Chapter 2
Fatty Acid Synthesis in African Trypanosomes
is neither Type I nor Type II
The content of this work has been distilled and submitted for publication. The authors
are Soo Hee Lee, Jennifer L. Stephens, Kimberly S. Paul, and Paul T. Englund.
34
SUMMARY
Bloodstream trypanosomes, parasites that cause sleeping sickness, synthesize
predominantly myristate to use for the GPI anchors of the surface glycoproteins that
mediate antigenic variation. In contrast, insect stage trypanosomes do not myristoylate
GPIs and make longer fatty acids (FAs) for phospholipids. We found in cell-free assays,
using gene knockout trypanosomes, that these FAs are synthesized by microsomal FA
elongases. In other eukaryotes, elongases do not make FAs de novo; instead they extend
pre-existing long-chain FAs. Trypanosome elongases 1–3 convert butyryl-CoA (C4)
stepwise to caproyl-CoA (C10), myristoyl-CoA (C14), and stearoyl-CoA (C18),
respectively. Pathway regulation results in different FA products in the bloodstream and
insect stages, and the entire pathway is up-regulated if exogenous FAs are limiting. Thus
this pathway is ideally suited for the trypanosome which at different life-cycle stages has
different needs for FAs. This is the first example of elongases mediating de novo FA
synthesis, and it represents a third class of FA biosynthetic pathway, distinct from
conventional type I and type II systems.
35
INTRODUCTION
Trypanosoma brucei, the sleeping sickness parasite, divides its life cycle between
the tsetse fly vector and the mammalian host. The bloodstream form (BSF) evades the
host's immune response by switching surface coats composed of 107 variant surface
glycoprotein (VSG) molecules (1). Each VSG molecule is tethered to the plasma
membrane by a glycosylphosphatidylinositol (GPI) anchor whose fatty acids (FAs) are
exclusively myristate, a 14-carbon saturated FA (2). Myristate, via myristoyl-CoA, is
incorporated into the GPI precursor in a microsomal FA remodeling reaction that replaces
other long chain FAs with myristate (3). In contrast, tsetse fly procyclic forms (PCF) do
not myristoylate the GPI anchors of their surface glycoproteins (4).
BSF trypanosomes need massive quantities of myristate for their GPI anchors. It
was long believed that they were unable to synthesize FAs de novo (5) and that they
acquired all FAs, including myristate, from the host blood. Although trypanosomes
salvage myristate efficiently (6), the myristate concentration in blood is hardly adequate
to support GPI myristoylation at high parasite densities (7). Several years ago we
reopened the question of whether trypanosomes synthesize FAs and discovered that
indeed they do (8). We developed a cell-free system for FA synthesis containing
trypanosome membranes (a source of enzymes), butyryl-CoA (the four carbon primer;
acetyl-CoA does not prime), malonyl-CoA (the two carbon donor), and NADPH. The
end product with BSF membranes is predominantly myristate which can be used for GPI
anchors. In contrast, the longest end product in PCFs is stearate (C18).
In other cell types, FAs are made by a type I or type II FA synthase (9, 10).
Type I enzymes, found in the cytosol of mammals and fungi, have multiple catalytic
36
activities residing on separate domains of one or two large polypeptides (9). Type II FA
synthases, found in bacteria, plants, and in eukaryotic organelles of prokaryotic origin,
have similar catalytic activities but each is on a separate polypeptide (10, 11). In both
systems the enzymes are soluble and the growing chain is esterified to acyl-carrier
protein (ACP) or an ACP-like domain.
So which system makes FAs in T. brucei? The T. brucei genome does not encode
a type I FA synthase, but does predict a type II system (7). Thus, we naturally assumed
that the synthase was type II. However, several observations taken together suggested
the type II pathway is not responsible for trypanosome bulk FA synthesis. First, typical
type II systems are composed of soluble proteins yet the trypanosome activity is
membrane-associated. Second, the type II -ketoacyl-ACP synthase (KAS) (K. Paul
unpublished) and ACP (J. Stephens unpublished) localize to the trypanosome's single
mitochondrion but the myristate product is needed in the ER where GPI remodeling
occurs. Third, triclosan, a type II enoyl-ACP reductase inhibitor, affects cell-free FA
synthesis only at high concentrations, suggesting non-specific inhibition (12). Fourth,
cerulenin, which inhibits the KAS (both type I and type II), does not affect cell-free
synthesis of intermediates up to C10, although it does inhibit extension beyond C10 (8).
This specificity of cerulenin could be explained by the existence of two KAS activities,
however the genome predicts only one. Finally, and most compellingly, RNA
interference (RNAi) silencing of ACP, a key component of type II systems, has no effect
on FA synthesis in the cell-free system (experiment performed by J. Stephens; Fig. 2.1a).
For these and other reasons we concluded that the quantity of type II pathway FA
products is minor and that the pathway is mitochondrial. Therefore, bulk FA synthesis in
37
trypanosomes must be unconventional, involving neither a type I nor a type II FA
synthase. As an alternative we considered the possibility that a microsomal FA elongase
system is involved.
Elongase systems extend the products of type I FA synthase (myristate in yeast
and palmitate in mammals) to longer-chain FAs (13). Saccharomyces cerevisiae ELO1–
3 extend C14 to C16, C16 to C24, and C24 to C26, respectively (14, 15). The chemistry
of the elongase reactions resembles that of type I and type II systems, except the growing
acyl chain is esterified to coenzyme A (CoA) instead of ACP. The ELOs condense
malonyl-CoA with the growing acyl chain, extending it by two carbon atoms. The
resulting -ketoacyl-CoA is then reduced (-ketoacyl-CoA reductase), dehydrated (hydroxyacyl-CoA dehydrase), and reduced again (trans-2-enoyl-CoA reductase)
producing a longer saturated acyl-CoA (16). The two reduction steps require NADPH.
