gene A Lx1 Lx2
left-hand IacZ gene wild We
right-hand /acZgene A Lx1
Clone I 4.2
F3
LN L4
-4.17 -3.86
-3.03 -2.41
5L20-F4
Fig. 4. Molecular analysis of La&? recombinants. (A) Restriction maps for unrearranged plrec, gene conversion of pLrec and reciprocal exchange/un~qual SCE of ptrec. Restriction enzyme sites are shown above and homologous target sequences for the PCR primers are shown below each map. Map positions, in base pairs from the vector‘s unique Hind111 site, are indicated in parentheses. B, Burnt-II: H, ~~~dIl1; Hp. Hpnl: X, Xkol. (B) PCR analysis: 5L20 is an A-T fibroblast clone that contains a single unrearranged copy of the pLrec vector while 5L20-F4 is a FACS-sorted LacZ+ derivative of 5L20. PCR primer pair A-C amplifies a 1.01.kb fragment specific to the left-hand IacZ gene; primer pair B-C amplifies a 1.14-kb fragment and is specific for the right-hand neo-/acZ junction. The Lx2 gene PCR product of SL20 yields 905 and 10%bp fragments upon Hpul digestion and 740- and 273-bp fragments upon XhoI digestion; indicating that the amplified DNA contains a single restriction site for each enzyme. XhoI and Hpul digests are diagnostic of 5L20-F4 having undergone a gene conversion event involving the Ieft-hand fucZ gene (Lx?); the Xhol linker, represented by an X&I site, has been replaced by sequence donated by &xl, thus regenerating the HpaI site found in wt In& and giving three fragments (624, 273 and 108 bp) on NPLII digest of the left-hand gene. (C) Southern analysis: LNL4 is a Lesch-Nyhan syndrome human fibroblast line that contains a single unrearranged copy of the pLrec vector while clones 4.2 and F3 are FACS-sorted LacZ’ derivatives of LNLC Genomic DNAs from these clones were obtained and digested with either Hind111 and BamHI or HindHI, XhoI and BurnHI, then electrophoresed, blotted and probed, as described in Table I, using the 4.2.kb BarnHI-BurnHI fragment from pLrec that contains both nco and 1acZ sequence. The probe detects 4.2- and 3.8.kb HindIII-BumHI fragments containing 1ucZ sequence in the parent cell line (LNL4) and clone 4.2, but the 3.8-kb fragment is resistant to XizuI-digestion in clone 4.2, indicating that clone 4.2 is the product of a gene conversion event. DNA from clone F3 contains only a single 3.8.kb icrcZ fragment that is resistant to Xhof digestion, consistent with F3 containing a wt lacZ gene that is the product of a reciprocal exchange or unequal SCE. Methods: FDG staining of trypsinized fibroblasts was as described by Nolan et al. (1988) and cells were brought to a final concentration of 2 x IO’ cells/ml. Fluorescent cells were sorted utilizing a modified Beckton-Dickinson fluorescence-activated cell sorter. Sorted cells were plated and grown without G418 selection. Clones were isolated and analyzed for BGal activity using XGal. Genomic DNA was then obtained
167 TABLE
I
Spontaneous
intrachromosomal
recombination
Total number
Cell line”
of LacZ+
rates of lacZ genes in mammalian
cells containing
Total number
cells detectedb
the pLrec vector Rate of conversion
of cells screened
(conversions/cell (x 10-y Human cell lines GM847-LNL6 GM5849-5L20
Murine ES cell line CCE-4-BGal-9
“Each cell line contains
0.72
10
2.10 x lo6 3.76 x lo3
443
43.0 x lo3
14 72
5.36 x lo5
1.3
7.23 x lo5
3.1
a single copy of the pLrec vector integrated
with 0.25% trypsin (Gibco) dSpontaneous recombination
152 225
into their genomic
‘As determined by microscopic examination following XGal staining. “Determined by direct microscopic examination if < 10000 cells. For populations
analysis tables Methods: For 500-2000 cells fibroblast line,
0.70
7.60 x lo6
206 41
and then counting the resulting cell suspension rates were derived from the number of LacZ+
to LacZ’ generation)d
DNA (see Methods
below).
