Proc. Natl. Acad. Sci. USA Vol. 77, No. 9, pp. 5192-5196, September 1980 Autoradiographic detection of animal mutants altered in synthesis (Chinese hamster ovary cells/immobilized colonies/filter paper replica plating/ metabolism/CDP- synthetase) JEFFREY D. ESKO AND CHRISTIAN R. H. RAETZ* Department of Biochemistry, College of Agricultural and Life Sciences, University of Wisconsin, Madison, Wisconsin 53706 Communicated by Eugene P. Kennedy, June 19,1980

ABSTRACT We have screened approximately 20,000 colo- tecting mutants of Chinese hamster ovary (CHO) cell colonies nies ofChinese hamster ovary cells immobilized on filter paper immobilized on discs of filter paper (8). This makes possible the [Esko, J. D. & Raetz, C. R. H. (1978) Proc NatL Acad. Sci. USA analysis of 104-105 colonies for specific biochemical or phe- 75, 1190-1193] for strains unable to incorporate [methl-"'CJ- choline into trichloroacetic acid-precipitable phospholipid at notypic defects by replica plating or in situ enzymatic assays. 40'C. Mutant 58, identified in this way, was specifically de- Previously, we isolated several myoinositol auxotrophs of CHO fective in choline incorporation, and other isolates were also cells without prior enrichment and demonstrated that myo- blocked in thymidine and leucine incorporation into DNA and starvation of these strains resulted in the virtual elimi- protein, respectively. Further analysis of mutant 58 revealed that nation of from subcellular membranes (8, the strain grew almost normally at 330C, the permissive tem- 9). Robbins (10) has described a histochemical assay for de- perature, but divided only once at 40'C, the restrictive tem- in CHO cells attached to filter perature. After a 20-hr incubation at 400C, the phosphatidyl- tecting a-mannosidase activity choline level dropped from 41% to 20% in the mutant whereas paper and has isolated several variants lacking this lysosomal other , including , continued to enzyme. accumulate. Wild-type cells contained approximately 50% We now report an autoradiographic screening technique for phosphatidyicholine at both temperatures. Anion-exchange identifying CHO colonies unable to generate choline-linked chromatography of the water-soluble choline metabolites ex- phospholipids in vivo. Strain 58, deficient in CDP-choline and tracted from mutant 58 revealed that accu- phosphatidylcholine synthesis at the nonpermissive tempera- mulation increased from 6 nmol/mg of protein at 330C to 42 nmol/mg of protein at 40'C whereas CDP-choline decreased ture, is strikingly defective in CDP-choline synthetase (cho- from 0.42 nmol to less than 0.07 nmol per mg of protein. Phos- linephosphate cytidylyltransferase; CTP:cholinephosphate phorylcholine also increased in wild-type cells shifed from 330C cytidylyltransferase, EC 2.7.7.15) assayed in vitro and accu- to 40°C (from 1.8 nmol to 16 nmol per mg of protein), but the mulates high levels of phosphorylcholine in vivo. Mutant 58 level of CDP-choline was not altered (from 0.52 nmol to 0.58 provides a new approach to studies of the regulation of mem- nmol per mg of protein). Enzymatic assays of extracts prepared brane phosphatidylcholine content and creates the possibility from mutant and wild-type cells revealed a reduction of CTP: phosphorylcholine cytidylyltransferase (EC 2.7.7.15) activity of isolating and mapping the genes involved in mammalian (CDP-choline synthetase) in the mutant to 1/40th that in the wild phosphatidylcholine metabolism. type, and mixing experiments excluded the production of an- tagonists to CDP-choline synthesis in the mutant. Thus, the in- EXPERIMENTAL PROCEDURES ability of the mutant to generate normal amounts of phospha- Materials. [32P]Orthophosphate (carrier-free), [methyl- tidylcholine in vivo was correlated with an enzymatic lesion 14C]choline, [methyl-14C]phosphorylcholine, [methyl-14C]- in the biosynthesis of CDP-choline in vitro. thymidine, and L-[1-14C]leucine were obtained from New Phosphatidylcholine is a