JOURNAL OF BACTERIOLOGY, August 1973, p. 543-551Copyright 0 1973 American Society for Microbiology

Vol. 115, No. 2Printed in U.S.A.

Quantitative Measurement of the Effectivenessof Unsaturated Fatty Acids Required for the

Growth of Saccharomyces cerevisiaeEUGENE D. BARBER AND WILLIAM E. M. LANDS

Department of Biological Chemistry, The University of Michigan, Ann Arbor, Michigan 48104

Received for publication 19 March 1973

The growth response of a mutant of Saccharomyces cerevisiae which is unableto synthesize unsaturated fatty acids has been measured in the presence ofvariable concentrations of exogenous unsaturated fatty acids. Final cell yields,doubling times, and lag times were all found to vary as a function of the initialconcentration of the added unsaturated acid. The cell yield was found to be aconvenient quantitative measurement to use in comparing the effectiveness ofvarious unsaturated acids. Values for the acids ranged from 1.7 to 11 cells perfemtomole with values for oleate and palmitoleate at 2.7 and 4.3 cells perfemtomole, respectively. In general, the effectiveness of unsaturated acids wasfound to increase with an increasing number of double bonds. Saturated fattyacids of a chain length of 5 to 18 carbon atoms were completely ineffective. Thevaried efficiencies of different unsaturated fatty acids indicate that unsaturationper se was not the basis of the nutritional requirement and indicate certain acidsthat would be useful in further studies of the role of unsaturated acids in cellfunction.

Large differences in lipid composition existamong naturally occurring biomembranes. Forexample, the lipids of rat liver plasma mem-brane contain about 20 to 30% sterol by weight(30), whereas the membrane lipids of Esche-richia coli and other bacteria lack sterol com-pletely (14, 17). Similarly, phosphatidyl cholinerepresents some 40% of the total phospholipidin rat liver membranes (30), but is absent fromE. coli cells (12, 14), which contain about 85%phosphatidyl ethanolamine (12). In spite of thelarge compositional differences, some generali-zations can be made concerning biologicalmembranes; one is that all of these membranescontain phospholipids. In general, the phos-pholipids have been found to contain bothsaturated and unsaturated fatty acyl esterswhich are distributed nonrandomly between theavailable positions (Table 1, 8). Reports duringthe past 10 years have provided a more detailedbase of information on how certain acids cometo be located in certain positions of tissuephosphoglycerides. Not only have the speciflcenzymic steps for selectivity been suggested,but also some concept of the chemical featuresrecognized by those enzymes is evolving. Fur-thermore, the selectivities for some acyltrans-

ferases measured in vitro may occasionallyallow a fairly accurate prediction of the compo-sitions of saturated and unsaturated fatty acidsin some membrane lipids (18, 19, 43). Most im-portantly, the enzymic selectivities for acidsmay not be simply classed as saturated or un-saturated, but may depend more upon ethylenicir-bonds at certain locations or upon configura-tion at certain locations depending upon thecatalyst being considered (26, 40).Although the compositional data always

showed a consistent presence of some un-saturated or low melting acids in membranephosphoglycerides, the recent isolation of mu-tants requiring exogenous unsaturated fattyacids of E. coli (5, 27, 35) and Saccharomycescerevisiae (15, 32) has more clearly demon-strated the essentiality of the presence of un-saturated fatty acids in membrane lipids.Nevertheless, though several groups have shownthat a variety of fatty acids can fulfill thenutritional requirement for an unsaturatedfatty acid (4, 31, 36, 46, 47), no quantitativecomparison of the effectiveness of various acidshas been provided and the structural featuresthat make an unsaturated acid beneficial toliving cells have not been identified. A method

543

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

TABLE 1. Effectiveness of some common fatty acids in supporting the growth of S. cerevisiae KD46"

Double~~ ~ ~ ~ ~ ~ 1MaximumFattv . bond E(cells Per fmol) T, (h) KS(,uM) K. (,uM) yieldacid position()per cell) (cells/ml

x 10 6

16:1 9 4.26 + 0.36 (11) 234 1.8 6.3 47 11018:1 9 2.73 0.16 (22) 366 2.4 11.:3 900 12518:2 9,12 5.02 0.43 (11) 199 2.0 4.6 22 12018:3 9, 12, 15 4.64 ± 0.24 (8) 216 1.6 5.4 40 12520:2 11,14 4.74 0.39 (7) 211 6.1 1.0 5 3020:3 11,14,17 4.36 0.49 (13) 230 3.6 5.0 17 4020:4 5,8,11,14 8.80 + 0.63 (20) 114 2.2 4.5 170 15020:5 5,8, 11, 14, 17 9.20 + 0.75 (7) 109 2.4 4.2 100 13022:1 13 1.67 0.30 (6) 600 6.0 1.2 (4) 3022:6 4, 7, 10, 13, 16, 19 10.99 ± 0.65 (10) 91 2.2 5.0 75 1105:0 0.007 > 140,0006:0 0.012 > 83,0008:0 0.062 > 16,00010:0 0.063 > 16,00016:0 0.122 ± 0.029 (10)18:0 0.044 0.082 (11) >23,000

a6 Values are given as mean + standard error (number of determinations). Tmin, K6, and K8' values aredetermined from intercepts of plots such as those shown in Fig. 4 and 5. Values for e and maximum yield foreach acid were determined from yield plots such as those shown in Fig. 3.

