Clli'iJni)hiO/(lg~Inrernurionul
Vol. 9. No. 3, pp. 222-230 0 1992 International Society of Chronobiology
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Circadian Rhythms in Neurospora crassa: The Role of Mitochondria
Summary: Energy metabolism and mitochondria have been discussed with respect to their role in the circadian rhythm mechanism for some time. Numerous examples of inhibitors that affect the mitochondria of plants and animals and microorganisms are known, which cause large phase shifts in the rhythms of these organisms. Analogous studies on the role of mitochondria in the Neiirosporu circadian rhythm mechanism habe also been reported and summarized. This communication differs from previous studies on other organisms in that it will focus on two lines of evidence derived from studies on Nczrrosporu strains carrying mutations affecting the mitochondria. (a) Strains whose growth rate is resistant to oligomycin (01;') owing to an altered protein in the F, sector ofthe mitochondrial ATPase, showed no phase shifts when pulsed with oligomycin. Control strains (oli') showed large phase shifts when pulsed with oligomycin. This indicates that the phase-shifting effect of oligomycin is due to the direct inhibition of the mitochondrial ATPase and not some side effect of this inhibitor. (b) In Nairosporu, many different strains are known that carry mutations in the nuclear o r mitochondrial genome that affect mitochondrially localized proteins. Some of these, such as di', [ M I - 3 ] ,or cyu-5. showed shorter ( - 19-h) periods compared with the normal (21.5-h) period. Others showed little or no change in period. Those mutant strains exhibiting shorter periods also contained ~ 6 0 % more mitochondria1 protein per gram total protein in extracts compared with the normal strains. Assays of the level of a mitochondrial-specific protein, acyl carrier protein, showed that the cellular content of this protein was approximately doubled. A parallel set of studies on the effects of antimycin or chloramphenicol on Neiirosporu demonstrated that these inhibitors also produced shorter periods as well as increased amounts of mitochondrial proteins. These two new lines of evidence may be interpreted to indicate that in Neurosporu either some part of the oscillator is localized to the mitochondria and/or that mitochondria exert their effect on the clock mechanism through their effects on biosynthetic pathways or by their contribution in determining ion gradients. Key Words: Mitochondria-Neziro.~poru crussu-Circadian rhythm.
Received December 23, 1991; accepted February 17, 1992. Addresscorrespondence and reprint requests to Dr. S. Brody at Department of Biology, 01 16, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-01 16, U.S.A. Presented in part at the 20th International Conference on Chronobiology, June 16-21, 1991, Tel Aviv, Israel.
222
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
MITOC’HONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
223
Inhibitors of mitochondrial function have long been known to lead to phase changes in many organisms. Many of these effects have been listed in books by Edmunds ( 1 ) and also by Bunning (2). In plants, these inhibitors have been given as pulses in the transpiration stream, and in microorganisms, they have been given as pulses in the growth medium. The effects of anoxia have also been reported for the Drosuphilu rhythm (3), but these effects may not be specific to mitochondria since there are a significant number of oxygen-requiring biosynthetic and degradative reactions in cells. Most of the mitochondrial inhibitors that have effects on the rhythms of these different organisms lead to phase changes, and phase-response curves are known for most of these compounds. The phase changes produced by pulses of these inhibitors are generally produced by concentrations of these inhibitors that are not unusually high or toxic to cells. However, the question of how much of an inhibitor is truly removed at the end of the pulse has rarely been examined in these studies. The inhibitors used are primarily those that affect energy metabolism, such as uncouplers or azide, cyanide, etc. Some of the inhibitors that have been uncritically employed and reported unfortunately have effects other than just on mitochondrial energy production. The group of inhibitors used does not include, in general, inhibitors of other mitochondrial reactions such as the arginine biosynthetic pathway or mitochondria1 protein synthesis. Only in Neurosporu has chloramphenicol been shown to have an effect on the period of the rhythm (4).As more is known about the capacity of mitochondria to make more than just ATP, it may well be that other inhibitors will be employed, and/or mutant strains with deficiencies in a given reaction will be usefully employed. The bulk of our knowledge about mitochondria and rhythms is due to phase shifts produced by inhibitors, whereas only a few studies reported period changes. Although it is widely acknowledged that phase shifts produced by chemicals could be due to effects on an input pathway to the oscillator, and not necessarily on the oscillator mechanism itself, it should be pointed out the period changes due to inhibitors could also be due t o effects on an input pathway, and not necessarily on the oscillator mechanism itself.
THE CIRCADIAN RHYTHMS OF NEUROSPORA The circadian rhythm of the fungus Neurosporu crussa is routinely assayed by its expression as a conidiation (asexual spore formation) rhythm on agar medium. This complex developmental process is triggered by a “clock mechanism” at some phase of the circadian cycle. The Nezirosporu rhythm shares the basic properties common to circadian rhythms of most organisms: (a) The period of the rhythm is close to but not equal to 24 h; (b) the rhythm is endogenous and self-sustaining; (c) the phase of the rhythm can be changed by pulses of light or temperature or entrained; and (d) the period of the rhythm varies little with ambient temperature, i.e., it is temperature compensated. A number of mutations that alter the rhythm properties have been reported and reviewed recently ( 5 ) . Mutations affecting the free-running period have been reported at many loci, and the structure of the.fiq gene has been reported.
224
S. BRODY
Mutations affecting the light response, such as poky and others, are known, and mutations leading to a loss of temperature compensation (6) have been reported also.
