Journal of General Microbiology (1984), 130, 789-796.
Printed in Great Britain
789
Heterocyst Differentiation in Cylindrospevmum Iichenqovme: Studies on the
Role of Transcription
By S U S A N V A N D E W A T E R ? A N D R O B E R T D . S I M O N * $
Department of Biology, Unicersitj7 qf' Rochester, Rochester, New York 14627, U S A
(Receiwd K October 1983)
Heterocysts of the cyanobacterium Cji/indrospermum licheniforme occur at the ends of the
filaments. These cells, specialized for aerobic N fixation, synchronously differentiate after the
fragmentation of filaments grown in a medium free of combined nitrogen. This study has
examined the role of transcription during and after heterocyst differentiation. Autoradiography
of intact filaments pre-labelled for 3 h with [3H]uracil revealed that R N A synthesis occurred at
similar rates in vegetative cells, developing (pro)heterocysts and mature heterocysts. Since
mature heterocysts no longer divide and contain a similar amount of D N A as vegetative cells,
transcription must continue under conditions where replication no longer occurs. Rifdmpicin,
when added to fragmented filaments at concentrations that reduced R N A synthesis by 95 % in
all cell types, inhibited proheterocyst formation but not the final morphological step of
differentiation (pore plug deposition). However, chloramphenicol blocked all stages of
heterocyst differentiation. Since the average m R N A half-life was 29 min, and the rifdmpicininsensitive stage lasted 3-6 h, the later stages of differentiation may depend on long-lived
mRNAs.
INTRODUCTION
Several groups of the filamentous cyanobacteria produce heterocysts, cells specialized for
aerobic N z fixation (Heselkorn, 1978). In Cj,/indrosperrnurn licheniforme Kiitz. the heterocysts
are located at the ends of the filaments. When cultures of this organism growing in nitrogen-free
medium are briefly sonicated, existing heterocysts are detached and the filaments are broken
into fragments of 3-5 cells long whose terminal cells differentiate synchronously (Van de Water
& Simon, 1982). In the first round of differentiation, new proheterocysts appear after 9-12 h and
reach a maximum number at 15 h. Mature heterocysts form after 15 & 1 h, and 12-1 5 %of all cells
in the culture become morphologically mature heterocysts with characteristic pore plugs within
26 h. The sequence in which several heterocyst-specific traits appear following filament
fragmentation has been characterized. For example, nitrogenase activity increases in parallel
with the mature heterocyst frequency between 15 and 26 h (Van de Water & Simon, 1982).
Heterocyst differentiation may be irreversible. In most species, the mature heterocyst cannot
divide (Fritsch, 1945), although rare cases of heterocyst division or germination have been
reported (Ladha & Kumar, 1975). Using microsurgical techniques, Wilcox et al. (1973) showed
that the destruction of adjacent vegetative cells caused an early proheterocyst to regress and
resume vegetative division. However, similar surgical isolation of a late proheterocyst or mature
heterocyst no longer induced regression. Electron microscopy has shown that the nuclear and
thylakoid regions of vegetative cells are reorganized during heterocyst differentiation (Lang &
Fay, 1971). Mature heterocysts are filled with contorted membrane, and a nucleoid region with
its fibrous DN A-containing material is not visible, although chemical measurements indicate
t Present
address: 25291 I H 10 W , Lot no 2. San Antonio, TX 78252, USA.
of Biology, State llniversity of New York, Geneseo, New York 14454, USA.
3 Present address: Department
Abbrrrratrons G6PDH, Glucose-6-phosphcite dehydrogenase. 6PDH, 6-phosphogluconate dehydrogenase.
