Symbiosis between Nitrogen-
Fixing Cyanobacteria and Plants
The establishment of symbiosis causes dramatic morphological
and physiological changes in the cyanobacterium
John C. Meeks
T
he cyanobacteria (blue-green
algae) are a morphologically
diverse group of Gram-negative prokaryotes within the domain
Bacteria (Giavannoni et al. 1988).
This group includes unicellular
forms, filamentous forms that contain only vegetative cells, and filamentous forms that differentiate specialized cells (Castenholz and
Waterbury 1989). Cyanobacteriaare
photoautotrophs with an oxygenevolving photosynthetic mechanism
that is identical to that of chloroplasts of algae and. higher plants.
Indeed, molecular genetic evidence
strongly supports the idea that a recent ancestor(s) in the cyanobacterial
evolutionary line gave rise to chloroplasts (Douglas 1994). Many cyanobacteria fix atmospheric nitrogen,
which is an oxygen-sensitive process; they fix nitrogen either under
low-oxygen growth conditions or,
when in air, in specialized cells called
heterocysts (Fay 1992). As a group,
oxygenic photoautotrophic cyanobacteria, especially those species that
fix nitrogen, are the most nutritionaUy independent organisms in the
biosphere. Because of this nutrition al
independence, as well as their tolerance to physicochemical extremes,
John C. Meeks (e-mail: jcmeeks@ucdavis.
edu) is a professor in the Section of Microbiology, Division of Biological Sciences,
University of California, Davis, CA 95616.
He studies the mechanisms of gene expression in filamentous cyanobacteria in the
context of the regulation of cellular differentiation in free-living and symbiotic
growth states. © 1998 American Institute
of Biological Sciences.
266
The interactions that
lead to a stable
symbiotic association
appear to be primarily
unidirectional, from
plant to cyanobacterium
cyanobacteria are ubiquitous in global aquatic and terrestrial habitats
that are illuminated.
Many unicellular and filamentous cyanobacteria grow in association with other prokaryotes, eukaryotic protists, metazoans, or plants
(Schenk 1992). All of these associations can be considered symbiotic
according to de Bary's (1879) broad
definition of the term; that is, they
are long-Iasting associations of two
or more different organisms. The
best studied cyanobacterial symbiotic associations are those that nitrogen-fixing genera of the order
Nostocales (Anabaena, Ca 10 thrix,
Nostoc, and Scytonema) form with
fungi, primarily in the establishment
of lichens, and those that they form
with representatives of four of the
major phylogenetic groups of plants.
The plant groups include the sporeproducing, nonvascular bryophytes,
especially the hornworts, such as
Anthoceros punctatus, and the liverwort Blasia pusilla; the spore-producing, vascularized ferns, specifically the aquatic genus Azolla; the
cycads, such as Macrozamia spp.,
which are gymnosperms; and the
angiosperm herbaceous genus Gunnera (Rai 1990a). In all of these
associations, the cyanobacteria remain extracellular, except for the
associations with Gunnera and with
the nonlichen fungus Geosiphon
pyriforme, in both of which the
cyanobacterium is intracellular. The
selective pressure on the formation
of these associations is presumably
the provision of fixed nitrogen by
the cyanobacterium for growth of
the plant partner.
The plant.partner in a symbiotic
association is a unique habitat with
specific physicochemical factors that
influence the growth and development of its cyanobacterial partner.
When nitrogen-fixing cyan acteria
enter into a symbiotic association,
their morphology and physiology
change dramatically. These changes
primarily involve a 5- to 10-fold enhancement in the degree of differentiation of motile filaments, caUed
hormogonia, and of heterocysts in
the vegetative filaments that develop
from the hormogonia. The hormogonia serve as the infective units in the
establishment of an association,
whereas nitrogen is fixed in the
heterocysts and made available to
both cyanobacterial and plant partners in the functional association.
In this article, I emphasize the
association between Nostoc species
and the gametophyte stage of the
hornwort A. punctatus (Figure 1) as
a context for exploring these changes
(Nostoc species do not infect the
sporophyte stage of bryophytes; Rodgers and Stewart 1977). My research
er
BioScience Vol. 48 No. 4
uses the A. punctatus-Nostoc association because the symbiotic partners can be cultured separately in the
laboratory. A. punctatus gametophyte tissue grows rapidly and can
be maintained in liquid suspension
pure culture much like plant tissue
cultures, and the association can be
reconstitutedunder liquid culture
conditions with a variety of Nostoc
isolates and mutant strains (Enderlin
and Meeks 1983). After colonization of preexisting cavities in the
gametophyte tissue and growth, the
Nostoc appear as dark, macroscopic
colonies of 0.4-1.0 mm in diameter
(Figure 1 b) that are highly active in Figure 1. Photomicrographs of field-collected Anthoceros punctatus. (a) Gametonitrogen fixation (Steinberg and phytes (G) grow in clusters; each typically bears a single sporophyte (S). Bar = 1 cm.
Meeks 1991). One particular iso- (b) Ventral surface of gametophyte tissue with macroscopic dark large and small
late, Nostoc punctiforme, which was colonies of Nostoc (N) located randomly throughout the tissue. Bar = 0.25 cm.
originally cultured from a .coralloid
root of the cycad Macrozamia sp. Rhizobia symbiont, all of these struc- sponding to chemical signals from
(Rippka et al. 1979) and can recon- tures preexist in the absence of the A. punctatus that affect the differenstitute both the A. -punctatus (Ender- cyanobacterial symbiont (Peters and tiation and behavior of hormogonia,
lin and Meeks 1983) and Gunnera Meeks 1989, Braun-Howland and leading to infection of the plant tismanicata (Johansson and Bergman Nierzwicki-Bauer 1990, Bergman et sue. The infection process is moni1994) associations, is amenable to al. 1992a).
tored by counting macroscopically
genetic analysis and now serves as
The interactions with A. punctatus visible Nostoc colonies in A. pu ncour experimental organism (Cohen that influence the morphology and tatus gametophyte tissue.
eta1.1994).
physiology of Nostoc and lead to a
The establishment of a symbiosis stable symbiotic association appear Formation of Nostoc infective units.
between cyanobacteria and their to be primarily unidirectional, from Motile hormogonia are essential for
plant partners is less complex than plant to cyanobacterium. They can establishment of plant-cyanobacthat between the Proteobacteria gen- be defined as two distinct stages, terial symbiotic associations-that
era Bradyrhizobium and Rhizobium with parallels to the Rhizobia- is, they appear to function as infec(Rhizobia) and leguminous plants. legurne symbiosis. The first stage is tive units (Campbell and Meeks
The Rhizobia-legume symbiosis in- the initial establishment of the asso- 1989, Rasmussen et al. 1994). Horvolves an extensive series of recipro- ciation (i.e., formation of hormogo- mogonia form transiently as part of
cal exchanges of chemical signals nia infective units, infection, and a naturallife cyde in some filamenbetween the partners (Stacey et al. dedifferentiation ofhormogonia); the tous cyanobacteria, induding some
1992). An exchange of signals that is second is its development into a func- genera that differentiate heterocysts
essential for recognition first occurs tional association (i.e., regulation of and others that do not; they are not
during infection of the plant by the growth and metabolism, and stimu- unique to symbiotic strains (Rippka et
Rhizobia. Formation of an effective lation of heterocyst differentiation al. 1979). Hormogonia are proposed
nitrogen-fixing association !nvolves coupled with release of fixed nitro- to function as units of dispers al for
signal exchange in the regulation of gen). The existence of two distinct colonization of new habitats (Tandeau
growth, metabolism, and cellular stages suggests that Nostoc respond de Marsac 1994). Hormogonia of
differentiation of both the Rhizobia to chemical signals produced by A. . Nostocales are morphologically disand legurne partners, leading to the punctatus, with different signals ac- tinct from vegetative filaments: They
formation of nodules. In plants that tivating different symbiotic-specific have a different cell shape, the cells
associate with cyanobacteria, by con- regulatory pathways. Similar stages are smaller, and the filaments lack
trast, the morphological changes are of interactions and activation of regu- heterocysts (Figure 2). Hormogonia
much less pronounced. In addition latory pathways presumably occur move by a gliding mechanism, but
to elaborating plant hair filaments in all other plant-cyanobacterial as- motility is often lost in culture
that function in nutrient transfer, sociations.