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
PREVIOUS STUDIES ON NEUROSPORA MITOCHONDRIA AND RHYTHMS Three types of results have been found in Neurospora. First, there are a number of inhibitors that lead to phase shifts when given as a pulse. Second, there are inhibitors that lead to a change in the period when given continuously. Third, there are mutations that affect mitochondria and lead to changes in period. An example ofthis latter type of mutant strain was found to be an oligomycin-resistant mutant, whose period was 18-19 h (7). Other examples are given below. Chemicals that lead to period changes in Neurospora are well known and are gathered together in a review ( 5 ) . One of these, chloramphenicol, leads to a 19-h period. Others, such as phenylethanol, also lead to shorter periods. However, the bulk of the work relating mitochondria and circadian rhythms in Neurospora comes from studies on inhibitor pulses. These inhibitors are antimycin A, azide, carbonyl cyanide chlorophenol, cyanide, dicyclohexylcarbodiimide, N-ethylmaleimide, and oligomycin. As in all studies on chemical pulses that lead to stable phase shifts, the questions often remain: What is the target(s) for any given inhibitor? How do we know for sure that the primary effect of the inhibitor leads to the observed phase shift, and not some secondary or side effect of the inhibitor? Although it can be argued that it is unlikely that all of the different mitochondrial inhibitors have the exact same side effect, it is more important to actually obtain experimental evidence on this point, rather than just argue it. Therefore, a study of the response of Neuruspora to oligomycin, a mitochondrial inhibitor, was done employing two strains, one sensitive to oligomycin and one resistant to it. These resistant mutations all affected the structure of a small, hydrophobic subunit in the Fo section of the mitochondria1 ATPase (8), a subunit thought to be involved in proton (H') transport into the mitochondria.
RESULTS Oligomycin Pulse Studies Figure 1 shows the phase-response curve to oligomycin for disks of Neurospora incubating in a liquid shaker culture. Similar phase-response curves (Type 0) have been previously reported by two other groups (9, lo), although there are some differences between the data obtained by the three groups with respect to the size of the phase shifts and the exact shape of the curve. In the cultures that were assayed to generate the data for Fig. 1 (top), the disks were extensively washed to get rid of the oligomycin. Upon transfer to the growth tube medium, after the oligomycin pulse, the disks gave rise to cultures growing at a rapid ( 1.3 mm/h) rate, a rate identical to that in cultures not pulsed with oligomycin. The immediate rapid growth rate indicates that the bulk of the oligomycin was no longer present after the pulse, even though some type of additional data, such as the complete loss of labeled oligomycin,
UironobiolInl. Vol. 9, No. 3, I992
MITOCHONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
225
Oligomycin Sensitive
5-
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
4-
FIG. 1. Mycelial disks of Neurosporu ( 1 5 ) in a D/D liquid shaker culture were pulsed with oligomycin for 2 h. The oligomycin (10 mg/ ml ethanol) was added to a final concentration of 10 pg/ml. The disks were washed and then placed on race tubes in D/D. Each time point represents the average of five race tubes. The phase relative to controls (ethanol only) was calculated for days 3-5 and averaged. Each point has a standard deviation for the phase difference of -2 h. Top: Phase-response curve for olis strain. Bottom: Phase-response curve for oli' strain.
0-
I
4 4
Oligomycin Resistant U-
d
0
would have been more definitive. The previous studies (9,lO) do not report the growth rates immediately after the pulse and transfer, so it is not clear how much of this lipid-soluble inhibitor was still present. This could account for the differences reported by the three groups. Figure 1 (bottom) shows the identical experimental conditions to Fig. 1 (top), with the exception that the oligomycin-resistant strain (bd csp- 1 oli') was employed. Few, if any, reliable effects on the phase of these cultures were observed. Higher concentrations of oligomycin (data not shown) also gave no effect on the phases of the cultures. These data indicate that oligomycin does not lead to phase shifts of Neuuospoua by any side effect in these cultures. Rather, it points out that oligomycin leads to the observed phase shifts by its primary effect on the mitochondria1 ATPase. The inhibition of the proton transport into the mitochondria by oligomycin would be expected to inhibit the synthesis of ATP and in turn affect many other cellular procedures, such as ionic gradients, membrane potentials, etc.
Chronohiol Inl, Vol. Y, N o . 3, 1992
S. BRODY
226
TABLE I . &/ficts on llial shorfcn tlic 22-11 pcriod Mutant 011
[MI-3] 1 vu-5 c 1.b-2
Period (h)
Chemical
Period ( h )
18-20 18-19
Chloramphenicol Antimycin Aurovcrtin Phenylethanol
19
19 18
19 19 18
Citations are given in ref. 5
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Studies on Amount of Mitochondria1 Protein and Period A number of mutant strains with altered mitochondria also exhibited effects on the circadian rhythms of N. crassa: oh*, [M1-3];cyu-5 all had slower growth rates but faster (18-19 h) clocks (5). In addition, various mitochondrial inhibitors such as antimycin or aurovertin had the same effects on growth rate and clock rate as these mutations (Table 1). Since the mitochondrial targets for these inhibitors and mutations were so different (Fig. 2), yet they all had the same clock effect, it was of interest to see what these mutations and inhibitors had in common. A study ofthe amount of mitochondrial protein per unit cell mass was undertaken as a first step. The amount of mitochondria was measured as the percentage of mitochondrial protein in the total protein in a cell-free extract as follows: Liquid cultures of N m r o spora were grown at 22°C in minimal medium containing either 2% glucose or 1%) acetate as the carbon source and were shaken until the exponential phase was reached. They were harvested by vacuum filtration through Whatman filters in a
ATPase Complex V
Complex IV
[M 1-31,
Oxidase
Venturicidin,
FIG. 2. A schematic diagram of mitochondrial metabolism indicating approximate sites of inhibitors by chemicals and/or mutations.
MITOCHONDRIA IN NEUROSPORA CIRCADIAN R H Y T H M
227
TABLE 2. Effectsof muiations and Erowth conditions on amount of‘mitochondriul moirin Strain
Growth medium
hd CSP- I hd c.sp- 1 oli‘ ( 16-16) hd C.SI)- I [MI-3] bd csp- I cyu-5 hd c.sp- I hd CSP- I !Id C‘sp-I
Glucose Glucose Glucose Glucose Acetate Glucose + chloramphenicol Antimycin
Protein in mitochondrial fraction (70) 4.0 t 0.5 7.5 f 1.2 6.3 ? 1.0 6.5 t 0.87 5.6 i 0.4 6.3 f 0.6 10.7 t 0.7
(9) (6) (8) (3) (20) (4) (8)
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Numbers in parentheses are individual samples assayed.