for quantitatively comparing the relative suita-bility of various unsaturated fatty acids wouldnot only provide more precise information onthe structure-function relationships of acids in agiven organism, but would also provide a com-parative measure of the acyl chain requirementsfor different organisms. We have, therefore,examined the ability of a variety of saturatedand unsaturated fatty acids to promote thegrowth of S. cerevisiae KD46, a mutant requir-ing unsaturated fatty acid (15). S. cerevisiaewas chosen because it is a eukaryote whosegrowth and genetic properties can be conven-iently studied in liquid culture. The resultsprovided reproducibly quantitative measures ofthe growth response as a function of the concen-tration of different added fatty acids.

MATERIALS AND METHODSYeast extract, agar, and peptone were all ob-

tained from Difco. Fatty acids were purchasedfrom the Hormel Institute or from Nu-Chek Preps andwere of the highest purity available. Tergitol NP-40, aproduct of Union Carbide Company, was provided tous by Alec Keith. Tween 80 and Tween 40 wereobtained from the Sargent Welch Scientific Com-pany. BF3 (14%) in methanol was obtained fromApplied Science Laboratories, State College, Pennsyl-vania.

All other chemicals were obtained from commercialsources and were reagent grade or better.Organism and medium used. S. cerevisiae, strain

KD46 (ole 2), was a generous gift from Alec Keith.The basal medium used is the complex yeast extract-peptone-dextrose (YEPD) broth described by Keith et

al. (15). The detergent Tergitol NP-40 was added at1% (wt/vol) only in a few experiments where indi-cated. A solid medium for cell growth was preparedby adding 2% (wt/vol) agar to the basal system. Forthese agar plates, either Tween 80 or Tween 40 (2%wt/vol for both) was included as a source of unsatu-rated or saturated fatty acids, respectively (32). TheKD46 mutant, because of its requirement for an un-saturated fatty acid, grows well on agar plates con-taining Tween 80 and very poorly on plates containingTween 40. A suspected revertant, isolated by itsability to grow in liquid medium which contained nounsaturated fatty acid supplement, was shown togrow equally well on both Tween 80 and Tween 40plates. In order to maintain the strain, fresh plateswere prepared every 2 to 3 days by dispersing the cellsfrom a single colony in sterile medium, diluting, andtransferring about 100 cells in 0.1 ml to a fresh plate.The purity of the mutant colonies was checked byreplicate plating onto both Tween 40 and Tween 80plates. Revertants that no longer require unsaturatedfatty acids as well as contaminants can be recognizedand avoided by this procedure.

Growth experiments were performed as follows.Twenty hours before an experiment was begun, a tubecontaining 8 ml of medium (with 50 gM oleate assupplement) was inoculated to a density of approxi-mately 10 x 106 to 15 x 106 cells per ml by using asterile loop to transfer colonies grown on solid me-dium. After 8 h of growth at 30 C, samples werediluted to provide several tubes with cell densitiesbetween 3 x 106 and 15 x 106 cells per ml. After 12additional h the tube that most nearly approximated3,4 log growth (approximately 101 cells per ml)provided the inoculum for the growth experiment.Growth tubes were then incubated at 30 ± 2 C on aNew Brunswick model G-25 gyratory shaker atshaker speed "9". The above procedure was designed

544 J. BACTERIOL.BARBER AND LANDS

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FATTY ACIDS REQUIRED FOR S. CEREVISIAE GROWTH

to insure uniformity among the inocula from experi-ment to experiment.Measurements of cell growth. Cell growth in

liquid culture was monitored turbidimetrically byreading the absorbance at 660 nm in a Bausch &Lomb Spectronic-20 spectrophotometer. Tubes, 18mm in diameter, containing 8 ml of medium wereused throughout. Cell cultures having absorbance at660 nm less than 1.0 were read directly, whereas thosehaving absorbance greater than 1.0 were evaluated bydiluting cultures 1 to 9 in 0.1 M NaCl and bydetermining absorbance at 660 nm of this dilutedsolution. Cell numbers (cells per milliliter) were alsodetermined by direct microscope observation andcounting in a Levy-Hausser hemacytometer (42). Inaddition, viable cell counts were obtained by plating aknown dilution of liquid culture on Tween 80 platesand counting the number of colonies developing after48 to 72 h.

