Eur. J. Biochem. 58. 177-184 (1975)
Control of Fatty-Acid-Synthetase Biosynthesis in Succhuromyces cerevisiue Grrhard DIETLEIN and Eckhart SCHWEIZER Institut fiir Biochemic der Universitat Wiir7burg (Received March 3iJune 30, 1975)
143 out of 308 fusl mutants (47%) and 139 out of the 443 fus2-mutants (32%) genetically studied in this laboratory fail to complement with any other fus-mutant (deficient in fatty acid synthetase) of the same gene locus. From these noncomplementing fus-mutants no mutant fatty acid synthetase can be isolated using the wild-type enzyme purification procedure. Furthermorc the noncomplementing fhs-mutants generally contain no material immunologically crossreacting with a specific fatty acid synthetase antiserum. However, subunits obtained after dissociation of the complex with sodium dodecylsulfate still cross react with this antiserum. Therefore, it is concluded that noncomplementing jbs-mutants contain no fatty acid synthetase component proteins, though one of the two jius-loci is mutationally unaffected. This conclusion was further confirmed by 14Clabeled amino acid incorporation studies which indicated that in noncomplementing jus-mutants, other than in wild type and complementing fhs-mutant cells, no label was incorporated into fatty acid synthetase subunits or precursor proteins. At nonpermissive temperature, the samc biochemical and immunological characteristics were observed with temperature-sensitive noncomplementingfus-mutants. These results suggest that noncornplementing ,fus-mutants either represent regulatory mutants unable to induce the mutationally unaffected other jibs-gene locus or that they are associationdefective mutants. In both cases the resulting individual subunits of the complex may be rapidly degraded by intracellular proteases. Available evidence indicates that in Succhuromyces cerevisiue the fatty acid synthetase multienzyme complex is composed of only two different protein subunits, each present in probably six copies per complex molecule [l 61. Consequently, both subunits represent multifunctional polypeptides chains accommodating several different fatty acid synthetase component enzyme activities, each [2]. This protein structure of the multienzyme complex ensures the stoichiometric biosynthesis of the component enzymes encoded by the same,fus-locus. On the other hand it is unknown, so far, whether the expression of the two genetically unlinked fatty acid synthetase loci jasl and ,fad is under coordinate control, too [3,4]. Of both~fus-locinoncomplementing mutants have been isolated. Characteristically, the few noncompleinenting mutants so far studied did not contain any fatty acid synthetase complex comparable to the wildtype enzyme [3]. Since many of these mutants represent chain-terminating nonsense mutations synthesizing only a fragmentary,fusl orfus2 gene product, these biochemical characteristics were expected [6]. However, besides these nonsense mutations, a sizable
portion of the noncomplementing fus-mutants has been identified as missense mutations [6]. Thus, it remained to be shown whether the failure of fatty acid synthetase production was a general characteristic of noncomplementingfus-mutants no matter whether they were nonsense or missense mutations. In addition, noncomplementingfas nonsense mutants may be used to answer the question whether the second, mutationally unaffected fus-locus is not expressed in these mutants either or whether the fatty acid synthetase subunits encoded in this locus are still synthesized. Therefore, these mutants are considered as especially useful objects for the study of control mechanisms involved in fatty acid synthetase biosynthesis. Experiments pertaining to this question will be reported in this study. MATERIALS AND METHODS Yeast Struins
The origin and characteristics of the Succhuromyces cerevisiae wild type and ,jus-mutant strains
Fatty-Acid-Synthetase Biosynthesis
178
have been described in previous publications IS,71. Mutants fasl-I55 (Ill), fasl-313 (III), fusl-42 (IIJ), 3I - ts (111), 60-ts (IX), 32-ts (IX),fas2-285 (IV), fas2-3Y4 (IV), and fas2-242 (IV) are noncomplementing fasmutants of the pleiotropic complementation groups 111, IV and IX, respectively [5].Mutantsfasl-201 (Va), Jus2-105 (VI), fus2-46 (VI), 38-ts (Vb) and 25-ts (VI) are complementing fas-mutants with their complementation groups indicated by roman numbers in brackets. Yeast Growth and Media
The fatty-acid-containing and the fatty-acid-free yeast extract-peptone-sucrose medium (called fatty acid medium and fatty-acid-free medium) have been described earlier [3,7]. For 14C-labeling of proteins, a synthetic medium was used containing 2 % sucrose, 0.7 7; yeast nitrogen base w/o amino acids (Difco), 1 % Tween 40 (Serva), 0.03 % myristic acid and 20 mg per liter of each of the following amino acids: alanine, arginine, aspartic acid, glutamic acid, glycine, leucine, isoleucine, lysine, phenylalanine, proline, serine, threonine, tyrosine, valine, and, per liter, 16 pCi I4C-labeled amino acid mixture (Amersham, 57 mCi/ matom carbon). Temperature-sensitive fas-mutants were grown each strain in 3 1 of 14C-labeled synthetic medium at 36°C for 2 days. Then, the cells were harvested by centrifugation at 22 "C and washed twice with 200 ml of 22°C sterile distilled water, each time, and, subsequently, they were resuspended in 16 1 of unlabeled fatty-acid-free medium at 22 "C. In this medium, growth was continued for about 2- 3 generations. Growth was followed turbidimetrically at 578 nm. Nonconditionalfus-mutants were labeled by growing each strain in 2 1 of 14C-labeled synthetic medium at 30 "C. In parallel cultures, haploid ,fas-mutants of opposite mating type and genetically defective in the other ,fus locus were grown in 2 1 of unlabeled fatty acid medium, each. At early stationary phase, labeled as well as unlabeled cells were harvested by centrifugation, washed twice with 200 ml of sterile distilled water, each time. Then, labeled and unlabeled strains of opposite mating type were thoroughly mixed in 200 ml of fatty-acid-free medium, spun down again and allowed to mate overnight at 30°C in the tightly packed wet pellet. Subsequently, the cells were suspended in 16 I of fatty-acid-free medium and the diploids formed were allowed to grow for another 2 - 3 generations. Growth was again followed turbidimetrically at 578 nm. Wild-type yeast cells were labeled accordingly. The labeled haploid wild-type strain was mated with another wild-type yeast strain of opposite mating type. Routinely, temperature-sensitive fas-mutant cultures were examined for revertants after their final
harvesting, whereas nonconditional fas-mutants were examined before the mating reaction. The fatty acid synthetase complex was isolated only from revertantfree cultures. The enzyme purification procedure followed the isolation scheme described earlier [I]. Fatty Acid Synthetase Antiserum Preparation and Immunod$fusion Assays Antibodies against the yeast fatty acid synthetase complex were raised in rabbits by intramuscular injection of 1.7 mg of the wild-type enzyme purified by sucrose density gradient centrifugation, together with incomplete Freund's adjuvant (Difco). Booster injections were performed, without adjuvant, intravenously after 4 and 6 weeks using 1.5-2.5 mg of antigen, each time. 10days after the last injection the rabbit was bled. After clotting, the blood serum was collected by centrifugation and kept at - 15°C. For immune double diffusion analysis, the antiserum was usually diluted five-fold with 0.85 % saline pH 6.8. Between 15 and 20 p1 of this solution were placed into the central antiserum well. Immunoprecipitates were allowed to develop during a 24-36 h incubation period at 4 "C.