(Castenholz and Waterbury 1989).
the gametophyte and leaf cavity in
The differentiation of hormogonia is
bryophytes and Azolla sp., the coral- Initial establishment of a
characterized by the immediate cesloid root in cycads, and the special- cyanobacterial symbiosis
sation of DNA replication, of proized cells of the stern gland in
tein synthesis (most notably of
Gunnera sp. all increase in size. Un- Interactions that result in the estab- phycobiliproteins; Figure 2b), and
like legurne nodules, however, which lishment of a symbiotic association of biomass increase, and by rapid
form only in the presence of the can be characterized as Nostoc re- septation (Tandea u de Marsac 1994).
April 1998
267
Figure 2. Photomicrographs of a Nostoc punctiforme culture containing a mass of
hormogonia and two vegetative filaments with heterocysts. (a) Culture as seen by phase
contrast microscopy. (b) Culture as seen by epifluorescence microscopy. Sampie was
excited with light at a wavelength of 510-560 nm (which is absorbed by phycoerythrin), and light was emitted at a wavelength of more than 600 nm, which passes emission
from phycocyanin and chlorophyll a. Vegetative cells show fluorescence from phycocyanin and chlorophyll a; heterocysts, which typically lack phycobiliproteins, do nOt
show fluorescence under these conditions. Note that cells in the hormogonia show
markedly less fluorescence than cells in vegetative filaments because of the decrease in
phycobiliprotein content. Ho, hormogonia; V, one of the vegetative filaments; H,
heterocysts. Bar = 10 \lIIl.
Many environmental factors increase the frequency of hormogonium differentiation, in particular
transfer to fresh medium, transfer
from liquid to solid medium, and,
for some Calothrix and Nostoc
strains, exposure to red light (Tandeau de Marsac 1994). Nitrogenlimited A. punctatus stimulates hormogonium differentiation through
an excreted hormogonium-inducing
factor (HIF; Campbell and Meeks
1989). Similar inducing activity appears to be present in the mucilage
secreted by Gunnera stern glands
(Rasmussen et al. 1994) and in the
root exudates of a traditionally nonsymbiotic host, wheat (Gantar et al.
1993). An active factor has not been
isolated and characterized from any
of these plants.
The transient hormogonium cyde
is best observed in nitrogen-fixing
cultures. When heterocyst-containing filaments differentiate into hormogonia, the filaments fragment at
the junction between the vegetative
cells and heterocysts, detaching the
heterocysts, which are now nonfunctional (Campbell and Meeks 1989).
Hormogonia, therefore, do not fix
nitrogen. Within 12 hours of exposure to A. punctatus-produced HIF,
essentially 100% of vegetative Nostoc filaments are converted to hormogonia. The hormogonia remain
268
in an actively gliding state for approximately 48 hours, after which
they differentiate heterocysts, beginning with the end cells of the filaments, and begin to fix nitrogen.
Reversion, or dedifferentiation, of
hormogonia to vegetative nitrogenfixing filaments is complete within
96 hours after exposure to HIF. Once
the Nostoc culture has completed a
cyde of HIF-induced hormogonium
formation and reversion back to vegetative filaments, there is a lengthy
delay before extensive hormogonium
formation can occur aga in in response to the presence of HIF-that
is, there appears to be aperiod of
immunity to HIF (Elsie Campbell
and John C. Meeks, unpublished results). Thus, the initial infection window for the A. punctatus association
is approximately 48 hours, and secondary infection of the growing gametophyte tissue is delayed. An initial infection window of the same
duration was observed microscopically in the liverwort B. pusilla-Nostoc association (Kimura and N akano
1990).
Infection. The actual infection process involves Nostoc hormogonia
gliding to the site of entry into the
plant structure that will be colonized
(except in associations with Azolla
spp.); the site of entry is most likely
identified by a chemotactic signaling
mechanism. In the bryophytes A.
punctatus and B. pusilla, the hormogonia enter a preexisting slimefilled cavity in the gametophyte tissue (Rodgers and Stewart 1977). The
actual infection process has not been
observed in A. punctatus, but hormogonia have been observed by light
microscopy to glide on the surface of
B. pusilla gametophyte tissue before
entering the cavity through a pore
(Kimura and Nakano 1990). AIthough there is no evidence of chemotaxis to the cavity pore in any bryophyte, hormogonia of a Nostoc strain
exhibit chemotactic behavior in response to exudate from B. pusilla
(Knight and Adams 1996).
Associations with Azolla spp. are
characterized by a cyanobacterial
inoculant that is carried within the
fern tissue throughout its life cyde,
and the cyanobacterium may not
have a free-living growth state (Peters and Meeks 1989). The cyanobacteria present in the fern apical
meristem enter the leaf cavity by
attaching hormogonium-like filaments to a fern hair filament. Growth
of the leaf tissue subsequently endoses the inoculum (Peters and
Meeks 1989).
The mecnanism by which cyanobacteria enter the coralloid roots of
cycads is unknown. Because the
Gunnera association is intracellular
and occurs several celllayers deep in
the stern gland tissue, the infection
process is more complex than it is for
the other associations. The Nostoc
hormogonia glide on Gunnera gland
tissue that forms an intercellular
channel from the exterior of the plant
stern to the channel base (Johansson
and Bergman 1992). Hormogonia of
most cyanobacterial strains are positively phototactic; because the hormogonia move against a light gradient in the channel of Gunnera sp.,
Johansson and Bergman (1992) proposed that a chemotactic mechanism
attracts the hormogonia to the channel base and that this mechanism
must override any phototactic response. Once hormogonia reach the
channel base, some Gunnera sp.
gland cell walls dissolve by an unknown mechanism and are released
into these nonphotosynthetic cells,
although they remain separated from
the host cell cytoplasm by a mem-
BioScience Vol. 48 No. 4
brane of host eell origin (Johansson
and Bergman 1992).
Additional indireet evidenee for a
ehemotactic response in infection
comes from the observation that not
all Nostoc strains that form hormogonia in response to HIF can infeet either A. punctatus (Enderlin
and Meeks 1983) or Gunnera spp.
(Johansson and Bergman 1994). It is
possible that these strains have genetic defects in a chemotactic system. For example, one Nostoc strain
forms extensive hormogonia in the
presence of A. punctatus but is unable to infect the gametophyte tissue
(Enderlin and Meeks 1983). Recent
experiments (Elsie Camp bell and
John C. Meeks, unpublished data)
show that this strain is described
more accurately as having a delayed,
low level of infection. The Nostoc
sp. colonies that form after approximately six weeks of co-culture with
A. punctatus are able to supply fixed
nitrogen to the host tissue, just like
other symbiotically competent Nostoc strains. When re-isolated from
A. punctatus gametophyte tissue,
however, the N ostoc strain shows
the same infection phenotype of A.
punctatus as the original eulture.
