Results
Isolation and sequence analysis of Atfpg
clones
The strict conservation of an 8-amino acid sequence at
the N-terminal end of FPG from several bacterial genera facilitated the search
for homologous genes in plants by providing the model that led us to identify
an Arabidopsis EST. A probe with the base sequence of the EST was used
to select two clones from an Arabidopsis flower cDNA library (Murphy
and Gao, 1998. These were substantially identical to two clones reported
by Nakabeppu et al (1998). A second round of selection
using the same probe and library obtained three additional clones, the 5' sequence
of which confirmed their relationship to the original Arabidopsis clones.
Fig. 1 shows the
relationship of the various Arabidopsis FPG cDNAs to each other and to
the sequence of the genomic template for the transcripts, which was determined
previously (Genbank accession AB010690; Ohtsubo et al., 1998).
The clones denoted AtMMH-1 and AtMMH-2 are those described by
Nakabeppu et al. (1998) and Ohtsubo et al. (1998);
AtFPG-1 and AtFPG-2 are the ones described by Murphy and Gao (1998).
There is one base difference between the cDNA sequences determined by Ohtsubo
et al. (1998), and those that we determined (Murphy and
Gao, 1998 and present work), a silent single-base difference
in a lysine codon (AAA in strain Columbia--Ohtsubo et al.,
1998; AAG in strain Landsberg erecta--Murphy and Gao,
1998).
Atfpg-1a,
Atfpg-3, and Atfpg-4 were obtained in the second
round of selection. Atfpg-1a differs from Atfpg-1
and AtMMH-1 only by the position of polyadenylation. Atfpg-3 is unique among the clones in lacking exon 4, and it
differs from Atfpg-1 in the selection of bases from the
region from 1950-2250. Atfpg-4 is probably incomplete,
since it is truncated relative to Atfpg-1, -1a,
and -3, and its predicted protein does not have a termination
codon, but it demonstrates an example of an mRNA containing exon 4 and
a selection of bases otherwise like that of Atfpg-3.
The seven reported clones represent four distinct combinations
of exons. Atfpg-2 and AtMMH-2 were unique in possessing bases
from the entire region from 1956 to 2443.
RT-PCR
RT-PCR was used to test the presence of mRNA in the various
organs (Fig. 2). The first-strand cDNA
synthesis employed two reverse primers (Fig. 2, primers R1 and R2), one designed
to copy mRNAs like Atfpg-1, -1a, and -3, with a base sequence
reaching near the end of the gene (2970), the other designed to copy mRNAs like
Atfpg-2, with sequences in the region 2220-2440. The same reverse primers,
together with a forward primer common to both clones (primer F in exon 5), were
used for PCR. Both sets of primers span introns, so the sizes of the fragments
that were produced distinguished between cDNA and any contaminating genomic
DNA templates. The products were visualized by hybridization with the same clone-specific
probes used with the DNA blots. The results demonstrated the presence of Atfpg-1-class
mRNA in inflorescences and roots and Atfpg-2-class mRNA in inflorescences
and stems. With the latter, there were three major bands. One of these was the
size expected of a product from mRNA matching the Atfpg-2 clone; one
was larger; and one was smaller. There were also minor bands, including one
at the size expected of a product from unspliced transcript. Mock reverse transcriptase
reactions (no enzyme) using the same RNA preparations confirmed that the bands
did not arise from DNA contamination.
The bands appearing in the RT-PCR performed
with Atfpg-2 primers were isolated from agarose gels, amplified,
and cloned; and the sequences of individual clones were determined.
The results confirmed that the middle band was Atfpg-2,
that is, it had the same sequence of bases between the two primers.
The larger and smaller bands had sequences between the two primers
that indicated the presence of mRNAs with selections of exons
that differed from those found in any of the selected clones, bringing the number of splicing alternatives to six. Of course, we cannot know the sequence of those mRNAS outside
the limits imposed by the primers used for PCR.
Southern blot
A Southern blot (Fig.
3) provided confirmatory evidence that the clones and RT-PCR fragments came
from a single gene, although there may be one or two other genes with the functionally
important exon 1 sequence. Hybridization of DNA blots with a probe representing
nucleotides 1 to 199 (Fig. 3a, probe 1) suggested that there are 2-3 genes that
share this common sequence . Hybridization of the same DNA blots with Atfpg-1-specific
and Atfpg-2-specific probes (Fig. 3b and c, probes 2 and 3, respectively)
indicated that both cDNAs represented mRNAs from only one of those genes. The
sizes of the bands were consistent with the restriction map of Arabidopsis
chromosome 1 BAC f6d8 (Accession AC008016), which shows that the entire Atfpg
gene is contained within one EcoR1 fragment of 11246 bp; the sequence
represented by probe 1 is within one HindIII fragment of 1999 bp; and
the sequences represented by both Atfpg-1 and -2 unique probes
are within one HindIII fragment of 1199 bp.