We now describe the trypanosome's unusual biochemical machinery involving FA
elongases for de novo FA synthesis.
38
RESULTS
Elongases are involved in bulk FA synthesis
T. brucei has four candidate ELO genes. ELO1–3 are in tandem on chromosome
seven (NCBI AAX70671, AAX70672, and AAX70673) while ELO4 is on chromosome
five (with allelic variants AAX69821 and AAX70768). Like S. cerevisiae ELOs, the
trypanosome ELOs have multiple predicted transmembrane domains and an HXXHH
amino acid motif (15). We used RNAi to silence ELO1 in procyclic trypanosomes and
then assayed membrane isolates for FA synthesis. In contrast to the effect of ACP RNAi,
ELO1 knockdown drastically reduced FA synthesis (Fig. 2.1b). RNAi also provided
preliminary evidence that other ELOs act on longer acyl chains (Fig. 2.2). These studies
indicated that the ELOs are involved in cell-free FA synthesis.
Sequential FA synthesis by ELO1–3
To evaluate more rigorously the role of each ELO in de novo FA synthesis, we
made knockout strains for each of the four ELO genes (elo) in BSF trypanosomes.
Figure 2.3 shows Southern blots confirming the knockout of each gene. We assayed
membranes from wild-type and knockout strains in the cell-free FA synthesis system (8)
using acyl-CoA primers of varying chain-length. We detected and quantitated the
products by thin layer chromatography (TLC) and phosphorimaging (Fig. 2.4), and we
determined the product chain lengths by reverse-phase TLC of their methyl esters (Fig.
2.5a). These data (presented below) show that ELO1 mainly extends C4 to C10, ELO2
extends C10 to C14, and ELO3 extends C14 to C18. Furthermore, the ELO specificities
39
are overlapping. For example, ELO2, which prefers C10- and C12-CoA primers, also
has low activity with C8- and C14-CoA.
Using wild-type membranes, we found robust synthesis with primer lengths from
4 to 12 carbons, though utilization of C4- and C12-CoA primers was half as efficient as
that of the intermediate primers (Fig. 2.5a). Wild-type membranes showed relatively low
extension of C14- and C16-CoA, as expected, because BSFs normally stop synthesis at
C14 (8). Chain-length analysis of the FA products from these wild-type reactions
revealed that C4-CoA is elongated efficiently, mostly to C14, with little accumulation of
intermediates (Fig. 2.5a). In contrast, elongation of C6- and C8-CoA was less
processive, with more accumulation of intermediates. A similar phenomenon was
reported for the S. cerevisiae ELOs (17).
In elo1 membranes, synthesis from primers ranging from C4- to C12-CoA was
almost completely lost, whereas synthesis from C14- and C16-CoA was less affected
(Fig. 2.5a). Despite these observations, other evidence indicates that ELO1 is mainly
involved in synthesis up to C10. As shown below (Figs. 2.4 and 2.5a), BSF elo2
membranes efficiently utilize primers up to C8 (forming C10), an activity that must be
due to ELO1. Furthermore, cerulenin does not affect synthesis in the cell free system up
to C10, but blocks further elongation (8), suggesting that it inhibits ELO2 but not ELO1.
We do not know the reason why BSF elo1 is defective in ELO2 activity; perhaps ELO1
protein is required for downstream ELO activity in vitro.
We used the same approach to study the specificities of ELO2–4. Membranes
from elo2 were mainly deficient in extending C10- and C12-CoA primers (Fig. 2.4).
As mentioned, chain length analysis of these products showed that elo2 stops FA
40
synthesis at C10 with shorter primers (Fig. 2.5a). The minor chain elongation of C10and C12-CoA in elo2 membranes is likely due to overlapping ELO1 and ELO3
activities. Similarly, elo3 membranes appeared deficient in extending C14- and C16CoA and chain elongation of shorter primers stopped at C14 (Figs. 2.4 and 2.5a). The
chain-length specificity of ELO3 was confirmed using PCF ELO3 RNAi membranes
(Fig. 2.2). Residual elongation of C14-CoA in elo3 membranes is likely due to
overlapping ELO activity. Products of elo4 membranes resembled those of wild-type,
indicating ELO4 is not involved in FA synthesis up to C18. ELO4 may elongate FAs
longer than those studied here or may function on unsaturated FAs. Substrate preferences
for ELO1–3 are summarized in Fig. 2.5b. Taken together, these data indicate that
ELO1–3 account for virtually all of the FA synthesis observed in the cell-free system.
In vivo FA synthesis
We next tested whether the ELO pathway is responsible for FA synthesis in vivo.
We identified a single enoyl-CoA reductase (EnCR; NCBI AAX80213) on chromosome
three by homology to that from S. cerevisiae, where it functions downstream of all three
yeast ELOs (18). We found EnCR RNAi cells defective for growth in both low lipid
(Fig. 2.6a) and normal media (not shown), indicating that the ELO pathway is essential
in PCF trypanosomes. The growth defect was reversed by stearate addition (Fig. 2.6a).
This rescue, signifying that inadequate stearate is produced in the RNAi cells,
demonstrates that the ELO pathway is involved in stearate synthesis in vivo. We then
assayed RNAi membranes for cell-free FA synthesis and found 77%, 63%, and 54%
reduced activity using C4-, C10-, and C14-CoA primers, respectively, compared to
41
uninduced control (Fig. 2.6b). As with many other trypanosome genes, EnCR RNAi was
incomplete. Based on the RNAi phenotype, it is likely, as in yeast, that ELO1–3 share a
single EnCR. Finally, we measured the effect of EnCR RNAi on FA synthesis in vivo.