0023-1287'84~00C)1-0349$02 00 0 1984 SGM
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
790
S . VAN DE W A T E R AN D R . D . S I M O N
that heterocysts contain as much as DNA as vegetative cells (Simon, 1980). Little is known
about the capacity of mature heterocysts to synthesize macromolecules, or about the relationship
between the 'reversibility' of differentiation and the potential for mature heterocysts to produce
new RNA and proteins. Recent studies suggest that mature heterocysts can synthesize protein
(Janaki & Wolk, 1982), and Thomas (1972) has shown that heterocysts in old cultures of
Anabaena L-3 1 can resynthesize the phycocyanin which initially disappears during
differentiation.
There have been few studies of the relative importance of transcription during heterocyst
differentiation and in the functioning of mature heterocysts, although the complex changing
pattern of protein synthesis which takes place during heterocyst differentiation (Fleming &
Haselkorn, 1974) suggests that significant changes in transcription probably occur. This paper
reports the results of studies on transcription during proheterocyst and heterocyst
differentiat ion.
ME T HO D S
Culture conditions. An axenic isolate of Cylindrospermum licheniforme Kiitz., derived from strain B 1828 of the
University of Texas culture collection (UTEX) (Fisher & Wolk, 1976) was grown in the nitrogen-free medium of
Allen & Arnon (1955) diluted 16-fold (AA/l6). Growth conditions and procedures for heterocyst induction were as
described by Van de Water & Simon (1982).
Measurement of macromolecular synthesis. R N A synthesis was measured as the rate of incorporation of [5,63H]uracil [44.5Ci (1.65 TBq) mmol-I ; 1 pCi ml-1 in the growth medium] into material precipitable with 576
(w/v) TCA at 4 "C. TCA precipitates were collected on premoistened Whatman G F j C filters, and were washed
three times with cold 10% (w/v) TCA and once with 957; (v/v) ethanol. Radioactivity was determined by liquid
scintillation counting in 2 ml scintillation fluid (0.25 g M2POPOP, 4.0 g PPO per litre toluene). Uracil was also
incorporated into D N A , and this incorporation was measured as the amount of [3H]uracil remaining TCAprecipitable after incubation in 0.6 M-NaOH at 37 "C for 20 h. Samples were neutralized with HCl prior to TCA
precipitation. The rate of R N A synthesis was corrected for the fraction incorporated into D N A .
Incorporation of [ 'T]phenylalanine into TCA-precipitable material was used as a measure of protein synthesis.
Precipitates were treated with pronase in order to verify their protein nature. [ I-ITIPhenylalanine
(536 mCi mmol-' ; 19.8 GBq mmol-I) was added at 0.1 pCi ml-I to cultures buffered with 0.1 M-TES. pH 8.0.
The pronase (Calbiochem B-grade, nuclease free) was predigested (20 mg ml- in 0.01 M-TrisiHC1 pH 7.6) for 3 h
at 37 "C in order to degrade any endogenous nucleases.
Solutions of inhibitors of macromolecular synthesis were freshly prepared and filter-sterilized (0.45 pm
Millipore HAWP-1300) for each experiment. To study the effects of inhibitors on macromolecular synthesis,
[3H]uracil was added to duplicate cultures at 1 pCi m l - l , and inhibitors were added to one of the cultures 3060 rnin later. Samples were removed at various times from both cultures. Inhibition of D N A and R N A synthesis
was determined by comparing the linear rates of incorporation between 3 and 6 h after inhibitor addition.
Mensurement o f m R N A h a l f - l ~ The
f ~ . half-life of m R N A was calculated from the rate at which [3H]uracil was lost
from TCA-precipitable material after pie-labelled cells were treated with rifampicin, an inhibitor of R N A
synthesis. f3H]Uracil was added to a culture of C. lichen(forme at 1 pCi m l - l , and rifampicin (10 pg m l - I ) was
added 60 min later. Samples were removed at various times, precipitated in TCA, and collected on GFIC filters.
Radioactivity was determined by liquid scintillation counting. The data were computer-fitted to a first-order
decay curve: fraction c.p.m. remaining = eLtwhere k is the decay constant and f is the time in min.
