Biology Reference
In-Depth Information
3 of NADH dehydrogenase (Hanson et al., 1999). The sunÞower toxin has similarities with subunit
8 of ATP synthase, and the alpha subunit of ATP synthase acts as the antidote (Horn, 2002). By
comparison with bacterial toxins, the mitochondrial proteins are on the large side with some 150
amino acids, and all are membrane proteins. With the exception of CMS in the bean, the toxinÏanti-
toxin units are constitutively expressed in all tissues of the plant without any obvious negative effects.
These systems show a high degree of polymorphism. In sunÞowers, 64 different systems are known.
All these systems have very likely originated from genetic recombination, which must be very common
in plant mitochondria (Budar and Pelletier, 2001; Knoop and Brennicke, 2002).
Jacobs proposed that subunit 8 of the mitochondrial ATP synthase (Atp8) originated from a
bacterial toxinÏantitoxin system. Atp8 has no known homolog in any prokaryotic or plastid ATP
synthase but shows similarity to members of the Hok family of toxins mentioned earlier, some of
which are involved in post-segregational killing in bacteria (Jacobs, 1991).
A good indication that these CMS systems are nothing else but selÝsh systems can be seen in
the fact that they have always elicited a host response that tries to overcome male sterility. This also
questions the role of gynodioecy in the evolution of separate sexes. The host response manifests in a
variety of nuclear restorers. The large majority of these nuclear-restorer genes presumably act on the
maturation of the mitochondrial mRNA of the toxinÏantitoxin unit. The molecular mechanisms are
not yet understood. However, one of the nuclear restorers of maize that does not act on the mRNA
level has recently been identiÝed as aldehyde dehydrogenase (ALDH). Soluble ALDH accumulates
in the mitochondrial matrix. This led to the hypothesis that the toxin in Texas CMS might cause an
increased inÞux of electrons reducing the redox components of the chain, which would lead to an
overproduction of reactive oxygen such as superoxide. Superoxide will react with, among other things,
polyunsaturated fatty acids in the membrane, and one of the end products is malondialdehyde, which
itself is reactive and, eventually, the mitochondrion will induce the killing of the cell. ALDH in the
mitochondrial matrix could prevent the accumulation of dangerous levels of malondialdehyde and
CMS (Schnable and Wise, 1998; Liu et al., 2001; M¾ller, 2001; Schnable, 2001).
I
NTERCELLULAR
M
ECHANISMS
Intercellular means here that one of two initially independent cells tries to kill the other if it does
not have the same toxinÏantitoxin system. With the example of the bacteriocins we would like to
stress again that it is an ancestral mechanism that has proven itself over time to such a degree that
most bacteria have retained it. For this reason, we should expect its action in intracellular symbiotic
bacteria as well, where it may contribute to quorum sensing or CI.
Bacteriocins
Bacteriocins are proteic toxins that are exported from the cell to kill bacteria of the same or related
species that do not carry that toxin system. Two classes are generally recognized. Class I members
are often encoded by a gene cluster on a large conjugative transposon, usually cysteine rich and
highly post-translationally modiÝed. They contain b-c.unsaturated amino acids like dehydroalanine
or dehydrobutyrine and typically intramolecular thioester bridges known as c.methyl lanthionines,
which are formed between the sulfhydryl group of a cysteine and the double bond of a dehydroamino
acid. This has led to the name of lantibiotics. They are common in Gram-positive bacteria. They
have a rather broad spectrum of antimicrobial activity, which has led to a wide range of applications
in food and feed preservation. Lantibiotics act through pore formation in the cytoplasmic membrane.
The antitoxin, in these cases often called the immunity factor or immunity protein, is a lipoprotein
encoded by the same gene cluster. It prevents pore formation. The expression of the system is
highly (auto)regulated in a way that allows quorum sensing in the bacterial population.
Class II bacteriocins are not post-translationally modiÝed; they are mainly plasmid coded and
show a much narrower host range and antimicrobial activity. Bacteria that carry a bacteriocinogenic