These results indicate that the rare
infections by this strain are not due
to selection of a more infective mutant strain; rather , the infections most
likely occur by chance. This strain is
an example of an infection-defective
phenotype in a symbiotically effective background (see below). Strains
with such a phenotype could be used
as recipients in complementation experiments to identify genetic information that is present in infective Nostoc
strains; unfortunately, this .Nostoc
strain is not readily amenable to genetic manipulations (Michael Cohen
andJohn C. Meeks, unpublisheddata).
Repression of hormogonium formation. Beeause hormogonia are in a
nongrowing state, continued entry
into the hormogonium cyde is presumably lethaI. One might predict
that Nostoc would have a mechanism to block either the sensing of
the continual presence of HIF (or an
equivalent signal) or the initiation of
hormogonium formation. A possible
hormogonium repression system was
identified by analysis of a transposoninduced mutant of N. punctiforme
Apri/1998
Figure 3. Photomicrographs of Nostoc punctiforme cultured with nitrogen gas as the
sole nitrogen source in both the free-living growth state and in symbiotic association
with Anthoceros punctatus. (a) Phase contrast microscopy of the free-living growth
state. (b) Epifluorescence microscopy of the free-living growth state using an excitation
wavelength of 510-560 nm and an emission wavelength of over 600 nrn. (c) Phase
contrast microscopy of the syrnbiotic association. (d) Epifluorescence microscopy of the
symbiotic association using an excitation wavelength of 510-560 nrn and an emission
wavelength of over 600 nm. Whereas the heterocysts (B) in (a) have a shape that is
distinct frorn that of the vegetative cells, a11 of the cells in (c) are large and irregularly
shaped, making it difficult to distinguish vegetative cells from heterocysts. Similarly,
whereas phycocyanin and chlorophyll a fluoresce in the vegetative cells but not the
heterocysts in (b), all of the cells in (d) fluoresce. Electron rnicrographs of sirnilar
sampIes of Nostoc frorn reconstituted syrnbiotic tissue show that 25-45% of the
irregularly shaped cells are heterocysts (Meeks 1990). These observations provide
indirect evidence for the presence of phycobiliproteins in heterocysts of N. punctiforme
when in association with A. punctatus. Bars = 10 Jll1l.
that is approximately SO-fold more .extract elimina tes HIF -stirn ula ted
infective of A. punctatus than its hormogonium formation in the wildparental culture (Cohen et a1. 1994). type strain. The mutant strain may
From this mutant, two genes were be more infective of A. punctatus
identified as part of an operon in N. beeause it immediately reenters the
punctiforme: one gene (hrmA) has hormogonium state in the presence
no similarity in major databases, of HIF, thereby extending the initial
whereas the second gene (hrmU) has infection window (Cohen and Meeks
a signature sequence belonging to a 1997). These observations suggest
family of NAD(P)H-dependent oxi- that the gene products of the hrm UA
doreduetases. Reporter gene con- operon may block hormogonium
structs in hrmU and hrmA are in- formation by produeing an inhibitor
duced by an aqueous extract of A. or by eatabolizing an activator.
punctatus tissue, but not by HIF;
Although hormogonia are essenindeed, the presence of the aqueous tial in establishment of a Nostoc
269
Table 1. Photosynthetic characteristics of cyanobacteria in symbiotic associations. The values for the symbiotic associations
are given as a percentage of the free-living values." In many instances, only qualitative results have been reported.
Light-dependent
CO z fixation
In vitro Rubisco
activity
Rubisco protein
in vegetative cells in heterocystsb Reference(s)
Lichen, Peltigera
canina"
NDd
ND
100%
100%
0
Bergman and
Rai 1989, Rai
1990b
Lichen, Peltigera
aphthosa'
12%
8%
100%'
100%1
01
Rai et al. 1981a,
Bergmanand
Rai 1989
Hornwort,
Anthoceros punctatus
12%
12%
100%
100%
oor
<100%
Steinberg and
Meeks 1989,
Rai et al. 1989,
Meeks 1990
Fern, Azolla spp.
85%
ND
mRNA
100%
approximately 10%g
100%
Ray et al. 1979,
Kaplan et al.
1986, NierzwickiBauer and
Haselkorn 1986
Cycad, Zamia,
Macrozamia spp.
Not detectable
100%
ND
100'%
0
Lindblad et al.
1987, Lindblad
and Bergman
1989
Angiosperm,
Gunnera spp.
<1%
100%
100%
100%
0
Soderback and
Bergman 1992,
1993
Association
PhIcobili~roteinsb
"The values for free-living NostoclAnabaena spp. are as follows: light-dependent COz fixation activity = 128 nmol· min-1 • mg-1 ; in vitro Rubisco
(ribulose-1,5-bisphosphate carboxylase/oxygenase) COz fixation activity =215-321 nmol . min-1 • mg-1; Rubisco protein (milligrams of Rubisco
protein per gram of cellular protein) =52; phycobiliproteins (milligrams of phycobiliproteins per gram of cellular protein) in vegetative cells =250,
and in heterocysts =O. From a compilation of various control experiments in the cited references.
b100% value means that phycobiliproteins are present at levels equal to free-living vegetative ceIls; less than 100% value means that phycobiliproteins
are present but at lower levels than in vegetative ceIls; 0 means phycobiliproteins are absent.
'Bipartite association.
dNot determined.
'Tripartite association with an additional green alga, Coccomyxa sp.
IResults from the tripartite lichen Nephroma arcticum with internal cephalodia containing the Nostoc.
gRubisco protein has not been measured in the Azolla association.
symbiotic association, their continued formation prevents the accumulation of vegetative filaments containing heterocysts. This behavior is
counterproductive to the formation
of a functional nitrogen-fixing association. The hormogonium repressive mechanism that is activated by
A. punctatus may, therefore, be important in symbiosis by shifting the
developmental direction of the associated Nostoc toward heterocyst differentiation and nitrogen fixation
simply by preventing hormogonium
differentiation.
Development of a
functional association
Interactions that lead to the development of a functional symbiotic association involve the provision by the
cyanobacterium of nitrogen gas-derived ammonium or organic nitro-
270
gen for growth of the plant partner.
The fixed nitrogen that is made available to the plant partner is in excess
of that required tor growth of the
associated Nostoc species. Indeed,
in its symbiotic state, Nostoc undergoes a metabolie decoupling such
that high er rates of heterocyst differentiation and nitrogen fixation are
accompanied by lower, rather than
higher, rates of vegetative growth,
photosynthetic carbon dioxide fixation, and ammonium assimilation.
Because the plant partners themselves
are photosynthetic and can assimilate ammonium, in most experiments
the Nostoc sp. is physically separated from the association and its
physiological and biochemical characteristics determined by in vivo or
in vitro assays. Nitrogen fixation,
which is unique to the associated
Nostoc, can be monitored in the intact association (in situ).
Regulation of cyanobacterial growth
and metabolism. Under optimallaboratory conditions, most heterocystforming cyanobacteria have doubling
times in the free-living growth state
that are substantially more rapid than
those of their symbiotic plant partners
(Enderlin and Meeks 1983, Peters and
Meeks 1989, Braun-Howland and
Nierzwicki-Bauer 1990). Thus, for a
stable symbiotic association to endure,
the growth rate of the previously
free-living cyanobacterium must be
slowed. Although slower growth of
symbiotically associated Nostoc has
been documented (Enderlin and Meeks
1983), nothing is known about the
mechanisms of cyanobacterial growth
control in any association.