As an additional measure of mRNA half-life, the decline in the rate of protein synthesis was determined after
the addition of an inhibitor of R N A synthesis (Smith, 1979). ['TIPhenylalanine (536 mCi mmol- ;
19.8 GBq mmol- I ) was added to a growing culture at 0.5 pCi ml- I , and rifampicin (10 pg ml- I ) was added after
60 min. Samples were removed at various times, precipitated in cold 57; (w/v) TCA, and collected on GFIC filters.
Radioactivity was determined by liquid scintillation counting.
R N A qxthesis in djfjerent cell types: ,fi/amwt autorariiographj~. Filaments were labelled with [3H]uracil
(10 pCi m l - ! ) for 3 h, with or without I h pretreatment with rifampicin (20 pg m1-I). Cell samples were
centrifuged at room temperature for 5 min at IOOOg, and resuspended in growth medium to lo8 cells ml- I . Drops
of these concentrated cell suspensions were fixed on subbed slides [0.17: (w/v) gelatin, 0.01 7" (wiv) chromic
potassium sulphate] with Carnoy's fixative (ethanol/chloroform/acetic acid, 6 : 3 : I , by vol.), and air dried at 40 "C.
The slides were dipped in acid alcohol (acetic acid!ethanol, 3 : I , v/v) for 5 min at room temperature, twice in 5 S ,
(w/v) TCA for 5 min at 4 'T,and then dehydrated through an ethanol series (500/",7096. so:;, 95",, by vol.) before
being air dried. After the acid alcohol treatment, control slides were treated with 0.20, (wlv) RNAase at 37 "C for
1 h for the removal of R N A , or with So/, (w/v) TCA at 90 "C for 15 min for the removal of all nucleic acids. The
slides were coated with Kodak NTB-2 radioautographic emulsion, and kept at 4 "C for 1 week to 3 months. They
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
79 1
Transcription and heterocyst differentiation
were developed using Kodak D19 developer (2.5 min), Kodak indicator stop bath (20 s), and Kodak rapid fix
( 5 min). Grains were scored under 650 x magnification with phase contrast optics on a Zeiss standard microscope.
Data were expressed as the number of grains per cell per day of emulsion exposure, and are given
the 95%
confidence limits which were computed as tous ( n - 1) x sin, where s = standard deviation, n = number of cells
scored, and t o u s ( n - 1) = two-sided t distribution.
Enzyme assays. Nitrogenase activity was measured using an acetylene reduction assay (Stewart er al., 1968). The
first two enzymes of the pentose phosphate shunt [glucose-6-phosphate dehydrogenase (G6PDH, EC 1.1.1.49)
and 6-phosphogluconate dehydrogenase (6PDH, EC 1.1.1.44)]were measured as previously described (Van de
Water & Simon, 1982).
RESULTS
Macromolecular synthesis
Cells of an exponentially growing culture incorporated up to 20% of uracil added to the
medium into TCA-precipitable material in 48 h, although the most rapid incorporation
occurred over the first 8-10 h. Uracil was slowly incorporated into DNA at a constant rate for at
least 24 h. During the first 8-10 h, 1-2% of the [3H]uracilwas found in DNA, increasing to 10%
at 24 h. Uridine, thymine, thymidine and cytosine were poorly incorporated into nucleic acids.
[ 14C]Phenylalaninewas specifically incorporated into proteins; up to 50% of the labelled
phenylalanine was TCA-insoluble after 24 h, and > 90% of this radioactivity was solubilized by
p ronase.
Whole filament autoradiography showed that vegetative cells, proheterocysts and heterocysts
incorporated uracil with approximately equal efficiency (Table 1). Treatment with RNAase or
hot TCA reduced the number of grains per cell by 9504. DNA synthesis, measured as the
difference between the grain counts after RNAase treatment and those after hot TCA
treatment, accounted for only 1-2% of the total incorporation. Although proheterocysts were
not morphologically distinguishable until 9 h after filament fragmentation, the terminal cells
destined to become proheterocysts incorporated [3H]uracilslightly more rapidly than intercalary
cells of the same filament.
