`
`Transformation of the Gene for Hypoxanthine
`Phosphoribosyltransferase
`
`Lloyd H. Graf, Jr., Gall Urlaub, and Lawrence A. Chasin
`
`Department of Biological Sciences, Columbia University, New York, New York 10027
`
`Received 18 July 1979
`
`Abstract--Purified DNA from wild-type Chinese ovary (CHO) cells has
`been used to transform
`three hypoxanthine phosphoribosyltransferase
`(HPRT) deficient murine cell mutants to the enzyme positive state. Trans-
`formants appeared at an overall frequency of 5 x I0 -s colonies/treated cell
`and expressed CHO H P R T activity as determined by electrophoresis. One
`gene recipient, B21, was a newly isolated mutant of LMTK- deficient in both
`H P R T and thymidine kinase (TK) activities. Transformation of B21 to
`H P R T + occurred at I/5 the frequency of transformation to TK+ ; the latter
`was, in turn, an order of magnitude lower than that found in the parental
`L M T K cells, 3 x 10 -6. Thus both clonal and marker-specific factors play a
`role in determining transformability. The specific activity of H P R T in
`transformant extracts ranged from 0.5 to 5 times the CHO level. The rate of
`loss of the transformant H P R T + phenotype, as measured by fluctuation
`analysis, was lO-4/cell/generation. While this value indicates stability
`compared to many gene transferents, it is much greater than the spontaneous
`mutation rate at the indigenous locus. The ability to transfer the gene for
`H P R T into cultured mammalian cells may prove useful for mutational and
`genetic mapping studies in this well-studied system.
`
`INTRODUCTION
`
`The transfer of active cellular genes by treatment of cultured mammal-
`ian cells with purified DNA (transformation) has recently been demon-
`strated. The initial experiments involved the transfer of a viral gene, that
`specifying the thymidine kinase (TK, EC 2.7.t.75) of herpes simplex virus
`(1-4). This was soon followed by the extension of this phenomenon to cellular
`tk (5) and then to the genes specifying adenine phosphoribosyltransferase (5)
`and dihydrofolate reductase (6). From this limited number of examples, it
`
`1031
`
`0o98-o366/79/11OO-lO315o3.oo/o 9 1979 Plenum Publishing Corporation
`
`Merck Ex. 1037, pg 1123
`
`
`
`1032
`
`Graf et al.
`
`appears that transformation may have general application as a method of
`gene transfer in cultured cells of higher organisms.
`Transformation can be used to detect the presence of specific genes in a
`preparation of DNA. This bioassay could be employed to monitor the
`purification of a particular gene by a combination of biochemical and cloning
`techniques. Until now, gene cloning has been limited to random fragments of
`genomic DNA or to those genes for which mRNA or cDNA probes have been
`available. The latter class is in turn limited to loci coding for an abundant
`gene product. The transformation bioassay does not depend on the abundance
`of a gene product; rather it depends on the availability of an appropriate
`selective system. Most loci that have been subjected to somatic cell genetic
`analysis encode nonabundant but selectable products. Transformation offers
`a possible approach to the cloning of such genes.
`One locus of considerable interest is that governing the enzyme hypoxan-
`thine phosphoribosyltransferase (HPRT, EC 2.4.2.8). Powerful methods have
`been developed for the selection of forward and reverse mutants at this locus.
`The X-linkage of the hprt gene facilitates the isolation of enzyme deficient
`mutants, since only one copy of each X-linked gene is functional in most
`mammalian cells (7). Consequently a large number of well characterized hprt
`mutants exists (for review see 8).
`Structural analysis of the effects of mutation at the hprt locus would'be
`aided by the availability of cloned DNA sequences representing the hprt
`wild-type and mutant genes. Such cloned sequences might also be useful for
`the study of X-inactivation at this locus. Transformation of the hrpt gene
`represents one approach to the cloning of this sequence. For these reasons we
`undertook the development of a transformation system for hprt in cultured
`mouse cells.
`While this work was in progress, Willecke and his colleagues (9)
`described the isolation of a clone of mouse cells with the characteristics of an
`intraspecific hprt transformant. We report here the interspecific transforma-
`tion of three HPRT-deficient mouse cell lines using purified Chinese hamster
`DNA. The HPRT-positive clones isolated arise at a low frequency, but
`exhibit the enzymatic and stability characteristics expected of transformants.