RESULTS Protein Chemical Characterization qf Complementing and of Noncomplementing fas-Mutants When fatty- acid -synt hetase -deficien t Saccharornyces cerevisiae mutants are subject to genetic complementation analysis, 11 major complementation patterns are observed [5]. The relative distribution of a representative sample of fasl and ,fbs2 mutants among these complementation groups is listed in Table 1. Groups 111 and IV are genetically pleiotropic and fail to complement with all other fus-1 andfus2 mutants, respectively [5]. Group IX mutants are genetically pleiotropic, too, but other than the group IV mutants, they complement with mutants of complementation group VII [ 5 ] .As it is evident from Table 1, 143 of the 308 fasl mutants studied (47%) and 32% (139 out of 443) of the $as2 mutants studied are strictly noncomplementing within the same gene locus. Similar results were obtained with temperaturesensitive fatty acid synthetase mutants (E. Schweizer, unpublished results). This shows again that not only nonsense but also missense mutants may exhibit the noncomplementing phenotype. Earlier results reported by Kiihn et al. [3] suggested that from noncomplementing &-mutants, other than from complementing ones, no mutant fatty acid synthetase complex may be isolated. This has now been confirmed for a sample of 46 randomly chosen group 111 and group IV mutants. Of each
G. Dietlein and E. Schweizer
179 A
I
0.2
F 0
% c
m
s m
e
2 0.1
4
10
m
10 Fraction number
I 20 0
10
20
Fig. 1. Sucrose density gradient profiles c?f fatty acid synthetase preparations purified from complementing and noncomplementing fas-mutants. About 200 g of each, the noncomplementing temperature-sensitive mutant 31-rs (A), grown at permissive (0) as well as at nonpermissive (0) temperature, the complementing mutant fas2-46 (B) and of the noncomplementing mutants f a d - 4 2 (C) and f a d - 2 8 5 (D) were used for enzyme purification. Experimental details of enzyme isolation and sucrose density gradient centrifugation were as described previously [l]
Table 1. Relative frequencies qf the various fasl and fas2 complementation group mutations
Gene locus
Complementation group
Complementation characteristics
Number of mutants isolated
/as I
I1 Ill Va Vb vc
complementing noncomplementing .complementing complementing complementing
92 143 31 30 12
fas2
1v VI VII VIII IX
noncomplementing complementing complementing complementing partially noncomplementing
139 185 40 55
24
mutant the cell homogenate was subject to the fatty acid synthetase isolation procedure adopted for the wild-type enzyme. Whereas the enzyme prepared from wild-type yeast or from a complementing fusmutant strain appears as a distinct protein peak at the final sucrose density gradient centrifugation step (Fig. 1 A and B), with all noncomplementing fusmutants studied, no protein is observed at the corresponding position in the gradient. This is indicated in Fig. 1C and D for two representative noncomplementing fasl and fus2 mutants. As it had been shown by Tauro et ul. [6] one of them, i.e. f a d - 4 2 , is an amber-type nonsense mutation. On the other hand, all of the more than 60 complementing fusmutants so far studied contained the mutant enzyme
at this position. In Fig. 1 A sucrose density gradient profiles of two fatty acid synthetase preparations are shown which were isolated from the noncomplementing temperature-sensitive fus-mutant 31-ts (111) grown at permissive and at nonpermissive temperature, respectively. Obviously, also in this noncomplementing missense mutant the amount of fatty acid synthetase synthesized at nonpermissive temperature is negligible compaired to the permissive conditions. These results indicate, that, quite generally, only the complementing fus-mutants synthesize a fatty acid synthetase complex of a structure like that of the wild type enzyme whereas in noncomplementing fas-mutants, irrespective whether they are missense or nonsense mutations, this complex cannot be demonstrated. Thus, these mutants either contain no fatty acid synthetase protein at all, or the protein structure of the multienzyme complex is grossly altered compared to the wild-type enzyme, thereby causing it to escape the standard fatty acid synthetase isolation procedure. Immunological Characterization of Complementing and of Noncomplementing fas-Mutants
To study the question whether noncomplementing fas-mutants, though inactive in producing the complete multienzyme complex, still synthesize the individual fatty acid synthetase subunits, extracts of a large number of noncomplementing fus-mutants were studied, by immune double diffusion analysis, for the presence of fatty acid synthetase-like cross-reacting material. Among 110 of them tested 85 mutants proved to be deficient of any cross-reacting material whereas 25 cell extracts developed, though mostly
Fatty-Acid-Synthetase Biosynthesis
Fig. 2. I , ) i i ~ i u , i o ~ / ~ ~ i r unalysis sio,l of' coinpkiiienting and of iioncornplemcnting ,/us-murants. Each strain was grown in 500 in1 of Fatty acid incdium at 30'C. At early stationary phase cells wcre harvested and broken with glass beads. The lioinogenatc was centrifuged at 12000 x g for 1 h at 4 ' C and 20 p1 (about 0.3 mg) of the supernatant were used for the immunological tests. The antiserum well contained 20 p1 of a 1 : 5 diluted fatty acid synthetase antiserum solution. After 18 - 36 h incubation at 4 'C thc immunoprecipitation lines formed were recorded. In (d) purified wild-type (WT) fatty acid synthetasc (about 5 mg/ml) was denatured with 204 sodium dodecylsulfate (SDS) and lox inercaptocthanol for 1 h at 22'C. 20 pl of this solution were placed into the antigen well and, as a control, another well contained the same amount of the untreated wild-type enzyme. Roman numbers indicate cornplemcntation groups, arabic numbers fus-mutant alleles
very weak, immunoprecipitation lines. On the other hand, all complementing fusl or .fusZ mutants tcsted exhibited distinct immunoprecipitation lines indistinguishable from those obtained with the wild-type enzyme. Of this series of experiments representative results for the various different fus complementation groups are indicated in Fig.2A and B. Interestingly, the cross-reacting material synthesized by many of the cross-reacting-material-positive noncomplementing fusl and fhs2 mutants apparently differs from wildtype fatty acid synthetase in size as well as in its antigenic determinant pattern. This is suggested by the formation of a spur which is observed when extracts of these mutants are immunologically compared to those of wild-type cells (Fig. 2 E). In contrast to the noncoinplementing group 111 and group IV mutants, almost half of the partially pleiotropic complementation group IX mutants studied were cross-reacting-material-positive. Therefore this group was subdivided into groups IXa (negative reaction) and I X b (positive reaction), cf Fig. 2B. At it is shown in Fig. 2C, complementing and non-comple-
menting temperature-sensitive ,fhs-mutants exhibit the same immunological characteristics as the corresponding non-conditionalfus-mutants. Control experiments performed with a series of increasingly diluted wild-type cell extracts indicated that the fatty acid synthetase immunoprecipitation lines were detectable even after ten-fold dilution. Thus, the results obtained with the negatively reacting fas-mutants suggest that these strains contain essentially no free fatty acid synthetase subunits nor any abnormal aggregates, although only one of the two ,fus-loci was mutationally affected in each mutant. This assumption is further supported by the finding that a sodium-dodecylsulfate-treated fatty acid synthetase, though dissociated, still forms immunoprecipitation lines with a specific fatty acid synthetase antiserum (Fig. 2D). Dissociation of the multienzyme complex, under these conditions, is confirmed by the demonstration that its sedimentation velocity in a sucrose density gradient is considerably reduced compared to the intact complex (Fig.3A and B). By sodium dodecylsulfate polyacrylamide gel electrophoresis of the
G. Dietlein and E. Schweizer 0.4
0.3
0.2
0.1
0
0.3
i
N L
2
0 -2
i
9 0.1
C
0
.:
0.i
0.'
