ARTICLE
Introduction
A large family of structurally related chemotactic cytokines (the chemokines)
is responsible for controlling the migration of leucocytes during an inflammatory
response or infection. Within this family are four subgroups that are
characterised depending on the arrangement of the cysteine residues occurring
near the N-terminus of their mature peptides. These groups are
commonly known as CXC (a), CC (b), C (g) and CX3C (d) [1].
The three dimensional structures of many chemokines belonging to the CXC
and CC groups have been determined and appear similar, comprising three
b-strands in a Greek key formation and an overlying a-helix stabilised
by two disulphide bonds between the four conserved cysteines [2, 3]. CXC
chemokines can be further classified depending on whether or not they
possess a Glu-Leu-Arg (ELR) motif adjacent to the CXC motif. Those CXC
chemokines in possession of an ELR sequence (e.g. interleukin-8) show
neutrophil-attracting ability [4, 5], whereas those lacking an ELR motif
specifically attract lymphocytes [6, 7].
Three non-ELR CXC chemokines that are closely related according to their
peptide sequences are monokine induced by interferon-g (Mig), interferon-inducible
protein 10 (IP-10) and interferon-inducible T-cell alpha chemoattractant
(I-TAC) known, under the new nomenclature, as CXCL9, CXCL10 and CXCL11
respectively [1]. These three chemokines selectively share a common receptor,
CXCR3, that is expressed on T cells, B cells and natural killer cells
with preferential expression apparent on Th1 type T cells. Activity of
these ligands and expression of the CXCR3 receptor is intensely increased
by T cell activation suggesting a role in inflammatory processes for these
molecules [1]. Furthermore, interferon-g (IFN-g), but not TNF-a or IL-1b
alone, nor Th2-type cytokines (IL-4, IL-10, and IL-13), has been shown
to greatly increase the expression of CXCL9, CXCL10 and CXCL11 consistent
with the association of these chemokines with Th1 responses [7, 8, 9].
Of this small subset of chemokines, CXCL10 was the first identified. Modulation
of CXCL10 mRNA expression by IFN-g has been observed chiefly in endothelial
cells, monocytes, fibroblasts, astrocytes, keratinocytes and neutrophils
[10, 11, 12, 13]. Later, CXCL9 was identified as a chemokine expressed
in IFN-g-induced THP-1 cells, PBMCs, endothelial cells, keratinocytes,
and fibroblasts with 37% amino acid identity to CXCL10 [8, 14]. CXCL11,
a chemokine expressed by activated monocytes and astrocytes, was discovered
most recently and is approximately 40% identical to CXCL9 and CXCL10 at
the amino acid level [7]. This latter chemokine appears to have the highest
affinity of the three for the CXCR3 receptor.
Little is known about the cytokine network of fish relative to that
known for mammals. However, recent progress has been made in isolating
the genes for important cytokines from several species of fish, including
interleukin-1b (IL-1b)[15, 16], tumour necrosis factor-a (TNF-a)[17, 18]
and three isoforms of transforming growth factor-b [19, 20, 21, 22]. In
addition, several chemokine genes have been isolated from fish which are
homologous to CXC [23,24,25,26] and CC chemokines [27, 28]. The presence
of similar cytokine genes in fish and mammals suggests some evolutionary
conservation of the complexities of the immune system although further
investigation into the activity of such genes within fish is required.
Although it is known that fish possess chemokines from both the CC and
CXC subgroups there are no published data concerning the CXCL9, CXCL10
or CXCL11 chemokines in non-mammalian vertebrates. However, whilst searching
for sequences with similarity to chemokines in the EST databases, a sequence
with highest amino acid identity to IP-10 (39%) was identified from a
catfish Ictalurus punctatus brain cDNA library (Accession
No. BE212851). Sequence information from this EST was used to isolate
a similar sequence from the rainbow trout Oncorhynchus mykiss that
is reported here. Analysis of the mRNA expression of the trout gene was
investigated in various tissues and the effect of poly I:C, and LPS upon
levels of expression studied. The implications for the presence of a chemokine
that controls the trafficking of activated T cells during an inflammatory
response in fish are discussed.
Methods and materials
Isolation of a gIP-10-like sequence from a trout
cDNA library
The full cDNA sequence of trout g-interferon inducible protein (gIP)
was isolated from a rainbow trout kidney leucocyte cDNA library, prepared
as described previously [20]. An oligonucleotide primer (IP10-R2) (Table
1) specific for an EST sequence of the catfish was initially used in an
anchored polymerase chain reaction (PCR) with the vector-specific T3 primer.
The sequence derived from the resulting product was used to design trout
specific primers that were used to amplify the remainder of the molecule.
A second product was amplified by semi-nested PCR using an initial round
of amplification with the IP10-67-F2 primer with the vector-specific primer
T7, followed by a further round of PCR using the T7 primer with the IP10-67-F1
primer. All primers are listed in Table 1 and their relative positions
depicted in Figure 1. To ensure
that both fragments were derived from an individual sequence, a third
product was amplified from the cDNA library using the IP10-67-F2 primer
and IP10-67-R2 primer that spanned the full coding sequence of the putative
trout gIP cDNA, and most of the 3' and 5' UTRs. The latter PCR was repeated
to amplify gIP cDNA from a rainbow trout macrophage (RTS-11) cell-line
that was derived from the spleen of sexually immature trout [29].