`
`MATERIALS AND METHODS
`
`Cells. The mouse L-cell derivatives LMTK- (10), lacking TK activity,
`and A9 (ll), lacking HPRT activity were provided by S. Silverstein and
`S. Shin, respectively. An HPRT-deficient derivative of mouse 3T6 cells,
`3T6TG8 (12), was provided by C. Basilico, and human KB (13) cells by
`G. Zubay. The K1 clon e of CHO cells (14) was used.
`Cell Culture. Cells were routinely grown as monolayers at 37 ~ in 5%
`
`Merck Ex. 1037, pg 1124
`
`
`
`hprt Transformation
`
`1033
`
`CO2 in a mixture of F12 and Dulbecco's modified Eagle's medium (15) with a
`final glucose concentration of 4 g/liter and lacking hypoxanthine and
`thymidine. This growth medium was supplemented with 10% (v/v) fetal calf
`serum.
`Selective media were formulated with the following modifications.
`Selection for the HPRT + phenotype was carried out in either HAT medium
`(100 uM hypoxanthine, 0.7 #M amethopterin, and 15 t~M thymidine) or
`HAS medium (30 IzM hypoxanthine, 20/zM azaserine, and 3 ~M thymidine).
`It was necessary to use HAS medium for selections using derivatives of
`LMTK-, since these cells cannot grow in HAT medium due to their TK
`deficiency. Selection for TK + was accomplished either in HAT or in AAT.
`The latter is identical to HAT except for the substitution of 50 izM adenine
`for hypoxanthine. AAT was used in all TK selections involving HPRT- cells.
`Adenine phosphoribosyltransferase is still present in these cells and allows the
`salvage of adenine as the sole purine source when de novo synthesis is blocked
`with amethopterin.
`Mutants of LMTK- cells deficient in HPRT activity were selected on
`the basis of their resistance to purine analogs as described by Sharp et al.
`(16). Cells were mutagenized with ethyl methanesulfonate (360 #g/ml) and
`allowed 6 days' growth (17) before being challenged in medium containing
`8-azaguanine and 6-thioguanine (3 izg/ml each). Fourteen million mutage-
`nized cells gave rise to 4 colonies that were recloned in selective medium. No
`drug-resistant colonies were obtained from 6 • 106 nonmutagenized cells.
`DNA Purification. Late log-phase suspension (spinner) cultures of
`CHO or KB cells, or nearly confluent monolayers of A9 cells, were harvested
`and DNA was extracted and purified as described by Pellicer et al. (18). The
`preparations were of high molecular weight as judged by slower migration in
`0.4% agarose gels than a Barn H 1-1inearized plasmid pLC7-21-ColE1 marker
`(15.7 kb, kindly provided by C. Squires).
`Transformation Procedure. Most experiments were carried out accord-
`ing to Wigler et al. (6), with the following modifications: Recipient cells,
`0.5-1.0 • 106/100-mm dish, were preplated in nonselective growth medium
`14-20 h before the addition of DNA. Four hours before DNA addition this
`medium was replaced with 9 ml of Dulbecco's modified Eagle's medium
`containing 0.4% glucose, 10 mM HEPES (pH 7.1), and supplemented with
`5% calf serum (exposure medium).
`DNA was coprecipitated with calcium phosphate (6) in siliconized glass
`tubes in a final volume of 1-3 ml at 30 tzg DNA/ml. Double-strength
`HEPES-buffered saline stocks were periodically tested for retention of pH
`7.05-7.10. Additions involving DNA were done with disposable plastic 1-ml
`or 5-ml pipettes. Dispersion of the DNA during initial mixing was accom-
`plished by a combination of gentle hand agitation and bubbling air through
`
`Merck Ex. 1037, pg 1125
`
`
`
`1034
`
`Graf et al.