In addition to a decreased growth
rate, the vegetative cells of symbiotically associated N ostoc increase twoto threefold in circumference (compare Figures 3a and 3c). The reasons
BioScience Vol. 48 No. 4
Table 2. Nitrogen metabolie characteristics of cyanobacteria in symbiotic associations. Because more data are available for
nitrogen than for carbon metabolism, quantitative values are reported as measured in the specific study, unless otherwise noted.
For clarity and comparison, the glutamine synthetase (GS) values are reported as a percentage of the free-living values. a
Association
Heterocyst
frequency
Free-living Nostocl 3-10%
Anabaena
Nitrogenase
specific activityb
Nitrogen released
as percentage of
fixed and form
in which released
Glutamine synthetase
Protein
Activity
2.7-6.3
Less than 10% as
organie nitrogen
100%
100%
Reference(s)
Lichen, Peltigera
canitutI
4.7%
5.3'
55% as ammonium
5%
100%
Sampaio et al.
1979, Rai et al.
1983, Rai 1990b
Lichen, Peltigera
aphthosaf
21.2%
14.0'
95% as ammonium
5%
100%
Stewart and Rowell
1977, Sampaio et
al. 1979, Rai et al.
1981b, Rai 1990b
23.5-185.7g
80% as ammonium
15%
100%
HiIl1975, Rodgers
and Stewart 1977,
Meeks et al. 1985,
Joseph and Meeks
1987, Steinberg
and Meeks 1991
Hornwort, Antho- 43-45%
eeros punetatus
Fern, Azolla spp.
30%
6.2'
40% as ammonium
30%
38%
HilI 1975, Ray et
al. 1978, Meeks et
al. 1987, Lee et al.
1988
Cycad, Zamia,
Maerozamia spp.
17-46%
26.7'
Unknown amount
as organie nitrogen
100%
100%
Lindblad et al.
1985,1987,
Lindblad and
Bergman 1986,
Perraju et al. 1986,
Pate et al. 1988
Angiosperm,
Gunnera spp.
20-60%
24.8
90% as ammonium
70%
100%
Silvester 1976,
Soderback et al.
1990, Bergman et
al. 1992b, Silvester
et al. 1996
aThe values for GS activity and protein levels in free-livingNostoclAnabaena are as folIows: GS activity, measured as,..glutamyl hydroxymate formation
in a transferase assay =0.85-1.8 x 1Q-6 mol· min-1 • mg-1; GS protein (milligrams GS per gram cellular protein) =6.8-7.6. From a compilation ofvarious
control experiments in the cited references.
bNitrogenase specific activity (C2H 4 reduction) is given as nmol . min-1 • mg-1 •
cFrom a compilation of various control experiments in the cited references.
dBipartite association.
'Converted from nanomoles per hour per microgram chlorophyll, assuming that chlorophyll makes up 4% of total cellular protein.
f'J'ripartite association with an additional green alga, Coeeomyxa.
gRate of 185.7 from laboratory tissu~ grown with approximately one-half the intensity of full sunlight and 5% (by volume) CO2; because these high
lightand CO2 values are not steady state laboratory culture conditions, this rate was not used in the general comparisons.
for, and consequences of, this decrease
in the surface-to-volume ratio are unknown.
Photosynthetic metabolism. In
most plant-cyanobacterial associations, photosynthetic carbon dioxide fixation and nitrogen-derived ammonium assimilation capacities
decrease in the associated Nostoc
(Tables 1 and 2). Both the extent and
mechanism of the declines vary
among the associations. Only in the
association with Azolla does the rate
of light-dependent carbon dioxide
fixation by the immediately sepa-
April 1998
rated symbiotic cyanobacterium ap- . 12 % of free-living cultures in the A.
pear to approach that of free-living punctatus and tripartite (cyanocultures. However, like most other bacterium-eukaryotic alga-fungus)
cyanobacterial symbionts, Anabaena lichen associations (Table 1). In both
maintains an insignificant photosyn- bipartite (cyanobacterium-fungus)
thetic rate in the intact Azolla asso- and tri partite lichen associations, the
ciation (Peters and Meeks 1989). cyanobacterium is assumed to conThese in vivo and in situ differences tribute photosynthate to the fungus
have not been reconciled. In other (Rai 1990b). Nevertheless, the rate
associations, the rate of photosyn- of photosynthetic carbon dioxide
thetic carbon dioxide fixation by fixation by the symbiotic Nostoc in
immediately separated symbiotic these lichens is lower than that in the
cyanobacteria ranges from essentially free-living growth state.
The symbiotic growth state does
undetectable in the cycad and Gunnera
associations to between 11 % and not appear to lead to regulation of
271
the transeription of eomponents of
the photoautotrophie machinery in
associated Nostoc, despite their diminished photosynthetic capaeity in
symbiosis. There is no marked decrease in the synthesis of light-harvesting phycobiliproteins in vegetative cells of cyanobaeteria in any
association (Table 1). Synthesis of
the primary carbon dioxide fixation
enzyme, ribulose-1,5-bisphosphate
carboxylase/oxygenase (Rubisco), is
similar in symbiotic and free-living
Nostoc cultures except~ perhaps, in
the Azolla association. Rubisco protein levels have not been quantified
in the Anabaena associated with
Azolla; however, mRNA eneoding
Rubisco in the separated symbiont
accumulates to only 10% ofthe level
in free-living cultures (NierzwickiBauer and Haselkorn 1989), which
is not eonsistent with the high rate of
in vivo carbon dioxide fixation.
In the A. punctatus (Steinberg and
Meeks 1989) and Peltigera aphthosa
(Rai et al. 1981a) associations, regulation of carbon dioxide assimilation occurs in part by inhibition of
Rubisco activity of the Nostoc partner. However, crude extracts of
Nostoc associated with cycads (Lindblad et al. 1987) and with Gunnera
sp. (Soderback and Bergman 1993)
have as much Rubisco activity as
free-living cultures. The insignificant
light-dependent in vivo carbon dioxide assimilation by Nostoc in association with cycads and with Gunnera sp. may, therefore, be a
consequence of impaired photoehemical reactions, of limited substrates in the reductive pentose phosphate eyde, or of a reversibly bound
inhibitor. Additional components of
the photosynthetic machinery of symbiotically associated eyanobacteria
have not been examined.
Nitrogen fixation and metabolism. Symbiotieally associated cyanobacteria tend to have an enhaneed
rate of nitrogen fixation (Table 2).
The electrons to reduee atmospheric
nitrogen via nitrogenase in heterocysts come from photosynthate provided by adjacent vegetative cells,
while the ATP can be derived in
heterocysts from oxidative phosphorylation or photophosphorylation
(Wolk et al. 1994). In the free-living
growth state, the rate of nitrogen
fixa tion by heterocyst-forming
272
eyanobacteria is relatively low under dark heterotrophie conditions
and is stimulated by light. (The role
of light is to generate photosynthate
in vegetative cells and ATP in heterocysts.) The diminished photosynthetie eapacity of symbiotic eyanobacteria implies that the enhanced
rate of symbiotic nitrogen fixation
probably depends on redueed carbon supplied by the plant tissues and
that symbiotic cyanobacteria have a
high potential for heterotrophie metabolism. These condusions are supported by the high rate of dark heterotrophie nitrogen fixation, which
is also equal to the light-dependent
rate, in the A. punctatus-Nostoc association (Steinberg and Meeks
1991).