Fatty acid supplements. The concentration andpurity of all fatty acid stock solutions was checked byusing standard quantitative gas chromatographictechniques with methyl esters prepared by modifica-tion of the method originally described by Metcalfeand Schmitz (24) with a known amount of pen-tadecanoic acid included as an internal standard ineach run.The purity of the acids used in these studies is

given below: palmitoleate (16:1), oleate (18:1), andeicosadienoic acid (20: 2) were all >99% pure with noimpurities detectable on the chromatograms. Linole-ate (18:2) contained about 1% oleate and the linole-nate (18:3) preparation was contaminated with 1%18:2 and 0.6% 18:1. Arachidonate (20:4) was alsofound to contain small amounts of lineolenate (0.8%)and oleate (1.5%). Eicosatrienoic (20:3) and eicosa-pentaenoic (20:5) acids each contained trace (<1%)impurities whose retention times did not correspondwith any common unsaturated fatty acid. Docosahex-aenoic acid (22:6) contained a 3.2% impurity whoseretention time was longer than that of 22:6 itself.Fatty acid stock solutions were prepared as soaps inethanol and stored at -14 C.

RESULTSThe relationship between cell abundance and

absorbance at 660 nm derived from the entirerange of values encountered in these studiesgave a value of (38.4 2.0) x 106 cells per ml fora culture of A = 1.0. A series of viable cell-countdata obtained by plating diluted cultures ontoplates containing Tween 80 gave a value of (35.0+ 2.6) x 106 cells per ml per A6.0-Effect of fatty acid concentration on the

growth of KD46. Growth curves obtained inthe presence of various concentrations of pal-mitoleic, arachidonic, oleic, and linoleic acidsare presented in Fig. 1. Results obtained whenno exogenous fatty acid was added or whenpalmitate was added are also included in Fig. 1for comparison. All of the curves can be charac-terized by the lag, log, and stationary growthphases outlined by Monod (25). The control

cultures which were unsupplemented, or sup-plemented with palmitate, exhibited somegrowth but generally failed to exceed one dou-bling the initial cell number. Jollow et al. (11)and Proudlock et al. (31) report similar growthof S. cerevisiae on unsupplemented complexgrowth medium and were able to trace thesource of this growth to small amounts ofunsaturated fatty acids present in yeast extract.The cells in our control experiments may wellhave been utilizing such contaminants.

It can be seen in Fig. 1 that the final cellnumber attained by a given culture dependedupon the initial concentration of the un-saturated fatty acid supplement. It can also beseen that the duration of the initial lag phasewas extended at increasing fatty acid concen-trations. This phenomenon seemed particularlystriking for palmitoleate (Fig. la).The rates of cell growth also varied with the

initial concentration of fatty acid supplement.For example, in Fig. lb the doubling times ofthe cultures growing on arachidonate are 4.5 hat 2.6 AM, 3.0 h at 5.2AM, and 2.0 h at 13.1 or26.1 AM. Similar dependencies were observedfor the other acids and, in addition, some acidswere found to slow the rate of growth at highconcentrations. Such a situation can be seen inFig. ld where the doubling times of culturesgrowing on oleate are 3.0 h at 5.9 AM 18: 2, 2.8 hat 11.7 ,uM, 2.5 h at 17.5 ,M, and then 3.3 h at58.5 gM. The dependencies of the cell yieldsand growth rates on fatty acid concentrationwill be presented in some detail to show thereproducible and quantitative nature of thephenomenon for added fatty acids.

Figure 2a shows the final cell yield plottedas a function of fatty acid concentration foroleate and arachidonate and for oleate in thesame medium containing added 1% Tergitol.The cell yields were calculated as the differencein final and initial cell numbers from growthcurves such as those shown in Fig. 1. The sameprocedure was applied to the control growthcurves obtained in the absence of any fatty acidsupplementation. This latter value was sub-tracted from those observed with added fattyacids. The data in Fig. 2, then, representchanges in cell density over and above that seenin the control experiments, and negative values(such as those for palmitate) are a result ofcultures which show less increase than thecontrols. This procedure corrects for growtharising from traces of oleate carried over fromthe inoculum or from materials other than theadded fatty acid that is under study. The datain Fig. 2a are composites of six experiments foroleate and five for arachidonate. Thus, thevariability in the data may reflect small varia-

VOL. 115, 1973 545

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

30

E0181_,1A8

4200

----~ A- NONE

o .A 2.5pM 96'0-./1*

5

10 20 30 40 50 10 20 30

TIME, hours

200 d

18:1 ~~~~~~~18:2£18:1 ~A,'~ £5. -.------..

m00 (t / 11rta

.17.4 I

(a)-/11.8 0n a 5.9

30

E~~~~~~~~~~~~20 - NONE

E~~~~~~~~

w 101

-~

I0 20 30 40 10 20 30 40

TIME hoursFIG. 1. Growth curves of KD46 on four different fatty acid supplements. The log of the cell number

(determined turbidimetrically) is plotted against the time after inoculation. All cultures were grown in YEPDmedium at 30 -±- 2 C. The number to the right of each curve is the micromolar concentration of the fatty acidsupplement. (a), Palmitoleate, cis A9 hexadecenoate; (b), Arachidonate, cis A5, 8, 11, 14 eicosatetraenoate; (c),Oleate, cis A.9 octadecenoate; (d), Linoleate, cis A9, 12 octadecadienoate. Control cultures containing palmitate(a) or no added fatty acid (b, c) are also shown.