Bucher funnel and the wet weights of the Ncwosporu mats were recorded. Samples from identical cultures were combined to yield a weight of at least 1 g, if necessary. The Ncwrosporu was disrupted with a bead beater for 1 min in a mitochondrial isolation buffer (0.25 A4 sucrose, 1 m M ethylenediaminetetraacetate, I0 m M Tris, pH 7.2). The suspension was then spun at 600 g for 5 min to remove cell debris. The supernatant was poured through cheesecloth and the total volume was recorded. One milliliter was saved for the Lowry protein assay ( 1 1). This “homogenate” represents the total protein in the sample. The supernatant was then spun at 10,000 g for 20 min to pellet organelles. The supernatant was poured off and the pellet was resuspended with a glass homogenizer in 5 ml of the mitochondrial isolation buffer. This was spun at 1,200 g for 10 min. The pellet was discarded and the supernatant was spun at 10,000 g for 20 min. This pellet was resuspended in 1 ml of the mitochondrial isolation buffer and represents the protein contained in the “mitochondrial” fraction. Lowry assays ( 1 1) were performed to determine the amount of protein contained in the homogenate and mitochondrial fractions using bovine serum albumin as a standard. The percentage mitochondria was calculated as a fraction of the amount of total protein (i.e., homogenate). The amount of protein in the mitochondrial fraction was 4.0 f 0.5% ofthe protein found in a total crude homogenate of Neurospora, employing the breakage procedure given above. This value was increased -90% by a mutation affecting ATPase (olir), 60% by a maternally inherited mutation affecting cytochrome oxidase ([MZ-3]), and by another mutant with low levels of cytochrome oxidase (cyu-5).The mitochondrial level was increased by -40% by growth of Nezirosporu on acetate as a carbon source instead of glucose and was increased 60% by chloramphenicol or 150% by the mitochondrial inhibitor antimycin (Table 2). Three additional studies of the amount of mitochondrial protein were also performed (data not shown). The first was in a variety of strains, such as (a) strains with different alleles at the oli‘ locus: (b) the [MI-5] strain, similar to [MI-3];( c )different isolates of the hd c.sp- 1 strain and of the hd strain; and (d) several wild-type strains such as RL3-8A and 740R23IVA. The presence or absence of the hd or csp-I markers had no measurable effect on the amount of mitochondrial protein observed. The [MI-5]strain gave the same values as the [MI-3] strain, and the values obtained
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
228
S. BRODY
for the different oli‘ strains correlated with their degree of oligomycin resistance, the least resistant showing little changes from the bd csp- 1 strain. Second, the amount of mitochondrial proteins were assayed for cultures grown on solid medium, such as bd csp and bd csp oh‘, and for bd csp grown in the presence of antimycin. The change from liquid to solid medium did not affect the values obtained. Third, the level of a mitochondrial-specificprotein (1 2), acyl carrier protein (ACP), was assayed by comparing the amount of radioactive pantothenate in this protein in a crude homogenate versus that found in fatty acid synthetase. This technique, sodium dodecyl sulfate polyacrylamide gel electrophoresis, showed that the oli‘ mutation increased the level of the ACP by -90%, comparable with the studies on amounts of protein (data not shown).
DISCUSSION When viewed together, these results can be interpreted as follows: Mitochondria Clock effect increased amount --* “faster clock” decreased rate + phase shift (inhibitor pulse) The increased amount could lead to increased rates of mitochondrial processes, and the “faster” clock refers to a shorter period. Of course, the shorter period could be due to either a decreased amplitude or a “faster” clock, which implies a rate change. Therefore, one of the possible interpretations would be that an increased rate of mitochondria1processes leads to an increased clock rate. The decreased rate of mitochondrial processes due to an inhibitor pulse is a temporary one and may be for only certain processes as well. It is well known that uncouplers decrease the rate of ATP synthesis, but increase the rate of electron transport and respiration, for example. These two findings taken together do not necessarily pinpoint the exact step or pathway in the mitochondria that leads to these clock changes. This is because the biochemical reactions of the mitochondria are so intimately tied in with the rest of the cellular metabolism that effects on one have serious “ripple” effects on the other. This interrelationship is true at many levels: First, any change in the rate of ATP generation could be expected to change the rate of the ATP-requiring reactions. These could be a sizable portion of the cellular machinery. Second, an increase in the amount of mitochondria in the cell would alter the ratio of certain isozymes found in both mitochondria and other cellular compartments. This possible imbalance in enzymes would also be true for any pathway, such as arginine synthesis or phospholipid synthesis (phosphatidyl serine decarboxylase),where only some of the steps are localized to the mitochondria. Third, the increase in mitochondrial levels observed here may very well be a reflection of the general increase in the levels of transcription of many genes, both nuclear and mitochondrial. A specific example of this is given by the fivefold increase in the mRNA for a small mitochondrial ribosomal protein in response to impaired mitochondrial function ( 1 3). It seems quite probable that some nonmitochondrial enzymes may also be increased in response to the same generalized cellular signal, so that a balanced metabolism can occur. It would be of interest
C‘hronobiolhi.Vol 9, No 3, 1992
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
MITOCHONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
229