Since trypanosomes do not take up [14C]acetate (8), a FA precursor, we radiolabeled
them with [14C]threonine which is catabolized to glycine and acetyl-CoA (19). The latter
compound is incorporated into FAs following conversion to malonyl-CoA (19, 20). We
detected all FA synthesis intermediates between C8 and C18 (C6 methyl ester is lost
during the hexane extraction) indicating that the whole pathway is active in vivo.
Because [14C]threonine labeling of FAs from C8 to C18 is reduced by more than half in
normal and low lipid medium by EnCR RNAi (Fig. 2.6c), and because RNAi was likely
incomplete, the ELO pathway must be responsible at least for the majority of in vivo FA
synthesis in T. brucei.
Regulation of the ELO pathway
It appears that the T. brucei ELO pathway is regulated in vivo. Comparison of
lanes 1 and 3 in Fig. 2.6c (uninduced for RNAi) shows that PCF in vivo FA synthesis is
twice as high when cells are grown in low lipid medium, indicating that the entire ELO
pathway is up-regulated in a low lipid environment. In a control experiment, we found
that [14C]threonine was taken up at the same or lower rate by cells grown in low lipid
medium than by those grown in normal medium (Fig. 2.7). Therefore, the higher rate of
FA synthesis in low lipid medium cells is not due to higher uptake of the [14C]threonine
precursor. Another mode of regulation involves mRNA levels. Curiously, ELO3 mRNA
was reduced in elo2 (Fig. 2.8). This reduction was not observed for ELO2 mRNA in
42
elo1. Based on this observation, we hypothesize limiting intracellular myristate downregulates BSF ELO3. Myristate, normally made by ELO2 for VSGs, must become scarce
in elo2. Further depletion of myristate by ELO3 activity would be undesirable and thus
cells reduce ELO3 activity via ELO3 mRNA levels. A correlating decrease in ELO3
activity was observed in vitro for elo2 (Fig. 2.4).
FA synthesis in other trypanosomatids
The genomes of related disease-causing trypanosomatids Leishmania major and
Trypanosoma cruzi also do not encode type I FA synthases but do predict type II and
ELO systems. Indeed, like T. brucei, their membranes efficiently synthesize FAs from
butyryl-CoA but not acetyl-CoA (Fig. 2.9). Although in vitro synthesis activity was
more robust in T. cruzi than in L. major, FA synthesis products ranging from C8 to C18
were observed in both systems.
43
DISCUSSION
The T. brucei elongase pathway as well as the GPI remodeling pathway and
enzymes for phospholipid biosynthesis all conveniently reside in the ER membrane. This
proximity minimizes the problem of transporting fatty acids from the first pathway to the
latter two—myristate for GPI remodeling and longer FAs for phospholipid synthesis.
The type II FA synthase localizes to the mitochondrial matrix (J. Stephens, unpublished)
and probably makes octanoic acid as a precursor for lipoate and longer FAs for
mitochondrial phospholipids. In light of these discoveries, it becomes interesting that the
GPI remodeling enzymes can utilize FAs shorter than myristate in vitro (21); the
remodeling enzymes utilize lauroyl-CoA and octanoyl-CoA more efficiently than
decanoyl-CoA (21), the end product of ELO1 dependent FA synthesis. Myristate
specificity in in vivo remodeling was previously assumed to be due to the lack of other
available short-chain fatty acyl-CoAs. However, we now know that these other
substrates (octanoyl-CoA to lauroyl-CoA) are present in the cell as intermediates in the
FA synthesis pathway. In general, enzyme active site pockets can sterically exclude
larger unwanted substrates but it is not as easy to exclude smaller substrates. It may be
that the elongase pathway faithfully channels the growing acyl-CoAs with no release, so
the myristoyl-CoA product could be directly transferred to the remodeling pathway
within the ER membrane. Or perhaps myristate exchange, a second and distinct VSG
myristoylation pathway (22), evolved in the cells for proof-reading purposes precisely
because of the availability of these substrates and the promiscuity of the remodeling
enzymes.
44
What is the source of precursors for FA synthesis? We can account for the source
of malonyl-CoA and NADPH but not butyryl-CoA that are required for ELO pathway FA
synthesis in vivo. The malonyl-CoA is formed from acetyl-CoA carboxylase (ACC)
within the cytosol (K. Paul, unpublished). Thus it is likely that the ER resident ELO
pathway enzymes face the cytosol. NADPH can be regenerated from NADP+ by malic
enzyme in the conversion of malate to pyruvate in the cytosol (19) and by the pentosephosphate pathway which appears both glycosomal and cytosolic in trypanosomes (23).
In addition, we know trypanosomes are able to take up C10 and longer exogenous fatty
acids (although C10 and C12 are not present in normal serum (24)) that can then be
activated by acyl-CoA synthases (25) and then fed into the FA elongase pathway.
L. major and T. cruzi have 13 and five ELO homologs, respectively. TbELO1–3
and its 11 L. major and 4 T. cruzi orthologs diverge from TbELO4 and its 2 L. major and
1 T. cruzi ortholog in a phylogenetic tree based on ClustalW protein alignment.
Interestingly, the ELOs from the three trypanosomatids were more similar to each other
than to any of the S. cerevisiae ELOs. This divergence is mirrored by differential drug
inhibition. Cerulenin inhibits extension beyond C10 in the T. brucei ELO system (8), but
it does not inhibit the yeast ELO system (26, 27). Plant fatty acid elongases (FAEs),
which do not share homology to animal ELOs, are inhibited by high doses of cerulenin
(28).