Isolation of the genomic sequence for trout gIP
Genomic DNA was isolated from RTS-11 cells using a proteinase K
method as described previously [30]. An extraction mixture containing
500 mL 10X TNE buffer [0.01 M Tris-HCl (pH 8.0); 0.1 M
NaCl; 0.01 M EDTA; all Sigma, UK], 50 mL 1M Tris-HCl (pH 8.0)
and 20 mL 25% sodium dodecyl sulphate (SDS; Sigma, UK) was added
to cell pellets and the mixture incubated at 55¡C in the presence of proteinase
K (20 mg.mL - 1; Promega, UK) with regular vortexing.
After 3 h, reactions were incubated at 37¡C with RNase (10 mg.mL - 1;
Sigma, UK) for 1 h and two phenol-chloroform extractions were performed.
Fifty mL sodium acetate solution (3M; Sigma, UK) and 1 mL cold ethanol
(100%; Sigma, UK) were added to the DNA fraction. Samples were allowed
precipitate overnight at - 20¡C and DNA recovered by centrifugation
at 10000 « g for 15 min. The DNA pellet was washed with
200 mL cold ethanol (70%) at 10000 « g for 5 min.
Excess ethanol was removed and the dried DNA pellet dissolved to a concentration
of 250 ng.mL - 1 in sterile milliQ H2O.
PCR was performed under the conditions described above where 2 mL
of genomic DNA was used as template utilising primers designed in the
5' (IP10-67-F2) and 3' (IP10-67-R2) ends of the rainbow trout gIP cDNA
sequence (see Table 1 and Figure
1). The IP10-67-F1 and IP10-67-R1 primers were used to assist the sequencing
through the central portion of the trout gIP genomic clone.
Cloning and sequencing
All products obtained by PCR were directly cloned into the pGEM-T Easy
vector (Promega, UK). Plasmid DNA was isolated from bacterial colonies
that had an appropriately sized insert using a Qiaprep spin miniprep kit
(QIAGEN, UK). Three randomly selected clones representing each product
were sequenced on an ABI 377 Automated Sequencer (Applied Biosystems)
using vector or gene-specific primers. Comparisons of nucleotide and amino
acid sequences with EMBL and SWISSPROT databases were performed using
the FASTA [31] program. Direct comparisons between two sequences were
performed using the NEEDLE global alignment program [32] within the EMBOSS
suite of programs (http://www.uk.embnet.org/Software/EMBOSS/) from which
overall percentage values were recorded. Multiple sequence alignments
were generated using CLUSTALW (version 1.74) [33]. Phylogenetic trees
were constructed from CLUSTAL generated alignments of CXC chemokine peptide
sequences using the Neighbour-joining method [34] within the MEGA 2.1 program
[35]. Distance options were set for Poisson Correction and pairwise deletion
was used for handling gaps in the alignment. Trees were bootstrapped 1000 times
to test phylogeny and then displayed using the TREEVIEW program [36].
The SIGNALIP program [37] was used to predict the position of signal peptide
cleavage.
Expression studies
Rainbow trout weighing 300-350 g were obtained from a local fish
farm and maintained in a recirculating freshwater system at 14¡C with
regular feeding. Three fish were killed and tissues (blood, brain, gill,
liver, spleen, heart and head kidney) were isolated immediately and snap
frozen in liquid nitrogen. Total RNA was extracted from the frozen tissues
using 0.5-1.0 mL RNAzol B (Biogenesis, UK) as previously described
[20]. First strand cDNA was synthesised from 5 mg total RNA using
Moloney murine leukemia virus (MMLV) reverse transcriptase (Gibco, UK)
at 42¡C for 50 min using the oligo-dT primer (Gibco) and PCR performed
on the resultant cDNAs using the IP10-67-F2 and IP10-67-R1 primers. These
primers were specific for trout gIP (Table
1) and spanned intron 1 of trout genomic gIP. PCR reagents were assembled
as described above and the cycling protocol was 1 cycle of 94¡C for
2 min, 35 cycles of 94¡C for 1 min, 60¡C for 1 min
and 72¡C for 10 s, with a final extension step of 5 min at 72¡C.
As a positive control for RT-PCR, b-actin was amplified using trout specific
b-actin FOR and b-actin REV primers listed in Table 1 using 30 PCR
cycles and an annealing temperature of 55¡C. Amplified products were analysed
on a 2% agarose gel containing ethidium bromide (100 ng.mL - 1).
In vitro regulation of gIP was also investigated using the RTS-11 cell-line.
RTS-11 cells were incubated in vitro at 18¡C with poly I:C
(50 mg.mL - 1; Sigma, UK) or with E. coli
0127:B8 LPS (5 mg.mL - 1; Sigma, UK) for 2,
4 or 8 h. Control cells were incubated with an equivalent volume
of phosphate-buffered saline (PBS, Gibco, UK). The cells were harvested
using trypsin-EDTA (Sigma, UK) and total RNA was extracted as above. Semi-quantitative
RT-PCR was used to compare the level of the gIP transcript in each sample
to the b-actin transcript level. Both gIP and b-actin genes were co-amplified
in the same PCR reaction using 30 cycles of 94¡C for 30 s, 55¡C
for 30 s and 72¡C for 10 s, with the primer combinations described
for expression PCR above. Similarly, the in vitro and in vivo
effects of recombinant trout TNF-a (rtTNFa) [17, 18] on gIP expression
in trout cells were investigated using recombinant proteins of both isoforms
discovered in this species. Thus, RTS-11 cells were incubated at
18¡C for 5 h in the presence of either rtTNFa1 or rtTNFa2 at 0.1,
1.0, 10.0 or 100.0 ng.mL - 1 in EB buffer (Qiagen
Ni-NTA kit, Qiagen). A control reaction was performed in which an equivalent
volume EB buffer was added to RTS-11 cells. Total RNA was extracted
and semi-quantitative PCR used to determine changes in gIP expression
as described above. To investigate the effect of rtTNF-a in vivo,
trout were injected intraperitoneally with either rtTNF-a (a mixture comprising
30 mg rtTNF-a1 plus 55 mg rtTNFa2) or PBS. After 24 h the
head kidneys were removed and RNA extracted as above. Semi-quantitative
RT-PCR for gIP was performed as above.