to see if the genes coding for gluconeogenic enzymes, for example, have some promoter sequences similar to those employed for the genes coding for mitochondrially localized proteins. There have been other studies on Neurosporu that are pertinent to this discussion. Nakashima (9) has studied ATP level, respiration, and ATPase inhibitors and their possible role in clock function. His conclusion, i.e., clocks still “run” even though respiration can be very low, does not negate the findings presented here. These two lines of evidence can be reconciled by restating some of the considerations listed above: Mitochondria do more than just electron transport and respiration. They have a role in Ca2+metabolism, in certain amino acid biosynthetic pathways, and even in fatty acid synthesis (14). It seems plausible that even when respiration is quite low, ion gradients are still maintained across the membranes and biosynthetic reactions are still proceeding at some rate. Other studies on inhibitor pulses have been summarized (5). These data on Neurosporu and the previous studies performed on other organisms suggest a role for the mitochondria in the clock mechanism. The studies, of course, do not allow a definitive enough interpretation as to whether the role is a primary or a secondary one. A primary role for the mitochondria would be if the entire clock mechanism, or some significant portion of it, were localized in the mitochondria. A secondary role would be if clocks merely required ATP to “run” or if effects on mitochondrial metabolism just had subsequent effects on a key clock process. Nevertheless, these results do have one novel aspect to them: the shorter periods associated with the increased mitochondrial dosage. There are few reports in the literature that cite shorter periods for organisms; most treatments lead to longer periods (lithium, D20, etc.), if any. This is to be expected if inhibitors are employed since they normally decrease the rate of some biochemical function. Therefore, the “increased clock rate” is an anomaly and could be indicative of a different kinetic role for the mitochondria than just supplying ATP for a clock process. The effect of mitochondria1 dosage also brings up a second point, namely, that almost all of the phase shift chemical treatments that are effective are metabolic inhibitors. There have been no reports yet of some type of “magic bullet” compound that would shift a clock mechanism without inhibiting severely some area of metabolism. Although light and temperature pulses are effective phase shifters, we do not know their primary biochemical effects in cells. Therefore, we are still looking for the key effector or the key pathway. It is proposed here that some type of dosage study may provide clues to the clock mechanism, by giving examples of treatments that speed up clocks rather than the numerous examples that can slow them down, directly or indirectly. These dosage studies can be any type, ranging from the use of animals or plants that are trisomic or having partial chromosomal duplications to those that have been genetically engineered to have an extra dose of a gene or small region. Perhaps this type of study could implicate key steps that are poised to make clocks run faster or slower as opposed to the studies now that show slower clocks associated with some severe inhibition listed somewhere in a metabolic wall chart. In summary, the studies reported here indicate that (a) a pulse of a mitochondria1 inhibitor, oligomycin, leads to a phase shift due to its effect on the mitochondria and not due to effects on any other area of cellular metabolism; (b) partial blocks in
Chronobiol In(. Vol. 9, N o . 3, 1992
230
S. BRODY
mitochondria1 metabolism, due either to mutations or inhibitors, surprisingly lead to shorter periods and increased levels of mitochondria; and (c) in Neurosporu, the mitochondria may play a direct role in the clock mechanism, and not just a “supporting” role, i.e., providing energy to run a clock. Acknowledgment: The author thanks Khoosheh Khomenian Gosink and Jennifer Germain for performing many of these experiments and Pat Lakin-Thomas and Gary Cot6 for comments on this manuscript.
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
REFERENCES 1. Edmunds LN Jr. Cellirlur and rnolwzrlur huws o/hiologicul clocks. Berlin, Heidelberg: Springer-Verlag, 1988. 2. Bunning E. The ph,vsiological clock. New York: Springer-Verlag. 1967.
3. Pittendrigh CS. Circadian oscillations in cells and the circadian organization of multicellular systems. In: Schmit FO, Worden FGs, eds. NeirroscienccJs:third stirdyprogrum. Boston: MIT Press, 1974:43758. 4. Frelinger JG. Motulsky H. Woodward DO. Effects of chlorarnphenicol on the circadian rhythm of Nezrrospora crussa. Pluni Physiol I976;58:592. 5. Lakin-Thomas P, Cot6 G, Brody S. Circadian rhythms in Neirrosporu cras.su: biochemistry and genetics. Crii Rev Microhiol 1990; 17:365-4 16. 6. Mattern DL, Forman LR, Brody S. Circadian rhythms in Nenrosporu crus~u:a mutation affecting temperature compensation. Proc Nut1 Acud Sci L‘SA 1982;79:825-9. 7. Dieckmann C, Brody S. Circadian rhythms in N(iitro.sporu crus.w: oligomycin-resistant mutations affect periodicity. Science 1980207:896. 8. Brody S. Dieckmann C, Mikolajczyk S. Circadian rhythms in Noirrosporu cru.$sc~: the effects of point mutations on the proteolipid portion of the mitochondrial ATP synthetase. Mol G m Gene! 1985;200:155-61, 9. Nakashima H. Effects of respiratory inhibitors on respiration, ATP contents, and the circadian conidiation rhythm of ,Vcwrosporu crussa. Plant Phj.siol 1984;76:6 12. 10. Schulz R, Pilatus U, Rensing L. O n the role of energy metabolism in Nezrrospora circadian clock function. Clironobiol In1 1985;2:223. 11. Lowry OH, Rosebrough NH, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. JBiol Chem 1951:193:265-75. 12. Brody S, Mikolajczyk S. Neitrosporu mitochondria contain an acyl carrier protein (ACP). Eirr J Biochm 1988; I73:353-9. 13. Kuiper MTR, Akins RA, Holtrop M, de Vries H, Lambowitz AM. Isolation and analysis of the Neirrospora crus.su Cyt-21 gene. .I Biol Chrrn 1988;263:2840-7. 14. Brody S, Mikolajczyk S. De novo fatty acid synthesis mediated by ACP in N~ztro,sporacra.s.w mitochondria. Eirr J Biochem 1990; 187:43 1-7. 15. Nakashima H. A liquid culture method for the biochemical analysis of the circadian clock of Nezrrosporu crussu. Plant Cell Phvviol 198 I :22:23 I .