In addition to saturated FAs, unsaturated FAs are major components of membrane
phospholipids. In T. brucei, in the chain-length analysis of FA products for the C16-CoA
primer (Fig. 2.5), we detected a minor radiolabeled band comigrating with C16 and an
even fainter band at C14. Argentation TLC of radiolabeled methyl esters of the FA
45
synthesis products (Fig. 2.10) revealed the presence of mono-, di-, tri-, and tetraunsaturated species. The singly and doubly unsaturated fatty acids were more abundant
and are likely to be mostly C18:1 and C18:2.
The data in Fig. 2.6a indicate that the ELO pathway is essential in PCF T. brucei.
The BSF ELO knockout lines grew normally in culture and were competent for infection
in rats, suggesting that this pathway may not be required for growth in lipid rich
environments. It is also possible that ELO specificies overlap sufficiently to compensate
for knockout of a single ELO. Our attempts to construct a BSF triple ELO1–3 knockout
failed to yield transformants perhaps because this pathway is in fact essential for BSFs.
We are currently making a conditional knockout for the EnCR gene in BSF trypanosomes
to evaluate the physiological relevance of this pathway in BSFs.
An interesting feature of the ELO pathway is that it is regulated to suit the
parasite's needs. PCFs up-regulated the pathway in low-lipid medium (Fig. 2.6c) and
BSFs down-regulated ELO3 mRNA (and ELO3 activity) in elo2 (Figs. 2.8a and 2.4) in
what we interpreted as the cell's attempt to conserve myristate in an otherwise lipid-rich
environment. In light of our understanding of trypanosome FA synthesis, three additional
observations support in vivo ELO pathway regulation in T. brucei. First, the rate of FA
synthesis in PCFs is 5.3 times higher than in BSFs as measured in the cell-free system
(8). Second, life-cycle stage-dependent regulation of ELO3 is apparent in that BSFs
make predominantly myristate whereas PCFs produce longer FAs (8). Third, we
previously reported BSFs grown in medium containing 5% serum efficiently elongate
exogenous [3H]myristate to palmitate and stearate whereas trypanosomes in whole blood
do not (6). Whole blood has 20 times more palmitate and stearate than media with 5 %
46
serum. Given our current knowledge of the elongase pathway, this finding indicates
ELO3 is not only up-regulated in PCFs but is up-regulated when needed in BSFs such as
when they are cultured in FA-depleted medium. Taken together, these observations
suggest that trypanosomes regulate ELO activity in response to environmental FAs. We
do not yet know the molecular mechanisms by which trypanosomes sense the level of
exogenous FAs or regulate the elongation pathway. Regulation of ELO mRNA levels by
exogenous FAs has been observed in S. cerevisiae (14).
In conclusion, T. brucei and its trypanosomatid kin are unique among eukaryotes
in that they synthesize FAs with a microsomal elongase pathway. Remarkably, it seems
that this pathway is regulated in T. brucei to satisfy the needs of the parasite in different
environments during its life cycle or even within a host. In BSFs, synthesis of myristate
for GPI anchors would be favored by down-regulation of ELO3. However, under certain
circumstances, such as when BSFs invade the cerebrospinal fluid, where the ratio of total
lipids to that in serum is ~0.003 (24), higher rates of synthesis of longer FAs as well as
myristate must be needed. This could simply be accomplished by selective regulation of
the ELOs. PCFs residing in the tsetse fly midgut might encounter a different situation.
As the tsetse absorbs the blood meal and BSFs differentiate into dividing PCFs, FAs may
become limiting and the entire ELO pathway may be up-regulated (as in Fig. 2.6c, lanes
1 and 3) to produce the palmitate and stearate needed for phospholipid synthesis. Thus,
the diverse needs of the trypanosome must have had a hand in the evolution of this
piecemeal de novo FA synthesis machinery that accommodates the parasite's need for
rapid adaptation to differing host environments.
47
EXPERIMENTAL PROCEDURES
Trypanosomes
Procyclic Trypanosoma brucei brucei strain 29-13 (29) and derived cells were cultured at
27 °C and 5% CO2 in SDM-79 medium (30). Bloodstream wild-type 427, 90-13, and
knockout strains were cultured at 37 °C and 5% CO2 in HMI-9 media (31). BSF lysates
were made from trypanosomes grown in rats. Strain 90-13 was grown in non-irradiated
retired male breeder rats while elo1–4 strains were grown in lethally irradiated (1000
rad) rats. Trypanosomes were harvested when parasitemias reached 1-2 x 109
trypanosomes/ml by cardiac exsanguination. Blood was transferred to an equal volume
of Percoll® (Sigma) containing 67 U/ml heparin, 8.6% sucrose, and 2.0% glucose (pH
adjusted to 7.5 with solid HEPES) (32). After centrifugation (28,000 x g, 4 °C, 15 min),
the upper layer of trypanosomes was purified further by DE53 chromatography (33).
BSF and PCF crude cell membranes were prepared by washing cells in BBSG (50 mM
bicine-Na, pH 7.5, 50 mM NaCl, 5 mM KCl, and 70 mM glucose). Cells were lysed at
109 cells/ml on ice in 1 mM HEPES-KOH, pH 7.5, 1 mM EDTA, 1 µg/ml leupeptin, and
0.1 mM TLCK. After 5 min, an equal volume of 100 mM HEPES-KOH, pH 7.5, 50 mM
KCl, 10 mM MgCl2, 300 mM sucrose, 1 µg/ml leupeptin, and 0.1 mM TLCK was added.
Lysate aliquots were snap-frozen and stored at –80 °C.