> 10 000, total cell number
was determined
by treating
using a hemocytometer. cells detected and the total number
the cells
of cells screened
using the fluctuation
provided by Capizzi and Jameson (1973). each experiment, dishes were seeded with enough cells to provide 100-500 colonies/dish, grown until individual colonies contained each, then stained for LacZ expression with XGal and LacZ+ and LacZ- cells counted. Cell culture: GM847, a human Lesch-Nyhan and GM5849, an A-T fibroblast line of complementation group D, were obtained from the Human Genetic Mutant Cell Repository,
Camden, NJ. The murine ES line CCE was obtained from Dr. J. Eisenstadt, Yale University. The pLrec vector was electroporated into these three cell lines, stable transformants isolated and their genomic DNA analyzed in order to identify clones that contained single copies of the pLrec vector integrated into their genomic DNA (e.g., GM847-LNL6, GM5849-5L20 and CCE-4-BGal-9) Cells were maintained to dMEM supplemented with 15520% fetal calf serum. ES cells were grown using feeder layers or supplemented with 50% buffalo rat liver cell conditioned media. Cell transformation: Cells were grown exponentially and passaged 2-3 days prior to electroporation. Cells at 50&75% confluence were trypsinized. washed twice in cold in PBS to a final concentration of (l-2) x 10’ cells/ml. Vector DNA, linearized using Seal, was added to PBS w/o Ca’+ or Mg’+ and resuspended a final concentration of 10 pg/ml, 0.8 ml of the mixture placed into a 0.4-cm gap cuvette (BioRad) and placed on ice for 10 min. Electroporation was performed on a BioRad Gene Pulser (settings 7OOV, 25 pF.) The cells were incubated on ice for a further 10 min and then plated in 15% DMEM. After 48 h, G418 (Gibco) at 400 ug/ml was added for selection. After 12-21 days growth, drug-resistant clones were isolated and grown in the presence of antibiotic. Genomic DNA was prepared according to Sambrook et al. (1989) and Southern hybridization using Hybond-N nylon membranes was carried out as recommended by the manufacturer (Amersham). Blots were probed lacZ sequence, labeled with 32P using a random-priming kit (Boehringer-Mannheim).
SV40-transformed fibroblast lines and for a murine embryonic stem cell (ES) line (Table I). Rates for GM847, a Lesch-Nyhan syndrome line, averaged 0.71 x 10m5 events/cell generation, which are comparable to rates determined for other mammalian fibroblast lines using conventional recombination vectors containing dominant selectable marker genes as recombination substrates (Bollag et al., 1989). Similar rates were also observed for CCE, a pluripotent mouse embryonic stem cell line. Rates for GM5849, a cell line derived from an A-T patient, averaged 1.9 x 10m3 events/cell generation, > 200-fold higher than those seen in the Lesch-Nyhan line, confirming the hyper-recombinogenic A-T phenotype that was previously observed using a dominant-selectable marker-
for Southern analysis and PCR amplification, Digests bromide-stained 0.7% agarose gel while PCR products
with the 3.8-kb HindIII-BamHI
fragment
of pLrec
that contains
based recombination vector (Meyn, 1993). We screened five independently derived stable transformants of GM5849 containing the negative control vector p(Lxl)Lrec and three containing the negative control vector p(Lx2)Lrec for the presence of LacZ+ cells. We found no LacZ+ cells in over lo8 cells screened for each stable transformant and conclude that reversion frequencies of the mutant lacZ genes are < 10m8. On average, lo-15% of cells in a population of G418-resistant fibroblasts or murine embryonic stem cells stably transformed with the positive control vectors were found to be LacZ+ by XGal staining, with some clones showing no detectable expression and many with >50% of the population producing detectable PGal.
of genomic DNA for Southern analysis were electrophoresed overnight through an ethidium were electrophoresed through a 1.3% agarose gel. Standard PCR: 1 pg of genomic DNA from
each recombinant was amplified enzymatically in a 25 ~1 reaction mixture which included 15 ng each of amplimer pairs A-C or B-C/200 pm of each deoxynucleotide triphosphate/l.25 units of Tuq polymerase/l.5 x Tay buffer (Boehringer-Mannheim). PCR was performed in a programmable thermal cycler (Perkin-Elmer Model 9600) for 30 cycles, each consisting of 30 s of denaturation at 94”C, annealing at 64°C and polymerization at 72°C. Primers: A, 5’-AAT CCG CCG TTT GTT CCC AC; B, 5’-TGC GAA GTG GAC CTG GGA CC; C, 5’-ACG ATC GCG CTG CAC CAT TC.