Chronobiol Ini. Vol. 9, N o 3. 1992
Vol. 9. No. 3, pp. 222-230 0 1992 International Society of Chronobiology
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Circadian Rhythms in Neurospora crassa: The Role of Mitochondria
Summary: Energy metabolism and mitochondria have been discussed with respect to their role in the circadian rhythm mechanism for some time. Numerous examples of inhibitors that affect the mitochondria of plants and animals and microorganisms are known, which cause large phase shifts in the rhythms of these organisms. Analogous studies on the role of mitochondria in the Neiirosporu circadian rhythm mechanism habe also been reported and summarized. This communication differs from previous studies on other organisms in that it will focus on two lines of evidence derived from studies on Nczrrosporu strains carrying mutations affecting the mitochondria. (a) Strains whose growth rate is resistant to oligomycin (01;') owing to an altered protein in the F, sector ofthe mitochondrial ATPase, showed no phase shifts when pulsed with oligomycin. Control strains (oli') showed large phase shifts when pulsed with oligomycin. This indicates that the phase-shifting effect of oligomycin is due to the direct inhibition of the mitochondrial ATPase and not some side effect of this inhibitor. (b) In Nairosporu, many different strains are known that carry mutations in the nuclear o r mitochondrial genome that affect mitochondrially localized proteins. Some of these, such as di', [ M I - 3 ] ,or cyu-5. showed shorter ( - 19-h) periods compared with the normal (21.5-h) period. Others showed little or no change in period. Those mutant strains exhibiting shorter periods also contained ~ 6 0 % more mitochondria1 protein per gram total protein in extracts compared with the normal strains. Assays of the level of a mitochondrial-specific protein, acyl carrier protein, showed that the cellular content of this protein was approximately doubled. A parallel set of studies on the effects of antimycin or chloramphenicol on Neiirosporu demonstrated that these inhibitors also produced shorter periods as well as increased amounts of mitochondrial proteins. These two new lines of evidence may be interpreted to indicate that in Neurosporu either some part of the oscillator is localized to the mitochondria and/or that mitochondria exert their effect on the clock mechanism through their effects on biosynthetic pathways or by their contribution in determining ion gradients. Key Words: Mitochondria-Neziro.~poru crussu-Circadian rhythm.
Received December 23, 1991; accepted February 17, 1992. Addresscorrespondence and reprint requests to Dr. S. Brody at Department of Biology, 01 16, University of California, San Diego, 9500 Gilman Dr., La Jolla, CA 92093-01 16, U.S.A. Presented in part at the 20th International Conference on Chronobiology, June 16-21, 1991, Tel Aviv, Israel.
222
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
MITOC’HONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
223
Inhibitors of mitochondrial function have long been known to lead to phase changes in many organisms. Many of these effects have been listed in books by Edmunds ( 1 ) and also by Bunning (2). In plants, these inhibitors have been given as pulses in the transpiration stream, and in microorganisms, they have been given as pulses in the growth medium. The effects of anoxia have also been reported for the Drosuphilu rhythm (3), but these effects may not be specific to mitochondria since there are a significant number of oxygen-requiring biosynthetic and degradative reactions in cells. Most of the mitochondrial inhibitors that have effects on the rhythms of these different organisms lead to phase changes, and phase-response curves are known for most of these compounds. The phase changes produced by pulses of these inhibitors are generally produced by concentrations of these inhibitors that are not unusually high or toxic to cells. However, the question of how much of an inhibitor is truly removed at the end of the pulse has rarely been examined in these studies. The inhibitors used are primarily those that affect energy metabolism, such as uncouplers or azide, cyanide, etc. Some of the inhibitors that have been uncritically employed and reported unfortunately have effects other than just on mitochondrial energy production. The group of inhibitors used does not include, in general, inhibitors of other mitochondrial reactions such as the arginine biosynthetic pathway or mitochondria1 protein synthesis. Only in Neurosporu has chloramphenicol been shown to have an effect on the period of the rhythm (4).As more is known about the capacity of mitochondria to make more than just ATP, it may well be that other inhibitors will be employed, and/or mutant strains with deficiencies in a given reaction will be usefully employed. The bulk of our knowledge about mitochondria and rhythms is due to phase shifts produced by inhibitors, whereas only a few studies reported period changes. Although it is widely acknowledged that phase shifts produced by chemicals could be due to effects on an input pathway to the oscillator, and not necessarily on the oscillator mechanism itself, it should be pointed out the period changes due to inhibitors could also be due t o effects on an input pathway, and not necessarily on the oscillator mechanism itself.
THE CIRCADIAN RHYTHMS OF NEUROSPORA The circadian rhythm of the fungus Neurosporu crussa is routinely assayed by its expression as a conidiation (asexual spore formation) rhythm on agar medium. This complex developmental process is triggered by a “clock mechanism” at some phase of the circadian cycle. The Nezirosporu rhythm shares the basic properties common to circadian rhythms of most organisms: (a) The period of the rhythm is close to but not equal to 24 h; (b) the rhythm is endogenous and self-sustaining; (c) the phase of the rhythm can be changed by pulses of light or temperature or entrained; and (d) the period of the rhythm varies little with ambient temperature, i.e., it is temperature compensated. A number of mutations that alter the rhythm properties have been reported and reviewed recently ( 5 ) . Mutations affecting the free-running period have been reported at many loci, and the structure of the.fiq gene has been reported.
224
S. BRODY
Mutations affecting the light response, such as poky and others, are known, and mutations leading to a loss of temperature compensation (6) have been reported also.