I
1
20 Fraction number
I 40
Fig. 3. Sucrore densit). gradient cenrrijiugation r f p q f i r d native (C) and s o d i u m - d o ~ / e c ~ . l ~ u ~ ~ ~ e - d(eAn ,oBf t)~j uriei dj , acid syrhetasr. 2.1 ml (about 10 mg) of wild-type fatty acid synthetdse purified by ammonium sulfate fractionation, calcium phosphate gel adsorption and 2 consecutive sedimentations at 1 O O O O O x g was divided into three equal portions. Samples (A) and (B) were incubated with 2y.i sodium dodecylsulfate and 10% mercaptoethanol for 1 h at 22°C. Subsequently, sample (B) was dialyzed against 2 I of 0.01 M sodium phosphate buffer pH 6.8 for 2 h at 20°C and then placed on top of a sodium-dodecylsulfate-free 10-3OX sucrose density gradient. Sample (A) was centrifuged without dialysis in a similar sucrose density gradient containing 0.1 "4 sodium dodecylsulfate. As a control, the native enzyme (C) was seditnented in a detergent-free sucrose density gradient along with the denatured samples. After 15 h centrifugation at 23000 rev./ min in a Spinco SW 25 rotor the banding patterns of the native and denatured fatty acid synthetases were determined (A-C). 0.02 rnl of thc fractions indicated by arrows were used for sodium dodecylsulfate polyacrylamide gel electrophoresis (D). Electrophorcvs was performed as described previously [I 1. The arrow indicates the direction of sedimentation
peak fractions of both the dissociated and the intact fatty acid synthetase it was shown that, as expected, not the size of the individual subunits but only that of the aggregate was affected by the detergent (Fig.3D). As it is evident from Fig.3B the sodium-dodecylsulfate-dissociated enzyme apparently has no tendency to reassociate when the detergent is removed by dialysis. Therefore it is unlikely, that the imrnunoprecipitation lines observed in Fig.2D were due to the reassociation of the multienzyme complex during
the irnmunodiffusion test. Nevertheless, partid reassociation of subunits to lower aggregates cannot be excluded completely by these experiments. Biosynthesis of Futfji Acid Synthetase Subunits in Noncomp lemen t ing fa s- Mu run ts
As it was mentioned above, the immunological cross-reaction still observed after sodium dodecylsulfate denaturation of the fatty acid synthetase
Fatty-Acid-Synthetase Biosynthesis
182 B
A
X
I
I
I
0 0
1
' 10
LI-i I
20 0 Fraction number
'
10
'0
20
Fig. 4. Incorporation oj ''C'-luheled umino ucids iiiro the furry acid svntherasev of complementing and of nancomplemen/ing teniperaturesensirive fas-mu/anrs. The complementing mutant 36-ts (A), the noncomplementing mutants 31-rs (B) and 60-ts (C) and. as a control. a haploid wild-type yeast strain (D) were first grown at 36 "C with 14 C-labeled amino acids and then transferred to an unlabeled medium at 2 2 T as described in Materials and Methods. Enzyme purification and sucrose density gradient centrifugation were performed as described previously [I]. ).( Absorbance; (0) radioactivity
complex indicates, though it does not conclusively prove, that free fatty acid synthetase subunits are serologically related to the intact complex. Therefore, by serological data alone it cannot be concluded definitely that negatively reacting-fas-mutants contain no free fatty acid synthetase subunits. The biosynthesis of fatty acid synthetase subunits in noncomplementing fus-mutants was therefore studied also by another method. By this method it was tried to incorporate mutationally unaffected free fatty acid synthetase subunits, supposed they were synthesized in noncomplementing ,fas-mutants cells, into the multienzyme complex under appropriate conditions. For instance, subunits present in a noncomplementing temperature-sensitive ,$as-mutant at nonpermissive temperature should be incorporated into the multienzyme complex which is made after these cells were shifted to a permissive growth temperature. Corre-
spondingly, subunits present in temperature-independent noncomplementing ,$as-mutants should be incorporated into the fatty acid synthetase complex synthesized after mating these cells with a wild-type strain or with another, complementing mutant. In Fig. 4A - C the results of an experiment are depicted in which a complementing (38-ts) and a noncomplementing (31-tslfasl-mutant as well as one noncomplementing (60-ts) ,$us2-mutant were first grown at 36°C in a 14C-labeled amino acid medium and, subsequently, allowed to continue growth for another 2- 3 generations at 22 "C in an unlabeled medium. After this, the fatty acid synthetase complex was isolated from all mutants and analyzed for radioactivity. Fig. 4B and C indicate that none of the noncomplementing jasl or f u d mutants had any label incorporated into their fatty acid synthetases under these conditions. On the other hand, the enzyme synthesized by the complementing ts-mutant (Fig. 4A) as well as the enzyme isolated from correspondingly treated wild-type cells (Fig. 4 D) were clearly radioactively labeled. In another series of experiments, nonconditional fus-mutants were labeled in the following way: The complementing mutants jas2-1USa and fusl-2Olx as well as the noncomplementing mutants .fasl-42a and fus2-394cr were grown in 14C-labeled amino acid synthetic medium. At the same time, the corresponding strains of opposite mating type were grown in unlabeled medium. The cells obtained were allowed to mate in the pair-wise combinations fas2-105*/fasl-20I, jasl-201*lfa.s2-105, fasl-42*lfa.s2-394 and fas2-394*/ jasl-42, the asterisks indicating the radioactively labeled strains. The resulting diploids were grown in unlabeled medium for another 2- 3 generations. After this, the fatty acid synthetase complex was isolated from all diploids and analyzed for radioactivity. The results shown in Fig. 5 indicate that in noncompleinenting ,furs1 or ,fus2 mutant cells no fatty acid synthetase subunits are synthesized which are later on incorporated into the fatty acid synthetase complex made within the diploid cells (Fig. 5C and D). On the other hand, when complementing fas-mutants were treated correspondingly a radioactively labeled synthetase was isolated from both mutants studied (Fig. 5A and B). These results, again, indicate that noncoinplementing fas-mutants contain no fatty acid synthetase subunits even not those encoded by the genetically intactfhs-locus. ,
,
DISCUSSION Results obtained previously in this laboratory suggested that in Saccharomjres cerevisicre both fatty acid synthetase gene loci encode only a single multifunctional polypeptide chain [1,6]. In this case, all complementation observed between mutants of the
G. Dietlein and E. Schwcizcr
183
2.0
1.5
1 .o
0.5
E
w w
m
m
O L
c m
e
51 $ 1.5
A
i'
1 .o
0.5
-Jn fas2-394 lfasl-42
0
10
2 0 0