Results
Cloning and characterisation of rainbow trout gIP
Anchored PCR on a rainbow trout cDNA library, using a primer designed
against an EST sequence of the catfish that resembled the CXCL10 chemokine,
generated a product containing 268 bp of a cDNA sequence. Comparison
of this sequence with the EMBL and Swissprot databases (using the FASTA
program) confirmed this sequence as a CXC type chemokine with highest
similarity to CXCL10. Further primers, specific for the amplicon, gave
rise to a second fragment when used in semi-nested PCR that contained
632 nucleotides. The two products overlapped to allow the deduction
of a contiguous sequence containing 787 nucleotides (Figure
2). Of these 787 bp, nucleotides 1-53 and 357-787 formed the
5' and 3' untranslated (UTR) regions respectively whilst the remaining
303 nucleotides formed an open-reading frame (ORF) predicted to encode
a peptide of 100 amino acids. A hydrophobic stretch incorporating
the first 24 residues was predicted to be a signal peptide using
the SignalIP program. Two polyadenylation signal sequences were present
within the 3' UTR at nucleotide positions 659-664 and 748-753 and a poly-A
sequence continued from nucleotide 769. The third product amplified, beginning
in the 5'UTR and ending in the 3'UTR, confirmed the two products were
derived from the same message. A similar sequence spanning the entire
coding region was amplified from RTS-11 cells. In addition, a second
cDNA sequence with 6 nucleotide differences in the coding region
(resulting in 3 amino acid changes), and several differences in the
3'UTR (mainly due to the presence of additional repeats of a 11 bp
stretch) was discovered in the RTS-11 cell-line. The amino acid substitutions
are Gln34 to Leu34, Lys95 to Asn95
and Gln96 to Arg96. The 5'UTRs of both transcripts
were identical. The original transcript will be referred to as gIP1 and
the second transcript identified from RTS-11 cells will be referred
to as gIP2.
The genomic sequence of trout gIP was isolated from the RTS-11 cell
line using the IP10-67-F2 and IP10-67-R2 primers. A product of 1934 nucleotides
was obtained and comparison to the cDNA sequence revealed the presence
of 3 introns (200 bp, 240 bp and 657 bp respectively)
between 4 exons (Figure
3). The genomic sequence isolated was identical to the gIP2 cDNA sequence
within its exons. Conventional intron splice sites gt-ag mark the boundaries
of all three introns.
Comparison of the trout gIP1 propeptide sequence with mammalian CXC
chemokines using the NEEDLE program showed approximately 30% amino acid
identity to a range of chemokines (CXCL1, CXCL2, CXCL3, CXCL8, CXCL10
and CXCL11), with 14-27% amino acid identity to other mammalian CXC chemokines
(Table 2). However, when comparing the predicted peptide regions
encoded by exon 2, CXCL10 is clearly the most closely related human CXC
chemokine to trout gIP, with over 37% amino acid identity; other chemokines
having between 17 and 31% amino acid identity within this region. Comparison
of the full length trout gIP translation with trout IL-8 [26] revealed
only 31% amino acid identity between the two trout chemokines, whereas
44% amino acid identity was shared between the catfish CXCL10-like sequence
and trout gIP.2
Multiple alignment of the putative translations of trout gIP1 and gIP2
with human CXCL9, CXCL10, and CXCL11 and the catfish CXCL10-like sequence
reveals the conserved positions of the 4 cysteine residues (Figure
4) and the predicted position of signal peptide cleavage that follows
a Gly residue in each case. Highest identity was observed in the central
portion of the chemokines, with more variation in peptide sequence at
both the N-terminal and C-terminal ends.
Phylogenetic analysis using the neighbour-joining program reveals that
the trout gIP1 peptide groups with the catfish CXCL10-like molecule and
human CXCL9, CXCL10 and CXCL11, although poor bootstrap values are given
(Figure 5). The clade of g-IFN
inducible chemokines groups distinctly from IL-8 molecules that include
trout and flounder IL-8.
Expression studies
RT-PCR showed tissue-specific expression of gIP mRNA in various tissues
of trout (Figure 6). Expression
of gIP was particularly noted in gill, with some low level expression
detected in the spleen, head kidney and liver. The brain and heart did
not appear to express gIP. Using a semi-quantitative RT-PCR, it was apparent
that poly I:C induced high expression levels of gIP in RTS-11 cells
after 4 h, which continued to be expressed at high levels at 8 h
(Figure 7). No induction
was observed above basal levels after only 2 h incubation with poly
I:C. By contrast, no induction of gIP was observed in RTS-11 cells
that were incubated over the same time range in the presence of LPS when
compared to control cells (Figure
7). No induction of gIP was observed in response to rtTNFa1 or rtTNFa2
in the RTS-11 cell-line even at relatively high concentrations (100 ng.mL - 1)(Figure
8A). However, in vivo expression of gIP could be induced in trout,
as seen in the head kidney following intraperitioneal injection with rtTNFa
(Figure 8B). In all RT-PCR experiments,
b-actin bands were successfully amplified.ÊÊ
Discussion
A sequence of 787 nucleotides, isolated from a rainbow trout leucocyte
cDNA library, was found to resemble CXCL10. This sequence (gIP1) translated
in a single open reading frame (ORF) to produce a putative peptide of
100 amino acids that contained four cysteine residues, characteristic
of most chemokines, and essential to formation of the tertiary structure.