`
`the delivery pipette with a Pipet-aid (Drummond). Precipitated DNA (1 ml)
`was added directly to the 9 ml of exposure medium. Cells were exposed to
`DNA for 3.5 h at 37% then were shifted to room temperature. After 30 min,
`2.5 ml of 30% (v/v) dimethyl sulfoxide [final concentration 6%, a modifica-
`tion of the procedure of Miller and Ruddle (19)] in exposure medium was
`added and incubation at room temperature continued for 30 rain. The
`medium was then replaced with growth medium and the dishes returned to
`the 37 ~ incubator. An atmosphere of 5% CO2 was maintained with the use of
`a gassed sealed box during room temperature manipulations. After a 24- to
`28-h expression period in nonselective medium, cells were challenged in
`selective medium either by direct replacement of the medium, or by trypsini-
`zation and replating (see Table 1). Selective medium was renewed twice a
`week. After 2-3 weeks' growth, transformant or revertant colonies were
`recloned in selective medium.
`Enzyme Measurements. Cell monolayers were harvested by trypsiniza-
`tion, washed twice with PBS, and stored as frozen cell pellets at - 2 0 ~ For
`enzyme extraction, cell pellets were thawed in 20 mM potassium phosphate,
`pH 7.0, containing 0.5% Triton X-100, at a concentration of about 1.5 • 108
`cells/ml. After centrifugation at 12,000 g for 25 rain, the resulting superna-
`tant solution was used as a crude cytoplasmic extract.
`HPRT was assayed by the conversion of [~4C]hypoxanthine to [14C]ino-
`sine monophosphate under reaction conditions previously described (20).
`Incubations were carried out in duplicate at 37 ~ in 50 #1 total volume and
`were stopped by the addition of 5 #1 of 100 mM EDTA and chilling in ice.
`Samples of 22 #1 were spotted onto 22 • 22-ram squares on a plastic-backed
`polyethyleneimine cellulose thin-layer chromatography sheet (Bakerflex PEI-
`F). The sheet was air dried, washed 5 rain with 0.5 mM ammonium formate,
`pH 7; washed 15 rain under running tap water, and dried under an infrared
`lamp. The squares were then cut out and counted in 5 ml of toluene-based
`scintillation fluid.
`Electrophoresis. A combination of previously described methods (21-
`23) yielded improved resolution of mouse and Chinese hamster HPRT
`activities in vertical slab polyacrylamide gels. The running gel contained 7.5%
`acrylamide, 0.188% N,N'-methylene-bis-acrylamide (bis), 0.06% N,N,N',N'-
`tetramethylethylenediamine (TEMED), 15 mM NaC1, 8% (w/v) sucrose,
`and 15 mM glycylglycine, pH 8.7. The spacer gel comprised 6% of the
`running gel and contained 4% acrylamide, 1% bis, 0.06% TEMED, and 15
`mM glycylglycine, pH 7.0. Polymerization was initiated by the addition of
`ammonium persulfate to a final concentration of 0.7 mg/ml. The reservoir
`buffer contained 5 mM glyclyglycine and 4 mM glycine, pH 8.7. After 20
`min preelectrophoresis at 3 mA/cm: at 4% samples containing 10-50 ug of
`protein were underlaid and electrophoresis carried out overnight at 4 ~ at
`
`Merck Ex. 1037, pg 1126
`
`
`
`hprt Transformation
`
`1035
`
`70-80 V for a 14-cm gel. Electrophoresis was terminated when a bromophe-
`nol blue-bovine serum albumin marker in a separate channel had run at least
`8 cm. Zones of HPRT activity were then detected using the radioactive assay
`method previously described (20).
`Estimation of HPRT + Segregation Rates by Fluctuation Analy-
`sis. Each clone tested was grown in mass culture in HAS medium for at least
`a week before the analysis. Twenty-four replicate cultures were then started
`for each clone, by inoculating 25 cells per culture in 24-well culture dishes
`(Linbro). Ten to 15 colonies formed in each well; these were dispersed after 1
`week and allowed to grow to near confluence (4.5-7 x 105 cells/culture,
`15-16 generations). Samples of 5 x 104 cells/culture (TB23 and TA28) or
`the total cell yield, 7 x 105 cells/culture (RD91) were then plated in 60-ram
`dishes in medium containing 25 #M 6-thioguanine. Reconstruction experi-
`ments indicated that the high cell density used in the case of RD91 selections
`does not affect the appearance of 6-thioguanine-resistant segregants. Selec-
`tion dishes were stained with crystal violet after 10 days.