Inhibitors qf' mucromolecular synthesis
Rifampicin, 5-fluorouracil and proflavin all effectively inhibited RNA synthesis (Table 2),
and had no short term effect on the uptake of labelled precursors (data not shown). Rifampicin
inhibited RNA synthesis equally well in vegetative cells, developing proheterocysts, and mature
heterocysts (Table 1). The effect of rifampicin and 5-fluorouracil on RNA synthesis in intact
filaments was irreversible after 15 min. Although rifampicin, actinomycin D, and proflavin are
destroyed by visible light, the effects of these compounds were not changed when cultures were
shielded from that portion of the visible spectrum which the inhibitors absorb.
DNA synthesis was totally blocked by actinomycin D, mitomycin C and novobiocin, but not
by nalidixic acid or 2'-deoxyadenosine (Table 2). Novobiocin was the only DNA synthesis
Table 1 . Effect of rlfampicin on R N A qxthesis in zwrious cell types of' Cylindrosperrnum
licheniforme
[ \H]Uracil incorporated (grains per cell d - I )
Cell type
Control
Intact filaments*
Heterocysts
Vegetative cells
Fragmented filaments?
Proheterocysts (terminal cells)
Vegetative cells (intercalary cells)
* Exponentially growing cells in
10.8
10.9
k 0.57
0.19
13.6 f 0.49
8.6 k 0.37
Rifampicin treated Percentage
(20 pg ml- I )
inhibition
0.345
0.370
0.087
0.014
96.8
96.6
0.358 i 0.060
0.180 f 0.010
97.4
97.9
3 d cultures
t Analysed 12 h after the filaments were fragmented. a period of maximal differentiation.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
792
S . VAN DE W A T E R AND
R. D. S I M O N
Table 2. Effect of' inhibitors on macromolecular synthesis in intact $laments of' Cvlindrospermum
licheniforme
Percentage inhibition
of synthesis
Inhibitor
Concn*
Rifampicin
2
5
5
50
5
50
5
50
5
50
5
50
5
100
10
100
5-Fluorouracil
Proflavin
Actinomycin D
Mitomycin C
Nalidixic acid
Novobiocin
2'-Deox yadenosine
* pg ml-' except for actinomycin D
f
3
RNA
DNA
80
95-100
89
95-100
41
100
18
80
74
100
54
61
6
21
33
60
100
100
81
100
76
100
0
31
100
100
0
78
ot
100
0
46
(p~).
Novobiocin had no effect on RNA synthesis for 6 h. Thereafter, inhibition was nearly complete.
inhibitor which at low concentrations did not have any immediate effect on RNA synthesis. The
effect of novobiocin on DNA synthesis remained reversible for up to 3 h.
Chloramphenicol at 50 pg ml- completely inhibited protein synthesis, and had no effect on
RNA synthesis for at least 12 h.
Effect of inhibitors on differentiation
To determine whether continued RNA synthesis was required for heterocyst differentiation,
rifampicin (20 pg ml- l ) was added at various times following filament fragmentation. At 30 h,
when heterocyst differentiation was complete in control samples, rifampicin-treated samples
were examined for the presence of proheterocysts and mature heterocysts. If rifampicin was
added during the first 12 h of differentiation, proheterocysts, which take 9-12 h to appear, did
not form. However, if rifampicin was added more than 12 h after fragmentation, a significant
percentage of existing proheterocysts continued differentiation to become mature heterocysts.
When compared to the plot of heterocyst frequency at the time of rifampicin addition, the plot of
heterocyst frequency at 30 h after the time of rifampicin addition is congruent but shifted by 36 h (Fig. 1). Similar results were obtained when rifampicin concentrations were varied from 10
to 100 pg ml- l , or when 5-fluorouracil (20 pg ml- l ) was used, although 5-fluorouracil caused
vegetative cells to disintegrate after 12 h of exposure. The heterocyst frequencies in rifampicintreated samples were similar when measured at 24,30 or 48 h after drug addition (Fig. 1). Thus,
rifampicin did not merely slow development. Chloramphenicol (50 pg ml-I) halted the
differentiation of heterocysts at all stages.