The first two cysteines were separated by one residue (Gln), thus adopting
the CXC-type chemokine arrangement. There was no ELR motif preceding the
CXC motif, consistent with observations of mammalian CXCL9, CXCL10 and
CXCL11 and other non-ELR CXC chemokines that are inactive toward neutrophils
[1]. Signal peptide cleavage was predicted to occur following residue
24 (Gly), which is in accord with the observed signal peptide cleavage
position of CXCL9, CXCL10 and CXCL11 [38, 39], and suggests that the trout
chemokine is also a secreted molecule. Within the 431 nucleotides
of the 3' UTR, two potential polyadenylation signal sequences (AATAAA)
[40] were observed, one 16 nucleotides and the other 105 nucleotides
upstream from the poly-A sequence. Whether both could be functional is
unclear, however, no transcripts ending at the first polyadenylation signal
sequence were isolated. An 11 nucleotide stretch within the 3'UTR
(gaagactgcct) is repeated 11 times, sometimes inexactly, and has
been observed, in other isolates of gIP1, to vary in frequency thus producing
transcripts of differing lengths. This phenomenon appears to be common
for rainbow trout cytokines, also occurring for both IL-8 and TNF-a molecules
(Laing, personal observation). A second copy of this sequence (gIP2) was
isolated from the RTS-11 cell-line in addition to a sequence identical
to gIP1 isolated from the leucocyte cDNA library. With 6 nucleotide
differences in the coding region, gIP2 potentially had 3 amino acid
substitutions within its peptide structure, all in positions that tend
to be relatively non-conserved between mammalian CXCL9, CXCL10 and CXCL11.
Although the complete 3'UTR was not obtained for this transcript (it ends
in the region of the IP10-67-R2 primer that resides 10 bp upstream
from the first polyadenylation signal sequence), it was apparent that
the length of the 3'UTR was substantially longer in gIP2 than in gIP1.
This was mainly due to the presence of more copies (27 versus 11)
of the 11 nucleotide repeat sequence mentioned above, but various
other differences between the Ôpolymorphic region' and the IP10-67-R2
primer position were also observed. The short 5'UTR sequences (containing
53 bp) of both transcripts were identical. Whether these two different
gIP sequences represent different functional genes or merely two alleles,
still requires verification. However, two genes for IL-1b and TNF-a have
been observed in rainbow trout [41, 42] that appear to be regulated via
different mechanisms, perhaps reflecting different functions of these
cytokines. Indeed, it has been noted that fish exhibiting polyploidy (as
is the case for rainbow trout), commonly possess multiple copies of some
genes [43] that do not become deleted from the genome because they adopt
new, fundamental roles within the organism. Single amino acid substitutions
in mammalian CXC chemokines have been shown to alter the specificity and/or
potency of a chemokine to a particular receptor [44], potentially altering
the action of the chemokine. Thus, the 3 different amino acids between
gIP1 and gIP2 could result in functional differences between these molecules.
The genomic sequence of gIP2 was sequenced from RTS-11 cells, and
was shown to contain 4 exons and 3 introns. The introns (which
contain 200, 240 and 657 nucleotides in introns 1-3 respectively)
divide the coding sequence in conserved positions when compared to mammalian
CXCL10 and CXCL11 genes. Exon 1 encodes the signal peptide region,
containing 70 nucleotides in trout gIP2 and 61 nucleotides in
mouse CXCL10 and human CXCL11 (Figure
9) accounting for the signal peptide of trout gIP containing 3 more
amino acids than both mammalian CXCL10 and CXCL11. Exon 2 of mammalian
CXCL10 and CXCL11, in both cases, contains 127 nucleotides, identical
to the number of nucleotides present in the second exon of trout gIP.
The lengths of the final two exons are more variable, reflecting the difference
observed in the C-terminal ends of CXC chemokines. Mouse CXCL10 contains
90 nucleotides in exon 3 and 19 in exon 4, whereas exons 3 and 4
of human CXCL11 are 73 and 25 bp long respectively. Exon 3 and 4
of trout gIP more closely resemble those of CXCL10, with 93 nucleotides
and 12 nucleotides in each respectively. The partial genomic sequence
was also obtained, from RTS-11 cells, for gIP1, revealing differences
in the lengths of introns between the two trout sequences in addition
to differences already observed in the cDNA sequences. The 2nd
intron of gIP1 contains 253 nucleotides compared to the 240 bp
of intron 2 of gIP2. This finding confirms that differences observed between
the two cDNA sequences could not be the result of sequencing error.ÊÊÊ
Comparisons of trout gIP1 with mammalian CXC chemokines were made using
the NEEDLE program. When the complete peptide sequences were compared,
no individual chemokine displayed an amino acid identity that was clearly
higher than others - human CXCL1, CXCL2, CXCL3, CXCL8, CXCL10
and CXCL11 all displayed approximately 30% amino acid identity to trout
gIP1 whilst the rest exhibited identity values of 14-27%. Amino acid sequences
of the predicted regions encoded by exon 2 were also compared. This
region has been shown to be the essential portion for CXC chemokine activity,
as removal of the other regions does not inhibit activity [5]. This comparison
revealed that, trout gIP1 clearly showed highest identity to human CXCL10
(37%) with the values of amino acid identity to exon 2 of other CXC chemokines
being lower by approximately 6% or more. These values suggest the functional
portion of trout gIP is more like CXCL10 than any other chemokine. When
comparing trout gIP1 to the trout IL-8 precursor, only 31% amino
acid conservation occurred between the 2 molecules. However, the
translated catfish CXCL10-like EST sequence shared 44% amino acid identity
to gIP1. This is indicative of a closer relationship between the catfish
sequence than trout IL-8 to trout gIP.
