Reconstitution of the pyridoxal 5′-phosphate (PLP) dependent enzyme serine palmitoyltransferase (SPT) with pyridoxal reveals a crucial role for the phosphate during catalysis

Article abstract of DOI:10.1039/c3cc43001d

The pyridoxal 5′-phosphate (PLP)-dependent enzyme serine palmitoyltransferase (SPT) is required for de novo sphingolipid biosynthesis. A previous study revealed a novel and unexpected interaction between the hydroxyl group of the l-serine substrate and the 5′-phosphate group of PLP. By using pyridoxal (PL), the dephosphorylated analogue of vitamin B₆, we show here that this interaction is important for substrate specificity and optimal catalytic efficiency.

Full text of DOI:10.1039/c3cc43001d

ChemComm
View Article Online
View Journal | View Issue
COMMUNICATION
Reconstitution of the pyridoxal 50-phosphate (PLP)
dependent enzyme serine palmitoyltransferase (SPT)
with pyridoxal reveals a crucial role for the phosphate
during catalysis†‡
Cite this: Chem. Commun., 2013,
49, 7058
Received 23rd April 2013,
Accepted 24th June 2013
DOI: 10.1039/c3cc43001d
Ashley E. Beattie, David J. Clarke, John M. Wadsworth, Jonathan Lowther,
Ho-Lam Sin and Dominic J. Campopiano*
www.rsc.org/chemcomm
The pyridoxal 50-phosphate (PLP)-dependent enzyme serine palmitoyl-
transferase (SPT) is required for de novo sphingolipid biosynthesis. A
previous study revealed a novel and unexpected interaction between
the hydroxyl group of the L-serine substrate and the 50-phosphate
group of PLP. By using pyridoxal (PL), the dephosphorylated analogue
of vitamin B6, we show here that this interaction is important for
substrate specificity and optimal catalytic efficiency.
Sphingolipid (SL) biosynthetic pathways have been extensively
studied in humans, yeast and bacteria.1 Despite their diversity, all
of the pathways use the same condensation of an acyl-CoA sub-
strate with L-serine to form 3-keto-dihydrosphingosine (KDS) to
initiate de novo SL synthesis. In all organisms, this key step is
catalysed by the pyridoxal 50-phosphate (PLP)-dependent enzyme
Scheme 1 Proposed SPT mechanism showing observed intermediates and the
H-bond between the 50-phosphate and serine hydroxyl in aldimine II.
The PLP cofactor contains four functional groups of interest;
serine palmitoyltransferase (SPT).2 PLP acts as a cofactor for a wide aldehyde, hydroxyl, pyridine and phosphate. Recently, the impor-
variety of enzyme-catalysed reactions as well as also acting as a tance of the pyridine nitrogen has been studied in three PLP
general organic catalyst.3 For example, PLP enzymes catalyse C–C enzymes from two fold types.11 By replacement of the natural
formation, amino-group transfer and racemisation reactions.4–6 cofactor by the carbocyclic analogue, 1-deazapyridoxal 50-phosphate
The PLP-dependent enzyme superfamily has been classified into (deazaPLP) it was demonstrated that not all PLP-dependent enzymes
five different fold-types based on 3D structure.7,8 SPT is a member require the protonated pyridine nitrogen to stabilise carbanion
of the a-oxoamine synthase family (AOS, fold type I) and a general formation of the PLP-substrate aldimine.12
mechanism is postulated for all members that begins with PLP
bound to an active site lysine residue as an internal aldimine/ been less well studied. Our interest in its role arose when we
Schiff’s base (Scheme 1, internal aldimine I).9
observed an unexpected H-bond interaction (length 3.9 Å) between
Surprisingly, the role of the 50-phosphate group of PLP has
This PLP-bound lysine is documented in 229 structures in the the hydroxyl group of L-serine and the 50-phosphate in the X-ray
Protein Data Bank (PDB).10 In Sphingomonas paucimobilis SPT structure of the SP SPT:PLP:L-serine external aldimine complex II
(SP SPT) the L-serine displaces the lysine residue (Lys265) to form (PDB code: 2W8J).13 As well as this H-bond the phosphate has a
the PLP:L-serine external aldimine (II). Deprotonation of II gives a number of interactions with the SPT that together make up a
carbanion/quinonoid species (III) which condenses in a Claisen-like ‘‘phosphate binding cup’’ (Fig. 1).