`Segregation rates were calculated using the mean method (equation 8)
`of Luria and Delbrtick (24) and the median method of Lea and Coulson (25)
`based on the estimated number of colonies per culture. Due to an expected
`delay in expression of an HPRT phenotype (17) as 6-thioguanine resistance,
`segregation events occurring in the last several days of nonselective growth
`are probably not represented as colonies. The actual rates of genetic events
`may be somewhat higher than those calculated in Table 3 as a consequence of
`this expression lag.
`
`RESULTS
`
`Transformation for Thymidine Kinase (TK). Exposure of LMTK- cells
`to DNA from hamster (CHO), human (KB), or mouse (A9) cells led to the
`appearance of colonies with a TK + phenotype (HAT + or AAT +) in more
`than half (67/122) of the treated dishes representing an overall frequency of
`3.7 x 10 -6 (Table 1, top). Control dishes, treated either with salmon sperm or
`bacterial DNA, or with calcium phosphate precipitates containing no DNA,
`only rarely yielded a colony (4 • 10-8). The few colonies observed may in
`fact represent spontaneous mutation to amethopterin resistance, as revertants
`to TK + have never been found using L M T K - (5), and the one colony tested
`from these experiments was indeed resistant to amethopterin (0.67 #M) and
`to 5-bromodeoxyuridine (0.1 mM). These results with tk transformation are
`in essential agreement with those previously reports by Wigler et al. (5) and
`attest to the effectiveness of the DNA preparations and transformation
`methods used.
`Transformation
`
`for Hypoxanthine Phosphoribosyltransferase
`
`Merck Ex. 1037, pg 1127
`
`
`
`1036
`
`Graf et al.
`
`Table 1. Summary of transformation experiments
`
`No.
`experi-
`Recipient ments a DNA
`
`No.
`cells
`exposed
`
`Selective
`medium"
`
`Positive
`Pheno- dishes/
`type
`total
`selected
`dishes
`
`Colonies/
`107 cells
`No.
`colonies d exposed
`
`LMTK
`
`22
`
`A9
`
`11
`
`B21
`
`14
`
`TTKB1
`
`3T6TG8
`
`3
`
`5
`
`control b 56 x 106 HAT & AAT TK
`CHO
`68 x 106 HAT
`TK
`CHO
`43 x 106 AAT
`TK
`KB
`16 x 106 AAT
`TK
`A9
`4 x 106 AAT
`TK
`control b 20 x 106 HAS
`HPRT
`CHO
`24 • 106 HAS
`HPRT
`KB
`19 • 106 HAS
`HPRT
`A9
`19 • 106 HAS
`HPRT
`control b 34 x 106 AAT
`TK
`CHO
`51 x 106 AAT
`TK
`KB
`14 x 106 AAT
`TK
`control b 42 x 106 HAS
`HPRT
`CHO
`64 x 106 HAS
`HPRT
`KB
`14 x 106 HAS
`HPRT
`control b 11 x 106 HAS
`HPRT
`CHO
`18 • 106 HAS
`HPRT
`control b 6 • 106 HAS
`HPRT
`CHO
`8 • 106 HAS
`HPRT
`KB
`4 • 106 HAS
`HPRT
`
`2/49
`39/83
`16/23
`10/13
`2/3
`0/15
`2/20
`0/14
`0/15
`1/26
`9/40
`0/7
`1/32
`3/52
`0/7
`0/10
`1/15
`1/9
`1/9
`0/5
`
`2
`256
`149
`63
`10
`0
`3
`0
`0
`1
`13
`0
`1
`3
`0
`0
`1
`1
`1
`0
`
`0.4
`37
`35
`39
`25
`<0.5
`1.2
`<0.5
`<0.5
`0.3
`2.5
`<0.7
`0.2
`0.5
`<0.7
`<0.9
`0,6
`1.6
`1.2
`<2.5
`
`aEach experiment represents the exposure of one cell population on a given day to experimental
`and control DNA.
`bFor simplicity a variety of control treatments have been placed under this one heading. In the
`majority of experiments, DNA from Microeoccus lysodeikticus (Sigma) was used at 30
`#g/dish. Salmon sperm DNA (30 tzg/dish, Sigma) was used in most of the remaining
`experiments. Occasionally, calcium phosphate precipitate alone was used as a control.