Alignment of trout gIP1 and gIP2 and the catfish CXCL10-like EST sequence
with the peptide sequences of human CXCL9, 10 and 11 revealed conservation
of the cysteine residue positions and the position of signal peptide cleavage.
The CXCL9 sequence is much longer at the C-terminal end (it contains 125 amino
acids compared to 100 residues of trout gIP) [45], and this is reflected
in the low values of overall identity relative to those of CXCL10 and
CXCL11 when compared to trout gIP. From the alignment it can be seen that
most conserved residues reside in the central portion of the chemokine,
with relatively less conservation at the N-terminal (signal peptide) and
C-terminal ends. This central portion corresponds to the "loop" region
speculated to contain the second binding site of CXC chemokines (the 3 amino
acids immediately upstream of the CXC motif being the first binding site)
that is more ordered than the N-terminus and contains residues essential
to creating the correct tertiary structure of the chemokine [2]. As most
of this region is encoded by exon 2, it follows that the sequence encoded
by this exon would be likely to remain highly conserved for each particular
chemokine, as suggested by the values for amino acid identity described
from global alignments above.
Phylogenetic comparison of mammalian and fish CXC chemokines provides
further evidence of the relationship between trout gIP and the CXCL9/CXCL10/CXCL11
group of chemokines. The putative trout precursor peptide (together with
the catfish EST) groups with human CXCL9, CXCL10 and CXCL11 in one clade,
branching more closely to CXCL11, although the bootstrap values are too
low to allow a definite statement as to which of the human chemokines
is most related. The trout IL-8 molecule groups along with other
published IL-8 peptide sequences in a separate clade, supporting
that gIP and IL-8 of trout are distinct CXC chemokines. Analysis of the
data using the PROTPARS program also grouped trout gIP with CXCL10 and
CXCL11, although bootstrap support was once again rather poor (data not
shown), increasing the confidence that the correct grouping of this molecule
has been obtained. The low bootstrap values most likely result from the
high divergence rate that is common for chemokines and their receptors
[46]. After the discovery that mouse CXCL11 resided in a tandem arrangement
with the CXCL9 and CXCL10 genes it was speculated that CXCL9, CXCL10 and
CXCL11 arose by gene duplication from a common precursor early in mammalian
evolution [39]. If this is the case, perhaps there is only one representative
of this subset of chemokines in fish (i.e. trout gIP may have evolved
from the precursor to CXCL9, CXCL10 and CXCL11). This question may only
be answered when the full genome of a fish species has been sequenced.
Expression studies reveal a widespread expression of gIP in trout tissues,
with constitutive expression mainly in the gill, and some low level expression
in spleen, head kidney and liver. Not all tissues expressed gIP, though,
as shown by an absence of this transcript in the brain and heart. Some
degree of variation could be observed in the levels of expression between
different fish individuals, although the exact reasons for this are unclear.
Primers used for these expression studies would amplify both sequences
reported in this study, hence differences are unlikely to result from
differential expression/polymorphic factors. Constitutive expression of
CXCL10 has also been observed in mammalian lymphoid organs, such as spleen
and thymus [47], although expression can be induced in a wide array of
tissues following an immune stimulus (e.g. brain, eye, lung, intestine,
liver, heart, muscle, blood vessels and lymph nodes) [48]. The CXCL11
message does not appear to be constitutively expressed yet is highly expressed
in the lung, heart, small intestine and kidney in humans during endotoxemia,
differing from expression patterns of CXCL9 and CXCL10 [39]. Endotoxemia-induced
CXCL9 expression is greatest in liver.
The regulation of trout gIP expression in response to various stimuli
was compared to the effects similar factors had on levels of mammalian
CXCL9, CXCL10 and CXCL11 expression. The effects of LPS and poly I:C upon
gIP expression in a rainbow trout macrophage cell-line (RTS-11) were investigated.
Semi-quantitative RT-PCR showed that gIP expression levels (relative to
b-actin levels) were increased dramatically above basal levels in RTS-11 cells
after 4 h of exposure to poly I:C (50 mg.mL - 1)
and was maintained at high levels after 8 h. This is consistent with
observations in cultured mammalian macrophages as poly I:C caused a dramatic
increase in CXCL10 expression in murine macrophages [49]. However, no
effect of LPS (5 mg.mL - 1) was observed in
the RTS-11 cells up to 8 h post-stimulation. This lack of response
cannot be attributed to a lack of receptors for LPS in this cell-line
since previous studies have shown induction of IL-1b expression in RTS-11
cells following LPS challenge [29]. In contrast, there have been some
reports of LPS induction of CXCL10 in mammalian macrophages [50, 51] that
are believed to be mediated via the release of IFN-a/b from macrophages
[48]. Peak expression of CXCL10 could be obtained by 4 h, declining
rapidly by 8 h and returning to basal levels after 16 h [51].