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
PREVIOUS STUDIES ON NEUROSPORA MITOCHONDRIA AND RHYTHMS Three types of results have been found in Neurospora. First, there are a number of inhibitors that lead to phase shifts when given as a pulse. Second, there are inhibitors that lead to a change in the period when given continuously. Third, there are mutations that affect mitochondria and lead to changes in period. An example ofthis latter type of mutant strain was found to be an oligomycin-resistant mutant, whose period was 18-19 h (7). Other examples are given below. Chemicals that lead to period changes in Neurospora are well known and are gathered together in a review ( 5 ) . One of these, chloramphenicol, leads to a 19-h period. Others, such as phenylethanol, also lead to shorter periods. However, the bulk of the work relating mitochondria and circadian rhythms in Neurospora comes from studies on inhibitor pulses. These inhibitors are antimycin A, azide, carbonyl cyanide chlorophenol, cyanide, dicyclohexylcarbodiimide, N-ethylmaleimide, and oligomycin. As in all studies on chemical pulses that lead to stable phase shifts, the questions often remain: What is the target(s) for any given inhibitor? How do we know for sure that the primary effect of the inhibitor leads to the observed phase shift, and not some secondary or side effect of the inhibitor? Although it can be argued that it is unlikely that all of the different mitochondrial inhibitors have the exact same side effect, it is more important to actually obtain experimental evidence on this point, rather than just argue it. Therefore, a study of the response of Neuruspora to oligomycin, a mitochondrial inhibitor, was done employing two strains, one sensitive to oligomycin and one resistant to it. These resistant mutations all affected the structure of a small, hydrophobic subunit in the Fo section of the mitochondria1 ATPase (8), a subunit thought to be involved in proton (H') transport into the mitochondria.
RESULTS Oligomycin Pulse Studies Figure 1 shows the phase-response curve to oligomycin for disks of Neurospora incubating in a liquid shaker culture. Similar phase-response curves (Type 0) have been previously reported by two other groups (9, lo), although there are some differences between the data obtained by the three groups with respect to the size of the phase shifts and the exact shape of the curve. In the cultures that were assayed to generate the data for Fig. 1 (top), the disks were extensively washed to get rid of the oligomycin. Upon transfer to the growth tube medium, after the oligomycin pulse, the disks gave rise to cultures growing at a rapid ( 1.3 mm/h) rate, a rate identical to that in cultures not pulsed with oligomycin. The immediate rapid growth rate indicates that the bulk of the oligomycin was no longer present after the pulse, even though some type of additional data, such as the complete loss of labeled oligomycin,
UironobiolInl. Vol. 9, No. 3, I992
MITOCHONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
225
Oligomycin Sensitive
5-
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
4-
FIG. 1. Mycelial disks of Neurosporu ( 1 5 ) in a D/D liquid shaker culture were pulsed with oligomycin for 2 h. The oligomycin (10 mg/ ml ethanol) was added to a final concentration of 10 pg/ml. The disks were washed and then placed on race tubes in D/D. Each time point represents the average of five race tubes. The phase relative to controls (ethanol only) was calculated for days 3-5 and averaged. Each point has a standard deviation for the phase difference of -2 h. Top: Phase-response curve for olis strain. Bottom: Phase-response curve for oli' strain.
0-
I
4 4
Oligomycin Resistant U-
d
0
would have been more definitive. The previous studies (9,lO) do not report the growth rates immediately after the pulse and transfer, so it is not clear how much of this lipid-soluble inhibitor was still present. This could account for the differences reported by the three groups. Figure 1 (bottom) shows the identical experimental conditions to Fig. 1 (top), with the exception that the oligomycin-resistant strain (bd csp- 1 oli') was employed. Few, if any, reliable effects on the phase of these cultures were observed. Higher concentrations of oligomycin (data not shown) also gave no effect on the phases of the cultures. These data indicate that oligomycin does not lead to phase shifts of Neuuospoua by any side effect in these cultures. Rather, it points out that oligomycin leads to the observed phase shifts by its primary effect on the mitochondria1 ATPase. The inhibition of the proton transport into the mitochondria by oligomycin would be expected to inhibit the synthesis of ATP and in turn affect many other cellular procedures, such as ionic gradients, membrane potentials, etc.
Chronohiol Inl, Vol. Y, N o . 3, 1992
S. BRODY
226
TABLE I . &/ficts on llial shorfcn tlic 22-11 pcriod Mutant 011
[MI-3] 1 vu-5 c 1.b-2
Period (h)
Chemical
Period ( h )
18-20 18-19
Chloramphenicol Antimycin Aurovcrtin Phenylethanol
19
19 18
19 19 18
Citations are given in ref. 5
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Studies on Amount of Mitochondria1 Protein and Period A number of mutant strains with altered mitochondria also exhibited effects on the circadian rhythms of N. crassa: oh*, [M1-3];cyu-5 all had slower growth rates but faster (18-19 h) clocks (5). In addition, various mitochondrial inhibitors such as antimycin or aurovertin had the same effects on growth rate and clock rate as these mutations (Table 1). Since the mitochondrial targets for these inhibitors and mutations were so different (Fig. 2), yet they all had the same clock effect, it was of interest to see what these mutations and inhibitors had in common. A study ofthe amount of mitochondrial protein per unit cell mass was undertaken as a first step. The amount of mitochondria was measured as the percentage of mitochondrial protein in the total protein in a cell-free extract as follows: Liquid cultures of N m r o spora were grown at 22°C in minimal medium containing either 2% glucose or 1%) acetate as the carbon source and were shaken until the exponential phase was reached. They were harvested by vacuum filtration through Whatman filters in a
ATPase Complex V
Complex IV
[M 1-31,
Oxidase
Venturicidin,
FIG. 2. A schematic diagram of mitochondrial metabolism indicating approximate sites of inhibitors by chemicals and/or mutations.
MITOCHONDRIA IN NEUROSPORA CIRCADIAN R H Y T H M
227
TABLE 2. Effectsof muiations and Erowth conditions on amount of‘mitochondriul moirin Strain
Growth medium
hd CSP- I hd c.sp- 1 oli‘ ( 16-16) hd C.SI)- I [MI-3] bd csp- I cyu-5 hd c.sp- I hd CSP- I !Id C‘sp-I
Glucose Glucose Glucose Glucose Acetate Glucose + chloramphenicol Antimycin
Protein in mitochondrial fraction (70) 4.0 t 0.5 7.5 f 1.2 6.3 ? 1.0 6.5 t 0.87 5.6 i 0.4 6.3 f 0.6 10.7 t 0.7
(9) (6) (8) (3) (20) (4) (8)
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
Numbers in parentheses are individual samples assayed.