major structural component of most England Nuclear. All other materials were reagent grade and cells possessing internal membranes (1, 2), representing about obtained from Sigma. Ham's F12 culture medium, trypsin, half of the membrane phospholipid. Its synthesis appears to be Mycostatin, Fungizone, and fetal bovine serum were obtained localized on the cytoplasmic face of the endoplasmic reticulum from GIBCO. Organic solvents were distilled prior to use. (3, 4), but little is known about the biochemical and genetic Methods. Strain CHO-K1 was cultured as described (8). For factors that control the relative amount and distribution of this mutant screenings, cells were treated with 500 gtg of ethyl . Most subcellular membranes contain different amounts methanesulfonate per ml at 370C for 18 hr (8) and shifted to of phosphatidylcholine (1) distributed unequally in their op- 330C, the permissive temperature, for several generations. After posing faces (5, 6), suggesting that mechanisms exist for its in- the cells were subcultured once at 330 C, they were stored in tracellular transport and insertion into preexisting membranes. liquid nitrogen and revived as needed. The isolation of the Because choline-linked phospholipids are also present in high mutant from this cell stock is described in the text. Filter paper abundance in serum lipoproteins (7), salvage of serum and glass beads were prepared as described (8). Trypsin was and secretion may function along with de novo synthesis and used routinely to subculture cells (11), but for biochemical ex- degradation to set the cellular choline-lipid content. Enzymatic periments, cells were scraped from plastic culture dishes with and genetic studies of animal cell mutants unable to synthesize a rubber policeman and centrifuged at 600 X gav for 10 min. phosphatidylcholine would help to unravel the complex me- To make a crude lysate, we washed cell pellets once, resus- tabolism of this lipid, but such mutants have not been ob- pended them in 0.3 M sucrose containing 10 mM Tris-HCl (pH tained. 7.6), and froze them at -200C. Recently we described a rapid screening procedure for de- Radioactive labeling, extraction, and analysis of phospho- lipids were done as described (9) or were modified as indicated The publication costs of this article were defrayed in part by page in the table legends. Protein was measured according to Lowry charge payment. This article must therefore be hereby marked "ad- vertisement" in accordance with 18 U. S. C. §1734 solely to indicate Abbreviation: CHO cells, Chinese hamster ovary cells. this fact. * To whom reprint requests should be addressed. 5192 Downloaded by guest on September 29, 2021 Biochemistry: Esko and Raetz Proc. Natl. Acad. Sci. USA 77 (1980) 5193 RESULTS Autoradiographic Detection of Mutant Colonies Unable To Incorporate Choline into Phospholipid. When CHO cells were labeled for several hours with radioactive choline, chlo- roform extraction revealed that the label was incorporated ef- ficiently into choline-containing phospholipids, primarily phosphatidylcholine. If labeled cells were treated with tri- chloroacetic acid, over 98% of the acid-precipitable radioac- tivity was also chloroform soluble, indicating that [methyl- A 14C]choline was not a precursor for macromolecules other than membrane lipids (data not shown). Because of this specificity, the incorporation of labeled choline into acid-precipitable ~~~~~B material permitted the autoradiographic detection of phos- phatidylcholine synthesis in colonies attached to filter paper (Fig. 1). When CHO colonies derived from mutagen-treated cells were transferred to filter paper at 330C and then incubated for several hours with radioactive choline at 400C, autoradi- ography revealed that all of the colonies present on the disc were radioactive except for an occasional variant, indicated by the C 0 arrow (compare Fig. 1 A and B). Unlabeled colonies of this kind occurred with an approximate incidence of 