formationally altered protein structure is the basis of phenotypic complementation among jasl and fax2 mutants, respectively. On the other hand, the noncomplementing mutants studied were either free of cross-reacting-material or they contained a material distinctly different from the wild-type enzyme. This is indicated by the spur formation observed with the extract of most cross-reacting-material-positive noncomplemcntingfizs-mutants when compared immunologically with wild-type synthetase. Thus, either the complete lack or a grossly altered protein structure of the fatty acid synthetase complex seems to be a prerequisite for the noncomplementing character of afus-mutant. The results reported in this study suggest the absence of fatty acid synthetase subunits in noncomplementing fus-mutants. This may be generally ascribed to two main reasons: to the repression of fatty acid synthetase biosynthesis or to the rapid proteolytic degradation of any jasl or fusZ gene product possibly synthesized. As indicated by the genetic characterization of a series of non-complementing fusl mutants only part of the noncomplementing fas-mutants represent amber or ochre-type nonense mutants whereas an appreciable number of them apparently are missense mutants [6]. These findings were further confirmed by the identification of numerous noncomplementing temperature-sensitivefusl and fusZ mutants (E. Schweizer, unpublished results). Thus, in the absence of any regulatory control mechanisms both the mutationally unaffected and the mutationally altered fus-gene locus should be translated in these mutants. However, if there exists a coordinate control for the twofus-loci this would have to be a mutual positive control mechanism of both loci by their respective gene products. In this case, a complete and correctly folded fusl gene product would be necessary for transcription or translation of $us2 and vice versa. Though conceptually complicated, this hypothesis cannot be excluded at present. To confirm it further, it would be necessary to isolate regulatory mutants. However, this has not been done, so far. Alternatively, in noncomplementing fas-mutants the mutationally altered fatty acid synthetase subunit may have lost the ability to associate with the other, genetically intact, subunit. Obviously, the majority of nonsense and frameshift mutations in cach of the two loci are likely to show this behaviour. In addition, many $zsl and fus2 missense mutations may have these characteristics, too. As a consequence, the individual and unassociated fusl and .fus2 gene products may undergo rapid proteolytic degradation, in these mutants. Since, usually, proteolytic degradation rates increase with the length of the polypeptide chain [9,10] the two abnormally large fatty acid synthetase subunits [ 13 probably represent exceptionally
10
XI
Fraction number
Fig. 5. Incorporation o f ''C-laheled umino acids into the f u t t v acid sFrithrtase of' coniplernivring and of noncomplernenring fas mutants. Experimental details of the labeling procedure are described in Materials and Methods. The asterisks indicate the mutants which wcre grown in radioactively labeled medium bcfore mating. Otherwise the experiments were performed as described in Fig. 4. (0)Absorbance; (0) radioactivity
samefus-locus represents intragenic complementation. This may be interpreted either as a mutual correction of two mutationally altered protein structures by interactions between homologous subunits within the complex [8] or by a crossfeeding of intermediates between different oligomers within the multimeric yeast fatty acid synthetase complex [2]. On the other hand, the lack of genetic complementation between allelic mutants is normal and, therefore, would be expected for the majority of jasf and fusZ missense mutants. Most of these noncomplementing missense mutants should be cross-reacting-material-positive since they are expected to produce a mutant fatty acid synthetase of wild type-like protein structure. The results reported in this study, however, indicate that with only few exceptions all positively reacting ,$us-mutants exhibit distinct allelic complementation characteristics. This findings suggests that the abovementioned crossfeeding of intermediates rather than the probably less common correction of two con-
184
sensitive targets for proteolysis. It may be assumed that only after incorporation into the complex the fatty acid synthetase subunits are protected against the protease action. This may be due either to the inaccessibility of sensitive peptide bonds within the complex or to the conformational transition, after association, into a structure which is resistant against proteolytic attack. In general, procaryotic as well as eukaryotic cells are known to contain proteases rapidly degrading various types of abnormal proteins, like nonsense fragments [lo, 111 or mutationally altered and denatured proteins [lo]. Similarly, the individual subunits of a protein aggregate may be protease-sensitive unless they are stabilized by association [12]. This may be true for the yeast fatty acid synthetase complex, too. According to currently available evidence the intact yeast fatty acid synthetase complex contains the various different component enzyme activities in a 1 : 1 molecular stoichiometry. This is achieved by fusion of the individual structural genes within each of the two complex gene loci fasf andfcrs2 and by association of the two different fatty acid synthetase polypeptide chains in equimolar amounts. As mentioned above, this association behaviour may either be a consequence of the coordinate expression of jasl and jhs2 or it reflects an inherent characteristic of the two protein components of the complex. In the latter case, any excess .fad or fas2 gene products, if synthesized at all, must be removed by degradation since besides the complete multienzyme complex no lower molecular weight precursors are detected immunologically in the yeast cell homogenate. Thus, no additional genetic control mechanism for the coordinated expression of the two unlinkedfcrs-loci would be necessary preventing the accumulation of excess
G. Dietlein and E. Schweizer:Fatty-Acid-Synthetase Biosynthesis
fusI or Jhs2 gene products in the cell. It has been suggested that in eukaryotic cells enzyme levels are probably more efficiently controlled by protein catabolism rather than at the site of protein biosynthesis [13,14]. The biosynthesis of the yeast fatty acid synthetase complex may be another example for this mechanism. This work has been supported by thc gernernschufr.
/ k i / / . /it, ~
/ oi u / i i u i y \ -
REFERENCES I . Schweizer, E., Kniep, B., Castorph, H. & Holzner, U . (1973) Eur. J . Biochem. 39. 353-362. 2. Schweizer, E., Holzner, U., Meyer, K.-H., Tauro. P. & Schweizer. M . (1974) in Coinpararive Biochemisrr)~and PIiysiology of Transport (Bolis, L.. Bloch, K., Luria, S. E. & Lynen, F.. eds) pp. 219- 244, North-Holland Publishing Company, Amsterdam. 3. Kiihn, L., Castorph, H. & Schweizer, E. (1972) Eur. J . Biuchern. 24,492 - 497. 4. Burkl, G., Castorph, H. & Schweizer, E. (1972) M u / . G m . Genet. 119. 315-322. 5 . Schweizer, E., Kiihn, L. & Castorph. H. (1971) Hoppe-Se,/er’s Z . Physiol. Chem. 352, 371- 384. 6. Tauro, P., Holzner, U.. Castorph, H., Hill, F. & Schweizer. E . (1974) Mol. Gen. Gener. 129. 131- 148. 7. Schweizer, E. & Bolling, H. (1970)Proc. Nut/ Acad. Sci. Lr.X.4. 67,660-666. 8 . Crick, F. H. C. & Orgel, L. E. (1964) J. Mu/. Bid. 8, 161 - 165. 9. Dehlinger, P. J., Schimkc, R. T. (1970) Biorhem. Rioph?~s.Res. Cummun. 40, 1473- 1480. 10. For review see: Goldberg, A. L. & Dice. J. I-‘. (1974) Anrnr. Rev. Biochem. 43, 835- 869. 11. Lin, S. & Zabin, J. (1972) J . Bid. Chem. 247, 2205-2211. 12. Siekevitz, P. (1972) J . Theor. Biol. 37, 321-334. 13. Schimke, R. T. (1973) Adv. Enzymol. 37. 135- 187. 14. Rechcigl. M., Jr (1971) in Enzyme Synthesis and Degradation in Mammalian Svsierns (Rechcigl, M.. cds.) pp. 236- 310, University Park Press, Baltimore.