168
A
B # of cells amplified
0 0
# of cells amplified 23312M
1211643
kb ALxl
1.14)
bP 864
_/
-
Lx2 1.01)
624
273
108
Fig. 5. PCR ampfification of LacZ’ cell clusters. (A) One to six cells were picked from the same LacZ” sector of a colony of 5L20 fibroblasts and amplified using the two-step PCR method described below. As shown, neither of the blank lanes contains l.Ol- and 1.14-kb fragments characteristic of ~LICZgene amplification, while two of three reactions containing a single LacZ’ cell produced the characteristic fragments, as did the rest of the reactions. (B) HpnI digest of DNAs obtained from PCR amplification of samples of one to three ceils taken from the same LacZ+ sector. The resultant bands (864,624,273, and 108 bp) are diagnostic for a gene conversion event. Abbreviations: C, control pLrec plasmid; M, size marker. 0, 1, 2, 3, 4 and 6 indicate the number of ceils contained in each PCR reaction mixture. Methods: Two-step PCR: Individual LacZ+ clusters were harvested using a glass micropipette, transferred into PCR tubes containing 10 ~1 standard PCR and 0.5 mg/ml protease K, and incubated at 37°C for 2 h. 2.5 ~1 were transferred to a reaction tube containing for 20 min at 94°C. Remaining reagents were then added to a finai volume of 25 d, and PCR performed for as for standard PCR (Fig. 4 legend). Aliquots of PCR products were run directly on an ethidium-stained 1.3% with HpaI; HpaI buffer and enzyme
were added
directly
to the PCR reaction
(c) FDG staining and FACS sorting of individual ZacZ+ recombinants XGal staining is toxic to mammalian cells and does not allow for clonal expansion of recombinant LacZ+ cells after staining. We therefore have made use of fluorescein di-P-o-galactopyranoside (FDG), a non-toxic galactose analog which is cleaved by E. coli PGal to yield fluorescein. This lack of toxicity has allowed us to use FDG staining to identify, isolate and clonally expand LacZ+ colonies from mixed populations of La&+ and LacZ- colonies (data not shown). FDG staining also allows for isolation of LacZ+ cells by cell sorting. E.g., we used a fluorescent activated cell sorter (FACS) to collect the 0.3% most fluorescent cells from a population of FDG-stained A-T fibroblasts containing a single integrant of pLrec (5L20), initially seeded at low density to lessen the recovery of jackpots of previously recombined LacZ+ cells which may have been present in the original stock. Sorted cells were plated and clones picked and expanded for analysis. Over 50% of the colonies arising from the sorted population were LacZ+ by histochemical staining. Although
The numbers cells and cell
buffer supplemented with 0.1% NP40 20 ~1 of 1 x PCR buffer and incubated 50 cycles, 1 min/cycie at temperatures agarose gel, or first digested overnight
mixture.
not required for isolation of LacZ’ A-T fibroblasts, two consecutive sorts were necessary for efficient recovery of IacZ’ recombinant Lesch-Nyhan fibroblasts due to low spontaneous recombination rates. High levels of expression in FACS-sorted lines, as indicated by > 75% of cells showing intense XGal staining, persisted after passage and expansion. All FACS-sorted LacZ+ clones analyzed proved to be re~ombinants (data not shown). (d) PCR amplification and molecular analysis of individual recombinant clusters The two mutant IacZ genes contained in the recombination vectors differ from each other and the wt lacZ gene in having distinct 5’ and 3’ polymorphisms which allow for molecular analysis of the recombination products contained in a LacZ’ cell by Southern blotting or PCR amplification followed by restriction enzyme digestion (e.g., Fig. 4A and B). Using the primers described in Fig. 4, we have successfully amplified both of the _ I-kb 1ac.Z target sequences of plrec, starting from suspensions that contained fewer than seven LacZ+ cells (data not shown). We have suc-
169 cessfully
analyzed
attached
cells as well (Fig. 5A), sug-
gesting that it will be possible lacZ gene fragments
directly
to PCR-amplify from clusters
diagnostic
of LacZ+
cells
found in tissues of transgenic animals. Once amplified, the 1acZ gene fragments can then be digested to help determine
the nature
to BGal production
of the recombinant
events that led
(e.g., Fig. 5B). Although
ACKNOWLEDGMENTS
PCR ampli-
fication can be carried out on single LacZ+ cells, we have restricted our analysis to cells obtained from LacZ+ clus-
We wish to thank R. Carbone
E. Brainerd
for technical
of the Yale Comprehensive
assistance,
Cancer
Center
for FACS sorting, G. Ruano for advice on PCR amplification and Dr. R.M. Liskay for invaluable advice and discussion.
This work
was supported
from the National
Institutes
American
Society (MG-15 ).