1:5000. To repurify FIG. 1. [methyl- 14CJCholine autoradiography of CHO cell the putative mutants from adjacent wild-type cells (Fig. 1A), colonies immobilized on filter paper. Mutagen-treated cells were we retrieved the desired colonies from the original master plates dispersed with trypsin (11) and placed in 100-mm-diameter tissue- culture dishes to yield approximately 200 colonies per plate at-330C. by using glass cloning cylinders and trypsin (14). The cells were After 1 day, the cells were overlayed with a disc of Whatman no. 50 replated at different dilutions at 330C and again transferred filter paper and glass beads (8) and incubated at 330C for another 16 to filter paper as described in the legend of Fig. 1. Because the days. After the medium was aspirated and the beads were decanted, majority of colonies present after the first cloning were defec- the filter paper was removed from the dish with sterile tweezers and tive, as judged by choline autoradiography (Fig. 1 C and D), placed cell side up on a sterile metal or glass pan tilted at a 600 angle. further was not The surface tension of the residual medium held the disc firmly repurification generally necessary. Single against the pan. The paper was then rinsed vigorously with a 30-ml colonies were picked and used for the experiments described stream of medium lacking serum to remove loose cells. Next, it was below. The autoradiogram and stained filter paper corre- placed cell side up on top ofan even layer ofglass beads in a dish filled sponding to the repurification of mutant 58 (the colony indi- with enough medium (containing 10%, vol/vol, dialyzed serum) to cated by the arrow in Fig. 1 A and B) are shown in Fig. 1 C and keep the paper moist. After 16 hr at 401C, the disc was placed on an D, respectively. absorbent paper towel to remove excess moisture and was transferred Macromolecular Synthesis in Putative Mutants. Because to another dish containing 1 ml ofgrowth medium supplemented with 0.1 mM [methyl- 14Cjcholine (10 Ci/mol; 1 Ci = 3.7 X 101° becquerels). genetic lesions other than those unique to phosphatidylcholine After 4 hr of further incubation at 400C, the paper was treated with synthesis (Fig. 2) could result in defective choline incorporation 1 ml of 10%1 (vol/vol) trichloroacetic acid, which resulted in precipi- in vivo (e.g., energy generation), we measured the synthesis of tation of the choline-linked phospholipids. Unincorporated radio- DNA and protein in all the mutants. As shown for two variants active precursor was removed by placing each disc in a Bachner funnel (Table 1), incorporation of thymidine and leucine into DNA and passing five 50-ml volumes of 2% trichloroacetic acid through the and protein, at 330C was similar to that observed paper under reduced pressure (13). The papers were then dried respectively, overnight at room temperature and exposed to KODAK XR-5 x-ray in wild-type cells. Under the same conditions, mutant 58 in- film for 4 days. After autoradiography the papers were stained over- corporated choline into phospholipid about 20% less efficiently night with 0.05% Coomassie brilliant blue G in 10% (vol/vol) acetic than wild-type cells. After 21 hr of incubation at 40'C, mutant acid to visualize all the colonies. The papers were destained by 11 synthesized less of all macromolecules when compared to stacking them in a beaker containing 300 ml ofmethanol/water/acetic wild-type cells, whereas mutant 58 retained a nearly normal acid, 45:45:10 (vol/vol), and stirring at 371C for about 1 hr. Several changes of destaining solution resulted in the appearance of bright capacity for DNA and protein synthesis, despite a 90% reduc- blue colonies on a virtually white background. Throughout these tion of choline incorporation into lipid relative to wild-type manipulations, the master plate was stored at 280C under otherwise normal growth conditions in medium supplemented with 10%1 serum, 20 units of Mycostatin per ml, and 2.5 ,g of Fungizone per