RNAi
A fragment from a region of the target open reading frame was cloned into pZJM, a
vector for tetracycline-inducible RNAi (34). For ACP, the pZJM insert was 447 bp
encoding the entire open reading frame. The ELO1 insert was 571 bp, starting with
48
5' TTGCCAATAA, ELO3 insert was 415 bp, starting with 5' ACACGGCCTT, and that
for EnCR was 416 bp, starting with 5' GGAGCTGGAG. pZJM transformants of T.
brucei 29-13 were selected and cloned by limiting dilution. RNAi induction and
Northern analyses were performed as described (34).
Generating ELO knockouts and Southern analysis
ELO genes in BSF 427 strain were replaced by homologous recombination with pKONEO
and pKOHYG cassettes containing ~500 bp each of 5' and 3' sequences surrounding the
open reading frame of the target gene (35). Knockout transformants were cloned and
drug selection was removed after gene deletions were confirmed by Southern analysis.
For the latter, genomic DNA was isolated using Puregene™ DNA isolation kit (Gentra
Systems), restriction digested, and fractionated. DNA was partially depurinated,
transferred to GeneScreen Plus Nylon membrane (Perkin Elmer), and hybridized at 40 °C
to digoxigenin-labeled ELO gene sequences according to instructions in the DIG Probe
Synthesis Kit and DIG Easy Hyb Kit (both from Roche).
Cell-free FA synthesis
Thawed cell lysates were washed twice in HKML buffer (50 mM HEPES-KOH, pH 7.4,
25 mM KCl, 5 mM MgCl2, 1 µg/ml leupeptin) and protein concentration was determined
by the Bradford assay (BioRad). Duplicate FA synthesis assays were conducted at
1 x 109 cell equivalents/ml in HKML, 1 mM DTT, 2 mM each of NADPH and NADH,
50 µM [2-14C]malonyl-CoA (American Radiolabeled Chemicals, 55 mCi/mmol), and
50 µM acyl-CoA primer. Incubations were at 37 °C for 30 min unless otherwise
49
indicated. Lipids were extracted and fractionated on silica gel 60 HPTLC plates (EMD
Chemicals Inc.) and visualized on Kodak BioMax MR film (8). Free FAs on TLC plates
were quantitated using a Fujifilm BAS-2500 Phosphorimager. FA synthesis, measured as
photo-stimulated luminescence, was corrected for local background and normalized for
protein concentration.
Fatty acid chain length analysis
Chloroform/methanol/water (CMW) 10:10:3 extracts from cell-free FA synthesis
reactions (~2.0 x 107 cell equivalents) or 3H-labeled FA standards were mixed with
carrier FAs (~0.2 nmol each of C8, C10, C12, C14, and C16). Samples were dried under
N2 gas, dissolved in 0.1 ml dry methanol containing 2% H2SO4 and 0.2% benzene, and
incubated at 65 °C for 2-16 h to form methyl esters. Reactions were stopped with 100 µL
water and FA methyl esters were extracted with 100 µL hexane (36). Methyl esters of
shorter fatty acids, especially C6, are lost during the hexane extraction. The hexane
extracts were fractionated on C18 RP-TLC plates (Analtech) using CMW 5:15:1 v/v as
mobile phase.
Measurement of in vivo FA synthesis
Cells were grown either in normal medium (SDM-79 containing 10% FBS) or in low
lipid medium (SDM-79 supplemented with 10% delipidated FBS from Cocalico
Biologicals). The latter medium contains ~20% of the lipids of normal medium. Stearate
(35 µM) with -cyclodextrin (350 µM) as carrier was delivered and RNAi was induced
on day zero. Prior to metabolic labeling, cells were maintained in either normal medium
50
or in low lipid medium. After five days, cells (3-5 x 106/ml) were washed and
resuspended at 1.2 x 108 cells/ml in delipidated FBS containing appropriate drugs. Cells
(6 x 107) were then treated with 1.2 µCi L-[UL-14C]threonine (Sigma, 155 mCi/mmol)
for 4.3 hr (27 °C, 5% CO2). Cells were centrifuged, washed with PBS, and extracted in
0.8 ml CMW 10:10:3. Part of the CMW phase (1.5 x 107 cell equivalents) was used for
chain length analysis. Cells had normal morphology and motility after radiolabeling.
[14C]Threonine uptake
Procyclic T. brucei grown five days in normal or low lipid medium was washed twice in
BBSG and resuspended at 1.2 x 108 cell/ml in delipidated FBS containing appropriate
drugs. For each 350 µl, 0.84 µCi [14C]threonine was added. Samples were incubated at
27 °C, under 5% CO2. At various times up to 4.5 hours, 50 or 100 µl of sample was
layered onto 100 µl of 95 % dibutyl phthalate and 5 % paraffin oil, and cells were
centrifuged through the oil mixture. The upper aqueous phase was removed and the tube
containing the remaining oil and cells was washed twice with 100 µl PBS. The oil
mixture was then removed, and the cell pellet was washed with fresh oil mixture. Cells
were resuspended in PBS, transferred to scintillation fluid, and counted.
FA synthesis in T. cruzi and L. major
T. cruzi CL Brener epimastigotes were grown in LIT medium supplemented with 10%
FBS at 27 C. L. major Friedlin promastigotes were grown in chemically defined M199+
medium (37) supplemented with 2 mM glutamine (Gibco) and 0.1 mM adenosine
(Sigma) at 26 C. Membranes were prepared by the method used for T. brucei except L.
51
major was lysed in double the volume of lysis buffer; thus lysates were half as
concentrated when frozen. Cell-free FA synthesis assays were performed as before with
NADPH (NADH was omitted) and in chain length analyses carrier FAs were omitted.