546

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FArTY ACIDS REQUIRED FOR S. CEREVISIAE GROWTH

tions in growth medium from day to day.The response of cell yield to fatty acid con-

centration was found to be reasonably linear atlow fatty acid concentrations, whereas the re-

sponse at higher concentrations of fatty acidwas somewhat variable. In some cases, the finalcell yield became independent of fatty acidconcentration (e.g., oleate, arachidonate); inother cases a decrease in the yield was observedas fatty acid concentration was raised (e.g.,palmitoleate, docosahexaeneoate). In the for-mer cases, we presume that some other compo-nent of the growth medium such as glucose or

amino acids become growth limiting at a celldensity of about 150 million cells per ml. In fact,we have calculated by using the constant pre-sented by Beauchop and Elsden (1) that 2.3 mg(dry weight) per ml is the expected cell yield forfermentative growth on 110 mM glucose (2%wt/vol). This value is similar to the yieldobtained by Clark-Walker and Linnane (2).Measurements taken from electron micrographs(2) reveal that petite mutant cells are about one-third the volume of the wild-type cells whichhave a dry weight of 37 x 10-12 g per cell (13).Thus, the yields of petite cell could be expectedto be limited to 186 million cells per ml (2.3mg/ml 12.3 x 10- 9 mg/cell) under ourgrowth conditions. Table 1 shows that the maxi-mum cell yields obtained with high levels of nu-trient acid were in fair agreement with this cal-culation. However, for three of the acids we

tested, namely 20:2, 20:3, and 22:1, the cellyield became independent of the fatty acidconcentration at 30 to 40 million cells per ml.The explanation of this phenomenon is notreadily apparent since we know that the me-

dium was usually capable of supporting moreextensive growth.The data in Fig. 2 show that for the low

ranges of fatty acid concentration there is alinear dependence of cell yield on fatty acidconcentration. Such linear dependencies havebeen observed for a variety of carbon sources(31) and growth factors (10) and represent thebasis for one type of bioassay (37, 38). Thedimensions of the slope of such a line are cells(produced) per mass of nutrient added and thisslope, which we have called epsilon (e), there-fore, can be thought of as an efficiency factor.The values of e for various unsaturated fattyacids are given in Table 1.The growth rate can be characterized by a

doubling time or generation time (T) which isrelated to the specific growth rate (k) by theequation: T = ln2/k = 0.69/k (29).The relationship between the doubling time

(T) and the reciprocal of the initial substrate

a

> e1_,.

4 =Z E

20 40 60

[PATTY ACID].,M80 100

FIG. 2. The difference in cell yield between experi-mental and control cultures is plotted as a function ofthe initial concentration of fatty acid supplement.The cell yields were determined from growth curvessuch as those shown in Fig. 2 by subtracting theoriginal cell density from the maximum observed celldensity. In each case the cell yield observed in controlcultures (no added fatty acid) has been subtracted.(a), Yield curves for oleate (0, 0) and arachidonate(A, A) in YEPD medium. Closed symbols are turbidi-metric data and open symbols are from direct hema-cytometer counts. The dashed lines extend the initiallinear phase used to assign the efficiency of thenutrient acid. The curve (x) represents yields witholeate in YEPD medium containing, 1% (wt/vol)Tergitol NP-40. (b), Yield curves for docasahexenoate(0, 0) and palmitoleate (A, A) and palmitate (x).Filled and open symbols have the same origin as inpart (a).

concentration is shown in Fig. 3 for palmitole-ate, oleate, arachidonate, and docosahexaenoicacids. The linearity of these plots suggest thatthe rate data are related to fatty acid concentra-tion by a hyperbolic equation. From these plots,the concentration of fatty acid giving one half ofthe maximum growth rate (ks) can be obtainedfor each fatty acid and the maximum growthrate can be estimated from the ordinate inter-cept. The kinetic constants K8 and Tmin aretabulated in Table 1 for each acid.