manner with the acyl-CoA thioester substrate to form a PLP:b-
The concept of this ‘‘cup’’ was proposed by Denesyuk et al.
ketoacid, which then decarboxylates to generate the PLP:KDS pro- after analysis of the X-ray structures from five different fold
duct quinonoid (IV). This then reprotonates to form PLP:KDS types.14,15 In SP SPT the cup is formed by a constellation of residues
aldimine (V) which is finally displaced by Lys265.
that bind the oxygen atoms but there appears to be little conserva-
tion of these across the PLP-superfamily. However, within the AOS
family (which display overall sequence homology of 30–40%) the
phosphate binding ligands are well conserved (Fig. S1 and S2, ESI‡).
In the SPT:PLP internal aldimine structure the seven cup-forming
ligands are provided by side-chains and backbones from both
EaStCHEM, School of Chemistry, University of Edinburgh, Edinburgh, EH9 3JJ,
Scotland, UK. E-mail: Dominic.Campopiano@ed.ac.uk; Tel: +44 (0)131-650-4712
† Celebrating 300 years of Chemistry at Edinburgh.
‡ Electronic supplementary information (ESI) available. See DOI: 10.1039/
c3cc43001d
c
7058 Chem. Commun., 2013, 49, 7058--7060
This journal is The Royal Society of Chemistry 2013

View Article Online
Communication
ChemComm
Fig. 1 (A) Structure of the PLP:L-serine external aldimine intermediate (II) of
SP SPT (PDB code: 2W8J). (B) Phosphate binding cup diagram (a) T294* (b) Y135
(c) G134 (d) water (e) T262; (f) S264 (g) A295* (h) L-serine-hydroxyl interaction
(* indicates residue from opposite subunit).
subunits and a water molecule. In forming the key SPT:PLP:L-serine
external aldimine intermediate all of these interactions are retained
and the only difference is the additional H-bond from the L-serine
(Fig. 1 and Fig. S1, ESI‡). To probe the function of this interaction we
prepared SPT with the PLP cofactor replaced by the dephosphory-
lated form, pyridoxal (PL). The UV-visible spectrum of the holo-form
of SPT (SPT:PLP, Fig. 2A) showed peaks at 335 nm and 420 nm,
characteristic of the enolimine and ketoenamine forms of PLP
bound to the conserved Lys265 residue.
The PLP was removed by incubation with the aminothiol
L-pencillamine (L-pen), which forms a non-covalent PLP:thiazolidine
adduct which absorbs at 333 nm (Fig. 2A).16 Upon dialysis the
resultant apo-enzyme was prone to precipitation, and the holo-
enzyme could not be recovered in high yield by simple addition of
PLP or dialysis against PLP-containing buffers. Therefore, both
regenerated SPT:PLP and SPT:PL were produced by this exchange
Fig. 2 (A) UV-vis spectra of SPT:PLP (solid line) and after incubation with L-pen
to remove the PLP (dashed line). (B) UV-vis spectra of apo-SPT (solid line), SPT:PLP
formed by incubation of apo-SPT with PLP (dotted line), and SPT:PL (dashed
line). Conditions: 20 mM Tris, pH 7.5 and 40 mM enzyme. (C) Activity of SPT
regenerated with PLP and PLP. Activity values are relative to the as-purified
enzyme (SPT:PLP).