In all but the cycad associations,
the symbiotic Nostoc species release
from 40% to 95% of their fixed
nitrogen as ammonium into the symbiotic cavity, where it is taken up by
the plant (Table 2). In the cyead
association, the Nostoc releases organic nitrogen, not ammonium (Pate
et al. 1988). The release of nitrogenderived ammonium by symbiotic
cyanobacteria inversely paralleis a
decrease in aetivity of glutamine synthetase (GS), the primary ammonium-assimilating enzyme in cyanobacteria. However, the rate of
nitrogen fixation, the catalytic activity of GS, and the amount of nitrogen-derived ammonium released are
not strongly correlated (Table 2).
The regulation of GS expression by
Nostoc appears to be variable in the
different associations. In all but the
cycad and Gunnera associations, the
in vitro GS catalytic activity is markedly less than in free-living cultures,
but it is still suffieient to assimilate
all of the nitrogen-derived ammonium. The relatively high er GS aetivity of Nostoc species in the cyead
association is consistent with their
release of organic nitrogen to the
plant. In all but the assoeiation with
Azolla sp., the cellular concentration of GS in symbiotie cyanobacteria
is similar to its concentration in the
free-living growth state (Table 2).
Anabaena in association with
Azolla synthesizes less GS pro tein
than free-living eultures (NierzwickiBauer and Haselkorn 1986, Lee et
al. 1988). Because the Anabaena associated with Azolla may not have a
free-living growth state (Peters and
Meeks 1989), its reduced synthesis
of GS may reflect the selection of a
constitutively low GS level, which
would be consistent with its slower
symbiotic growth rate, rather than a
transcriptional or translational regulatory mechanism.
Generalities of symbiotic growth
and metabolism. In general, the rates
of carbon dioxide and ammonium
assimilation appear to decrease in
symbiotic cyanobacteria approximately in proportion to the decreased
rates of growth. Moreover, the enzymes catalyzing their assimilation
are negatively regulated by posttranslational mechanisms, or by secondary photochemieal or substrateproducing reactions, rather than by
control of enzyme synthesis. Whether
the lower catalytic activities cause
the lower growth rates, or the converse, is not known. The catalytic
inactivations are presumably the consequenee of biochemical reactions
(i.e., modifieation systems) thatcould
regulate other metabolie proeesses
in the free-living growth state but
that are activated differently or have
altered substrate specificity in the
symbiotic growth state. These unknown modification systems eould
act as sensors of the plant symbiotic
signals.
Heteroeyst differentiation and nitrogenase expression. In all symbiotic
associations involving a eukaryotic
photosynthetic partner, the heterocyst frequency of the symbiotic
cyanobaeterium is increased from
twofold to more than sixfold relative
to the free-living growth state (Table
2). The increased heterocyst frequency parallels an elevated rate of
nitrogen fixation. The regulatory
mechanism(s) responsible for the initiation of heterocyst differentiation
in symbiosis may not be identical to
those operative in the free-living
growth state.
Heterocyst differentiation in the
free-living growth state. When laboratory cultures of Nostoc and related genera are grown in the presence of eombined inorganic nitrogen,
such as ammonium or nitrate, heterocysts do not form. By contrast,
when cultures are deprived of combined nitrogen, heterocysts differentiate. These observations imply that
BioScience Vol. 48 No. 4
a combined nitrogen sensing and signaling pathway is involved in the
initiation of heterocyst differentiation. As many as 1000 genes could
be involved in heterocyst formation
and maintenance (Lynn et al. 1986);
some of the regulatory and structural genes have been isolated and
characterized and their transcription
has been placed in epistatic order,
primarily in the nonsymbiotic isolate Anabaena strain PCC 7120
(Buikema andHaselkorn 1993, Wolk
et al. 1994, Wolk 1996).
The differentiation of a vegetative
cell results in a heterocyst that is
specialized for oxygen-sensitive nitrogen fixation in an oxic environment. Heterocysts lack the photosystem 11 oxygen-producing re action;
instead, they have an increased rate
of respiratory oxygen uptake, and
they synthesize an outer envelope of
polysaccharide and glycolipid, which
collectively impede the entry and accumulation of oxygen (Wolk et al.
1994). Although there is an exception in which nitrogenase genes are
expressed in vegetative cells in the
absence of oxygen (Theil et al. 1995),
nitrogenase synthesis and its expression are generally confined to heterocysts in all free-living species that
form them (Wolk et al. 1994).
Immunoelectron microscopic experiments also show that the nitrogenase
is confined to heterocysts of Nostoc
in various symbiotic associations
(Bergman et al. 1992a, 1992b), thus
indicating that nitrogenase activity
correlates with heterocyst frequency
in symbiosis.
An immediate physiological response to deprivation of combined
nitrogen in cyanobacteria is the mobilization of stored nitrogen. Cyanobacteria store nitrogen in a unique
copolymer of aspartate and arginine
called cyanophycin and in phycobiliproteins, which are photosynthetic accessory pigments, as well as
in other proteins that are present at
lower cellular concentrations. Specific proteases cleave amino acids
from these polymers that are presumably used for synthesis of new
proteins that are characteristic of
heterocysts (Wolk et al. 1994). Synthesis of cyanophycin and phycobiliproteins ceases during nitrogen
deprivation; thus, a vegetative cell
that differentiates into a heterocyst
April 1998
in the free-living growth state initially has little or no phycobiliproteins. Such cells can be detected
by fluorescence microscopy as being
nonfluorescent when using exeitation light absorbed by the phycobiliproteins. The phycobiliproteins
are resynthesized in vegetative cells
that do not differentiate into heterocysts, but not in heterocysts of freeliving cultures (Figure 3; for exceptions, see Wolk et al. 1994).
Heterocyst differentiation in the
symbiotic growth state. The enhanced level of heterocyst differentiation and nitrogenase expression
by Nostoc in assoeiation with A.
punctatus may be induced in response
to a symbiotic sensing and signaling
system. This hypothetical system
could be activated by a plant product in a mechanism that is distinct
from the combined nitrogen deprivation sensing and signaling pathway that operates in the free-living
growth state. Several lines of evidence support this hypothesis. First,
electron micrographs show distinctly
the presence of cyanophycin and
phycobiliproteins in vegetative cells
of cyanobacteria in assoeiations with
the hornwort A. punctatus, the water fern Azolla caroliniana, the cycad
Zamia skinneri, and the angiosperm
G. manicata (see Braun-Howland
and Nierzwicki-Bauer 1990, Meeks
1990, Bergman et al. 1992a, 1992b
for reviews). These polymers are
characteristic of nitrogen -replete, not
nitrogen-starved, cells, implying that
the vegetative cells in symbiotic
cyanobacteria that differentiate into
heterocysts are not extremely nitrogen starved and may be responding
to a different signal in the initiation
of differentiation.
Second, heterocysts of cyanobacteria in association with Azolla
species (Kaplan et al. 1986), and
possibly with A. punctatus (Meeks
1990), contain phycobiliproteins, in
contrast to their ne ar-universal absence in heterocysts of free-living
cultures. There is some disagreement
as to the presence of phycobilipro teins in heterocysts of N ostoc
associated with A. punctatus, in
which it is difficult to distinguish
heterocysts and vegetative cells by
light microscopy (Figure 3c). More
than 95% of the cells of a symbiotic
nitrogen-fixing Nostoc colony asso-
ciated with laboratory cultures of A.
punctatus fluoresce orange to red
(more than 600 nm) when excited
with green (510-550 nm) light, which
is absorbed by the phycobiliprotein
phycoerythrin; all vegetative cells of
free-living cultures fluoresce in a similar fashion (Figures 3c, 3d). Electron
micrographs of parallel preparations,
however, show heterocyst frequeneies ranging from 25 % to 45 % of the
total cyanobacterial cells (Meeks
1990). The difference between the
less than 5% nonfluorescent cells
and the 25-45% frequency of identified heterocysts may reflect the presence of phycobiliproteins in heterocysts of symbiotic Nostoc. However,
Rai et al. (1989) were unable to
detect phycoerythrin by immunoelectron microscopy in heterocysts
of Nostoc in assoeiation with field
sampies of A. punctatus that had an
unknown physiological history .