Also notable in Fig. 1, 3c, and 3d is a distinct

[FATTY ACIDJ,pM150o b

/

II0 ,

100 *, *

fa50£

0 5 0

0FoiV I5 XX - I i ;

VOL. 115, 1973 547

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BARBER AND LANDS

x

T, hr. 0

0A] 0.2 0.3 0.41 P -1

[20-4]

6-

T, hr. j--i ~I I

0.1 0.2

Iw ' Jm -'

16-

12

T,hr 0

4 1>

0.5 1.0

[22:6]

4 K

T,hr

2_0

__ ___ _-

0.1 0.2

[6:]

FIG. 3. The doubling times (T) in hours (per gener-

ation) of various cultures are shown plotted againstthe reciprocal of the concentration of the fatty acid

increase in doubling time (inhibition of thegrowth rate) by high concentrations of fattyacid. When the data are replotted as T versus

(S), by analogy with the procedure used forexcess substrate inhibition of enzymatic veloci-ties (Ref. 3, p. 78), a substrate inhibition con-

stant (K8') can be determined for each acid.Examples of such plots are shown in Fig. 4 andthe resulting K8' values determined in this wayfor all acids tested are tabulated in Table 1.

100 200 300 400 500

[20:4],"M

FIG. 4. Determination of the excess (nutrient) sub-strate inhibition constants (K.') on the growth rate ofKD46. The doubling times are plotted against theconcentrations of oleate (a) and arachidonate (b). K8'can be determined from the abscissa intercepts.

supplement for arachidonate (a), oleate (b), docosa-hexaenoate (c), and palmitoleate (d). Solid circlesrepresent data obtained in YEPD medium, whereasthe data denoted by x were obtained in YEPDmedium containing 1% (wt/vol) Tergitol NP-40. Thesubstrate activation constant, K8, can be determinedfrom each abcissa intercept and the minimum dou-bling time (Tmin) can be determined from the ordinateintercepts.

548

b 10 L I I I I I

.

6

T, hr.

4 0

J. BACTERIOL.

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FATTY ACIDS REQUIRED FOR S. CEREVISIAE GROWTH

DISCUSSION

Included in Fig. 2 and 3 are some data ob-tained in the presence of 1% Tergitol NP-40, a

non-ionic detergent. This detergent was in-cluded in some early experiments as a solubi-lizing agent for added fatty acids. In each case

the growth response of the mutant is different inthe presence of Tergitol than that observed inits absence. In general, we feel that thesedifferences in response can be explained by theformation of hydrophobic complexes (mixedmicelles) which render a variable portion of theadded fatty acid unavailable to the organism. Insupport of this hypothesis is the observationthat although higher levels of added fatty acidwere needed to attain a given growth rate in thepresence of detergent, the Tmin values obtainedwere independent of the presence of detergent(Fig. 3a, b).Measurements of microbial growth have been

used for many years in the assay of vitaminsand other nutrients (37, 38) and a wide varietyof essential metabolites have been recognizedand purified using such assays. For example,McIlwain (23) used streptococcal growth tostudy the metabolic interrelationship betweenpantoyltaurine and pantothenate. More re-

cently Johnston et al. (10) used the growth of S.carlsbergensis to demonstrate the biologicalactivity of several oxidation products of myo-

inositol. Proudlock et al. (31) and Korman-cikova et al. (16) and many others have usedtotal growth to measure the efficiency of utiliza-tion of various energy sources by yeast cells.

In our studies, we have used measurements ofthe total growth of KD46 to calculate a relativeeffectiveness for each fatty acid tested. Appre-ciable amounts of carbohydrate and amino acidare present in the growth medium and anyneeded saturated fatty acids are presumablyformed from these sources. Since KD46 does nothave the mitochondrial oxidative pathway tooxidize the acids, the limiting amounts of thenutritionally required unsaturated acids thatwere added seem most likely to be incorporatedinto elements needed in cell function ratherthan representing in themselves a limit to theavailable energy. If so, the reciprocal of theefficiency term (1/e) represents, in part, a mini-mal requirement for forming cellular compo-nents. This concept was also presented byMonod (25) as the "amount of limiting nutrientused up in the formation of a standard cell."Values for each fatty acid (Table 1) rangebetween 100 amol (attomoles) per cell for 22:6and 600 amol per cell for 22:1, the most andleast effective fatty acids tested, respectively.

It is of interest to compare the amounts of

acids necessary for growth with the reportedcontent of lipids in S. cerevisiae. Using thereport that this organism contains about 3%phospholipid on a dry weight basis (11) and thatthe dry weight per cell may be 37 x 10-12 g (13),and assuming a molecular weight of 750 for anaverage phospholipid molecule, it can be calcu-lated that each cell may contain about 1.4 x10- 15 mol of phospholipid. In other words, about2,800 amol of total fatty acids might be requiredto form the phospholipids present in one wild-type cell.

Experimental results with wild-type cellsgrown in rich media show that palmitoleate andoleate normally represent 50 and 25%, respec-tively, of the yeast fatty acids (9, 20, 39).However, the values for 16: 1 and 18: 1 obtainedin this study indicate that the minimal amountsof these unsaturated acids needed is much lessthan that accumulated in cells with an abun-dant supply. In this way, studies with restrictednutrient supplies provide a clearer index of thedegree of essentiality and efficiency of certainacids for cell growth and function. In thisregard, Proudlock et al. (31) reported thatmitochondrial function was not appreciablyaltered by a shift of unsaturated fatty acidcontent from 70 to 20%.