To determine whether transcription must continue for G6PDH, 6PDH and nitrogenase
activities to increase during heterocyst development, rifampicin (20 pg ml- ) or chloramphenicol(50 pg ml- I ) was added at various times during the first 24 h after filament fragmentation.
The activities (mg protein)-' of 6PDH or G6PDH remained constant or declined slightly
between an assay at the time of addition and a second assay at 30 h; i.e. if the data are plotted as
in Fig. 1, the rifampicin curve coincides with the control curve. Thus, synthesis of new enzyme
was rapidly inhibited (see measurements of mRNA half-life below), and no rifampicininsensitive period of 3-6 h was apparent in the increasing activities of these enzymes during
differentiation.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
793
Transcription crnd heterocyst dgferentiation
I
10
20
30
Time after filament fragmentation ( h )
I
1
I
100 150
Time (min)
50
1
I
200
Fig. 2
Fig. 1
Fig. 1. Effect of rifampicin on heterocyst differentiation in C . fichengforme.Samples were removed
from a culture at various times following filament fragmentation, and rifampicin was added at a final
concentration of 20 pg ml- The frequency of mature heterocysts in each sample was determined at the
time of rifampicin addition
and 30 h after fragmentation (0).
(a),
Fig. 2. Effect of rifampicin, 5-fluorouracil and chloramphenicol on nitrogenase activity in C.
licheniforme. Samples (10 ml) of cell suspension were placed in 50 ml Erlenmeyer flasks. Immediately
following the addition of rifampicin (2 pg m l - 1: A),5-fluorouraci1(20 pg ml- ; O),
or chloramphenicontained
col(50 pg ml- ; O ) ,acetylene was added at loo/, of the total flask volume. Control flasks
no inhibitors. The flasks were incubated in the light, and at various times 0-3 ml gas samples were
removed and analysed for nitrogenase-catalysed ethylene production by gas chromatography.
(a)
90 80 70 -
00
I
I
I
I
I
I
1
2
I
I
I
I
5
6
d
60 60
120
180
Time (min)
240
Fig. 3
3 4
Time (h)
Fig. 4
Fig. 3. Effect of rifampicin on t3H]uracil incorporation into TCA-precipitable material in intact
filaments of C . licheniforme. Filaments were labelled for 1 h with [3H]uracil, and then rifampicin
(10 pg ml- l ) was added. Radioactivity in TCA-insoluble material is presented as the percentage of
c.p.m. incorporated at 60 min. The data given are a composite of three separate experiments.
Fig. 4. Effect of rifampicin on [ “Tlphenylalanine incorporation into TCA-precipitable material in
intact filaments of C. licheniforme. Filaments were labelled with [ 14C]phenylalanine, and rifampicin
(10 pg ml- ’) was added at the time indicated by the arrow. The total residual capacity for protein
synthesis after the addition of rifampicin is designated by T.
Unlike the enzymes of the pentose phosphate shunt, nitrogenase activity of exponentially
growing, intact filaments was directly inhibited by rifampicin (Fig. 2). If the filaments were
incubated under N 2 rather than under air, rifampicin had a less severe effect on nitrogenase
activity. For example, 10 pg rifampicin ml- inhibited nitrogenase activity by 90% under air
but by only 10% under N 2 . Chloramphenicol had no effect on nitrogenase activity for at least
12 h, while 5-fluorouracil, at concentrations sufficient to inhibit RNA synthesis, had no
immediate effect on nitrogenase activity (Table 2, Fig. 2).