ml. A fresh piece of Whatman no. 42 filter paper and glass beads were used to prevent formation of satellite colonies during storage. Mutants, identified as blue-staining colonies lacking an autoradiographic halo (arrow indicates position of mutant 58), were retrieved with glass cloning cylinders (14) from the master plates. All candidates were Choline m Choline t Phosphorylcholine passed through the above cloning procedure one more time to achieve their complete purification. (A and C) Autoradiograms from the original mutant screening and the subsequent repurification, re- gPDiglyceride m brn spectively; (B and D) the corresponding stained filter papers. Xoline et al. (12), with bovine serum albumin as standard. A simplified assay for CDP-choline synthetase in cell extracts is described Pasma membrane in the legend of Table 4. [methyl-14C]Phosphorylcholine was treated with charcoal before use to remove a minor radio- FIG. 2. CDP-choline-dependentphosphatidylcholine synthesis chemical contaminant present in commercial preparations. in mammalian cells. Downloaded by guest on September 29, 2021 5194 Biochemistry: Esko and Raetz Proc. Natl. Acad. Sci. USA 77 (1980) Table 1. Macromolecular synthesis in mutants defective in choline, and (Fig. 2), we investigated the possibility choline incorporation that the lesion present in mutant 58 was in the metabolism of Temp., cpm/106 cells these precursors by examining the phospholipid composition Strain OC [14C]Thymidine ['4C]Leucine [14C]Choline of this strain (Table 2). The content of phosphatidylcholine was lower at 33'C in the mutant compared to wild-type cells (41% Parent 33 2.0 X 104 2.5 X 104 5.2 X 103 compared to 47%), and after 20 hr of incubation at 400C, CHO-K1 40 2.8 x 104 2.8 X 104 7.5 x 103 phosphatidylcholine was reduced further to 21%. The levels Mutant 33 2.4 X 104 3.1 X 104 3.9 X 103 of other phospholipids were not grossly affected, except that 58 40 2.0 X 104 2.6 X 104 0.6 103 phosphatidylinositol accumulated somewhat more than the X other phospholipids. Sphingomyelin formation was not im- Mutant 33 1.8 X 104 2.6 X 104 8.0 X 103 mediately affected under nonpermissive conditions. These 11 40 0.1 X 104 0.6 X 104 3.3 X 103 results suggested that altered ATP, CTP, or diglyceride syn- thesis was not responsible for the reduction in phosphatidyl- Strains judged defective for radioactive choline incorporation by filter paper colony autoradiography (see legend to Fig. 1) were repu- choline content in mutant 58. rified and plated into multiple 60-mm dishes to 105 cells per plate at Defective CDP-Choline Synthesis in Mutant 58 at 40'C. 330C. After 2 days, some cultures were shifted to 40'C for 18 hr. The To determine at what step choline incorporation into lipid was growth medium was then removed and 1 ml of medium containing altered (Fig. 2), we labeled mutant and wild-type cells uni- 0.1 mM [methyl-14K+choline (5 Ci/mol), 30,gM [methyl-14C]thy- formly with radioactive choline and measured the radioactivity midine (33 Ci/mol), or 0.1 mM [1-14C]leucine (5 Ci/mol) was added. present in both water-soluble (choline, phosphorylcholine, and After 3 hr at the indicated temperature, 10 ml of 10% trichloroacetic CDP-choline) acid was added, the precipitate was collected on 0.4-.um Millipore and chloroform-soluble (phosphatidylcholine, filters, and radioactivity was measured by liquid scintillation spec- sphingomyelin, and ) material at 40'C trometry. In order that the amount of each precursor incorporated (Table 3). The mutant contained 13 nmol and 48 nmol of could be normalized to cell number, a second dish was harvested with water-soluble choline metabolites per mg of protein at 330C trypsin (11), and the cells were counted on a Coulter Model B Counter. and 40'C, respectively, whereas wild-type cells contained 18 Each value is from a single determination. nmol/mg of protein at 330C and 63 nmol/mg of protein at 40°C. Considerably less chloroform-soluble radioactivity was present in mutant