G. Dietlein and E. Schweizer, Lehrstuhl fur Biochemie der Friedrich-Alexander-Universitat, D-8520 Erlangen. Egerlandstrane 7, Federal Republic of Germany
Control of Fatty-Acid-Synthetase Biosynthesis in Succhuromyces cerevisiue Grrhard DIETLEIN and Eckhart SCHWEIZER Institut fiir Biochemic der Universitat Wiir7burg (Received March 3iJune 30, 1975)
143 out of 308 fusl mutants (47%) and 139 out of the 443 fus2-mutants (32%) genetically studied in this laboratory fail to complement with any other fus-mutant (deficient in fatty acid synthetase) of the same gene locus. From these noncomplementing fus-mutants no mutant fatty acid synthetase can be isolated using the wild-type enzyme purification procedure. Furthermorc the noncomplementing fhs-mutants generally contain no material immunologically crossreacting with a specific fatty acid synthetase antiserum. However, subunits obtained after dissociation of the complex with sodium dodecylsulfate still cross react with this antiserum. Therefore, it is concluded that noncomplementing jbs-mutants contain no fatty acid synthetase component proteins, though one of the two jius-loci is mutationally unaffected. This conclusion was further confirmed by 14Clabeled amino acid incorporation studies which indicated that in noncomplementing jus-mutants, other than in wild type and complementing fhs-mutant cells, no label was incorporated into fatty acid synthetase subunits or precursor proteins. At nonpermissive temperature, the samc biochemical and immunological characteristics were observed with temperature-sensitive noncomplementingfus-mutants. These results suggest that noncornplementing ,fus-mutants either represent regulatory mutants unable to induce the mutationally unaffected other jibs-gene locus or that they are associationdefective mutants. In both cases the resulting individual subunits of the complex may be rapidly degraded by intracellular proteases. Available evidence indicates that in Succhuromyces cerevisiue the fatty acid synthetase multienzyme complex is composed of only two different protein subunits, each present in probably six copies per complex molecule [l 61. Consequently, both subunits represent multifunctional polypeptides chains accommodating several different fatty acid synthetase component enzyme activities, each [2]. This protein structure of the multienzyme complex ensures the stoichiometric biosynthesis of the component enzymes encoded by the same,fus-locus. On the other hand it is unknown, so far, whether the expression of the two genetically unlinked fatty acid synthetase loci jasl and ,fad is under coordinate control, too [3,4]. Of both~fus-locinoncomplementing mutants have been isolated. Characteristically, the few noncompleinenting mutants so far studied did not contain any fatty acid synthetase complex comparable to the wildtype enzyme [3]. Since many of these mutants represent chain-terminating nonsense mutations synthesizing only a fragmentary,fusl orfus2 gene product, these biochemical characteristics were expected [6]. However, besides these nonsense mutations, a sizable
portion of the noncomplementing fus-mutants has been identified as missense mutations [6]. Thus, it remained to be shown whether the failure of fatty acid synthetase production was a general characteristic of noncomplementingfus-mutants no matter whether they were nonsense or missense mutations. In addition, noncomplementingfas nonsense mutants may be used to answer the question whether the second, mutationally unaffected fus-locus is not expressed in these mutants either or whether the fatty acid synthetase subunits encoded in this locus are still synthesized. Therefore, these mutants are considered as especially useful objects for the study of control mechanisms involved in fatty acid synthetase biosynthesis. Experiments pertaining to this question will be reported in this study. MATERIALS AND METHODS Yeast Struins
The origin and characteristics of the Succhuromyces cerevisiae wild type and ,jus-mutant strains
Fatty-Acid-Synthetase Biosynthesis
178
have been described in previous publications IS,71. Mutants fasl-I55 (Ill), fasl-313 (III), fusl-42 (IIJ), 3I - ts (111), 60-ts (IX), 32-ts (IX),fas2-285 (IV), fas2-3Y4 (IV), and fas2-242 (IV) are noncomplementing fasmutants of the pleiotropic complementation groups 111, IV and IX, respectively [5].Mutantsfasl-201 (Va), Jus2-105 (VI), fus2-46 (VI), 38-ts (Vb) and 25-ts (VI) are complementing fas-mutants with their complementation groups indicated by roman numbers in brackets. Yeast Growth and Media
The fatty-acid-containing and the fatty-acid-free yeast extract-peptone-sucrose medium (called fatty acid medium and fatty-acid-free medium) have been described earlier [3,7]. For 14C-labeling of proteins, a synthetic medium was used containing 2 % sucrose, 0.7 7; yeast nitrogen base w/o amino acids (Difco), 1 % Tween 40 (Serva), 0.03 % myristic acid and 20 mg per liter of each of the following amino acids: alanine, arginine, aspartic acid, glutamic acid, glycine, leucine, isoleucine, lysine, phenylalanine, proline, serine, threonine, tyrosine, valine, and, per liter, 16 pCi I4C-labeled amino acid mixture (Amersham, 57 mCi/ matom carbon). Temperature-sensitive fas-mutants were grown each strain in 3 1 of 14C-labeled synthetic medium at 36°C for 2 days. Then, the cells were harvested by centrifugation at 22 "C and washed twice with 200 ml of 22°C sterile distilled water, each time, and, subsequently, they were resuspended in 16 1 of unlabeled fatty-acid-free medium at 22 "C. In this medium, growth was continued for about 2- 3 generations. Growth was followed turbidimetrically at 578 nm. Nonconditionalfus-mutants were labeled by growing each strain in 2 1 of 14C-labeled synthetic medium at 30 "C. In parallel cultures, haploid ,fas-mutants of opposite mating type and genetically defective in the other ,fus locus were grown in 2 1 of unlabeled fatty acid medium, each. At early stationary phase, labeled as well as unlabeled cells were harvested by centrifugation, washed twice with 200 ml of sterile distilled water, each time. Then, labeled and unlabeled strains of opposite mating type were thoroughly mixed in 200 ml of fatty-acid-free medium, spun down again and allowed to mate overnight at 30°C in the tightly packed wet pellet. Subsequently, the cells were suspended in 16 I of fatty-acid-free medium and the diploids formed were allowed to grow for another 2 - 3 generations. Growth was again followed turbidimetrically at 578 nm. Wild-type yeast cells were labeled accordingly. The labeled haploid wild-type strain was mated with another wild-type yeast strain of opposite mating type. Routinely, temperature-sensitive fas-mutant cultures were examined for revertants after their final
harvesting, whereas nonconditional fas-mutants were examined before the mating reaction. The fatty acid synthetase complex was isolated only from revertantfree cultures. The enzyme purification procedure followed the isolation scheme described earlier [I]. Fatty Acid Synthetase Antiserum Preparation and Immunod$fusion Assays Antibodies against the yeast fatty acid synthetase complex were raised in rabbits by intramuscular injection of 1.7 mg of the wild-type enzyme purified by sucrose density gradient centrifugation, together with incomplete Freund's adjuvant (Difco). Booster injections were performed, without adjuvant, intravenously after 4 and 6 weeks using 1.5-2.5 mg of antigen, each time. 10days after the last injection the rabbit was bled. After clotting, the blood serum was collected by centrifugation and kept at - 15°C. For immune double diffusion analysis, the antiserum was usually diluted five-fold with 0.85 % saline pH 6.8. Between 15 and 20 p1 of this solution were placed into the central antiserum well. Immunoprecipitates were allowed to develop during a 24-36 h incubation period at 4 "C.