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
794
S . V A N DE W A T E R AND
R. D . S I M O N
Determination of m R N A half-life
Since the rifampicin-insensitive period during heterocyst differentiation could depend on
continued translation from long-lived mRNA, the average half-life of mRNA was determined.
The rate of decline in the amount of [3H]uracil in TCA-insoluble material induced by the
addition of rifampicin indicated that the half-life of unstable mRNA was approximately 29 min
(Fig. 3).
As an alternative approach to determining the half-life of functional mRNA, the decline in
the rate of protein synthesis following the addition of rifampicin was also measured (Smith,
1979). Three hours after the addition of rifampicin to a pre-labelled culture, the amount of
[’ “Clphenylalanine in TCA-precipitable material reached a plateau (Fig. 4). The difference
between the amount of incorporated phenylalanine at the plateau and the amount at any given
time declined exponentially. From this exponential decline, the half-life of functional mRNA in
3 d cultures was calculated as 29.5 and 31.7 min in two independent experiments.
DISCUSSION
As in other species of cyanobacteria, few labelled nucleotides or nucleosides were effective
precursors of macromolecular synthesis when added to the culture medium of C. licheniforme,
and uracil was the choice for labelling RNA (Glaser et al., 1973). On the basis of whole-filament
autoradiography, heterocysts and vegetative cells incorporated uracil at similar rates, and thus
probably synthesized RNA at similar rates (Table 1). This finding is in agreement with
observations that heterocysts continue to synthesize proteins (Janaki & Wolk, 1982). The
amount of RNA in heterocysts is several times that of vegetative cells (Simon, 1979), a result
expected for continued RNA synthesis in a non-dividing cell. Why transcription continues in
heterocysts under conditions where DNA accumulation stops is not known; however, the
dramatic changes which take place in the ultrastructural arrangement of the cyanobacterial
nucleoid during heterocyst differentiation (Lang, 1965; Lang & Fay, 1971) clearly do not make
the DNA inaccessible to proteins involved in RNA synthesis. Indeed, transcription probably
continues from the whole genome, since preliminary measurements of RNA complexity in
heterocysts and vegetative cells (data not shown) suggest that both types of cell transcribe about
70% of the total genome.
Rifampicin was used as an inhibitor of RNA synthesis because it had the least effect on DNA
synthesis, and no immediate effect on protein synthesis. The inhibition of nitrogenase by
rifampicin was probably a direct inactivation of the enzyme, since : (1) chloramphenicol had no
effect on nitrogenase; (2) the inhibition of nitrogenase activity by rifampicin was partially
relieved by incubation under N,, a treatment with no effect on transcriptional inhibition; and
(3) 5-fluorouracil, another RNA synthesis inhibitor, had no immediate effect on nitrogenase
activity. The protection from rifampicin afforded by performing the nitrogenase assay under N
would suggest that rifampicin may act to inhibit one of the heterocyst systems operating to
protect nitrogenase from O2 inactivation (Haselkorn, 1978). This direct inhibition of
nitrogenase, and the resulting nitrogen starvation, may explain the observations that long chains
of heterocysts form in the presence of low concentrations of rifampicin in Anabaena cj.lindrica
grown on agar plates (Wolk, 1975) and in a Cylindrospermum sp. following transfer from
nitrogen-containing to a nitrogen-free medium (Grover & Puri, 1979). However, the effect of
rifampicin may be more complex since in these latter two cases the final pattern of heterocyst
formation was also altered.
Even though rifampicin has a direct effect on nitrogenase, it is a suitable drug for examining
the pattern of transcription during heterocyst formation. Morphologically mature heterocysts
with a normal pattern can form in the absence of active nitrogenase, for example when
differentiation occurs under an argon atmosphere (Wolk, 1975). In addition, rifampicin
inhibition of the first 12-1 5 h of heterocyst differentiation occurs before active nitrogenase
appears (Van de Water & Simon, 1982). Rifampicin was not a general cellular poison because
protein synthesis continued for 3 h after the addition of the drug, treated cells remained intact
for at least 24 h, and the final 3-6 h of heterocyst differentiation occurred in the presence of the
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
Transcription and hcterocyst dgfjreerentiation
795
drug. Rifampicin insensitivity was not a result of decreased permeability of late proheterocysts
to the drug (e.g. due to the deposition of the heterocyst envelope around the vegetative wall)
since RNA synthesis was equally inhibited in vegetative cells, proheterocysts and heterocysts.