Cancer
in part by grants
of Health (ES05327)
and the
ters of two or more cells. In that way, we avoid analyzing cells that contain a situation
both sister chromatids
that may complicate
(e) Conclusions (I ) We have constructed
of unequal
molecular
SCEs,
analysis.
two sets of E. coli &Z-based
vectors designed for detection of wt BGal activity in mammalian cells which have undergone a recombinational event. These are the first &Z-based vectors to be described that can be used in studies of general genetic recombination in a variety of mammalian cell lines in vitro and tissue types in vivo. (2) These vectors yield rates and frequencies of mitotic intrachromosomal recombination in human and rodent cell lines which are similar to rates reported using other vector systems. (3) Molecular analysis of recombinational events involving the lacZ-based vectors can be carried out on genomic DNA isolated from clonally expanded populations. It also may be possible to analyze both products of gene conversions and unequal sister chromatid exchanges, events that do not involve production of nonreplicating episomes, by means of FDG staining to identify colonies that contain both LacZ’ and LacZ- cells followed by isolation and clonal expansion of both cell types from such sectored colonies and analysis of their genomic DNA. Unlike dominant-selectable markerbased vectors, recombinant cells containing pLrec or pArec do not have to be clonally expanded prior to analysis, as individual LacZ+ cells and small clusters of cells can be studied using PCR amplification. (4) Advantages and potential applications of these vectors in studying mammalian mitotic recombination include: rapid determination of recombination frequencies in cell lines as well as cells with limited proliferative capacity (e.g., primary fibroblasts); detection and analysis of individual recombinant cells; detection of recombination in non-dividing cells; and in situ detection of recombination in the tissues of transgenic animals.
REFERENCES Austerberry, C.F., Allis, C.D. and Yao, M.C.: Specific DNA rearrangements in synchronously developing nuclei of Tetrahymena. Proc. Natl. Acad. Sci. USA 81 (1984) 7383-7387. Bajaj, M. and Blundell, T.: Evolution and the tertiary structure tein. Annu. Rev. Biophys. Bioeng. 13 ( 1984) 453--492.
of pro-
Bollag, R., Waldman, A.J. and Liskay, R.M.: Homologous recombination in mammalian cells. Annu. Rev. Genet. 23 (1989) 1999225. Capizzi, R.L. and Jameson, J.W.: A table for the estimation of the spontaneous mutation rate of cells in culture. Mutation Res. 17 (1973) 1477148. Cavanee,
W.K.: Current
knowledge
of heritable
tumors.
Birth Defects
25 (1989) 585-599. H., Sullivan, M., Willingham, M., Khew-Voon, C., Galski. Gottesman, M., Pastan, I. and Merlino, G.: Expression of a human multidrug transgenic
resistance cDNA (MDRl) in the mice: resistance to daunomycin-induced
bone marrow of leukopenia. Mol.
Cell. Biol. 9 (1989) 4357-4363. Gunning, P., Leavitt, J., Muscat, G., Ng, S.Y. and Kedes, L.: A human B-actin expression vector system directs high-level accumulation of antisense transcripts. Proc. Natl. Acad. Sci. USA 84 (1987) 4831-4835. Lim, K. and Chae, C-B.: A simple assay for DNA transfection by incubation of the cells in culture dishes with substrates for betagalactosidase. Matsuoka, Muller, somatic
Biotechniques
7 ( 1989) 5766579.
M., Nagawa, F., Okazaki, K., Kingsbury, L., Yoshida, K. U., Larue, D.T., Winer. J.A. and Sakano, H.: Detection of DNA recombination in the transgenic mouse brain. Science
254(1991)81-86. Meyn, M.S.: Elevated rates of spontaneous intrachromosomal recombination in ataxia telangiectasia. Science 260 ( 1993) 132771330. Murti, J.R., Bubulis, gene conversion
M. and Schimenti, J.C.: High-frequency in transgenic mice. Mol. Cell. Biol.
2545-2552. Nolan, G.P., Fiering, Fluorescence-activated on beta-o-galactosidase
germ line 12 (1992)
S., Nicolas, J.-F. and Herzenberg, L.: cell analysis and sorting of viable cells based activity after tranduction of Escherischia
coli lacZ. Proc. Natl. Acad. Sci. USA 85 (1988) 2603-2607. Sambrook, J., Fritsch, E.F. and Maniatis, T.: Molecular Laboratory Manual. Cold Spring Harbor Laboratory
Cloning. A Press, Cold
Spring Harbor, NY, 1989. Thomas, K., Folger, K. and Capecchi, M.: High frequency targeting of genes to specific sites in the mammalian genome. Cell 44 (1987) 419-428.