Download The cloning and overexpression of E coli acyl carrier protein (ACP)

202s Biochemical Society Transactions ( 1 993) 21
The cloning and overexpression of E coli acyl carrier protein
( A W
A. Lesley Jones, Peter Kille, *Jane E. Dancer and John L.
Hanvood
Dept. of Biochemistry, UWCC, PO Box 903, Cardiff CFl 1ST
and Schering Agrochemicals Ltd, Chesterford Park Research
Station, Saffron Walden, Essex CBlO 1XL
Acyl carrier protein (ACP) is a small (9kDa) protein which
is essential for fatty acid synthesis. In bacteria [ l ] and plants [2]
which use type I1 dissociable fatty acid synthetases, acyl chains are
esterified to ACP through the 4'-phosphopantetheineprosthetic
group, and ACP acts as a carrier during fatty acid synthesis.
ACPs from bacteria and plants are very similar. All are small
acidic proteins with conserved regions particularly around the
serine residue through which the prosthetic group is attached.
The similarity of E. coli ACP to many plant ACPs is such that E.
coli ACP can be used as a substitute for plant ACP in many assays
for plant fatty acid synthetase activity; indeed it sometimes proves
more active in vitro than the native plant ACP [3]. This is useful
because such large quantities of plant material are required to
isolate relatively small amounts of protein. It is therefore easier to
isolate E. coli ACP. E . coli contains only one ACP isoform and
also has the advantage of being active in many plant system,
whereas plant ACPs often do not work in other plant systems, and
different and specific isoforms occur in different tissues [4].
It is possible to isolate 20-30mg of ACP from 200g of E.
coli wet cell paste in 3 days. However, assaying plant fatty acid
synthesis enzymes requires large quantities of ACP and. therefore,
1"
CCC-GAGCACCATCG-GTGTGMetSerThrIleGluGluArgValLysLysIleIle
50
10
6"
80
90
GGCG?&&GCTGGGCGTTAAGCAGGAAGAAGTTACCAACAATGCT
GlyGluGlnLeuGlyValLysGlnGluGluValThrAsnAsnAla
100
130
120
110
it would be useful to have a more readily available supply. With
this in mind we have attempted to overexpress E. coli ACP in E .
coli.
Initially no nucleotide sequence data was available for E.
coli ACP. Therefore degenerate oligonucleotide primers were
designed on the basis of available protein sequence data [5,6]
encoding the N- and C-terminal regions of the protein with
additional 5' extensions containing appropriate cloning sites (Fig.
1). When genomic E. coli DNA was used as a template for a
PCR containing these primers a product whose size corresponded
to that predicted for the ACP gene was generated. Cloning and
sequencing of the PCR product showed it to encode an ACP type
protein. However in all clones sequenced a frame shift was
identified due to the incorporation of an additional base in the Nterminal region of the gene. The latter addition was removed by
PCR overlap extension mutagenesis. The product of this reaction
was cloned into the expression vector pKK 223-3 but
overexpression of the ACP could not be detected. Furthermore, a
protein product of the appropriate size was not detected when in
virro translation of the plasmid was performed. Sequencing of the
PCR mutagenesis product in M13 revealed two families of clones,
the full length sequence (Fig. 1) and a deletion mutant in which
nucleotide 41 was absent. It is possible that the failure to
overexpress the gene was caused by the cloning of the latter
mutant into pKK 223-3.
The DNA sequence of E. coli ACP was published earlier
this year [7] and agrees exactly with the sequence we have
reported apart from the areas predetermined by our PCR primers.
Furthermore, translation of both open reading frames produces
identical protein products which disagree with the two previously
reported protein sequences at residues 24 and 43. Vanaman et al..
[5] reported Asp (24) and Val (43) and Jackowski and Rock [6]
Asn (24) and Ile (43). We found position 24 to be Asn and
position 43 to be Val (Fig. 1) as did Rawlings and Cronan [7].
One other very interesting observation of Rawlings and
Cronan [7] is that the natural gene when cloned into M13 had a
high frequency of spontaneous deletions. This is consistent with
the high frequency of mutations seen in the fragment we isolated
containing the ACP gene. They also suggest that DNA segments
encoding ACP are toxic to E. coli and their synthetic contruct
with high levels of ACP expression did indeed prove lethal to the
organism. [8].
Other expression systems to overexpress E. coli ACP are
currently being evaluated.
TCTTTCGTTGAAGACCTGGGCGCGGATTCTCTTGACACCGTTGAG
SerPheValGluAspLeuGlyAlaAsp.S.exLeuAspThrValGlu
140
160
150
LBO
I70
This is a LINK project with the SERC, AFRC and Schering
Agrochemicals Ltd.
CTGGTAATGGCTCTGGAAGAAGAGTTTGATACTGAGATTCCGGAC
LeuValMetAlaLeuGluGluGluPheAspThrGluIleProAsp
190
210
200
220
GAAGAAGCTGAGAAAATCACCACCGTTCAGGCTGCCATGTCTAT
GluGluAlaGluLysIleThrThrValGlnAlaAlaIleAspTyr
21"
240
250
255
The primer regions are heavily underlined and the restriction enzyme
sites built onto the end of the primers are single underlined.
Rawlings and Cronan [7] confirmed our original assumption that
the initial Met residue would be lost after translation but that
no other amino acid residues would be processed from the
N-terminus. The Ser residue through which the 4' phosphopantetheine group is attached is underlined.
Vagelos, P.R. (1971) Curr. Top Cell Reg. 3. 119
Ohlrogge, J.B., Browse, J., and Somerville, C.R. (1991)
Biochim. Biophys. Acta 1082, 1-26
Simoni, R.D., Criddle, R.S. and Stumpf, P.K. (1967)J.
Biol. Chem. 242,573-581
Ohlrogge, J.B. (1987) In The Biochemistry ofPIants (P.K.
Stumpf and E.E. Conn eds.) 9: pp 137-157. Academic
Press, New York.
Vanaman, T.C., Wakil, S.J. and Hill, R.L. (1968)J. B i d .
Chem. 243,6420-6436
Jackowski, S . and Rock, C.O. (1987) J. Bacreriol. 169,
1469-1473
Rawlings, M.and Cronan, J.E. (1992) J . Biol. Chem. 267,
5751-5754
Rawlings, M.and Cronan, J.E. (1988) FASEB J . 2, A1559