cells shifted to 400C, resulting in at least a cells. With the exception of mutant 58, all strains isolated by 50% reduction in phosphatidylcholine content and confirming colony autoradiography behaved like mutant 11. By this cri- the results shown in Tables 1 and 2. terion only mutant 58 was defective specifically in choline The water-soluble radioactivity was then fractionated by utilization, and so we characterized this strain further anion-exchange column chromatography to determine if the Temperature-Sensitive Growth of Mutant 58. Mutant 58 distribution of the choline metabolites was abnormal in the divided more slowly than wild-type cells at 330C, doubling mutant at 400C (Table 3). The majority of the radioactivity not every 24 hr as opposed to 20 hr (compare the lower curve in Fig. bound by the resin was identified as choline by ascending paper 3A to the upper curve in Fig. SB). After cells were shifted to chromatography (18) and was reduced in the mutant regardless 400C, the doubling time of wild-type cells was reduced to ap- of temperature. Both phosphorylcholine and CDP-choline were proximately 13 hr (Fig. 3A, upper curve), whereas the mutant retained by the resin and eluted with formic acid (17). Their divided only once and then stopped growing (Fig. 3B, lower identities were established by chromatography with authentic curve). Although more than 80% viability was retained for 30 radioactive compounds and chemical standards (data not hr at 40°C (data not shown), visible cell lysis began by 36 hr. shown). Phosphorylcholine accumulation increased about 8-fold Reduced Phosphatidylcholine Content in Mutant 58. Be- when either the mutant or wild-type cells were shifted from cause the formation of phosphatidylcholine requires ATP, CTP, Table 2. Phospholipid composition of wild-type cells and mutant 58 Temp., % of total phospholipid* *A B Strain 0C PC SPH PE PI PS PG Othert Parent 33 47.0 10.2 25.6 6.7 6.1 0.7 3.7 510 CHO-K1 40 50.4 9.3 22.9 7.0 6.6 0.7 3.1 Mutant 33 41.0 11.1 29.0 7.8 6.9 0.4 2.9 Shif 58 40 21.0 13.0 37.1 14.4 10.2 0.7 4.6 Cells were grown for several generations at 33°C in medium sup- plemented with 32P3 (2 ,uCi/ml) to label the phospholipids to constant los ftShift specific activity. The cells were then replated at 3 X 105 cells per 100-mm dish at 33°C in medium containing 32P at the same specific I I AnI I I * I I I I I activity. After 1 day, duplicate cultures were shifted to 40°C for ap- uA ZU 4U UU BU 100 12U 0 20 4U UU BU WU IZ0 proximately 20 hr. The cells were extracted with chloroform/methanol Time, hr (9, 15), and the phospholipids were analyzed by two-dimensional thin-layer chromatography as described (9). Average data are shown FIG. 3. Growth of mutant 58 and wild-type cells at 330C (@--) from at least two determinations, which varied by less than 10%. La- and 40'C (O--- 0). Multiple 60-mm-diameter dishes were inoculated beling of cells under these conditions nearly reflects the chemical with approximately 2 X 105 wild-type or mutant cells and incubated composition (9). at 330C. Twenty-four hours after the start of the experiment, one set * PC, phosphatidylcholine; SPH, sphingomyelin; PE, phoAphati- of cultures was transferred to 400C. At the indicated times cells from dylethanolamine; PI, phosphatidylinositol; PS, ; duplicate dishes were dispersed with trypsin (11) and counted without PG, . centrifugation on a Coulter Model B Counter. (A) Wild-type cells; t Other lipids include lysophosphatidylcholine, lysophosphati- (B) mutant 58. dylethanolamine, , and . Downloaded by guest on September 29, 2021 Biochemistry: Esko and Raetz Proc. Natl. Acad. Sci. USA 77 (1980) 5195