RESULTS Protein Chemical Characterization qf Complementing and of Noncomplementing fas-Mutants When fatty- acid -synt hetase -deficien t Saccharornyces cerevisiae mutants are subject to genetic complementation analysis, 11 major complementation patterns are observed [5]. The relative distribution of a representative sample of fasl and ,fbs2 mutants among these complementation groups is listed in Table 1. Groups 111 and IV are genetically pleiotropic and fail to complement with all other fus-1 andfus2 mutants, respectively [5]. Group IX mutants are genetically pleiotropic, too, but other than the group IV mutants, they complement with mutants of complementation group VII [ 5 ] .As it is evident from Table 1, 143 of the 308 fasl mutants studied (47%) and 32% (139 out of 443) of the $as2 mutants studied are strictly noncomplementing within the same gene locus. Similar results were obtained with temperaturesensitive fatty acid synthetase mutants (E. Schweizer, unpublished results). This shows again that not only nonsense but also missense mutants may exhibit the noncomplementing phenotype. Earlier results reported by Kiihn et al. [3] suggested that from noncomplementing &-mutants, other than from complementing ones, no mutant fatty acid synthetase complex may be isolated. This has now been confirmed for a sample of 46 randomly chosen group 111 and group IV mutants. Of each
G. Dietlein and E. Schweizer
179 A
I
0.2
F 0
% c
m
s m
e
2 0.1
4
10
m
10 Fraction number
I 20 0
10
20
Fig. 1. Sucrose density gradient profiles c?f fatty acid synthetase preparations purified from complementing and noncomplementing fas-mutants. About 200 g of each, the noncomplementing temperature-sensitive mutant 31-rs (A), grown at permissive (0) as well as at nonpermissive (0) temperature, the complementing mutant fas2-46 (B) and of the noncomplementing mutants f a d - 4 2 (C) and f a d - 2 8 5 (D) were used for enzyme purification. Experimental details of enzyme isolation and sucrose density gradient centrifugation were as described previously [l]
Table 1. Relative frequencies qf the various fasl and fas2 complementation group mutations
Gene locus
Complementation group
Complementation characteristics
Number of mutants isolated
/as I
I1 Ill Va Vb vc
complementing noncomplementing .complementing complementing complementing
92 143 31 30 12
fas2
1v VI VII VIII IX
noncomplementing complementing complementing complementing partially noncomplementing
139 185 40 55
24
mutant the cell homogenate was subject to the fatty acid synthetase isolation procedure adopted for the wild-type enzyme. Whereas the enzyme prepared from wild-type yeast or from a complementing fusmutant strain appears as a distinct protein peak at the final sucrose density gradient centrifugation step (Fig. 1 A and B), with all noncomplementing fusmutants studied, no protein is observed at the corresponding position in the gradient. This is indicated in Fig. 1C and D for two representative noncomplementing fasl and fus2 mutants. As it had been shown by Tauro et ul. [6] one of them, i.e. f a d - 4 2 , is an amber-type nonsense mutation. On the other hand, all of the more than 60 complementing fusmutants so far studied contained the mutant enzyme
at this position. In Fig. 1 A sucrose density gradient profiles of two fatty acid synthetase preparations are shown which were isolated from the noncomplementing temperature-sensitive fus-mutant 31-ts (111) grown at permissive and at nonpermissive temperature, respectively. Obviously, also in this noncomplementing missense mutant the amount of fatty acid synthetase synthesized at nonpermissive temperature is negligible compaired to the permissive conditions. These results indicate, that, quite generally, only the complementing fus-mutants synthesize a fatty acid synthetase complex of a structure like that of the wild type enzyme whereas in noncomplementing fas-mutants, irrespective whether they are missense or nonsense mutations, this complex cannot be demonstrated. Thus, these mutants either contain no fatty acid synthetase protein at all, or the protein structure of the multienzyme complex is grossly altered compared to the wild-type enzyme, thereby causing it to escape the standard fatty acid synthetase isolation procedure. Immunological Characterization of Complementing and of Noncomplementing fas-Mutants
To study the question whether noncomplementing fas-mutants, though inactive in producing the complete multienzyme complex, still synthesize the individual fatty acid synthetase subunits, extracts of a large number of noncomplementing fus-mutants were studied, by immune double diffusion analysis, for the presence of fatty acid synthetase-like cross-reacting material. Among 110 of them tested 85 mutants proved to be deficient of any cross-reacting material whereas 25 cell extracts developed, though mostly
Fatty-Acid-Synthetase Biosynthesis
Fig. 2. I , ) i i ~ i u , i o ~ / ~ ~ i r unalysis sio,l of' coinpkiiienting and of iioncornplemcnting ,/us-murants. Each strain was grown in 500 in1 of Fatty acid incdium at 30'C. At early stationary phase cells wcre harvested and broken with glass beads. The lioinogenatc was centrifuged at 12000 x g for 1 h at 4 ' C and 20 p1 (about 0.3 mg) of the supernatant were used for the immunological tests. The antiserum well contained 20 p1 of a 1 : 5 diluted fatty acid synthetase antiserum solution. After 18 - 36 h incubation at 4 'C thc immunoprecipitation lines formed were recorded. In (d) purified wild-type (WT) fatty acid synthetasc (about 5 mg/ml) was denatured with 204 sodium dodecylsulfate (SDS) and lox inercaptocthanol for 1 h at 22'C. 20 pl of this solution were placed into the antigen well and, as a control, another well contained the same amount of the untreated wild-type enzyme. Roman numbers indicate cornplemcntation groups, arabic numbers fus-mutant alleles
very weak, immunoprecipitation lines. On the other hand, all complementing fusl or .fusZ mutants tcsted exhibited distinct immunoprecipitation lines indistinguishable from those obtained with the wild-type enzyme. Of this series of experiments representative results for the various different fus complementation groups are indicated in Fig.2A and B. Interestingly, the cross-reacting material synthesized by many of the cross-reacting-material-positive noncomplementing fusl and fhs2 mutants apparently differs from wildtype fatty acid synthetase in size as well as in its antigenic determinant pattern. This is suggested by the formation of a spur which is observed when extracts of these mutants are immunologically compared to those of wild-type cells (Fig. 2 E). In contrast to the noncoinplementing group 111 and group IV mutants, almost half of the partially pleiotropic complementation group IX mutants studied were cross-reacting-material-positive. Therefore this group was subdivided into groups IXa (negative reaction) and I X b (positive reaction), cf Fig. 2B. At it is shown in Fig. 2C, complementing and non-comple-
menting temperature-sensitive ,fhs-mutants exhibit the same immunological characteristics as the corresponding non-conditionalfus-mutants. Control experiments performed with a series of increasingly diluted wild-type cell extracts indicated that the fatty acid synthetase immunoprecipitation lines were detectable even after ten-fold dilution. Thus, the results obtained with the negatively reacting fas-mutants suggest that these strains contain essentially no free fatty acid synthetase subunits nor any abnormal aggregates, although only one of the two ,fus-loci was mutationally affected in each mutant. This assumption is further supported by the finding that a sodium-dodecylsulfate-treated fatty acid synthetase, though dissociated, still forms immunoprecipitation lines with a specific fatty acid synthetase antiserum (Fig. 2D). Dissociation of the multienzyme complex, under these conditions, is confirmed by the demonstration that its sedimentation velocity in a sucrose density gradient is considerably reduced compared to the intact complex (Fig.3A and B). By sodium dodecylsulfate polyacrylamide gel electrophoresis of the
G. Dietlein and E. Schweizer 0.4
0.3
0.2
0.1
0
0.3
i
N L
2
0 -2
i
9 0.1
C
0
.:
0.i
0.'