It should be noted that the substrate satura-tion constants (K8) in Table 1 for the variousfatty acids are relatively similar. With theexceptions of 18: 1, 20: 2, and 22: 1, the value isabout 5 MM for all the acids. Thus, the rate-limiting effects were less selective than theextent-limiting properties. The values for in-hibitory effects (K8') were more variable thanthe K8 constants and, again, the value for oleatewas very atypical.

Stimulation of the rate of growth at lowessential nutrient concentrations can, as a firstapproximation, be fitted by the equation pre-sented by Monod (25, 42): v = VJ/(l + KJS).Recently, Shehata and Marr (34) have shownthat the response of the growth rate of E. coli toglucose, inorganic phosphate, and amino acidsshows a biphasic activation which cannot befitted by the simple Monod equation, but re-quires a more complex description involving asummation of individual terms. Our rate datacould be explained by such a summation inwhich one or more of the substrate terms is ofopposite sign. Although the inhibitory effectcould be due to a detergent action of the addedsubstrate, we find it hard to explain the greatvariances in inhibition by such structurallysimilar acids.The minimum T value (maximum growth

rate) was reasonably constant throughout theseries except for 20:2, 20:3, and 22: 1 which

VOL. 115, 1973 549

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

BARBER AND LANDS

gave higher T values. This observation can beinterpreted to mean that some reaction stepinvolved in the conversion of these acids fromfree exogenous acids to lipid products still limitsthe rate of growth of the organism even whenfatty acid is fully saturating.

It is not clear whether the magnitude of theefficiency term, E, is due to the suitability of theacid as a precursor of cellular materials or to thesuitability of the resultant products containingthat acid for growth and membrane function.Since this measure of effectiveness (e) wasdetermined by total growth and was independ-ent of the rate of growth, the suitability of anacid as a substrate could be important only ifsome reaction step showed a very high discrimi-nation against that acid. Conversely, if all acidstested have a similar rate of incorporation(exogenous acid - - lipid products), then thestructural or functional nature of the lipidproducts must be the principal determinant ofthe value of e. In the case of short- and medium-chain saturated acids, the esterified productsmight be expected to resemble those of typicalmonounsaturated acids (41). The fact that nosaturated acid tested could fulfill the require-ment for growth suggests that exogenous sup-plies of those acids are not suitable substrates.We expect that the different efficiencies ob-served will in some cases reflect both kineticfeatures of utilization and the subsequent fea-tures of the products formed.

Several groups have shown that the proper-ties of the lipid products correlate with observedcellular properties. Fox and co-workers (6, 44,45), using mutants of E. coli K-12 that requireunsaturated fatty acids, have shown that thethermal transition temperature for glycosidetransport is dependent upon the fatty acidincorporated into the lipid during induction.Using a different mutant of E. coli, Overath andco-workers (27, 28, 33) have correlated transi-tion temperatures of lipid monolayers with anumber of physiological transitions includingthose for growth and respiration. Silbert andVagelos (35) and Mavis and Vagelos (22) havedescribed fatty acid incorporation and the ther-mal behavior of three membrane-bound en-zymes, also in E. coli. Proudlock et al. (31) haveshown that the content of unsaturated fattyacids affects the efficiency of oxidative phos-phorylation in yeast mitochondria. Thus, thereis a growing body of evidence which implicatesthe properties of the acids of membrane lipidsas playing an important role as determinants ofmembrane function. Such functions may deter-mine the viability of KD46 with various fattyacids and thereby cause the quantitatively

different efficiency noted for each different acid.The effectiveness data in Table 1 show that

KD46 will grow on a variety of unsaturated fattyacids. No particular chain length requirement isevident nor can any general statement be madeas to a positional requirement for the doublebond(s). Acids containing 16 to 22 carbon atomsand acids containing double bonds located invarious positions from the A4 position throughA19 all support growth. Thus, no single locationof the unsaturated group is apparent as arequired feature of the nutrient acid in thesestudies, although it has been suggested else-where (46), and further modified (47). Theeffectiveness of these acids, however, does varyand inefficient acids can be distinguished fromthose of moderate or high efficiency. In general,there was a direct correlation between thenumber of double bonds present and the effec-tiveness of the acid. Esfahani et al. (4), Silbertet al. (36), and Overath et al. (27), showed thatthe mole percent of unsaturated fatty acidsaccumulated in lipids of E. coli mutants wasinversely proportional to the number of doublebonds in the acid. Our data suggest that asimilar phenomenon may be operating in yeastand that it can be regulating the yield of cells ina culture.