LPS does not induce the expression of all three IFN-g-inducible chemokines
in all mammalian cells, however. LPS failed to induce CXCL9 release from
human monocytes after 8 h [45], and only a small increase in CXCL11
expression was observed in the mouse RAW (macrophage) cell-line. In
vivo expression also varies in mammals following challenge with LPS.
Expression of human CXCL9 and CXCL11 in the lung, following injection
of LPS, is delayed compared with that of CXCL10 [39]. CXCL10 expression
rises rapidly and peaks 4 h after LPS injection, whereas CXCL9 and
CXCL11 mRNA levels continue to increase between 4 and 8 h.
Various cytokines and mitogens have been investigated as to their effect
on CXCL9, CXCL10 and CXCL11 expression levels in vitro. In murine
macrophages, IFN-a, IFN-b and IFN-g have been shown to greatly induce
the level of CXCL10 expression [49]. With IFN-g, this induction peaked
between 3-6 h post-stimulation. Indeed, murine CXCL9, CXCL10 and
CXCL11 are all strongly induced by IFN-g in RAW macrophage cells [39].
However, the three genes have differing responses to IFN-a, IFN-b and
to LPS in mammals. For example, murine CXCL11 is weakly induced by IFN-a,
IFN-b and LPS relative to induction by IFN-g [39]. CXCL9 can be upregulated
in human monocytes after 8 h and in peripheral blood mononuclear
cells by IFN-g, but not by IFN-a or LPS [45]. CXCL11 levels increased
in human keratinocytes following exposure to IFN-g for 6-8 h, yet,
as for CXCL10, no induction was induced by TNF-a [52]. Whilst some cytokines
alone appear to have little effect on CXCL9, CXCL10 and CXCL11 expression
(including IL-1a, IL-3, IL-4, GM-CSF, CSF-1 and TNF-a), the combination
of IFN-g with LPS, IL-1b, or TNF-a produces synergistic increases in induction
of all three chemokines in a wide variety of cell types [39]. Similarly,
combining TNF-a with IFN-g induced expression of murine CXCL11 in Swiss
3T3 cells > 50-fold higher than either cytokine alone. The
in vivo effect of TNFa on gIP expression in trout head kidney was
observed to show an increase of gIP expression above control levels after
24 h. Whilst no similar experiments have been reported for mammalian
systems, evidence from in vitro studies suggest TNF-a alone has
no effect of CXCL10 or CXCL11 expression. This is consistent with the
lack of gIP induction by TNF-a in the RTS-11 cell-line observed in
this study and suggests that TNFa alone has little or no effect on gIP
production by macrophages in trout. Perhaps the effects seen in vivo
result from the activation of other cell types either in the production
of gIP itself or the secretion of other cytokines that may induce gIP
in trout macrophages. Thus, it would be interesting to determine which
mechanisms and cells are involved in the increase of gIP expression observed
in response to TNFa injection in trout. To date, no IFN-g peptide (or
nucleotide sequence) has been isolated from a fish species, hence the
effect of this molecule upon the expression of the gIP chemokine could
not be assessed during this study.
CONCLUSION
In conclusion, a second trout CXC chemokine gene has been discovered
that appears to be a homologue of the subfamily of non-ELR chemokines
containing CXCL9, CXCL10 and CXCL11. That this gene may be virally regulated,
as with its mammalian counterparts, was shown by the ability of poly I:C
(a viral mimic) to greatly increase its expression levels in a trout macrophage
cell-line. To verify the biological effects of the trout chemokine, however,
will require production of the recombinant protein, which should help
establish whether trout gIP, like CXCL9, CXCL10 and CXCL11, is chemotactic
for activated T lymphocytes or some other target.
This work was supported by a grant from the Wellcome Trust (059180).
Thanks go to Mr Steve Bird and Miss Pauline Fotheringham for preparing
head kidney derived cDNA from trout injected with rtTNF-a and Dr Jun Zou
for performing in vitro stimulations with rtTNF-a.
Accepted on 25 August 2002REFERENCES
Murphy PM, Baggiolini M, Charo IF, Hebert CA, Horuk R, Matsushima K,
Miller LH, Oppenheim JJ, Power CA. 2000. International union of pharmacology.
XXII. Nomenclature for chemokine receptors (Review). Pharmacol. Rev.
52: 145.
Clark-Lewis I. 1999. Chemokine structure and function: models for receptor
interaction. Eur. Cytokine Netw. 10: 257.
Rajarathnam K, Sykes BD, Dewald BD, Baggiolini M, Clark-Lewis I. 1999.
Disulphide bridges in interleukin-8 probed using non-natural disulphide
analogues: dissociation of roles in structure from function. Biochem.
38: 7653.
Herbert CA, Vitangcol RV, Baker JB. 1991. Scanning mutagenesis of interleukin-8 identifies
a cluster of residues required for receptor binding. J. Biol. Chem.
266: 18989.
Clark-Lewis I, Schumacher C, Baggiolini M, Moser B. 1991. Structure-activity
relationships of interleukin-8 determined using chemically synthesized
analogues. Critical role of NH2-terminal residues and evidence
for uncoupling of neutrophil chemotaxis, exocytosis, and receptor binding
activities. J. Biol. Chem. 266: 23128.
Loetscher M, Gerber B, Loetscher P, Jones SA, Piali L, Clark-Lewis I,
Baggiolini M, Moser B. 1996. Chemokine receptor specific for IP-10 and
mig: structure, function, and expression in activated T-lympocytes. J.