Table 3. Quantitation of choline metabolites in mutant 58 and Table 4. Enzymatic assay of CDP-choline synthetase in extracts wild-type cells prepared from mutant and wild-type cells Chloroform- Specific activity, soluble Strain nmol/min per mg Water-soluble metabolites* metabolites* Total Parent CHO-K1 7.6 Temp., Not Phosphoryl- CDP- choline Mutant 58 0.2 Strain 0C bound choline choline lipids Mixing experiment 3.6 Parent 33 16 1.8 0.52 CDP-choline synthetase was measured in freeze-thaw lysates of 128 mutant and wild-type cells grown at 331C. Approximately 10-30 ,g CHO-K1 40 46 16 0.58 134 ofcrude cell protein was mixed with 20 mM MgCl2, 50 mM Tris-HCl (pH 8.0), 5 mM CTP, and 3 mM [methyl- 14C]phosphorylcholine (free Mutant 33 6.6 6.1 0.42 107 acid, 0.2 Ci/mol) in 0.1 ml and incubated for 15 min at 40'C. After 58 40 5.6 42 0.07 55 dilution to 1 ml with 1 Mmol of CDP-choline as carrier in cold 10 mM phosphate buffer (pH 7.5) containing 10 mM choline chloride, Mutant and wild-type cells were incubated in medium with 0.1 mM the mixture was applied to a pasteur pipet minicolumn plugged with [methyl- 14C]choline (10-20 Ci/mol) for approximately five cell gen- glass wool, containing approximately 200 mg ofcharcoal (80/100 mesh, erations at 330C in order to label choline-containing compounds to Anspec, Ann Arbor, MI). First, the column was eluted with 30 ml of constant specific activity. After trypsin treatment, approximately 3 the Pi/choline buffer to remove unreacted phosphorylcholine. Next, X 105 cells were placed in duplicate 100-mm-diameter dishes con- the bound CDP-choline was eluted with 10 ml- of ethanol/H20/ taining 15 ml ofgrowth medium and radioactive choline at the same NH40H, 50:50:1 (vol/vol), containing 10 mM sodium phosphate, 10 specific activity. When approximately 2 X 106 cells were present in mM choline chloride, and 10 mM cytidine. The recovery of CDP- the dishes, duplicate cultures were shifted to 40'C for 12-18 hr. The choline was consistently 65%. The samples containing CDP-choline cells were harvested with a rubber policeman and mixed with 6 ml of were dried and suspended in 2.0 ml of water and 13 ml of scintillation chloroform/methanol, 1:2 (vol/vol), containing 1 Amol each of bovine fluid (19), and radioactivity was measured by liquid scintillation phospholipid, phosphorylcholine, CDP-choline, and choline as spectrometry. Formation of CDP-choline was linear with time for at chemical carriers. After extraction under Bligh-LQyer conditions (15), least 60 min and with protein between 4 and 80 ,g. Specific activities the upper water/methanol phase was removed and adjusted to pH 8 varied by less than 10%o between duplicates. In the mixing experiment, with a few drops of 1.5 M NH40H dissolved in methanol. Water-sol- equal amounts of protein from mutant and wild-type cells were uble choline metabolites present in this fractionwere analyzed by combined and assayed. The same results were obtained with dCTP ion-exchange chromatography on a 0.9 X 25 cm column of 200-400 as the substrate. mesh of AG 1-X2 (formate). Nonadsorbing metabolites were first removed with 100 ml of water, and negatively charged metabolites were sequentially eluted with a nonlinear formic acid gradient (16, hibitor synthesis in the mutant 17). Five-milliliter fractions were collected (0.5 ml/min) and analyzed and also eliminated the presence for radioactivity. Recovery of radioactivity was greater than 95% in of an activating factor in wild-type extracts that was missing all experiments. The radioactivity of an aliquot of each chloroform in the mutant. phase from the original extraction was also determined as a measure of choline-linked lipids. The CDP-choline fraction is a mixture of ribo and deoxyribo CDP-choline. DISCUSSION * Values are given as nmol/mg of protein. All values were normalized to the amount of protein extracted, as determined from duplicate Although the incidence of specific variants like mutant 58 was cultures. Shown are the averages from two independent determi- about 1:20,000, autoradiography of colonies immobilized on nations, which varied by about 20%. filter paper permitted the detection of this mutant without an enrichment technique. Additional phosphatidylcholine mutants 33°C to 400C, but the mutant contained 2- to 3-fold more might be found