Bucher funnel and the wet weights of the Ncwosporu mats were recorded. Samples from identical cultures were combined to yield a weight of at least 1 g, if necessary. The Ncwrosporu was disrupted with a bead beater for 1 min in a mitochondrial isolation buffer (0.25 A4 sucrose, 1 m M ethylenediaminetetraacetate, I0 m M Tris, pH 7.2). The suspension was then spun at 600 g for 5 min to remove cell debris. The supernatant was poured through cheesecloth and the total volume was recorded. One milliliter was saved for the Lowry protein assay ( 1 1). This “homogenate” represents the total protein in the sample. The supernatant was then spun at 10,000 g for 20 min to pellet organelles. The supernatant was poured off and the pellet was resuspended with a glass homogenizer in 5 ml of the mitochondrial isolation buffer. This was spun at 1,200 g for 10 min. The pellet was discarded and the supernatant was spun at 10,000 g for 20 min. This pellet was resuspended in 1 ml of the mitochondrial isolation buffer and represents the protein contained in the “mitochondrial” fraction. Lowry assays ( 1 1) were performed to determine the amount of protein contained in the homogenate and mitochondrial fractions using bovine serum albumin as a standard. The percentage mitochondria was calculated as a fraction of the amount of total protein (i.e., homogenate). The amount of protein in the mitochondrial fraction was 4.0 f 0.5% ofthe protein found in a total crude homogenate of Neurospora, employing the breakage procedure given above. This value was increased -90% by a mutation affecting ATPase (olir), 60% by a maternally inherited mutation affecting cytochrome oxidase ([MZ-3]), and by another mutant with low levels of cytochrome oxidase (cyu-5).The mitochondrial level was increased by -40% by growth of Nezirosporu on acetate as a carbon source instead of glucose and was increased 60% by chloramphenicol or 150% by the mitochondrial inhibitor antimycin (Table 2). Three additional studies of the amount of mitochondrial protein were also performed (data not shown). The first was in a variety of strains, such as (a) strains with different alleles at the oli‘ locus: (b) the [MI-5] strain, similar to [MI-3];( c )different isolates of the hd c.sp- 1 strain and of the hd strain; and (d) several wild-type strains such as RL3-8A and 740R23IVA. The presence or absence of the hd or csp-I markers had no measurable effect on the amount of mitochondrial protein observed. The [MI-5]strain gave the same values as the [MI-3] strain, and the values obtained
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
228
S. BRODY
for the different oli‘ strains correlated with their degree of oligomycin resistance, the least resistant showing little changes from the bd csp- 1 strain. Second, the amount of mitochondrial proteins were assayed for cultures grown on solid medium, such as bd csp and bd csp oh‘, and for bd csp grown in the presence of antimycin. The change from liquid to solid medium did not affect the values obtained. Third, the level of a mitochondrial-specificprotein (1 2), acyl carrier protein (ACP), was assayed by comparing the amount of radioactive pantothenate in this protein in a crude homogenate versus that found in fatty acid synthetase. This technique, sodium dodecyl sulfate polyacrylamide gel electrophoresis, showed that the oli‘ mutation increased the level of the ACP by -90%, comparable with the studies on amounts of protein (data not shown).
DISCUSSION When viewed together, these results can be interpreted as follows: Mitochondria Clock effect increased amount --* “faster clock” decreased rate + phase shift (inhibitor pulse) The increased amount could lead to increased rates of mitochondrial processes, and the “faster” clock refers to a shorter period. Of course, the shorter period could be due to either a decreased amplitude or a “faster” clock, which implies a rate change. Therefore, one of the possible interpretations would be that an increased rate of mitochondria1processes leads to an increased clock rate. The decreased rate of mitochondrial processes due to an inhibitor pulse is a temporary one and may be for only certain processes as well. It is well known that uncouplers decrease the rate of ATP synthesis, but increase the rate of electron transport and respiration, for example. These two findings taken together do not necessarily pinpoint the exact step or pathway in the mitochondria that leads to these clock changes. This is because the biochemical reactions of the mitochondria are so intimately tied in with the rest of the cellular metabolism that effects on one have serious “ripple” effects on the other. This interrelationship is true at many levels: First, any change in the rate of ATP generation could be expected to change the rate of the ATP-requiring reactions. These could be a sizable portion of the cellular machinery. Second, an increase in the amount of mitochondria in the cell would alter the ratio of certain isozymes found in both mitochondria and other cellular compartments. This possible imbalance in enzymes would also be true for any pathway, such as arginine synthesis or phospholipid synthesis (phosphatidyl serine decarboxylase),where only some of the steps are localized to the mitochondria. Third, the increase in mitochondrial levels observed here may very well be a reflection of the general increase in the levels of transcription of many genes, both nuclear and mitochondrial. A specific example of this is given by the fivefold increase in the mRNA for a small mitochondrial ribosomal protein in response to impaired mitochondrial function ( 1 3). It seems quite probable that some nonmitochondrial enzymes may also be increased in response to the same generalized cellular signal, so that a balanced metabolism can occur. It would be of interest
C‘hronobiolhi.Vol 9, No 3, 1992
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
MITOCHONDRIA IN NEUROSPORA CIRCADIAN RHYTHM
229