Exp. Med. 184: 963.
Cole KE, Strick CA, Paradis TJ, Ogbourne KT, Loetscher M, Gladue RP,
Lin W, Boyd JG, Moser B, Wood DE, Sahagan BG, Neote K. 1998. Interferon-inducible
T cell alpha chemoattractant (I-TAC): a novel non-ELR chemokine with potent
activity on activated T cells through selective high affinity binding
to CXCR3. J. Exp. Med. 187: 2009.
Farber JM. 1997. Mig and IP-10: CXC chemokines that target lymphocytes
(Review). J. Leukoc. Biol. 61: 246.
Sauty A, Dziejman M, Taha RA, Iarossi AS, Neote K, Garcia-Zepeda EA,
Hamid Q, Luster AD. 1999. The T cell-specific CXC chemokines IP-10, Mig,
and I-TAC are expressed by activated human bronchial epithelial cells.
J. Immunol. 162: 3549.
Luster AD, Unkeless JC, Ravetch JV. 1985. g-Interferon transcriptionally
regulates an early-response gene containing homology to platelet proteins.
Nature 315: 672.
Luster AD, Ravetch JV. 1987. Biochemical characterization of a g-interferon-inducible
cytokine (IP-10). J. Exp. Med. 166: 1084.
Ransohoff RM, Hamilton TA, Tani M, Stoler MH, Shick HE, Major JA, Estes
ML, Thomas DM, Tuohy VK. 1993. Astrocyte expression of mRNA encoding cytokines
IP-10 and JE/MCP-1 in experimental autoimmune encephalomyelitis.
FASEB J. 7: 592.
Cassatella MA, Gasperini S, Calzetti F, Bertagnin A, Luster AD, McDonald
PP. 1997. Regulated production of the interferon-g-inducible protein-10
(IP-10) chemokine by human neutrophils. Eur. J. Immunol. 27: 111.
Farber JM. 1990. A macrophage mRNA selectively induced by g-interferon
encodes a member of the platelet factor 4 family of cytokines. Proc.
Natl. Acad. Sci. USA 87: 5238.
Zou J, Grabowski PS, Cunningham C, Secombes CJ. 1999. Molecular cloning
of interleukin 1 b of rainbow trout Oncorhynchus mykiss reveals
no evidence of an ICE cut site. Cytokine 11: 552.
Fujiki K, Shin DH, Nakao M, Yano T. 2000. Molecular cloning and expression
analysis of carp (Cyprinus carpio) interleukin-1b, high affinity
immunoglobulin E Fc receptor g subunit and serum amyloid A. Fish Shellfish
Immunol. 10: 229.
Laing KJ, Wang T, Zou J, Holland J, Hong S, Bols N, Hirono I, Aoki T,
Secombes CJ. 2001. Cloning and expression analysis of rainbow trout Oncorhynchus
mykiss tumour necrosis factor-a. Eur. J. Biochem. 268: 1315.
Zou J, Wang T, Hirono I, Aoki T, Inagawa H, Honda T, Soma GI, Ototake
M, Nakanishi T, Ellis AE, Secombes CJ. 2002. Differential expression of
two tumor necrosis factor genes in rainbow trout, Oncorhynchus mykiss.
Dev. Comp. Immunol. 26: 161.
Sumathy K, Desai KV, Kondaiah P. 1997. Isolation of a transforming growth
factor-b2 cDNA from a fish, Cyprinus carpio by RT-PCR. Gene
191: 103.
Hardie LJ, Laing KJ, Daniels GD, Grabowski PS, Cunningham C, Secombes
CJ. 1998. Isolation of the first piscine transforming growth factor b
gene: analysis reveals tissue specific expression and a potential regulatory
sequence in rainbow trout (Oncorhynchus mykiss). Cytokine
10: 555.
Laing KJ, Pilström L, Cunningham C, Secombes CJ. 1999. TGF-b3 exists
in bony fish. Vet. Immunol. Immunopathol. 72: 45.
Laing KJ, Cunningham C, Secombes CJ. 2000. Genes for three different
isoforms of transforming growth factor-b are present in plaice (Pleuronectes
platessa) DNA. Fish Shellfish Immunol. 10: 261.
Najakshin AM, Mechetina LV, Alabyev BY, Taranin AV. 1999. Identification
of an IL-8 homolog in lamprey (Lampetra fluviatilis): early
evolutionary divergence of chemokines. Eur. J. Immunol. 29: 375.
Long Q, Quint E, Lin S, Ekker M. 2000. The zebrafish scyba gene encodes
a novel CXC-type chemokine with distinctive expression patterns in the
vestibulo-acoustic system during embryogenesis. Mech. Dev. 97:
183.
Lee EY, Park HH, Kim YT, Choi TJ. 2001. Cloning and sequence analysis
of the interleukin-8 gene from flounder (Paralichthys olivaceous),
Gene 274: 237.
Laing KJ, Zou JJ, Wang T, Bols N, Hirono I, Aoki T, Secombes CJ. 2002.
Identification and analysis of an interleukin 8-like molecule in rainbow
trout Oncorhynchus mykiss. Dev. Comp. Immunol. 26: 433.
Dixon B, Shum B, Adams EJ, Magor KE, Hedrick RP, Muir DG, Parham P.
1998. CK-1, a putative chemokine of rainbow trout (Oncorhynchus mykiss)
(Review). Immunol. Rev. 166: 341.