These conflicting results can be
reconciled if the vegetative cells differentiating into heterocysts in symbiosis initially contain phycobiliproteins that have not yet been
degraded in response to a nitrogenstarvation signal. Except perhaps in
the Azolla association, there is little
or no synthesis of phycobiliproteins
in heterocysts; therefore, in the course
of normal turnover, phycobiliproteins decrease to low or undetectable levels in aging heterocysts of
symbiotically assoeiated N ostoc. The
presence of phycobiliproteins in developing heterocysts in symbionts is
consistent with initiation of their differentiation through a sensing and
signaling system that does not activate the nitrogen-starvation response
characteristic of free-living cultures.
The third li ne of evidence for a
symbiotic sensing and signaling pathway is that heterocysts of symbiotically associated cultures are physiologically different from those
formed in the free-living growth state.
In particular, the linkage of carbon
catabolism to nitrogen assimilation
is different. In all but the cycad assoeiations, heterocysts of symbiotic
Nostoc primarily translocate ammonium to adjacent vegetative cells
(Table 2), whereas heterocysts of freeliving cultures translocate glutamine
to adjacent vegetative cells (Thomas
et al. 1977). Therefore, heterocysts
of symbiotic cyanobacteria have a
273
lower demand than heterocysts of
free-living cyanobacteria for glutamate, or its precursor a-ketoglutarate,
in the assimilation of nitrogen-derived
ammonium, but they also maintain a
high demand for hexose-derived reductant to support the high rate of
heterotrophie nitrogenase activity
and oxidative phosphorylation.
As a consequence of the different
metabolie demands, free-living and
symbiotic cyanobacteria must differ
substantially in the end products of
the oxidative pentose phosphate
pathway, which is the primary route
of hexose catabolism and nonphotosynthetic reductant generation in
cyanobacteria (Summers et al. 1995).
Cyanobacteria have an incomplete
citric acid cyde, lacking a-ketoglutarate dehydrogenase (Smith 1982);
thus, when hexose is catabolized in a
linear oxidative pentose phosphate
pathway coupled to the citric acid
cyde, the end product is a-ketoglutarate, in addition to reductant as
NADPH. To avoid the accumulation
of a-ketoglutarate in excess of that
required for ammonium assimilation
while generating reductant, the symbiont must shift to a cydic pentose
phosphate pathway, whose end products are carbon dioxide and NADPH.
The extent of this change in carbon
flux in symbiotically associated Nostoc is not known, nor has the mechanism that controls it been identified.
Such a metabolie change would imply that different genes or gene products are activated following initiation of heterocyst differentiation in
free-living and symbiotic growth states,
perhaps as a consequence of different sensing and signaling systems.
Fourth, symbiotic nitrogen fixation is enhanced in proportion to the
increase in heterocyst frequency,
whereas the rate of assimilation of
ammonium is decreased (Table 2).
This observation indieates that vegetative cells of symbiotic Nostoc are
exposed to a substantial concentration of nitrogen-derived ammonium
before it is assimilated by the plant.
Nevertheless, these cells continue to
differentiate into heterocysts. Conversely, exogenously supplied ammonium or nitrate represses nitrogenase expression (and, presumably,
heterocyst differentiation) of Nostoc associated with A. punctatus
(Enderlin and Meeks 1983). Symbi274
otically associated Nostoc is unlikely
to respond to exogenously supplied
combined nitrogen by repression of
differentiation; after all, it does not
respond to its own excess of nitrogen-derived ammonium. Rather, the
plant partner may sense the exogenous combined nitrogen and cease
to produce the chemie al signals that
influence heterocyst differentiation
of the associated Nostoc.
As an initial test of this suggestion, a chemically induced mutant of
N. punctiforme that is defective in
nitrate assimilation was isolated
(Campbell and Meeks 1992). Nitrate neither supported growth nor
repressed heterocyst differentiation
and nitrogenase expression in the
free-living growth state of the mutant strain, as it did in the wild-type
strain; by contrast, ammonium was
an effective repressor of both mutant
and wild-type cultures, indicating the
continued presence and operation of
a combined nitrogen deprivation
sensing and signaling pathway. The
mutant strain also established a typical functional symbiotic association
with A. punctatus. When nitrogenfixing A. punctatus associations with
either the wild-type N. punctiforme
or the mutant strain were exposed to
exogenous nitrate, nitrogenase expression (and, presumably, heterocyst differentiation) by the associated N. punctiforme strains was
repressed. The response to exogenous
nitrate by the symbiotically associated mutant strain is consistent with
the plant partner ceasing to produce
signals controlling cellular differentiation in the associated cyanobacterium; it is not consistent with a
direct cyanobacterial response to the
exogenous combined nitrogen. These
results therefore support the presence of a combined nitrogen-independent sensing and signaling pathway in the initiation of symbiotic
heterocyst differentiation.
This symbiotic sensing and signaling pathway may supersede the
combined nitrogen deprivation sensing and signaling system that operates in the free-living growth state so
that heterocysts continue to differentiate in the presence of nitrogenderived ammonium. Nostoc mutant
strains or natural isolates that lack
the symbiotic sensing and signaling
pathway would remain competitive
under nitrogen-fixing conditions in
the free-living growth state and could
infect the plants (i.e., retain infectiveness). However, such strains would
fail to form a highly effective nitrogen-fixing symbiotic association because they could not respond to the
plant signals that initiate an enhanced
level of heterocyst differentiation.
Summary and future questions
The symbiotic interactions between
cyanobacteria of the genus Nostoc
and plants are characterized by a
predominantly unidirectional signaling from the plant to control the
activities of the cyanobacterium. At
least five distinct steps in the interaction between the bryophytes B.
pusilla or A. punctatus and Nostoc
can now be identified in which specific chemical signals from the plant
influence symbiotically competent
Nostoc species. These plant-produced
signals indude hormogonium-inducing factors, which yield a high frequency of infective units and increase
the probability of an infection;
chemotactic factors, which influence
the direction of hormogonium gliding for more efficient colonization of
gametophyte tissue; hormogoniumrepressing fa~tors, which are released
into the gametophyte symbiotic cavity and may shift developmental alternatives away from hormogonium
and toward heterocyst differentiation; growth- and metabolism-inhibiting factors that are essential for the
establishment of a stable association
and that mayaiso contribute to the
distinct physiologieal characteristics
of symbiotic heterocysts; and factors
that activate a symbiotic-specific
sensing and signaling pathway that
not only supersedes the nitrogen deprivation pathway operative in the
free-living growth state but also converges into a common regulatory
pathway of heterocyst development
and maturation.
Physiological genetic techniques
have recently been adapted to symbiotically competent N. punctiforme,
leading to its molecular genetic characterization (Cohen et al. 1994).