Interestingly, the unsaturated fatty acids nor-mally found in yeast, palmitoleate and oleate(20, 39), were among the least effective of thefatty acids tested. On the other hand, thepolyenoic fatty acids (20:4, 20:5, and 22:6),which had the highest observed efficiencies, arenot normal constituents of yeast cell lipids (20).In fact wild-type yeasts should be expected tohave little exposure to such acids, given theirnormal ecological niche (21). It is thus of someinterest that yeasts can so effectively utilize thepolyunsaturated acids that are found moreabundantly in cells of higher eukaryotic orga-nisms.

ACKNOWLEDGMENTSThis investigation was supported in part by Public Health

Service grant AM-05310 from the National Institute ofArthritis and Metabolic Diseases.

E. Barber was the recipient of a fellowship (no.2-02-NS48504-02) from the National Institute of NeurologicalDiseases.

LITERATURE CITED1. Beauchop, T., and S. R. Elsden. 1960. The growth of

micro-organisms in relation to their energy supply. J.Gen. Microbiol. 23:457-469.

2. Clark-Walker, G. D., and A. W. Linnane. 1967. Thebiogenesis of mitochondria in Saccharomyces cerevi-siae. J. Cell Biol. 34:1-14.

3. Dixon, M., and E. C. Webb. 1964. Enzymes, (2nd ed.).Academic Press Inc., New York.

4. Esfahani, M., E. M. Barnes, and S. J. Wakil. 1969.

550 J. BACTERIOL.

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from

FATTY ACIDS REQUIRED FOR S. CEREVISIAE GROWTH

Control of fatty acid composition in phospholipids ofEscherichia coli: response to fatty acid supplements ina fatty acid auxotroph. Proc. Nat. Acad. Sci. U.S.A.64:1057-1064.

5. Esfahani, M., T. Ioneda, and S. J. Wakil. 1971. Studieson the control of fatty acid metabolism. J. Biol. Chem.246:50-56.

6. Fox, C. F. 1969. A lipid requirement for induction oflactose transport in Escherichia coli. Proc. Nat. Acad.Sci. U.S.A. 63:850-855.

7. Gelvarg, C., and Y. Levin. 1972. Response of Escherichiacoli to ornithyl peptides. J. Biol. Chem. 247:543-549.

8. Hill, E. E., and W. E. M. Lands. 1970. Phospholipidmetabolism, p. 185-277. In S. Wakil (ed.), Lipid me-tabolism. Academic Press Inc., New York.

9. Hunter, K., and A. H. Rose. 1972. Lipid composition ofSaccharomyces cerevisiae as influence by growth tem-perature. Biochim. Biophys. Acta 260:639-653.

10. Johnston, J. A., R. C. Ghadially, R. N. Roberts, and B.W. Fuhr. 1962. Myo-inositol activity of oxidationproducts derived from myo-inositol. Arch. Biochem.Biophys. 99:537-538.

11. Jollow, D., G. M. Kellerman, and A. W. Linnane. 1968.The biogenesis of mitochondria. J. Cell Biol.37:221-230.

12. Kaneshiro, T., and A. G. Marr. 1962. Phospholipids ofAzotobacter agilis, Agrobacterium tumefaciens, andEscherichia coli. J. Lipid Res. 3:184-189.

13. Katchman, B. J., and W. 0. Fetty. 1954. Phosphorusmetabolism in growing cultures of Saccharomycescerevisiae. J. Bacteriol. 69:607-615.

14. Kates, M. 1964. Bacterial lipids. Adv. Lipid Res. 2:17-90.15. Keith, A. D., M. R. Resnick, and A. B. Haley. 1969. Fatty

acid desaturase mutants of Saccharomyces cerevisiae.J. Bacteriol. 98:415-420.

16. Kormancikova, V., L. Kovac, and M. Vidova. 1969.Oxidative phosphorylation yeast. Biochim. Biophys.Acta 180:9-17.

17. Korn, E. D. 1969. Current concepts of membrane struc-ture and function. Fed. Proc. 28:6-11.

18. Lands, W. E. M., M. L. Blank, L. J. Nutter, and 0. S.Privett. 1966. A comparison of acyltransferase activi-ties in vitro with the distribution of fatty acids inlecithins and triglycerides in vivo. Lipids 1:224-229.

19. Lands, W. E. M., and P. Hart. 1966. The control of fattyacid composition in glycerolipids. J. Amer. Oil Chem.Soc. 43:290-295.

20. Longley, R. P., A. H. Rose, and B. A. Knights. 1968.Composition of the protoplast membrane from Sac-charomyces cerevisiae. Biochem. J. 108:401-412.

21. Lund, A. 1958. Ecology of yeasts. In A. H. Cook (ed.),The chemistry and biology of yeasts. Academic PressInc., New York.

22. Mavis, R. D., and P. R. Vagelos. 1972. The effect ofphospholipid fatty acid composition membranous en-zymes in Escherichia coli. J. Biol. Chem. 247:652-659.