`CHAT = hypoxanthine + aminopterin + thymidine; AAT = adenine + amethopterin +
`thymidine; HAS = hypoxanthine + azaserine. See Materials and Methods.
`din most experiments involving B21 as a recipient, monolayers were trypsinized 48 h after
`exposure to DNA and divided between two parallel dishes containing AAT or HAS selective
`medium. Since the cell population quadruples during this 48-h expression period, transforma-
`tion frequencies could be overestimated by a factor of two in these experiments. In practice in
`the case of HAS selections with B21, no dish contained more than one colony, so that generation
`of viable sister colonies was not a factor. In most experiments not involving B21 as a recipient,
`monolayers were challenged in situ by changing to selective medium 24 h after exposure to
`DNA. Control experiments with LMTK- showed that trypsinization and replating did not
`significantly alter the tk transformation frequency compared to in situ selection.
`
`(HPRT). F o u r d i f f e r e n t H P R T - d e f i c i e n t m o u s e cell
`lines w e r e u s e d as
`r e c i p i e n t s for t h e t r a n s f e r of t h e hprt gene. T h r e e w e r e L-cell d e r i v a t i v e s ,
`c h o s e n b e c a u s e of t h e p r e v i o u s l y d e m o n s t r a t e d c o m p e t e n c e of this t y p e of cell
`for D N A m e d i a t e d
`t r a n s f o r m a t i o n . In a n initial series of e x p e r i m e n t s , t h e
`widely s t u d i e d A 9 line was used. T w e n t y - f o u r h o u r s a f t e r e x p o s u r e to D N A ,
`as d e s c r i b e d in M a t e r i a l s a n d M e t h o d s , cells w e r e c h a l l e n g e d in t h e s a m e
`
`Merck Ex. 1037, pg 1128
`
`
`
`hprt Transformation
`
`1037
`
`dishes with hypoxanthine + azaserine medium (HAS) to select for H P R T +
`transformants. As can be seen in Table 1, only three colonies were isolated
`using hprt + DNA, representing a frequency about 30 times lower than that
`obtained in transforming L M T K - to TK +. Control DNA (nonmammalian)
`as well as D N A from the A9 recipient itself yielded no H P R T + colonies. A9
`DNA was capable of transforming L M T K - to TK + (Table 1, top).
`When HAT medium was used in this selection, colonies appeared at a
`higher frequency (2 x 10 -7) on both control and experimental dishes. Many
`of these colonies did not breed true, and others proved to be resistant to
`amethopterin (1 #M). One transformant was isolated from this series (see
`below). HAS medium proved to be more specific and was used in all
`subsequent selections.
`The low frequency of hprt + transformation relative to tk could be due to
`the fact that the A9 line is less competent for transformation than the
`L M T K - line. While both are L-cells, they have been carried as separate lines
`for more than 15 years. Alternatively, the hprt gene may be innately more
`difficult to transfer (for example, if it were very large).
`In an attempt to distinguish between these possibilities, we constructed
`HPRT-deficient mutants of the L M T K - line (see Materials and Methods).
`One of these doubly deficient mutants, clone B21, was chosen for use as a
`recipient for transformation. This mutant had been induced by ethyl methane
`sulfonate, and it exhibited a low reversion frequency. For most of these
`experiments, the B21 cells were trypsinized 48 h after exposure to DNA and
`one half of the population challenged in HAS medium for H P R T + and the
`other half challenged
`in adenine + amethopterin + thymidine medium
`(AAT) for TK +. While the B21 cells are unable to utilize hypoxanthine as a
`purine source they remain capable of using adenine; thus AAT medium
`selects only for the transfer of the tk gene. The frequency of tk transforma-
`tion provides a standard for quantitating hprt transformation in the very same
`experiment.
`Transformation of hprt was again rare using B21 cells, with a frequency
`of less than 10 -7 . This low frequency may not be significantly different from
`the revertant frequency. However, as will be seen below, HAS + colonies
`selected after DNA treatment are in fact transformants. The frequency of tk
`transformation in B21 is five times higher than that of hprt, but the greater
`difference lies in the lower competence of the B21 clone: the tk transforma-
`tion frequency in L M T K - is ten times higher than in B21 (Table 1). This
`suggests that interclonal variability is the predominant factor in the lower
`frequency of hprt transformation. Three other HPRT-deficient mutants of
`L M T K - were also tested on a smaller scale for high frequencies of tk or hprt
`transformation. No transformants of either type were found after treatment
`of 4 x 106 cells.