Argentation TLC
Silver nitrate (Sigma; 1.5 g) was dissolved in 12 ml of 50 % acetonitrile and sprayed onto
a 10 x 10 cm2 silica gel 60 plate using an atomizer. After drying in the hood, the plate
was baked 30 minutes at 95 C. Methyl esters of C18:1, C18:2, C18:3, and C20:4 were
used as standards and in a carrier mix (3.8 % each of 16:1, 18:2, 18:3, and 20:4 methyl
esters and 5% 18:1 methyl ester in hexane) that was spotted on the radiolabeled samples
at the origin (1 µl/lane). The TLC was run 8 cm in dichloromethane. Standards were
visualized by incubating the dried plate in a chamber equilibrated with iodine crystals.
The plate was then sprayed with En3Hance (NEN), dried, and visualized on film.
52
ACKNOWLEDGMENTS
We thank Steve Beverly, Teresa Dunn, Dennis Grab, Yasu Morita, Shilpi Paul, Terry
Shapiro, and Paul Watkins for discussions. We thank Barbara Burleigh and Anne
Chessler for T. cruzi lysates and Dennis Dwyer for L. major cells and media. We thank
Gokben Yildirir for technical assistance and members of our laboratory for support. We
thank Wade Gibson for use of the phosphorimager.
I would also like to thank Jennifer Stephens for use of her data (Fig. 2.1a) and Kimberly
Paul for two vectors (pZJM-ELO1 and pZJM-ELO3) along with the ELO project which
she did not have time to explore before leaving for Clemson. Finally, I would like to
thank Paul Englund for his help in writing this chapter.
This work was supported by a grant from the National Institutes of Health (AI21334).
53
Figure 2.1. Effect of ACP and ELO1 RNAi silencing on FA synthesis. a, Upper
panel shows effect of ACP RNAi for two days on cell-free FA synthesis (1.3 x 108 cell
equivalents/lane) in procyclic trypanosomes. RNAi (+), uninduced control (-). Middle
panel is northern blot showing effect of RNAi on ACP mRNA (~0.9 kb). Load control
(load) is ethidium-stained rRNA. b, Same as a except ELO1 RNAi was for four days
(1.2 x 107 cell equivalents/lane). No lysate (nl) reaction does not contain cell
membranes. ELO1 mRNA is ~2.0 kb. fa, free fatty acid products, PL, phospholipids, o,
origin.
54
Figure 2.2. Effect of ELO3 RNAi on cell-free FA synthesis. ELO3 RNAi in procyclic
trypanosomes was for two days. a, Free FA products from FA synthesis (3.5 x 107 cell
equivalents/lane) in uninduced (-) and RNAi induced (+) membranes. b, Northern blot
showing effect of RNAi on ELO3 mRNA (~2.1 kb). Load control (load) is ethidiumstained rRNA.
a
RNAi
acyl -CoA
primer length
12 14 16 18
+
b
RNAi
mRNA
load
55
- +
Figure 2.3. Southern analysis of BSF ELO1–4 genomic knockouts (elo). Genomic
DNA from wild-type (wt), heterozygous (+/-), and homozygous knockout clones (elo)
was restriction digested with MluI/NotI/XhoI, NdeI/SacI/XhoI, KpnI/PacI/SacI, and
MluI/NotI/XhoI, respectively, for ELO1–4 blots. Fragments were detected with
appropriate probes. ACP 5' UTR sequence was used to control for loading (load).
probe: ELO1
load
wt
elo4+/elo4
elo3+/elo3
wt
elo2+/elo2
kb
85-
wt
elo1+/elo1
Wild-type sample for ELO3 was lost.
ELO2 ELO3 ELO4
56
Figure 2.4. FA synthesis by BSF elo1–4 membranes from various acyl-CoA
primers. The segment of the TLC plate with free FAs is shown (1.8 x 107 cell
equivalents/lane). The minor upper component is neutral lipid (e.g. triglyceride). Protein
per lane ranged from 7.5 to 10.3 µg. The no primer control for elo3 was lost (*), but in
a duplicate set the product level was similar to that of wild-type. Phosphorimager
quantitation of FAs are indicated in arbitrary units (au) for wild-type and as a percentage
of wild-type for the knockouts (% wt).
57
Figure 2.5. Chain length analysis of FA synthesis products. a, FA methyl esters
(~5000 DPM/lane when available) were fractionated on reverse-phase TLC. Thus,
intensities are not proportional to products formed. The uppermost band is an aqueous
phase contaminant that increases with volume of hexane phase loaded. The band labeled
with * in the C16-CoA panel is C18:1 (determined by argentation TLC in Fig 2.10). We
have not characterized the faint ladder of C8 to C18 products observed in elo2 and
elo3 lanes (C16-CoA primer) and that is sometimes observed in no primer samples. b,
Substrate preferences for ELO1–3.
58
Figure 2.6. Effect of RNAi knockdown of enoyl-CoA reductase (EnCR) in PCFs. a,
Growth of cells in low lipid medium in the presence or absence of 35 µM stearate.
Uninduced control, no stearate (closed circles); RNAi, no stearate (closed diamonds);
Uninduced control plus stearate (open triangles); RNAi plus stearate (open squares).
Inset is northern blot showing reduced EnCR mRNA (~3.1 kb) after six days of RNAi.
Load control (load) is ethidium-stained rRNA. b, Portion of TLC showing free FA
products formed in 10 minutes with C4, C10, and C14-CoA in uninduced (-) and five day
RNAi induced (+) membranes (1.6 x 107 cells/lane). Protein per lane was 15.7 µg and
23.1 µg, respectively. c, Chain-length analysis of FAs labeled in vivo with [14C]threonine
after five (+) or no (-) days of RNAi (1.5 x 107 cell equivalents/lane). Cells were cultured
either in normal (100%) or low lipid medium (20%). Lane 5 is a zero time point control.