process. The SPT:PL (Fig. 2B) displayed absorbance maxima at 320 binding (but not palmitoyl-CoA binding) and severely reducing
and 420 nm which are shifted B20 nm compared with SPT:PLP. It is the catalytic rate of the SPT:PL enzyme. This data supports our
interesting to note that the 320 nm peak is the most dominant, hypothesis that the L-serine:phosphate interaction plays a critical
indicating that a different tautomer predominates in the SPT:PL role in catalysis. In addition, we monitored KDS formation directly
complex. In order to ensure that the PL was bound to active site using a stopped assay (ESI‡). No KDS was produced by the apo-SPT
residue Lys265, chemical reduction of the SPT:PL-imine, followed by but the SPT regenerated with PLP and the PL-regenerated enzyme
enzymatic digest and mass spectrometry confirmed the sequence of produced 46% and 13% KDS relative to the as-purified enzyme.
the peptide (Fig. S3, ESI‡). We measured the relative enzyme activity This analysis provides further evidence that the removal of the
using a continuous assay that monitors the rate of formation of the phosphate impairs the catalytic activity of the SPT.
coenzyme A thiolate released during the C–C condensation reaction.
We then sought a mechanistic basis for the requirement of
The apo-SPT displayed a much reduced enzymatic activity (3%, the 50-phosphate for optimal SPT activity. Addition of L-serine to
Fig. 2C) and the regenerated SPT:PLP displayed 83% activity. In SPT:PL led to the formation of a peak at 425 nm suggesting that
contrast, the regenerated SPT:PL complex displayed only 6% activity the absence of the 50-phosphate did not prevent the formation of
showing that catalysis in this complex was severely compromised.
the SPT:PL:L-serine external aldimine intermediate. A KdL-Ser of
Using Michaelis–Menten kinetics we obtained kinetic constants B12.7 mM was obtained for the SPT:PL complex (Fig. S4, ESI‡)
for the regenerated SPT:PL enzyme (Table, ESI‡). The Km values for which is an 8-fold increase when compared with the L-serine Kd
L-serine and palmitoyl-CoA (1.6 mM and 35.6 mM, respectively) and for SPT:PLP (1.1 mM).17 This is evidence that removing the
the kcat of 1.54 sꢀ1 and kcat/Km of 962.5 Mꢀ1 sꢀ1 for L-serine are all phosphate from the cofactor impacts on serine binding and
in good agreement with published data.13 The SPT:PLP regenerated SPT:PLP:L-serine external aldimine formation and suggests that
enzyme displayed similar Km and kcat values to the as-purified SPT, the H-bond between the serine and phosphate may play a role in
with only a small increase in the Km for L-serine (2.7 mM). In substrate specificity. However, this reduced Kd for L-serine cannot,
contrast, the Km for L-serine increased to 5 mM for the SPT:PL by itself, account for the dramatic reduction in activity. Therefore
enzyme. The Km for palmitoyl-CoA was unaffected (33 mM) but we also investigated if the SPT:PL:L-serine intermediate could be
we observed significantly lower L-serine kcat (0.13 sꢀ1) and kcat/Km deprotonated by removal of the L-serine Ca-H. This step generates
(26.3 Mꢀ1 ꢀ1) values for the SPT:PL which are B12 and 37-fold the key carbanion/quinonoid species (III) which reacts with the
s
lower compared with the as-purified SPT. The removal of the PLP incoming acyl-CoA thioester. We used the C16-CoA thioether
phosphate has had a combined impact by weakening L-serine analogue, S-(2-oxoheptadecyl)-CoA (Fig. S6, ESI‡) which has
c
This journal is The Royal Society of Chemistry 2013
Chem. Commun., 2013, 49, 7058--7060 7059

View Article Online
ChemComm
Communication
Since SPT is the first enzyme in SL biosynthesis it should be
highly specific for L-serine as a substrate and not accept
L-alanine or glycine which would generate deoxySL products.