Fujiki K, Shin DH, Nakao M, Yano T. 1999. Molecular cloning of carp
(Cyprinus carpio) CC chemokine, CXC chemokine receptors, allograft
inflammatory factor-1, and natural killer cell enhancing factor by use
of suppressive subtractive hybridization. Immunogenetics 49: 909.
Brubacher JL, Secombes CJ, Zou J, Bols NC. 2000. Constitutive and LPS-induced
gene expression in a macrophage-like cell line from the rainbow trout
(Oncorhynchus mykiss). Dev. Comp. Immunol. 24: 565.
Johnson N, Stet RJM, Pike AW, Secombes CJ. 1998. Use of molecular techniques
to assess levels of polymorphism in fish major histocompatibility complex
class II B genes. In: Barnes A, Davidson G, Hiney M, McIntosh Eds Methodology
in Fish Diseases Research, Aberdeen: Fisheries Research Services.
217.
Pearson WR, Lipman DI. 1988. Improved tools for biological sequence
comparison. Proc. Natl. Acad. Sci. USA 85: 2444.
Needleman SB, Wunsch CD. 1970. A general method applicable to the search
for similarities in the amino acid sequence of two proteins. J. Mol.
Biol. 48: 443.
Thompson JD, Higgins DG, Gibson TJ. 1994. Clustal W: improving the sensitivity
of progressive multiple sequence alignment through sequence weighting,
position-specific gap penalties and weight matrix choice. Nucleic Acids
Res. 22: 4673.
Saitou N, Nei M. 1987. The neighbour-joining method - a new
method for reconstructing phylogenetic trees. Mol. Biol. Evol.
4: 406.
Kumar S, Tamura K, Jakobsen IB, Nei M. 2001 MEGA2: Molecular Evolutionary
Genetics Analysis software, Arizona State University, Tempe, Arizona,
USA.
Page RDM. 1996. TREEVIEW: An application to display phylogenetic trees
on personal computers. Computer Applications in the Biosciences
12: 357.
Nielsen H, Engelbrecht J, Brunak S, von Heijne G. 1997. Identification
of prokaryotic and eukaryotic signal peptides and prediction of their
cleavage sites. Protein Eng. 10: 1.
Vaddi K, Keller M, Newton RC. 1997. The Chemokine Factsbook,
London: Academic Press.
Widney DP, Xia YR, Lusis AJ, Smith JB. 2000. The murine chemokine CXCL11
(IFN-inducible T cell a chemoattractant) is an IFN-g- and lipopolysaccharide-inducible
glucocorticoid-attenuated response gene expressed in lung and other tissues
during endotoxemia. J. Immunol. 164: 6322.
Proudfoot NJ, Brownlee GG. 1976. 3' non-coding region sequences in eukaryotic
messenger RNA. Nature 263: 211.
Pleguezuelos O, Zou J, Cunningham C, Secombes CJ. 2000. Cloning, sequencing,
and analysis of expression of a second IL-1 beta gene in rainbow
trout (Oncorhynchus mykiss). Immunogenetics 51: 1002.
Secombes CJ, Wang T, Hong S, Peddie S, Crampe M, Laing KJ, Cunningham
C, Zou J. 2001. Cytokines and innate immunity of fish (Review). Dev.
Comp. Immun. 25: 713.
Wittbrodt J, Meyer A, Schartl M. 1998. More genes in fish ? Bioessays
20: 511.
Oppenheim JJ, Howard OMZ, Goetzl E. 2000. Chemotactic factors, neuropeptides,
and other ligands for seven transmembrane receptors In: Oppenheim JJ,
Feldmann M, Durum SK, Hirano T, Vilcek J, Nicola NA. Eds Cytokine reference:
a compendium of cytokines and other mediators of host defence. USA:
Academic Press, 985.
Farber JM. 1993. HuMig: a new member of the chemokine family of cytokines.
Biochem. Biophys. Res. Commun. 192: 223.
Shields DC. 2000. Gene conversion among chemokine receptors. Gene
246: 239.
Gattass CR, King LB, Luster AD, Ashwell JD. 1994. Constitutive expression
of interferon g-inductible protein in lymphoid organs and inducible expression
in T cells and thymocytes. J. Exp. Med. 179: 1373.
Luster AD. 2000. IP-10 In: Oppenheim JJ, Feldmann M, Durum SK,
Hirano T, Vilcek J, Nicola N. Eds Cytokine Reference: A compendium
of cytokines and other mediators of host defense. USA: Academic press,
1103.
Vanguri P, Farber JM. 1994. IFN and virus-inducible expression of an
immediate early gene, crg-2/IP- 10, and a delayed gene, I-A alpha in astrocytes
and microglia. J. Immunol. 152: 1411.
Ohmori Y, Hamilton TA. 1990. A macrophage LPS-inducible early gene encodes
the murine homologue of IP-10. Biochem. Biophys. Res. Commun. 168:
1261.
Tebo JM, Kim HS, Gao J, Armstrong DA, Hamilton TA. 1998. Interleukin-10 suppresses
IP-10 gene transcription by inhibiting the production of class I
interferon. Blood 92: 4742.
Tensen CP, Flier J, Van Der Raaij-Helmer EM, Sampat-Sardjoepersad S,
Van Der Schors RC, Leurs R, Scheper RJ, Boorsma DM, Willemze R. 1999.
Human IP-9: A keratinocyte-derived high affinity CXC-chemokine ligand
for the IP-10/Mig receptor (CXCR3). J. Invest. Dermatol. 112: 716.
|
Version imprimable