The deposition of the pore plug at the junction between the new heterocyst and an adjacent
vegetative cell, an event diagnostic of heterocyst maturation, requires the presence of functional
mRNA. The pore plug and the nitrogen storage polypeptide cyanophycin [multi-L-arginylpoly(aspartic acid)] appear to have a similar composition (Fogg, 1951; Lang et al., 1972). Since
cyanophycin is synthesized enzymically rather than on ribosomes (Simon, 1973), the final
maturation of heterocysts could represent m RN A-independent cyanophycin synthesis,
However, chloramphenicol prevented pore plug formation, and thus heterocyst maturation
must require continuing ribosome-dependent protein synthesis.
The half-life of mRNA in C . lichcntfbrme was comparable to that measured in other
cyanobacteria. In Anabaena cariabilis, the half-life of unstable RNA varied from 12 to 26 min as
the doubling-time of the cells varied from 5 to 25 h (Leach & Carr, 1974). The half-life of
functional mRNA capable of supporting protein synthesis varied from 25 to 50min as the
generation time of Anacystis nidulans varied from 6.7 to 33 h (Smith, 1979). In cells of C.
licheniforme which were doubling every 30 h, the half-life of unstable RNA was estimated to be
about 29 min. mRNA half-life was thus about 1.6% of the generation time, in accordance with
measurements using Anabaena variabilis (Leach & Carr, 1974) and Escherichia coli (Gray &
Midgley, 1970). Because chloramphenicol halts heterocyst differentiation at all stages, and
because an mRNA decaying with a 29 min half-life would have declined to 1.5:/, of its original
level after the cell had been in rifampicin for 3 h, the 3-6 h rifampicin-insensitive period in
heterocyst differentiation many consist of steps which depend on relatively long-lived mRNAs.
We wish to thank Dr C. P. Wolk for supplying a sample of C . licheniforme. These experiments were done by
S. D. V. in partial fulfilment of the Ph.D. requirements of the University of Rochester.
This work was supported by grant PCM77-202322 from the National Science Foundation. S. D. V . was
supported by National Science Foundation Graduate Fellowship SM 176-22877 and by National Institutes of
Health Graduate Traineeship in Cell Biology GM07230-05.
REFERENCES
ALLEN,M. B. & ARNON,D. I. (1955). Studies on
nitrogen-fixing blue-green algae. I. Growth and
nitrogen fixation by Anabaena cylindrica Lemm.
Plant Physiology 30. 336-372.
FISHER,R. W. & WOLK, C. P. (1976). Substance
stimulating the differentiation of spores of the bluegreen alga Cylindrospermum. Nature. Lmdon 259,
394-395.
FLEMING,
H. & HASELKORN,
R. (1974). The program of
protein synthesis during differentiation in nitrogen
fixing blue-green algae. Cell 3. 159- 170.
FOGG,G . E. (1951). Growth and heterocyst production
in Anabaena cylindrica Lemm. 111. The cytology of
heterocysts. Annals of Botanjs 15, 23-35.
FRITSCH,
F. E. (1 945). The Structure and Reproduction of
Algae. vol. 11. Cambridge, U K : Cambridge University Press.
GLASER,V. M., AL-NURI,M. A., GROSHEV,
V. V . &
SHESTAKOV,
S. V . (1973). The labeling of nucleic
acids by radioactive precursors in the blue-green
algae Anacystis nidulans and Sjwechocystis aquatilis
Sanv. Archit. .fir Mikrobiologie 92, 217-226.