Indeed, L-alanine is a very poor substrate for SPT:PLP, with no
detectable activity even at high substrate concentrations. Such
L-alanine- and glycine-derived deoxySLs are generated by mutant
forms of SPT and have recently been reported to be toxic to
mammalian cells.22 Our work provides evidence that amino acid
substrate specificity is not only controlled by residues from SPT
but also by an interaction between the phosphate of the PLP
cofactor and the L-serine side-chain. L-serine is used as a substrate
by other PLP-dependent enzymes such as serine racemase, serine
dehydratase and serine hydroxymethyltransferase so it will be
interesting to investigate whether the 50-phosphate also plays a
catalytic role in these important metabolic proteins.
Fig. 3 (A) UV-vis spectra of SPT:PLP (solid line), and after addition of; 40 mM
L-serine (dashed line), 1.5 mM S-(2-oxoheptadecyl)-CoA (dotted line). (B) UV-vis
spectra of SPT:PL (solid line), and after addition of 40 mM L-serine (dashed line)
and 1.5 mM S-(2-oxoheptadecyl)-CoA (dotted line). Conditions: 20 mM Tris,
pH 7.5 and 40 mM SPT.
been shown to bind to the SPT:PLP:L-Ser complex but cannot
condense with the L-serine.18 Addition of the analogue to the
L-serine external aldimine form of SPT:PLP led to the appearance of a
peak at 495 nm (Fig. 3A) which is thought to be the substrate
quinonoid (intermediate III). In contrast, the PL-reloaded enzyme
forms a broad shoulder at 495 nm under the same conditions
(Fig. 3B). These results suggest that the PL cofactor may sit in
a potentially ‘non-native’ environment due to the lack of the
50-phosphate. It can still produce the external aldimine (II) with
reduced affinity but crucially, it is severely compromised in its
ability to form and/or stabilise the key quinonoid (III) species.
Upon addition of the product KDS to the SPT:PLP enzyme we
observed changes at 335 and 420 nm, and the appearance of a peak
at 505 nm (Fig. S5A, ESI‡) which is thought to be the SPT:PLP:KDS
product quinonoid (intermediate IV). In contrast, the SPT:PL enzyme
did not form the 505 nm peak (Fig. S5B, ESI‡). Instead, we observed
a broad signal between 330 nm and 500 nm with a shoulder at
400 nm which suggests the product binds but the PL:KDS aldimine
We thank The Derek Stewart Trust (AEB), Lilly (JMW) and
the School of Chemistry (AEB and JMW) for funding. DJCla and
JL are funded by a BBSRC grant (BB/I013687/1). We thank Prof.
Stephen Fry and Dr. Lenka Frankova for help with KDS assays
and Prof. Bob Baxter for helpful discussions.
Notes and references
1 A. H. Merrill, Jr., Chem. Rev., 2011, 111, 6387–6422.
2 J. Lowther, J. H. Naismith, T. M. Dunn and D. J. Campopiano,
Biochem. Soc. Trans., 2012, 40, 547–554.
3 J. P. Richard, T. L. Amyes, J. Crugeiras and A. Rios, Biochim.
Biophys. Acta, 2011, 1814, 1419–1425.
4 A. C. Eliot and J. F. Kirsch, Annu. Rev. Biochem., 2004, 73, 383–415.
5 M. D. Toney, Arch. Biochem. Biophys., 2005, 433, 279–287.
6 M. D. Toney, Biochim. Biophys. Acta, 2011, 1814, 1407–1418.
7 J. N. Jansonius, Curr. Opin. Struct. Biol., 1998, 8, 759–769.
8 R. Percudani and A. Peracchi, EMBO Rep., 2003, 4, 850–854.
9 B. A. Yard, L. G. Carter, K. A. Johnson, I. M. Overton, M. Dorward,
H. Liu, S. A. McMahon, M. Oke, D. Puech, G. J. Barton, J. H. Naismith
and D. J. Campopiano, J. Mol. Biol., 2007, 370, 870–886.
then adopts an unusual conformation. We interpret these combined 10 L. Roglin, P. Thiel, O. Kohlbacher and C. Ottmann, Proc. Natl. Acad.