to see if the genes coding for gluconeogenic enzymes, for example, have some promoter sequences similar to those employed for the genes coding for mitochondrially localized proteins. There have been other studies on Neurosporu that are pertinent to this discussion. Nakashima (9) has studied ATP level, respiration, and ATPase inhibitors and their possible role in clock function. His conclusion, i.e., clocks still “run” even though respiration can be very low, does not negate the findings presented here. These two lines of evidence can be reconciled by restating some of the considerations listed above: Mitochondria do more than just electron transport and respiration. They have a role in Ca2+metabolism, in certain amino acid biosynthetic pathways, and even in fatty acid synthesis (14). It seems plausible that even when respiration is quite low, ion gradients are still maintained across the membranes and biosynthetic reactions are still proceeding at some rate. Other studies on inhibitor pulses have been summarized (5). These data on Neurosporu and the previous studies performed on other organisms suggest a role for the mitochondria in the clock mechanism. The studies, of course, do not allow a definitive enough interpretation as to whether the role is a primary or a secondary one. A primary role for the mitochondria would be if the entire clock mechanism, or some significant portion of it, were localized in the mitochondria. A secondary role would be if clocks merely required ATP to “run” or if effects on mitochondrial metabolism just had subsequent effects on a key clock process. Nevertheless, these results do have one novel aspect to them: the shorter periods associated with the increased mitochondrial dosage. There are few reports in the literature that cite shorter periods for organisms; most treatments lead to longer periods (lithium, D20, etc.), if any. This is to be expected if inhibitors are employed since they normally decrease the rate of some biochemical function. Therefore, the “increased clock rate” is an anomaly and could be indicative of a different kinetic role for the mitochondria than just supplying ATP for a clock process. The effect of mitochondria1 dosage also brings up a second point, namely, that almost all of the phase shift chemical treatments that are effective are metabolic inhibitors. There have been no reports yet of some type of “magic bullet” compound that would shift a clock mechanism without inhibiting severely some area of metabolism. Although light and temperature pulses are effective phase shifters, we do not know their primary biochemical effects in cells. Therefore, we are still looking for the key effector or the key pathway. It is proposed here that some type of dosage study may provide clues to the clock mechanism, by giving examples of treatments that speed up clocks rather than the numerous examples that can slow them down, directly or indirectly. These dosage studies can be any type, ranging from the use of animals or plants that are trisomic or having partial chromosomal duplications to those that have been genetically engineered to have an extra dose of a gene or small region. Perhaps this type of study could implicate key steps that are poised to make clocks run faster or slower as opposed to the studies now that show slower clocks associated with some severe inhibition listed somewhere in a metabolic wall chart. In summary, the studies reported here indicate that (a) a pulse of a mitochondria1 inhibitor, oligomycin, leads to a phase shift due to its effect on the mitochondria and not due to effects on any other area of cellular metabolism; (b) partial blocks in
Chronobiol In(. Vol. 9, N o . 3, 1992
230
S. BRODY
mitochondria1 metabolism, due either to mutations or inhibitors, surprisingly lead to shorter periods and increased levels of mitochondria; and (c) in Neurosporu, the mitochondria may play a direct role in the clock mechanism, and not just a “supporting” role, i.e., providing energy to run a clock. Acknowledgment: The author thanks Khoosheh Khomenian Gosink and Jennifer Germain for performing many of these experiments and Pat Lakin-Thomas and Gary Cot6 for comments on this manuscript.
Chronobiol Int Downloaded from informahealthcare.com by McMaster University on 12/29/14 For personal use only.
REFERENCES 1. Edmunds LN Jr. Cellirlur and rnolwzrlur huws o/hiologicul clocks. Berlin, Heidelberg: Springer-Verlag, 1988. 2. Bunning E. The ph,vsiological clock. New York: Springer-Verlag. 1967.
3. Pittendrigh CS. Circadian oscillations in cells and the circadian organization of multicellular systems. In: Schmit FO, Worden FGs, eds. NeirroscienccJs:third stirdyprogrum. Boston: MIT Press, 1974:43758. 4. Frelinger JG. Motulsky H. Woodward DO. Effects of chlorarnphenicol on the circadian rhythm of Nezrrospora crussa. Pluni Physiol I976;58:592. 5. Lakin-Thomas P, Cot6 G, Brody S. Circadian rhythms in Neirrosporu cras.su: biochemistry and genetics. Crii Rev Microhiol 1990; 17:365-4 16. 6. Mattern DL, Forman LR, Brody S. Circadian rhythms in Nenrosporu crus~u:a mutation affecting temperature compensation. Proc Nut1 Acud Sci L‘SA 1982;79:825-9. 7. Dieckmann C, Brody S. Circadian rhythms in N(iitro.sporu crus.w: oligomycin-resistant mutations affect periodicity. Science 1980207:896. 8. Brody S. Dieckmann C, Mikolajczyk S. Circadian rhythms in Noirrosporu cru.$sc~: the effects of point mutations on the proteolipid portion of the mitochondrial ATP synthetase. Mol G m Gene! 1985;200:155-61, 9. Nakashima H. Effects of respiratory inhibitors on respiration, ATP contents, and the circadian conidiation rhythm of ,Vcwrosporu crussa. Plant Phj.siol 1984;76:6 12. 10. Schulz R, Pilatus U, Rensing L. O n the role of energy metabolism in Nezrrospora circadian clock function. Clironobiol In1 1985;2:223. 11. Lowry OH, Rosebrough NH, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. JBiol Chem 1951:193:265-75. 12. Brody S, Mikolajczyk S. Neitrosporu mitochondria contain an acyl carrier protein (ACP). Eirr J Biochm 1988; I73:353-9. 13. Kuiper MTR, Akins RA, Holtrop M, de Vries H, Lambowitz AM. Isolation and analysis of the Neirrospora crus.su Cyt-21 gene. .I Biol Chrrn 1988;263:2840-7. 14. Brody S, Mikolajczyk S. De novo fatty acid synthesis mediated by ACP in N~ztro,sporacra.s.w mitochondria. Eirr J Biochem 1990; 187:43 1-7. 15. Nakashima H. A liquid culture method for the biochemical analysis of the circadian clock of Nezrrosporu crussu. Plant Cell Phvviol 198 I :22:23 I .
Chronobiol Ini. Vol. 9, N o 3. 1992