by screening 5-10 times the number of colonies phosphorylcholine at both temperatures. In contrast, the level (i.e., 100,000 colonies or approximately 500 plates). A specific of CDP-cholinet was about 20% lower in the mutant (0.42 selection for phosphatidylcholine mutants would be invaluable, nmol/mg of protein) at 330C compared to wild-type cells (0.52 but unfortunately our results demonstrate that DNA and pro- nmol/mg of protein) and was strikingly reduced at 40°C from tein synthesis continued almost to the point of cell lysis (30-36 0.58 nmol/mg of protein in wild-type cells to 0.07 nmol/mg hr. Table 1). This suggests that classical enrichment methods of protein in the mutant. using nucleoside analogs like BrdUrd (20) might not be feasible. These results suggested that mutant 58 was not defective in In view of the selectivity with which exogenous choline is me- choline transport or in the synthesis of phosphorylcholine (be- tabolized to membrane lipids in CHO cells, the possibility of cause the latter was present in greater than normal amounts), enriching by radiation suicide (21), by using tritiated choline, but rather in the formation of CDP-choline. To test this hy- deserves serious consideration. pothesis, we prepared extracts from mutant and wild-type cells Mutant 58 is the first mammalian cell variant with a specific grown at 330C and measured CDP-choline synthesis in vitro lesion in phosphatidylcholine synthesis. Our evidence strongly at 40°C by a modified charcoal-binding assay (Table 4). The suggests that the reduction of the CDP-choline pool observed product of the reaction was identified by comparison with in uivo (Table 3) results from a specific defect in CDP-choline authentic CDP-choline and by anion-exchange chromatogra- synthetase, as measured in nitro (Table 4). These results provide phy as described in Table 3. In extracts of the mutant, the direct genetic evidence that CDP-choline functions in mam- specific activity of CDP-choline synthetase measured in vitro malian cells as the primary source of phosphorylcholine was reduced to almost 1/40th that of the wild-type extracts moieties for de novo phosphatidylcholine synthesis, as first (Table 4). When equal amounts of protein from mutant and suggested by the classical work of Kennedy and Weiss (22). The wild-type cells were mixed and assayed for CDP-choline syn- possibility of polar head group exchange reactions (23, 24) and thetase, half the wild-type specific activity was obtained (3.6 (25) as additional compared to 7.6 nmol/min per mg of protein). This excluded sources of phosphatidylcholine cannot be excluded by our the possibility of increased CDP-choline degradation or in- findings, but these pathways are not sufficient for maintaining adequate levels of phosphatidylcholine in CHO cells. t About one-third of this fraction is deoxyribo CDP-choline. Mutants of Saccharomyces cerevisiae (26) and Neurospora Downloaded by guest on September 29, 2021 5196 Biochemistry: Esko and Raetz Proc. Natl. Acad. Sci. USA 77 (1980)

crassa (27, 28) requiring exogenous choline for growth syn- 9. Esko, J. D. & Raetz, C. R. H. (1980) J. Biol. Chem. 255,4474- thesize less phosphatidylcholine in the absence of supplemen- 4480. tation. These strains are defective either in CDP-diglyceride- 10. Robbins, A. R. (1979) Proc. Nati. Acad. Sci. USA 76, 1911- 1915. dependent phosphatidylserine sypthesis (26) or in phosphati- 11. Litwin, J. (1973) in Tissue Culture-Methods and Applications, dylethanolamine methylation (27, 28), but neither of these eds. Kruse, P. F. & Patterson, M. K. (Academic, New York), pp. pa.thways apparently plays a significant role in CHO cells. In 671-673. this regard, most tnammalian lines, including CHO cells, re- 12. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J. quire choline for growth (29, 30). Lipid alterations resembling (1951) J. Biol. Chem. 193,265-275. thqse observed in mutant 58 can be induced in wild-type cells 13. Raetz, C. R. H. (1975) Proc. Natl. Acad. Sci. USA 72, 2274- extensively starved for choline (29, 30). However, in mutant 2278. 