Future work on the A. punctatus
association will focus on genetie
analysis of N. punctiforme to identify genes involved in hormogonium
formation and behavior and to test
BioScience Vol. 48 No. 4
the working hypothesis of a symbiotic sensing and signaling initiation
pathway of heterocyst differentiation. Because of the broad symbiotic
competence of N. punctatus, mutants and genes identified in this speeies can also be characterized in the
cycad and Gunnera associations.
Essentially all of the research effort on symbioses between nitrogenfixing cyanobacteria and the plants
with which they associate has been
on the cyanobacterial symbiont; little
is known about how the plant partners respond to the symbiotic growth
state. Identification of the plant-produced factors that act as signals has
proven elusive. Assays for active factors would be simplified by the use
of reporter gene fusions in the Nostoc target genes. Such studies will
surely follow the identification of
target genes; subsequently, the regulation of the synthesis and release of
the factors from plant tissues can be
examined.
Acknowledgments
I thank Eisie Campbell, Michael
Cohen, Kari Hagen, Tom Hanson,
and Francis Wong for unpublished
data and critical review of the manuscript and an anonymous reviewer
for comments that vastly improved
its presentation. Work in my laboratory is sJIPported by grants from the
National Science Foundation (currently grant no. 95-14787).
References cited
de Bary A. 1879. Die erscheinung der
symbiose. Strass burg (Germany): Verlag
van K. J. Trubner.
Bergman B, Rai A. 1989. The Nostoc-Nephroma symbiosis: Localization,. distribution pattern and levels of key proteins
involved in nitrogen and carbon metabolism of the cyanobiont. Physiologia
Plantarum 77: 216-224.
Bergman B, Johansson C, Soderback E. 1992a.
The Nostoc-Gunnera symbiosis. New
Phytologist 122: 379-400.
_ _. 1992b. Cyanobacterial-plant symbioses. Symbiosis 14: 61-81.
Braun-Howland EB, Nierzwicki-Bauer SA.
1990. Azolla-Anabaena symbiosis: Biochemistry, physiology, ultrastructure, and
molecular biology. Pages 65-119 in Rai
AN, ed. Handbook of Symbiotic Cyanobacteria. Boca Raton (FL): CRC Press.
Buikema WJ, Haselkorn R. 1993. Molecular
genetics of cyanobacterial development.
Annual Review of Plant Physiology and
Plant Molecular Biology 44: 33-52.
Campbell EL, Meeks JC. 1989. Characteris-
April 1998
tics of hormogonia formation by symbiotic Nostoc spp. in response to the presence of Anthoceros punctatus or its extracellular products. Applied and
Environmental Microbiology 55: 125131.
_ _. 1992. Evidence for plant-mediated
regulation of nitrogenase expression in
the Anthoceros-Nostoc symbiotic association. Journal of General Microbiology
138: 473-480.
Castenholz RW, Waterbury JB. 1989. Oxygenic photosynthetic bacteria (section 19),
group 1. Cyanobacteria. Pages 1710-1728
in Staley JT, Bryant MP, Pfenning N, Holt
JG, eds. Bergey's Manual of Systematic
Bacteriology. Vol. 3. Baltimore: Williams
and Wilkins.
Cohen MF, Meeks JC. 1997. A hormogonium regulating locus, hrmUA, of the
cyanobacterium Nostoc punctiforme
strain ATCC 29133 and its response to an
extract of a symbiotic plant partner
Anthoceros punctatus. Molecular PlantMicrobe Interactions 10: 280-289.
Cohen MF, Wallis JG, Campbell EL, Meeks
JC. 1994. Transposon mutagenesis of
Nostoc sp. strain ATCC 29133, a filamentous cyanobacterium with multiple
cellular differentiation alternatives. Microbiology 140: 3233-3240.
Douglas SE. 1994. Chloroplast origins and
evolution. Pages 91-118 in Bryant DA,
ed. The Molecular Biology of Cyanobacteria. Boston: Kluwer Academic Publishers.
Enderlin CS, Meeks JC. 1983. Pure culture
and reconstitution of the AnthocerosNostoc symbiotic association. Planta 158:
157-165.
Fay P. 1992. Oxygen relations of nitrogen
fixation in cyanobacteria. Microbiological Reviews 56: 340-373.
Gantar M, Kerby NW, Rowell P. 1993. Colonization ofwheat (Triticum vulgare L.) by
N 2-fixing cyanobacteria: 111. The role of a
hormogonia-promoting factor. New
Phytologist 124: 505-513.
Giavannoni S, Turner S, Olsen GJ, Barns S,
Lane DJ, Pace NR. 1988. Evolutionary
relationships among cyanobacteria and
green chloroplasts. Journal of Bacteriology 170: 3584-3592.
Hili DJ. 1975. The pattern of development of
Anabaena in the Azolla-Anabaena symbiosis. Planta 122: 179-184.
Johansson C, Bergman B. 1992. Early events
during the establishment of the GunneraNostoc symbiosis. Planta 188: 403-413.
_ _. 1994. Reconstitution of the symbiosis of Gunnera manicata Linden: Cyanobacterial specificity. New Phytologist 126:
643-652.
Joseph CM, Meeks JC. 1987. Regulation of
expression of glutamine synthetase in a
symbiotic Nostoc strain associated with
Anthoceros punctatus. Journal of Bacteriology 169: 2471-2475.
Kaplan D, Calvert HE, Peters GA. 1986. The
Azolla-Anabaena azollae relationship.
XII. Nitrogenase activity and phycobiliproteins of the endophyte as a function of
leaf age and cell type. Plant Physiology
80: 884-890.
Kimura J, Nakano T. 1990. Reconstitution
of a Blasia-Nostoc symbiotic association
under axenic conditions. Nova Hedwigia
50: 181-200.
Knight CD, Adams DG. 1996. A method for
studying chemotaxis in nitrogen fixing
cyanobacterium-plant symbioses. Physiological and Molecular Plant Pathology
49: 73-77.
Lee KY, Joseph CM, Meeks JC. 1988.
Glutamine synthetase specific activity and
protein concentration in symbiotic Anabaena associated with Azolla caroliniana.
Antonie van Leeuwenhoek 54: 345-355.
Lindblad P, Bergman B. 1986. Glutamine
synthetase: Activity and localization in
cyanobacteria of the cycads Cycas revoluta
and Zamia skinneri. Planta 169: 1-7.
_ _.1989. Occurrence and localization of
phycoerythrin in symbiotic Nostoc of
Cycas revoluta and in the free-living isolate Nostoc 7422. Plant Physiology 89:
783-785.
Lindblad P, Hallbom L, Bergman B. 1985.
The cyanobacterium-Zamia symbiosis:
C2H 2 -reduction and heterocyst frequency.
Symbiosis 1: 19-28.
Lindblad P, Rai AN, Bergman B. 1987. The
Cycas revoluta-Nostoc symbiosis: Enzyme
activities of nitrogen and carbon metabolism in the cyanobiont. Journal of General
Microbiology 133: 1695-1699
Lynn ME, Bande JA, Ownby JD. 1986. Estimation of gene expression in heterocysts
of Anabaena variabilis by using DNARNA hybridization. Journal of Bacteriology 136: 2057-2065.
Meeks JC. 1990. Cyanobacterial-bryophyte
associations. Pages 43-63 in Rai AN, ed.
Handbook of Symbiotic Cyanobacteria.
Boca Raton (FL): CRC Press.
MeeksJC, Enderlin CS,Joseph CM, Chapman
JS, Lollar MWL. 1985. Fixation of [13N]N2
and transfer of fixed nitrogen in the
Anthoceros-Nostoc symbiotic association.
Planta 164: 406-414.