GRAY,W. J . H. & MIDGLEY,
J . E. M. (1970). The
control of ribonucleic acid synthesis in bacteria.
Steady state content of messenger ribonucleic acid in
Escherichia coli MRE 600. Biochemical Journal 120,
279-288.
GROVER.1. S. & PURI, S. (1979). Alteration of
heterocyst pattern in Cylindrospermum sp. by rifampicin. Indian JournalojExperimental Biology 17, 109111.
HASELKORN.
R. (1978). Heterocysts. Annual Rezliew of’
Plant Physiologj 29, 3 19-344.
JANAKI,S. & WOLK, C. P. (1982). Synthesis of
nitrogenase by isolated heterocysts. Biochimica et
biophysica acta 696, 187- 192.
LADHA,J . K . & KUMAR,H. D. (1975). Heterocyst
division in two blue-green algae. Archives of’ Microbiology 102, 171-173.
LANG,N . J . (1965). Electron microscopic study of
heterocyst development in Anabaena azollae Strasburger. Journal of’ P h j d o g j . 1, 127-1 34.
LANG,N. J . & FAY,P. (1971). The heterocysts of bluegreen algae. 11. Details of ultrastructure. Proceedings
of the Royal Society B178, 193-203.
LANG,N . J., SIMON,R. D. & WOLK,C. P. (1972).
Correspondence of cyanophycin granules with structured granules in Anabaena cylindrica. Archir .fir
Mikrobiologie 83. 3 13-320.
LEACH,C. K . & CARR.N. G . (1974). In-vitro protein
synthesis and measurement of the stability of
messenger RNA in the blue-green alga, Anabaena
r7ariabili.y. Journal of’ General Microbiology 81, 4758.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11
796
S . VAN DE WATER A N D R . D . S I M O N
SIMON.R. D. (1973). The effect of chloramphenicol on
the production of cyanophycin granule polypeptide
in the blue-green alga Anabaena cylindrica. Archit.fur
Mikrobiologie 92, 1 15-1 22.
SIMON,R. D. (1979). Macromolecular composition of
heterocysts and spores from Anabaena variabilis.
Abstracts of the III International Symposium on
Photosynthetic Procaryotes, Oxford, England, p. 104.
Edited by J . M. Nichols. Liverpool, U K : University
of Liverpool.
SIMON,R . D. (1980). DNA content of heterocysts and
spores of the filamentous cyanobacterium Anabaena
iiariabilis. F E M S Microbiology Letters 8, 241-245.
SMITH,R. J. (1979). Variation in the functional half-life
of mRNA in Anacystis nidulans with respect to
growth rate. F E M S Microbiology LRtters 5, 9-1 1.
STEWART,
W. D. P., FITZGERALD,
G . P. & BURRIS,
R . H. (1968). Acetylene reduction by nitrogen-fixing
blue-green algae. Archit. jiir Mikrobiologie 62, 336348.
THOMAS,
J . (1972). Relationship between age of culture
and occurrence of the pigments of photosystem I1 of
photosynthesis in heterocysts of a blue-green alga.
Journal of Bacteriology 110, 92-95.
VANDE WATER,S. & SIMON,R. D. (1982). Induction
and differentiation of heterocysts in the filamentous
cyanobacterium Cylindrospermumlicheniforme. Journal of General Microbiology 128, 91 7-925.
WILCOX,M., MITCHISON,
G. J . & SMITH,R. J . (1973).
Pattern formation in the blue-green alga Anabaena.
11. Controlled proheterocyst regression. Journal qf
Cell Science 13, 637-649.
WOLK, C. P. (1975). Differentiation and pattern
formation in filamentous blue-green algae. In Spores
VI, pp. 85-96. Edited by P. Gerhardt, R. N.
Costilow & H. L. Sadoff, Washington, D C : American Society for Microbiology.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Mon, 19 Jun 2017 02:09:11