I
1
20 Fraction number
I 40
Fig. 3. Sucrore densit). gradient cenrrijiugation r f p q f i r d native (C) and s o d i u m - d o ~ / e c ~ . l ~ u ~ ~ ~ e - d(eAn ,oBf t)~j uriei dj , acid syrhetasr. 2.1 ml (about 10 mg) of wild-type fatty acid synthetdse purified by ammonium sulfate fractionation, calcium phosphate gel adsorption and 2 consecutive sedimentations at 1 O O O O O x g was divided into three equal portions. Samples (A) and (B) were incubated with 2y.i sodium dodecylsulfate and 10% mercaptoethanol for 1 h at 22°C. Subsequently, sample (B) was dialyzed against 2 I of 0.01 M sodium phosphate buffer pH 6.8 for 2 h at 20°C and then placed on top of a sodium-dodecylsulfate-free 10-3OX sucrose density gradient. Sample (A) was centrifuged without dialysis in a similar sucrose density gradient containing 0.1 "4 sodium dodecylsulfate. As a control, the native enzyme (C) was seditnented in a detergent-free sucrose density gradient along with the denatured samples. After 15 h centrifugation at 23000 rev./ min in a Spinco SW 25 rotor the banding patterns of the native and denatured fatty acid synthetases were determined (A-C). 0.02 rnl of thc fractions indicated by arrows were used for sodium dodecylsulfate polyacrylamide gel electrophoresis (D). Electrophorcvs was performed as described previously [I 1. The arrow indicates the direction of sedimentation
peak fractions of both the dissociated and the intact fatty acid synthetase it was shown that, as expected, not the size of the individual subunits but only that of the aggregate was affected by the detergent (Fig.3D). As it is evident from Fig.3B the sodium-dodecylsulfate-dissociated enzyme apparently has no tendency to reassociate when the detergent is removed by dialysis. Therefore it is unlikely, that the imrnunoprecipitation lines observed in Fig.2D were due to the reassociation of the multienzyme complex during
the irnmunodiffusion test. Nevertheless, partid reassociation of subunits to lower aggregates cannot be excluded completely by these experiments. Biosynthesis of Futfji Acid Synthetase Subunits in Noncomp lemen t ing fa s- Mu run ts
As it was mentioned above, the immunological cross-reaction still observed after sodium dodecylsulfate denaturation of the fatty acid synthetase
Fatty-Acid-Synthetase Biosynthesis
182 B
A
X
I
I
I
0 0
1
' 10
LI-i I
20 0 Fraction number
'
10
'0
20
Fig. 4. Incorporation oj ''C'-luheled umino ucids iiiro the furry acid svntherasev of complementing and of nancomplemen/ing teniperaturesensirive fas-mu/anrs. The complementing mutant 36-ts (A), the noncomplementing mutants 31-rs (B) and 60-ts (C) and. as a control. a haploid wild-type yeast strain (D) were first grown at 36 "C with 14 C-labeled amino acids and then transferred to an unlabeled medium at 2 2 T as described in Materials and Methods. Enzyme purification and sucrose density gradient centrifugation were performed as described previously [I]. ).( Absorbance; (0) radioactivity
complex indicates, though it does not conclusively prove, that free fatty acid synthetase subunits are serologically related to the intact complex. Therefore, by serological data alone it cannot be concluded definitely that negatively reacting-fas-mutants contain no free fatty acid synthetase subunits. The biosynthesis of fatty acid synthetase subunits in noncomplementing fus-mutants was therefore studied also by another method. By this method it was tried to incorporate mutationally unaffected free fatty acid synthetase subunits, supposed they were synthesized in noncomplementing ,fas-mutants cells, into the multienzyme complex under appropriate conditions. For instance, subunits present in a noncomplementing temperature-sensitive ,$as-mutant at nonpermissive temperature should be incorporated into the multienzyme complex which is made after these cells were shifted to a permissive growth temperature. Corre-
spondingly, subunits present in temperature-independent noncomplementing ,$as-mutants should be incorporated into the fatty acid synthetase complex synthesized after mating these cells with a wild-type strain or with another, complementing mutant. In Fig. 4A - C the results of an experiment are depicted in which a complementing (38-ts) and a noncomplementing (31-tslfasl-mutant as well as one noncomplementing (60-ts) ,$us2-mutant were first grown at 36°C in a 14C-labeled amino acid medium and, subsequently, allowed to continue growth for another 2- 3 generations at 22 "C in an unlabeled medium. After this, the fatty acid synthetase complex was isolated from all mutants and analyzed for radioactivity. Fig. 4B and C indicate that none of the noncomplementing jasl or f u d mutants had any label incorporated into their fatty acid synthetases under these conditions. On the other hand, the enzyme synthesized by the complementing ts-mutant (Fig. 4A) as well as the enzyme isolated from correspondingly treated wild-type cells (Fig. 4 D) were clearly radioactively labeled. In another series of experiments, nonconditional fus-mutants were labeled in the following way: The complementing mutants jas2-1USa and fusl-2Olx as well as the noncomplementing mutants .fasl-42a and fus2-394cr were grown in 14C-labeled amino acid synthetic medium. At the same time, the corresponding strains of opposite mating type were grown in unlabeled medium. The cells obtained were allowed to mate in the pair-wise combinations fas2-105*/fasl-20I, jasl-201*lfa.s2-105, fasl-42*lfa.s2-394 and fas2-394*/ jasl-42, the asterisks indicating the radioactively labeled strains. The resulting diploids were grown in unlabeled medium for another 2- 3 generations. After this, the fatty acid synthetase complex was isolated from all diploids and analyzed for radioactivity. The results shown in Fig. 5 indicate that in noncompleinenting ,furs1 or ,fus2 mutant cells no fatty acid synthetase subunits are synthesized which are later on incorporated into the fatty acid synthetase complex made within the diploid cells (Fig. 5C and D). On the other hand, when complementing fas-mutants were treated correspondingly a radioactively labeled synthetase was isolated from both mutants studied (Fig. 5A and B). These results, again, indicate that noncoinplementing fas-mutants contain no fatty acid synthetase subunits even not those encoded by the genetically intactfhs-locus. ,
,
DISCUSSION Results obtained previously in this laboratory suggested that in Saccharomjres cerevisicre both fatty acid synthetase gene loci encode only a single multifunctional polypeptide chain [1,6]. In this case, all complementation observed between mutants of the
G. Dietlein and E. Schwcizcr
183
2.0
1.5
1 .o
0.5
E
w w
m
m
O L
c m
e
51 $ 1.5
A
i'
1 .o
0.5
-Jn fas2-394 lfasl-42
0
10
2 0 0