`
`Merck Ex. 1037, pg 1129
`
`
`
`1038
`
`Graf et al.
`
`Cells transformed for a given marker might have been selected from a
`small subset of highly competent recipient cells in the original population. If
`competence is heritable, then a once-transformed cell should be highly
`competent for a second transformation. Therefore a TK + transformant of
`B21, clone TTKBI, was used as a recipient for the subsequent transformation
`of hrpt. Table 1 shows that again the transformation frequency for hprt was
`very low: only one HPRT + colony was isolated, and this proved to be a
`revertant.
`HPRT-deficient mutants of cell lines other than mouse L-cells were also
`tested as recipients for the hprt gene. The last section of Table 1 shows the
`results using 3T6TG8, an H P R T - mutant of mouse 3T6 cells. Again the
`frequency was low, with only one transformant isolated. Similar experiments
`using two other rodent lines, the rat hepatoma FU5-5 (26) and a Chinese
`hamster ovary cell line, have failed to yield any hprt transformants so far
`(data not shown).
`Finally, while human (KB cell) DNA was able to transform the tk gene
`using L M T K - , it was ifleffective in hprt transformation.
`Characterization of HPRT Activity. Extracts of HAS + clones derived
`from control as well as experimental dishes were subjected to electrophoresis
`under conditions (see Materials and Methods) that allow the distinction of
`Chinese hamster and mouse HPRT activities (Fig. 1A, channels 1-3). Figure
`1A shows the results for two HAS + clones isolated after treatment of mouse
`cells with CHO DNA. Clone TA28 (channels 4-6) was derived from mouse
`A9 cells; clone TB23 (channels 7-9) was derived from the double mutant
`B21. In both cases the electrophoretic mobility of the H P R T activity is
`clearly that of the Chinese hamster, the DNA donor. An additional two
`B21-derived clones and the single transformant obtained from 3T6TG8 cells
`similarly contain CHO HPRT.
`In one case a HAS + clone that arose after treatment with CHO DNA
`proved to be a revertant. As can be seen in Fig. 1 B, an H P R T + clone derived
`from line TTKB1 (itself a tk transformant) contains mouse H P R T activity.
`Three HAS + clones isolated in experiments with control DNA also proved to
`be revertants with mouse wild-type H P R T electrophoretic mobility.
`Quantitation of enzyme levels in transformants and revertants that were
`identified by electrophoretic analysis is shown in Table 2. HPRT-specific
`activities in five transformants vary from about one half to five times the
`activity of wild-type CHO (donor DNA) cells or L-cells (recipient). The two
`revertants analyzed exhibit H P R T 16vels similar to wild-type cells.
`In general, colonies that arose under selective conditions (HAS) in these
`experiments bred true when cloned into the same selective medium. In some
`cases, however, colonies were isolated that grew well in nonselective medium
`but poorly or not at all in HAS. The reason for the initial colony formation in
`
`Merck Ex. 1037, pg 1130
`
`
`
`hprt Transformation
`
`1039
`
`Fig. 1. Electrophoretic mobility of HPRT activities. (A) Transformants: (1) CHO (Chinese
`hamster); (2) a physical mixture of CHO + L-cell (mouse); (3) L-cell; (4) transformant
`TA28 + CHO; (5) TA28 + L-cell; (6) TA28; (7) transformant TB23; (8) TB23 + CHO; (9)
`TB23 + L-cell, (B) Revertant: (l) L~cell; (2) L-cell + CttO; (3) CHO; (4) revertant RTB1; (5)
`L-cell; (6) RTBI + L-cell; (7) RTBL + CHO. Migration was toward the anode (bottom).
`
`H A S is not clear in these cases. Perhaps these colonies represent abortive
`transformation events.
`Stability of the HPRT + Phenotype in Transformants. There is a wide
`range in the stabilities of genes that have been introduced into cultured
`mammalian cells via the uptake of metaphase chromosomes or D N A .
`Estimates for the rate of loss of donor traits range from 10 -4 to 10 -1 per cell
`per generation (6, 27). The stability of two of the hprt transformants
`described above was estimated by fluctuation analysis of the segregation of
`the H P R T + phenotype, as indicated by resistance to 6-thioguanine (TG). As
`
`Merck Ex. 1037, pg 1131
`
`
`
`1040
`
`Graf et al.