Meeks JC, Steinberg NA, Enderlin CS, Joseph CM, Peters GA. 1987. Azolla-Anabaena relationship. XIII. Fixation of
[13N]N2 • Plant Physiology 84: 883-886.
Nierzwicki-Bauer SA, Haselkorn R. 1986.
Differences in mRNA levels in Anabaena
living freely or in symbiotic association
with Azolla. EMBO JournalS: 29-35
Pate JS, Lindblad P, Atkins CA. 1988. Pathways of assimilation and transfer of fixed
nitrogen in coralloid roots of cycad-Nostoc symbioses. Planta 176: 461-471.
Perraju BTVV, Rai AN, Kumar AP, Singh
HN. 1986. Cycas circinalis-Anabaena
cycadeae symbiosis: Photosynthesis and
the enzymes of nitrogen and hydrogen
metabolism in symbiotic and cultured
Anabaena cycadeae. Symbiosis 1: 239251.
Peters GA, Meeks Je. 1989. The AzollaAnabaena symbiosis: Basic biology. Annual Review ofPlant Pilysiology and Plant
Molecular Biology 40: 193-210.
Rai AN. 1990a. Handbook of Symbiotic Cyanobacteria. Boca Raton (FL): CRC Press.
_ _. 1990b. Cyanobacteria-fungal symbioses: The cyanolichens. Pages 9-41 in
Rai AN, ed. Handbook of Symbiotic
Cyanobacteria. Boca Raton (FL): CRC
Press.
Rai AN, Rowell P, Stewart WDP. 1981a.
Nitrogenase activity and dark CO 2 fixa-
275
tion in the lichen Peltigera aphthosa Willd.
Planta 151: 256-264.
_ _. 1981b. 15Nz Incorporation and metabolism in the lichen Peltigera aphthosa
Willd. Planta 152: 544-552.
Rai AN, Rowell P, Stewart WDP. 1983. Interactions between cyanobacterium and
fungus during 15Nz-incorporation and
metabolism in the lichen Peltigera canina.
Archives of Microbiology 134: 136-142.
Rai AN, Borthakur M, Singh S, Bergman B.
1989. Anthoceros-Nostoc symbiosis:
Immunoelectronmieroscopie localization
of nitrogenase, glutamine synthetase, phycoerythrin and ribulose-l,5 bisphosphate
carboxylase/oxygenase in the cyanobiont
and the cultured (free-living) isolate Nostoc 7801. Journal of General Microbiology 135: 385-395.
Rasmussen U, J ohansson C, Bergman B. 1994.
Early communication in the GunneraNostoc symbiosis: Plant induced cell differentiation and protein synthesis in the
cyanobacterium. Molecular Plant-Microbe Interactions. 7: 696-702.
Ray TB, Peters GA, Toia RE Jr., Mayne BC.
1978. Azolla-Anabaena relationship. VII.
Distribution of ammonia-assimilating enzymes, protein, and chlorophyll between
host and symbiont. Plant Physiology 62:
463-467.
Ray TB, Mayne BC, Toia Jr. RE, Peters GA.
1979. Azolla-Anabaena relationship. VIII.
Photosynthetic characterization of the
association and individual partners. Plant
Physiology 64: 791-795.
Rippka R, Deruelles J, Waterbury JB,
Herdman M, Stanier RY. 1979. Generie
assignments, strain histories and properties of pure cultures of cyanobacteria. Journal of General Microbiology 111: 1-61.
Rodgers GA, Stewart WDP. 1977. The
cyanophyte-hepatie symbiosis 1. Morphology and physiology. New Phytologist 78:
276
441-458.
Sampaio MJAM, Rai AN, Rowell RP, Stewart
WDP. 1979. Occurrence, synthesis and
activity of glutamine synthetase in N zfixing lichens. FEMS Mierobiology Letters 6: 107-110.
Schenk HEA. 1992. Cyanobacterial symbioses. Pages 3819-3854 in Ballows A, Truper
HG, Dworkin M, Harder W, Schleifer KH, eds. The Prokaryotes. 2nd edition.
New York: Springer-Verlag.
Silvester WB. 1976. Endophyte adaptation in
Gunnera-Nostoc symbiosis. Pages 521538 in Nutman PS, ed. Symbiotic Nitrogen Fixation in Plants. Cambridge (UK):
Cambridge University Press.
Silvester WB, Parsons R, Watt PW. 1996.
Direct measurement of release and assimilation of ammonia in the Gunnera-Nostoc symbiosis. New Phytologist 132: 617625.
Smith AJ. 1982. Modes of cyanobacterial
carbon metabolism. Pages 47-85 in Carr
NG, Whitton BA, eds. The Biology of
Cyanobacteria. Oxford: Blackwell.
Soderback E, Bergman B. 1992. The NostocGunnera magellanica symbiosis: Phycobiliproteins, carboxysomes and rubisco in
the cyanobiont. Physiologia Plantarum 84:
425-432.
___.1993. The Nostoc-Gunnera symbiosis: Carbon fixation and translocation.
Physiologia Plantarum 89: 125-132.
Soderback E, Lindblad P, Bergman B. 1990.
Developmental patterns related to nitrogen fixation in the Nostoc-Gunnera
magellanica Lam. symbiosis. Planta 182:
355-362.
Stacey G, Burris RH, Evans HJ. 1992. Biologieal Nitrogen Fixation. New York:
Chapman & Hall.
Steinberg NA, Meeks JC. 1989. Photosynthetic CO z fixation and ribulose bisphosphate carboxylase/oxygenase activity of
Nostoc sp. strain UCD 7801 in symbiotic
association with Anthoceros punctatus.
Journal of Bacteriology 171: 6227-6233.
_ _.1991. Physiologieal sources of re ductant for nitrogen fixation activity in Nostoc sp. strain UCD 7801 in symbiotic
association with Anthoceros punctatus.
Journal of Bacteriology 173: 7324-7329.
Stewart WDP, Rowell P. 1977. Modifications of nitrogen-fixing algae in lichen
symbioses. Nature 265: 371-372.
Summers ML, Wallis JG, Camp bell EL, Meeks
JC. 1995.Genetic evidence of a major role
for glucose-6-phosphate dehydrogenase
in nitrogen fixation and dark heterotrophie
growth of the cyanobacterium Nostoc sp.
strain ATCC 29133. Journal of Bacteriology 177: 6184-6194.
Tandeau de Marsac N. 1994. Differentiation
of hormogonia and relationships with
other biologieal processes. Pages 825-842
in Bryant DA, ed. The Molecular Biology
of Cyanobacteria. Boston: Kluwer Academie Publishers.
Theil T, Lyons EM, Erker JC, Ernst A. 1995.
A second nitrogenase in vegetative cells of
a heterocyst-forming cyanobacterium.
Proceedings of the National Academy of
Sciences of the Uni ted States of America
92: 9358-9362.
Thomas J, Meeks JC, Wolk CP, Shaffer PW,
Austin SM, Chien W-S. 1977. Formation
of glutamine from [13N]ammonia,
(13N]dinitrogen and (14C]glutamate by heterocysts isolated from Anabaena cylindrica. Journal of Bacteriology 129: 15451555.
Wolk CP. 1996. Heterocyst formation. Annual Review of Genetics 30: 59-78.
Wolk CP, Ernst A, Elhai J. 1994. Heterocyst
metabolism and development. Pages 769823 in Bryant DA, ed. The Molecular
Biology of Cyanobacteria. Boston: Kluwer
Academic Publishers.
BiaScience Val. 48 No. 4