58 phosphatidylcholine depletion occurs rapidly in the presence 14. Jacobs, L. & DeMars, R. (1977) in Handbook of Mutagenicity of exogenous choline and serum, and the intracellular phos- Test Procedures, eds. Kilbey, B. J., Legator, M., Nichols, W. & phorylcholine pool, which may have other metabolic fates, is Ramel, C. (Elsevier, New York), pp. 193-220. not depleted. 15. Bligh, E. G. & Dyer, W. J. (1959) Can. J. Biochem. Physiol. 37, Despite the rapid inhibition of phosphatidylcholine synthesis 911-917. 16. Hurlbert, R. B., Schmitz, H., Brumm, A. F. & Potter, V. R. (1954) at 40°C in mutant 58 (Tables 1 and 2), the synthesis of other J. Biol. Chem. 209, 23-39. phospholipids continued for about 20 hr. Because sphin- 17. Kennedy, E. P. (1956) J. Biol. Chem. 222, 185-191. gomyelin was also formed under conditions of CDP-choline 18. Schneider, W. C., Fiscus, W. G. & Lawler, J. B. (1966) Anal. depletion, the phosphorylcholine moiety of sphingomyelin may Biochem. 14, 121-134. not be derived primarily from CDP-choline (31). Some evi- 19. Patterson, M. S. & Green, R. C. (1965) Anal. Chem. 37, 854- dence exists that the polar head group of sphingomyelin can 862. arise from phosphatidylcholine (32, 33), a possibility that is 20. Kao, F.-T. & Puck, T. T. (1968) Proc. Natl. Acad. Sci. USA 60, consistent with our findings. 1275-1281. 21. Baker, R. M. & Ling, V. (1978) in Methods in Membrane Biology, We thank Ms. Mary Wermuth for excellent technical assistance in ed. Korn, E. D. (Plenum, New York), pp. 337-384. growing the cells. This research-was supported in part by a Basil 22. Kennedy, E. P. & Weiss, S. B. (1956) J. Biol. Chem. 222, 193- O'Connor Grant from the National Foundation, The March of Dimes, 214. by a Harry and Evelyn Steenbock Career Advancement Award, and 23. Dils, R. R. & Hubscher, G. (1961) Biochim. Biophys. Acta 46, by Grants AM 21722 and 1K04-AM00584 from the National Institute 505-513. of Arthritis, Metabolism and Digestive Diseases to C.R.H.R. The re- 24. Bjerve, K. S. & Bremer, J. (1969) Biochim. Biophys. Acta 176, search described here forms part of a dissertation of J.D.E. to be sub- 570-583. mitted to the University of Wisconsin-Madison in partial fulfillment 25. Bremer, J., Figard, P. H. & Greenberg, D. M. (1960) Biochim. of the requirements of the Ph.D. degree. J.D.E. was supported by a Biophys. Acta 43,477-488. competitive fellowship from the Wisconsin Alumni Research Foun- 26. Atkinson, K. D., Jensen, B., Kolat, A. I., Storm, E. M., Henry, S. dation. A. & Fogel, S. (1980) J. Bacteriol. 141, 558-564. 27. Hubbard, S. C. & Brody, S. (1975) J. Biol. Chem. 250, 7173- 1. Ansell, G. B., Dawson, R. M. C. & Hawthorne, J. N., eds. (1973) 7181. Form and Function of Phospholipids (Elsevier, New York), p. 28. Scarborough, G. A. & Nyc, J. F. (1967) J. Biol. Chem. 242, 494. 238-242. 2. Van den Bosch, H. (1974) Annu. Rev. Biochem. 43,243-277. 29. Horwitz, A. F. (1977) in Dynamic Aspects of Cell Surface Or- 3. Vance, D. E., Choy, P. C., Farren, S. B., Lim, P. H. & Schneider, ganization, eds. Poste, G. & Nicolson, G. L. (North-Holland, New W. J. (1977) Nature (London) 270,268-269. York), Vol. 3, pp. 295-305. 4. Coleman, R. & Bell, R. M. (1978) J. Cell Biol. 76,245-253. 30. Glaser, M., Ferguson, K. A. & Vagelos, P. R. (1974) Proc. Natt. 5. Bergelson, L. D. & Barsukov, L. I. (1977) Science 197, 224- Acad. Sci. USA 71, 4072-4076. 230. 31. Scribney, M. & Kennedy, E. P. (1958) J. Biol. Chem. 233, 6. Op den Kamp, J. A. F. (1979) Annu. Rev. Biochem. 48, 47- 1315-1322. 71. 32. Ullman, M. D. & Radin, N. S. (1974) J. Biol. Chem. 249, 7. Scanu, A. M. (1972) Biochim. Blophys. Acta 265,471-508. 1506-1512. 8. Esko, J. D. & Raetz, C. R. H. (1978) Proc. Natl. Acad. Scd. USA 33. Diringer, H., Marrgraf, W. D., Koch, M. A. & Anderer, F. A. 75, 1190-1193. (1972) Biochem. Biophys. Res. Commun. 47, 1345-1352. Downloaded by guest on September 29, 2021