`
`Table 2. HPRT activities
`HPRT-
`deficient
`parent
`(DNA
`recipient)
`- -
`- -
`A9
`B21
`B21
`B21
`3T6TG8
`TTKB 1
`3T6TG8
`
`Cell
`line
`CHO-K 1
`LMTK-
`TA28
`TB23
`TB31
`TB41
`T313
`RTB 1
`R31
`
`HPRT
`electrophoretic
`mobility
`CH b
`M b
`CH
`CH
`CH
`CH
`CH
`M
`M
`
`HPRT
`specific
`activity a
`4.18
`3.02
`19.40
`7.60
`1.95
`4.44
`4.12
`3.65
`2.53
`
`Wild-type cells
`
`Transformants
`
`Revertants
`
`amlU/mg protein.
`b"Wild-type" enzymes used for definition of Chinese hamster (CH) and murine (M) electro-
`phoretic mobilities.
`
`is shown in Table 3, the two transformants TB23 and TA28 lose the H P R T +
`phenotype at rates of 8 x 10 -5 and 4 x 10-4/cell/generation. These rates are
`among the lowest of those reported in gene transfer experiments. However,
`this rate of generation of T G resistance is still far greater than that expected
`for the spontaneous mutation of a resident hprt gene. To illustrate this point,
`we compared
`the transformants to an H P R T §
`revertant clone. This
`revertant, RD91, had been isolated from an H P R T - T K - L-cell m u t a n t in
`the course of a transformation experiment. It contains H P R T with an
`electrophoretic mobility and specific activity characteristic of wild-type
`mouse cells (data not shown). The last column in Table 3 shows that no
`TG-resistant colonies appeared spontaneously, indicating a mutation rate at
`least three orders of magnitude lower than the rate at which the H P R T §
`phenotype is lost from transformants.
`The rate of loss of H P R T § from transformants was comparable to the
`rate of segregation of this marker from pseudotetraploid heterozygous hybrid
`
`Table 3. Rate of loss of HPRT § phenotype by fluctuation analysis
`TB23
`TA28
`22
`7 x 105
`5 • 104
`70
`204
`583
`2.2 • 10 -4
`4.3 x 10 -4
`
`RD91
`24
`7 • 105
`7 x 105
`0
`0
`
`<6 x 10 -s
`
`No. of cultures
`Mean no. of cells/culture
`Mean no. of cells/sample
`Median no. TG r colonies/sample
`Mean no. TG r colonies/sample
`Variance/mean
`Segregation rates
`Luria-Delbr/ick Equation 8
`
`23
`4.5 • 105
`5 x 104
`11
`38
`89
`5.2 x 10 -5
`8.1 x 10 -s
`
`Merck Ex. 1037, pg 1132
`
`
`
`hprt Transformation
`
`1041
`
`cells (28). If transformation preferentially occurs in a minority of tetraploid
`cells in the population, then the relative instability of the transferred gene
`could simply reflect loss of a supernumerary chromosome with which the gene
`was associated. However, karyotype analysis of the two transformants and
`one revertant listed in Table 3 yielded chromosome numbers of 48, 45, and
`44, all less than the 52 found for the original LMTK- cells (10 metaphase
`spreads counted, standard deviation 3%).
`The loss of the transformed TK + phenotype has been shown to be
`reversible in some cases (29), indicating that gene inactivation, rather than
`physical loss, is taking place. Two HPRT- segregants of the two transfor-
`mants described in Table 3 were tested for the reappearance of the HPRT §
`phenotype. No HAS + colonies were found among 6 x 10 6 cells screened, a
`result consistent with the physical loss of the donated CHO hprt § gene.
`
`DISCUSSION
`
`We have obtained interspecific transformants for the hprt gene by
`treatment of HPRT-deficient mouse cells with purified high molecular
`weight CHO cell DNA. Hprt is thus the fourth mammalian cellular gene for
`which transformants have been selected, following tk (5), adenine phosphori-
`bosyltransferase (6), and dihydrofolate reductase (6). This result confirms
`and extends the experiment recently reported by Willecke et al. (9)
`describing the isolation of a clone of mouse A9 cells that had been
`transformed for hprt using DNA from a mouse revertant line with an altered
`form of the enzyme. Transformation of the hprt gene may be especially useful
`because of the extensive genetic study to which this locus has been subjected
`in cultured mammalian cells.