59
Cell density (log cells/ml)
a
11
- +
9
RNAi
mRNA
load
7
5
b
RNAi
Q uickTim e™ and a
G r aphics decom pr essor
ar e needed t o see t his pict ur e.
0
0
3
6
3
6
Days of RNAi
acyl-CoA
primer
4 10 14
c
1 2
9
9
3 4 5
C8-
-
C18-
+
o-
RNAi
- + - +
lipids 100%
60
20%
Figure 2.7. [14C]Threonine uptake and incorporation into PCF cells. Cells grown in
normal (closed circles) or low lipid medium (open triangles) were incubated with
[14C]threonine just as in the experiment shown in Fig. 2.6c. Cells were collected at timed
intervals up to 4.5 hours, centrifuged through oil, and scintillation counted. Average
disintegrations per minute (DPM) values from two separate labelings and standard
DPM per 10 5 cells
deviations are shown.
600
400
200
0
0
2.5
Time (hours)
61
5
Figure 2.8. Northern blot of BSF elo1–3 and map of ELO1–3 loci. a, Wild-type (wt)
and elo1–3 mRNA probed with ELO1–3 gene sequence. Load control (load) is an
ethidium-stained rRNA. ELO1 blots were reprobed for ELO3 mRNA. b, Map of ELO1–
3 genes in tandem on chromosome 7. Boxes represent open reading frames and arrow
wt
elo1
elo2
elo3
wt
elo1
elo2
elo3
wt
elo1
elo2
elo3
marks the direction of transcription. Ticks are spaced at 0.5 kb intervals.
a
kb
4.4
2.4
1.4
probe: ELO1 ELO2
load
ELO3
b
e lo3
e lo2
e lo1
ELO3
ELO2
ELO1
62
Figure 2.9. Cell-free FA synthesis in T. cruzi epimastigotes and L. major
promastigotes. T. cruzi and L. major membranes synthesize FAs from C4-CoA (4) and
longer primers but not from acetyl-CoA (2) similar to the T. bruci ELO system. No
primer (no) lane activity is likely due to endogenous primers. Upper panel is silica gel 60
TLC analysis of free fatty acid products (fa). PL, phospholipids, o, origin. Lower panels
show chain lengths of products. Unsaturated FAs such as C18:1 and C18:2 elute with
C16:0 and C14:0. For all panels, 1.3 x 107 cell equivalents/lane were loaded.
63
acyl -CoA primer length
Trypanosoma cruzi
no 2
Leishmania major
4 10 14
no 2 4 10 14
faPL-
oC8
12
14
18
64
Figure 2.10. Argentation TLC of FA methyl ester products. A silica TLC plate was
sprayed with silver nitrate which retards the migration of unsaturated species. The FA
synthesis products shown are from wild-type membranes [14C]labeled in the presence of
C6-CoA (6) and C16-CoA (16) primers. Nonradiolabeled unsaturated FA methyl ester
standards were visualized by iodination. Origin, o; saturated species, 0;
monounsaturated, 1; diunsaturated, 2; triunsaturated, 3; and tetraunsaturated species, 4.
6 16
unsaturated
01234o-
65
REFERENCES
1.
Donelson JE. Antigenic variation and the African trypanosome genome. Acta
Trop 2003;85(3):391-404.
2.
Ferguson MA, Cross GA. Myristylation of the membrane form of a Trypanosoma
brucei variant surface glycoprotein. J Biol Chem 1984;259(5):3011-5.
3.
Masterson WJ, Raper J, Doering TL, Hart GW, Englund PT. Fatty acid
remodeling: a novel reaction sequence in the biosynthesis of trypanosome
glycosyl phosphatidylinositol membrane anchors. Cell 1990;62(1):73-80.
4.
Field MC, Menon AK, Cross GA. A glycosylphosphatidylinositol protein anchor
from procyclic stage Trypanosoma brucei: lipid structure and biosynthesis. Embo
J 1991;10(10):2731-9.
5.
Dixon H, Ginger CD, Williamson J. The lipid metabolism of blood and culture
forms of Trypanosoma lewisi and Trypanosoma rhodesiense. Comp Biochem
Physiol B 1971;39(2):247-66.
6.
Doering TL, Pessin MS, Hoff EF, Hart GW, Raben DM, Englund PT.
Trypanosome metabolism of myristate, the fatty acid required for the variant
surface glycoprotein membrane anchor. J Biol Chem 1993;268(13):9215-22.
7.
Paul KS, Jiang D, Morita YS, Englund PT. Fatty acid synthesis in African
trypanosomes: a solution to the myristate mystery. Trends Parasitol
2001;17(8):381-7.
8.
Morita YS, Paul KS, Englund PT. Specialized fatty acid synthesis in African
trypanosomes: myristate for GPI anchors. Science 2000;288(5463):140-3.
66
9.
Smith S. The animal fatty acid synthase: one gene, one polypeptide, seven
enzymes. Faseb J 1994;8(15):1248-59.
10.
White SW, Zheng J, Zhang YM, Rock CO. The Structural Biology of Type II
Fatty Acid Biosynthesis. Annu Rev Biochem 2004.
11.
Zhang L, Joshi AK, Hofmann J, Schweizer E, Smith S. Cloning, expression, and
characterization of the human mitochondrial beta-ketoacyl synthase.
Complementation of the yeast CEM1 knock-out strain. J Biol Chem
2005;280(13):12422-9.
12.
Paul KS, Bacchi CJ, Englund PT. Multiple triclosan targets in Trypanosoma
brucei. Eukaryot Cell 2004;3(4):855-61.
13.
Moon YA, Shah NA, Mohapatra S, Warrington JA, Horton JD. Identification of a
mammalian long chain fatty acyl elongase regulated by sterol regulatory elementbinding proteins. J Biol Chem 2001;276(48):45358-66.