Sci. U. S. A., 2012, 109, E1051–E1053; author reply E1054.
11 W. R. Griswold and M. D. Toney, J. Am. Chem. Soc., 2011, 133,
results as evidence that the PL-containing enzyme cannot efficiently
catalyse the formation of the key substrate quinonoid (III). There-
14823–14830.
fore, it appears that the interaction between the 50-phosphate of PLP 12 W. R. Griswold and M. D. Toney, Bioorg. Med. Chem. Lett., 2010, 20,
1352–1354.
and the L-serine hydroxyl group is crucial for the formation of key
intermediates in the catalytic mechanism and in controlling the
13 M. C. Raman, K. A. Johnson, B. A. Yard, J. Lowther, L. G. Carter,
J. H. Naismith and D. J. Campopiano, J. Biol. Chem., 2009, 284,
optimal orientation of the reactive quinonoid/carbanion that attacks
the acyl-CoA thioester. This explanation lends further weight to the
‘‘Dunathan hypothesis’’ which was put forward over 40 years ago to
explain the stereoelectronic control that the cofactor provides during
PLP-dependent catalysis.19
SPT:PLP has been reported to have extremely poor affinity for
L-phosphoserine with a Kd > 100 mM.20 We sought to re-introduce
the interaction using L-phosphoserine as a substrate for the
SPT:PL enzyme. However, no substrate binding, nor detectable
17328–17339.
14 A. I. Denesyuk, K. A. Denessiouk, T. Korpela and M. S. Johnson,
J. Mol. Biol., 2002, 316, 155–172.
15 A. I. Denesyuk, K. A. Denessiouk, T. Korpela and M. S. Johnson,
Biochim. Biophys. Acta, 2003, 1647, 234–238.
16 J. Lowther, A. E. Beattie, P. R. R. Langdridge-Smith, D. J. Clarke and
D. J. Campopiano, MedChemComm, 2012, 3, 1003–1008.
17 M. C. Raman, K. A. Johnson, B. A. Yard, J. Lowther, L. G. Carter,
J. H. Naismith and D. J. Campopiano, J. Biol. Chem., 2009, 284,
17328–17339.
18 H. Ikushiro, S. Fujii, Y. Shiraiwa and H. Hayashi, J. Biol. Chem.,
2008, 283, 7542–7553.
activity was observed, suggesting that the key phosphate-hydroxyl 19 H. C. Dunathan, Proc. Natl. Acad. Sci. U. S. A., 1966, 55, 712–716.
20 H. Ikushiro, H. Hayashi and H. Kagamiyama, Biochemistry, 2004, 43,
interaction cannot be formed in this complex (data not shown).
1082–1092.
Older studies had looked at whether a sulphate at the 50-position
21 E. Groman, Y. Z. Huang, T. Watanabe and E. E. Snell, Proc. Natl.
of pyridoxal could replace the phosphate.21 Of the three enzymes
tested only arginine decarboxylase was active with this modified
cofactor (23% activity compared with the PLP-containing enzyme)
further emphasising the importance of the 50-phosphate in catalysis.
Acad. Sci. U. S. A., 1972, 69, 3297–3300.
22 A. Penno, M. M. Reilly, H. Houlden, M. Laura, K. Rentsch,
V. Niederkofler, E. T. Stoeckli, G. Nicholson, F. Eichler,
R. H. Brown, Jr., A. von Eckardstein and T. Hornemann, J. Biol.
Chem., 2010, 285, 11178–11187.
c
7060 Chem. Commun., 2013, 49, 7058--7060
This journal is The Royal Society of Chemistry 2013