23. McIlwain, H. 1944. Biochemical characterization of theactions of chemotherapeutic agents. Biochem. J.38:187-195.

24. Metcalfe, L. D., and A. A. Schmitz. 1961. The rapidpreparation of fatty acid esters for gas chromatographicanalysis. Anal. Chem. 33:363-364.

25. Monod, J. 1949. The growth of bacterial culture. Ann.Rev. Microbiol. 3:371-394.

26. Okuyama, H., W. E. M. Lands, F. D. Gunstone, and J. A.Barve. 1972. Selective transfers of trans-ethylenic acidsby acyl coenzyme A. Phospholipid acyltransferases.Biochem. 11:4392-4398.

27. Overath, P., H. V. Schairer, and W. Stoffel. 1970b.

Correlation of in vivo and in vitro phase transitions ofmembrane lipids in Escherichia coli. Proc. Nat. Acad.Sci. U.S.A. 67:606-612.

28. Overath, P., F. F. Hill, and I. Lamnek-Hirsch. 1971.Biogenesis of E. coli membrane: evidence for randomi-zation of lipid phase. Nature N. Biol. 234:264-267.

29. Painter, P. R., and A. G. Marr. 1968. Mathematics ofmicrobial populations. Ann. Rev. Microbiol.22:519-548.

30. Pfleger, R. C. P., N. G. Anderson, and F. Synder. 1968.Lipid class and fatty acid composition of rat liverplasma membranes isolated by zonal configuration.Biochem. 7:2826-2833.

31. Proudlock, J. W., J. M. Haslam, and A. W. Linnane.1971. Biogenesis of mitochondria 19. Bioenergetics2:327-349.

32. Resnick, M. A., and R. K. Mortimer. 1966. Unsaturatedfatty acid mutants of Saccharomyces cerevisiae. J.Bacteriol. 92:597-600.

33. Schairer, H. V., and P. Overath. 1969. Lipids containingtrans-unsaturated fatty acids change the temperaturecharacteristic of thiomethylgalactoside accumulationin Escherichia coli. J. Mol. Biol. 44:209-213.

34. Shehata, T., and A. G. Marr. 1971. Effect of nutrientconcentration on the growth of Escherichia coli. J.Bacteriol. 107:210-216.

35. Silbert, D. F., and P. R. Vagelos. 1967. Fatty acidmutant of E. coli lacking a ,B-hydroxydecanoyl thioesterdehydrase. Proc. Nat. Acad. Sci. U.S.A. 58:1579-1586.

36. Silbert, D. F., F. Ruch, and P. R. Vagelos. 1968. Fattyacid replacements in a fatty acid auxotroph of Esche-richia coli. J. Bacteriol. 95:1658-1665.

37. Snell, E. E. 1950. Microbiological methods in vitamin re-search. In P. Gyorgy (ed.), Vitamin methods, vol. 1.Academic Press Inc., New York.

38. Stanier, R. Y., M. Doudoroff, and E. A. Adelberg. 1963.The microbial world, p. 707-780, (2nd ed.). Prentiss-Hall, Inc., New Jersey.

39. Suomalaninen, H., and A. J. A. Keranen. 1968. The fattyacid composition of baker's and brewer's yeast. Chem.Phys. Lipids 2:296-315.

40. Tamai, Y., W. E. M. Lands, J. A. Barve, and F. D.Gunstone. 1973. Selective transfers of acetylenic acidsto form lecithins. Biochim. Biophys. Acta 296:563-

41. Van Deenen, L. L. M. 1966. Some structural and dynamicaspects of lipids in biological membranes. Ann. N. Y.Acad. Sci. 137:717-730.

42. Van Uden, N. 1971. Kinetics and energetics of yeastgrowth. In A. H. Rose and J. S. Harrison, (ed.). TheYeasts, vol. 2. Academic Press Inc., New York.

43. Waku, K., and W. E. M. Lands. 1968. Control of lecithinbiosynthesis in erythrocyte membranes. J. Lipid. Res.9:12-18.

44. Wilson, G., S. P. Rose, and C. F. Fox. 1970. The effect ofmembrane lipid unsaturation on glycoside transport.Biochem. Biophys. Res. Comm. 38:617-623.

45. Wilson, G., and C. F. Fox. 1971. Biogenesis of MicrobialTransport Systems: evidence for coupled incorporationof newly synthesized lipids and proteins into mem-brane. J. Mol. Biol. 55:49-60.

46. Wisnieski, B. J., A. D. Keith, and M. A. Resnick. 1970.Double-bond requirement in a fatty acid desaturasemutant of Saccharomyces cerevisiae. J. Bacteriol.101:160-165.

47. Wisnieski, B. J., and R. K. Kiyomoto. 1972. Fatty aciddesaturase mutants of yeast: growth requirements andelectron spin resonance spin-label distribution. J. Bac-teriol. 109:186-195.

551VOL. 115, 1973

on June 26, 2020 by guesthttp://jb.asm

.org/D

ownloaded from