14.
Toke DA, Martin CE. Isolation and characterization of a gene affecting fatty acid
elongation in Saccharomyces cerevisiae. J Biol Chem 1996;271(31):18413-22.
15.
Oh CS, Toke DA, Mandala S, Martin CE. ELO2 and ELO3, homologues of the
Saccharomyces cerevisiae ELO1 gene, function in fatty acid elongation and are
required for sphingolipid formation. J Biol Chem 1997;272(28):17376-84.
16.
Leonard AE, Pereira SL, Sprecher H, Huang YS. Elongation of long-chain fatty
acids. Prog Lipid Res 2004;43(1):36-54.
17.
Dittrich F, Zajonc D, Huhne K, et al. Fatty acid elongation in yeast--biochemical
characteristics of the enzyme system and isolation of elongation-defective
mutants. Eur J Biochem 1998;252(3):477-85.
67
18.
Kohlwein SD, Eder S, Oh CS, et al. Tsc13p is required for fatty acid elongation
and localizes to a novel structure at the nuclear-vacuolar interface in
Saccharomyces cerevisiae. Mol Cell Biol 2001;21(1):109-25.
19.
van Weelden SW, van Hellemond JJ, Opperdoes FR, Tielens AG. New functions
for parts of the Krebs cycle in procyclic Trypanosoma brucei, a cycle not
operating as a cycle. J Biol Chem 2005;280(13):12451-60.
20.
Klein RA, Linstead DJ. Threonine as a perferred source of 2-carbon units for lipid
synthesis in Trypanosoma brucei. Biochem Soc Trans 1976;4(1):48-50.
21.
Morita YS, Englund PT. Fatty acid remodeling of glycosyl phosphatidylinositol
anchors in Trypanosoma brucei: incorporation of fatty acids other than myristate.
Mol Biochem Parasitol 2001;115(2):157-64.
22.
Buxbaum LU, Milne KG, Werbovetz KA, Englund PT. Myristate exchange on
the Trypanosoma brucei variant surface glycoprotein. Proc Natl Acad Sci U S A
1996;93(3):1178-83.
23.
Duffieux F, Van Roy J, Michels PA, Opperdoes FR. Molecular characterization
of the first two enzymes of the pentose-phosphate pathway of Trypanosoma
brucei. Glucose-6-phosphate dehydrogenase and 6-phosphogluconolactonase. J
Biol Chem 2000;275(36):27559-65.
24.
Lentner C. Geigy Scientific Tables: Units of Measurement, Body Fluids,
Composition of the Body, Nutrition. Eighth ed. West Caldwell: Medical
Education Division, Ciba-Geigy Corporation; 1981.
68
25.
Jiang DW, Englund PT. Four Trypanosoma brucei fatty acyl-CoA synthetases:
fatty acid specificity of the recombinant proteins. Biochem J 2001;358(Pt 3):75761.
26.
Awaya J, Ohno T, Ohno H, Omura S. Substitution of cellular fatty acids in yeast
cells by the antibiotic cerulenin and exogenous fatty acids. Biochim Biophys Acta
1975;409(3):267-73.
27.
Rossler H, Rieck C, Delong T, Hoja U, Schweizer E. Functional differentiation
and selective inactivation of multiple Saccharomyces cerevisiae genes involved in
very-long-chain fatty acid synthesis. Mol Genet Genomics 2003;269(2):290-8.
28.
Schneider F, Lessire R, Bessoule JJ, Juguelin H, Cassagne C. Effect of cerulenin
on the synthesis of very-long-chain fatty acids in microsomes from leek seedlings.
Biochim Biophys Acta 1993;1152(2):243-52.
29.
Wirtz E, Leal S, Ochatt C, Cross GA. A tightly regulated inducible expression
system for conditional gene knock-outs and dominant-negative genetics in
Trypanosoma brucei. Mol Biochem Parasitol 1999;99(1):89-101.
30.
Brun R, Schonenberger. Cultivation and in vitro cloning or procyclic culture
forms of Trypanosoma brucei in a semi-defined medium. Short communication.
Acta Trop 1979;36(3):289-92.
31.
Hirumi H, Hirumi K. Continuous cultivation of Trypanosoma brucei blood stream
forms in a medium containing a low concentration of serum protein without
feeder cell layers. J Parasitol 1989;75(6):985-9.
32.
Grab DJ, Bwayo JJ. Isopycnic isolation of African trypanosomes on Percoll
gradients formed in situ. Acta Trop 1982;39(4):363-6.
69
33.
Lanham SM, Godfrey DG. Isolation of salivarian trypanosomes from man and
other mammals using DEAE-cellulose. Exp Parasitol 1970;28(3):521-34.
34.
Wang Z, Morris JC, Drew ME, Englund PT. Inhibition of Trypanosoma brucei
gene expression by RNA interference using an integratable vector with opposing
T7 promoters. J Biol Chem 2000;275(51):40174-9.
35.
Lamb JR, Fu V, Wirtz E, Bangs JD. Functional analysis of the trypanosomal
AAA protein TbVCP with trans-dominant ATP hydrolysis mutants. J Biol Chem
2001;276(24):21512-20.
36.
Fosbrooke AS, Tamir I. A modified method for the preparation of methyl esters
of a mixture of medium-chain and long-chain fatty acids. Application to the
determination of serum triglyceride and non-esterified fatty acid composition and
concentration by gas-liquid chromatography. Clin Chim Acta 1968;20(3):517-22.
37.
McCarthy-Burke C, Bates PA, Dwyer DM. Leishmania donovani: use of two
different, commercially available, chemically defined media for the continuous in
vitro cultivation of promastigotes. Exp Parasitol 1991;73(3):385-7.
70