`The overall frequency of appearance of HPRT § colonies following
`exposure to hprt + DNA was relatively low (5 x 10 -8) compared to results
`obtained for the three other markers mentioned above and compared to our
`own results for tk transformation in LMTK- cells (3.7 x 10-6). Also, DNA
`from KB cells did not produce an hprt transformant, although it was capable
`of yielding tk transformants. In fact, the frequency of hprt transformants was
`not significantly higher than the revertant frequency found overall (experi,
`mental plus control) with the three recipient mouse cell lines used
`1.3 x 10-8). Therefore it was necessary to establish the nature of each
`putative transformant by the electrophoretic identification of the HPRT
`activity as that of the CHO cell DNA donor. Two factors have been shown in
`these experiments to contribute to the low frequency of hprt transformation.
`The more important seems to be a difference in competence among different
`cell lines and among different subclones of the same cell line. Thus the double
`mutant (TK HPRT-) B21 cells are transformable for tk at only one tenth
`
`Merck Ex. 1037, pg 1133
`
`
`
`1042
`
`Graf et al.
`
`the frequency of the L M T K - cells from which they were derived (Table 1).
`An attempt to isolate a more competent HPRT- subclone, by selecting first
`for the transformation of tk, was not successful (clone TTKB 1, Table 1). This
`result suggests that competence is not inherited in an all-or-none manner, in
`agreement with a previous study by Wigler et al. using different markers
`(30).
`The second factor influencing transformation frequency appears to be
`intrinsic to the marker being transferred. Thus transformation of the same
`cells (B21) using the same DNA preparations yielded five times as many tk
`transformants as hprt transformants. There could be many reasons for an
`inherently lower transformation frequency for the hprt gene, including: a
`greater size, making it more susceptible to breakage during DNA isolation; a
`more fastidious requirement for an integration site consistent with gene
`expression; or a requirement for more complete expression before a newly
`transformed cell can survive in the selective medium. On the other hand, the
`transformability of the hprt gene may not be very different than tk on a per
`gene basis: It is likely that there is only one functional copy of the X-linked
`hprt gene per CHO cell (see refs. 8, 20), while there are two or more copies of
`active tk genes. [It is unlikely that the autosomal genes (3l) specifying
`mitochondrial TK, active in LMTK- cells (32), contribute to tk transforma-
`tion, since LMTK- DNA is ineffective as a t k + donor (5).]
`The state of the assimilated hprt gene in the transformed mouse cells is
`not known. It is clear that it is not behaving as a resident wild-type locus since
`it spontaneously mutates
`to an
`inactive form or
`is
`lost at a
`rate
`(10-4/cell/generation) far higher than the indigenous hprt gene of either
`CHO cells or L-cells (Table 3). On the other hand, the rate of loss is much
`lower than that associated with transgenome instability in several previous
`examples (1, 3, 4, 6) and is closer to the "stabilized" state reported after
`propagation of gene transferents isolated via the uptake of metaphase
`chromosomes .(27). In fact, the rate observed here of 10 -4/cell/generation is
`~imilar to the segregation rate of markers from heterozygous hybrid cells
`(28). It is unlikely that this rate is a composite reflecting the presence of both
`completely stable and unstable populations (27), since the calculated variance
`(Table 3) is close to that expected for a homogeneous population (24). The
`loss of the donated hprt gene could be occurring by the loss of a supernumer-
`ary chromosome into which it had been integrated, by the preferential
`excision of the foreign DNA, or by loss of a self-replicating, possibly acentric,
`DNA fragment. The ultimate description of the state of the donated hprt
`gene will probably require a probe for the physical presence of the DNA
`sequence.
`Transformation of the hprt gene using purified DNA represents a
`bioassay for the presence of this gene. One obvious use for such a bioassay is
`
`Merck Ex. 1037, pg 1134
`
`
`
`hprt Transformation
`
`1043
`
`the detection of bacterial or bacteriophage clones carrying this gene as part of
`a recombinant DNA molecule. However, i