formationally altered protein structure is the basis of phenotypic complementation among jasl and fax2 mutants, respectively. On the other hand, the noncomplementing mutants studied were either free of cross-reacting-material or they contained a material distinctly different from the wild-type enzyme. This is indicated by the spur formation observed with the extract of most cross-reacting-material-positive noncomplemcntingfizs-mutants when compared immunologically with wild-type synthetase. Thus, either the complete lack or a grossly altered protein structure of the fatty acid synthetase complex seems to be a prerequisite for the noncomplementing character of afus-mutant. The results reported in this study suggest the absence of fatty acid synthetase subunits in noncomplementing fus-mutants. This may be generally ascribed to two main reasons: to the repression of fatty acid synthetase biosynthesis or to the rapid proteolytic degradation of any jasl or fusZ gene product possibly synthesized. As indicated by the genetic characterization of a series of non-complementing fusl mutants only part of the noncomplementing fas-mutants represent amber or ochre-type nonense mutants whereas an appreciable number of them apparently are missense mutants [6]. These findings were further confirmed by the identification of numerous noncomplementing temperature-sensitivefusl and fusZ mutants (E. Schweizer, unpublished results). Thus, in the absence of any regulatory control mechanisms both the mutationally unaffected and the mutationally altered fus-gene locus should be translated in these mutants. However, if there exists a coordinate control for the twofus-loci this would have to be a mutual positive control mechanism of both loci by their respective gene products. In this case, a complete and correctly folded fusl gene product would be necessary for transcription or translation of $us2 and vice versa. Though conceptually complicated, this hypothesis cannot be excluded at present. To confirm it further, it would be necessary to isolate regulatory mutants. However, this has not been done, so far. Alternatively, in noncomplementing fas-mutants the mutationally altered fatty acid synthetase subunit may have lost the ability to associate with the other, genetically intact, subunit. Obviously, the majority of nonsense and frameshift mutations in cach of the two loci are likely to show this behaviour. In addition, many $zsl and fus2 missense mutations may have these characteristics, too. As a consequence, the individual and unassociated fusl and .fus2 gene products may undergo rapid proteolytic degradation, in these mutants. Since, usually, proteolytic degradation rates increase with the length of the polypeptide chain [9,10] the two abnormally large fatty acid synthetase subunits [ 13 probably represent exceptionally
10
XI
Fraction number
Fig. 5. Incorporation o f ''C-laheled umino acids into the f u t t v acid sFrithrtase of' coniplernivring and of noncomplernenring fas mutants. Experimental details of the labeling procedure are described in Materials and Methods. The asterisks indicate the mutants which wcre grown in radioactively labeled medium bcfore mating. Otherwise the experiments were performed as described in Fig. 4. (0)Absorbance; (0) radioactivity
samefus-locus represents intragenic complementation. This may be interpreted either as a mutual correction of two mutationally altered protein structures by interactions between homologous subunits within the complex [8] or by a crossfeeding of intermediates between different oligomers within the multimeric yeast fatty acid synthetase complex [2]. On the other hand, the lack of genetic complementation between allelic mutants is normal and, therefore, would be expected for the majority of jasf and fusZ missense mutants. Most of these noncomplementing missense mutants should be cross-reacting-material-positive since they are expected to produce a mutant fatty acid synthetase of wild type-like protein structure. The results reported in this study, however, indicate that with only few exceptions all positively reacting ,$us-mutants exhibit distinct allelic complementation characteristics. This findings suggests that the abovementioned crossfeeding of intermediates rather than the probably less common correction of two con-
184
sensitive targets for proteolysis. It may be assumed that only after incorporation into the complex the fatty acid synthetase subunits are protected against the protease action. This may be due either to the inaccessibility of sensitive peptide bonds within the complex or to the conformational transition, after association, into a structure which is resistant against proteolytic attack. In general, procaryotic as well as eukaryotic cells are known to contain proteases rapidly degrading various types of abnormal proteins, like nonsense fragments [lo, 111 or mutationally altered and denatured proteins [lo]. Similarly, the individual subunits of a protein aggregate may be protease-sensitive unless they are stabilized by association [12]. This may be true for the yeast fatty acid synthetase complex, too. According to currently available evidence the intact yeast fatty acid synthetase complex contains the various different component enzyme activities in a 1 : 1 molecular stoichiometry. This is achieved by fusion of the individual structural genes within each of the two complex gene loci fasf andfcrs2 and by association of the two different fatty acid synthetase polypeptide chains in equimolar amounts. As mentioned above, this association behaviour may either be a consequence of the coordinate expression of jasl and jhs2 or it reflects an inherent characteristic of the two protein components of the complex. In the latter case, any excess .fad or fas2 gene products, if synthesized at all, must be removed by degradation since besides the complete multienzyme complex no lower molecular weight precursors are detected immunologically in the yeast cell homogenate. Thus, no additional genetic control mechanism for the coordinated expression of the two unlinkedfcrs-loci would be necessary preventing the accumulation of excess
G. Dietlein and E. Schweizer:Fatty-Acid-Synthetase Biosynthesis
fusI or Jhs2 gene products in the cell. It has been suggested that in eukaryotic cells enzyme levels are probably more efficiently controlled by protein catabolism rather than at the site of protein biosynthesis [13,14]. The biosynthesis of the yeast fatty acid synthetase complex may be another example for this mechanism. This work has been supported by thc gernernschufr.
/ k i / / . /it, ~
/ oi u / i i u i y \ -
REFERENCES I . Schweizer, E., Kniep, B., Castorph, H. & Holzner, U . (1973) Eur. J . Biochem. 39. 353-362. 2. Schweizer, E., Holzner, U., Meyer, K.-H., Tauro. P. & Schweizer. M . (1974) in Coinpararive Biochemisrr)~and PIiysiology of Transport (Bolis, L.. Bloch, K., Luria, S. E. & Lynen, F.. eds) pp. 219- 244, North-Holland Publishing Company, Amsterdam. 3. Kiihn, L., Castorph, H. & Schweizer, E. (1972) Eur. J . Biuchern. 24,492 - 497. 4. Burkl, G., Castorph, H. & Schweizer, E. (1972) M u / . G m . Genet. 119. 315-322. 5 . Schweizer, E., Kiihn, L. & Castorph. H. (1971) Hoppe-Se,/er’s Z . Physiol. Chem. 352, 371- 384. 6. Tauro, P., Holzner, U.. Castorph, H., Hill, F. & Schweizer. E . (1974) Mol. Gen. Gener. 129. 131- 148. 7. Schweizer, E. & Bolling, H. (1970)Proc. Nut/ Acad. Sci. Lr.X.4. 67,660-666. 8 . Crick, F. H. C. & Orgel, L. E. (1964) J. Mu/. Bid. 8, 161 - 165. 9. Dehlinger, P. J., Schimkc, R. T. (1970) Biorhem. Rioph?~s.Res. Cummun. 40, 1473- 1480. 10. For review see: Goldberg, A. L. & Dice. J. I-‘. (1974) Anrnr. Rev. Biochem. 43, 835- 869. 11. Lin, S. & Zabin, J. (1972) J . Bid. Chem. 247, 2205-2211. 12. Siekevitz, P. (1972) J . Theor. Biol. 37, 321-334. 13. Schimke, R. T. (1973) Adv. Enzymol. 37. 135- 187. 14. Rechcigl. M., Jr (1971) in Enzyme Synthesis and Degradation in Mammalian Svsierns (Rechcigl, M.. cds.) pp. 236- 310, University Park Press, Baltimore.
G. Dietlein and E. Schweizer, Lehrstuhl fur Biochemie der Friedrich-Alexander-Universitat, D-8520 Erlangen. Egerlandstrane 7, Federal Republic of Germany
