Marine teleosts, unlike their freshwater counterparts, have a repressed ability to
synthesise long chain highly unsaturated fatty acids (HUPA). In competent species,
the A6 and A5 fatty acid desaturases are critical in the biosynthetic pathway that
produces the HUFA’s arachidonic acid (20:4/z-6; AA), eicosapentaenoic acid (20:5n3; EPA) and docosahexaenoic acid (22:6/z-3; DHA) from the Ci8 polyunsaturated
fatty acids (PUFA), linoleic acid (18:2/2-6) and a-linolenic acid (18:3/2-3). The
deficiency in HUFA biosynthesis in marine fish is of considerable practical
significance because, in consequence, farmed marine species require a dietary source
of presynthesised HUFA. This is provided by processed products from “industrial”
species of marine fish such as sand eel, sardine, capelin and anchovies which
themselves obtain HUFA through the food chain. Indicators suggest that the wild
fishery supporting the aquaculture feed industry is unsustainable at current levels of
exploitation. This has consequential effects on human health as fish, especially
marine fish, are the predominant dietary source of HUFA that are crucial for
maintaining cell membrane integrity as well as being central to eicosanoid
metabolism.
Therefore, the primary aims of this project were to further our understanding
of the molecular differences in HUFA biosynthesis between marine and freshwater
teleosts. This was achieved by comparing the fatty acid desaturase genes of
representative marine and freshwater fish. The desaturases are enzymes involved in
the biosynthesis of HUFA from PUFA and have been considered as one of the steps
that may be compromised in marine fish. The desaturase genes were studied with a
view to relating structural, and potential functional differences with different HUFA
synthesis phenotypes.
During the course of this project sequences of putative desaturase genes were
cloned from two freshwater (zebrafish and carp), two marine (turbot and cod) and
one anadromous fish species (Atlantic salmon). Once translated, the protein
sequences of all the gene products contained all the necessary domains and motifs
shown to be required for efficient desaturase function including an N-terminal
cytochrome bs domain, and three catalytically important histidine boxes conserved in
all members of the gene family. They all included the variant third histidine box that
seems typical of A5 and A6 desaturase genes described to date. All of the protein
sequences from the fish species had greatest homology to the mammalian
desaturases, specifically the human A6 desaturase.
The cDNAs of salmon, carp and zebrafish were functionally characterised in
Saccharomyces cerevisiae. Three carp transcripts were sequenced and functionally
characterised. Two had no A5 or A6 desaturase activity, while the third efficiently
desaturated 18:3/2-3 at the A6 position. Of the two functionally characterised salmon
transcripts one had no A5 or A6 activity whereas the third efficiently desaturated
20:4/2-3 at the A5 position. The transcripts that had no desaturase activity were
considered either non-functioning alleles or pseudogenes acquired as a result of a
genome doubling event. It is believed that other A5/6 like desaturases probably exist
for both carp and salmon as salmon is known to have high levels of A6 desaturase
activity. However, neither the cod nor the turbot cDNAs were functionally
characterised in yeast.
The most significant result of the functional characterisation study concerned
the zebrafish (Danio rerio). The 1590 bp transcript has close similarity to
mammalian A6 desaturase. However, the clone encodes a novel desaturase. When
expressed in yeast the zebrafish gene confers the ability to convert 18:2/2-6 and
18:3/2-3 to their corresponding A6 desaturase products, 18:3/2-6 and 18:4/2-3. In
addition, it confers the ability to convert 20:3/2-6 and 20:4/2-3 to their A5 desaturase
products, 20:4/2-6 and 20:5/2-3, respectively. Therefore, the zebrafish gene encodes a
bi-functional enzyme having both A6 and A5 desaturase activity. This was the first
report of a functionally characterised desaturase of fish, and, in particular, of a fatty
acid desaturase with both A6 and A5 activity.
The structure of the primary sequences of the fish desaturases were analysed
in relation to function and some interesting and potentially highly significant
relationships were discovered. However, it was not possible to determine which
residue or residues were responsible for the differing substrate specificities between
the transcripts.
In summary, the results presented in this thesis indicate that (i) all the fish
species used in this study possessed desaturase-like sequences (ii) the zebrafish
contains a novel, unique desaturase enzyme with both A6 and A5 desaturase activity
(iii) marine fish possess A5/6 desaturase-like transcripts (iv) there is some evidence
that fish species that have undergone tetraploidy or recent genome duplication appear
to have duplicated genes, possibly pseudogenes and/or non-functioning alleles (v)
significant differences in primary structure which may have important consequences
for function were observed although unequivocal identification of residues
responsible for determining function or specificity was not possible.
In conclusion, this study has produced results that not only further our
understanding of the fatty acid genes of fish but which also furthered our knowledge
of the fatty acid desaturases in general. The data will facilitate studies of how fatty
acid desaturase primary structures relate to function. Information from this and other
studies will lead to complete knowledge of how sequence and structure contribute to
confer substrate specificity and how the fatty acid desaturase gene family has
evolved