(Received for publication, March 10, 1997, and in revised form, April 23, 1997)
From the Departments of Biochemistry,
¶ Microbiology, and § Internal Medicine, University of
Iowa and the
Veterans Affairs Medical Center, Iowa
City, Iowa 52242
GP46 is an abundant glycoprotein of 46 kDa on the
surface of the promastigote form of most Leishmania
species. We show that the steady state level of GP46 mRNA increases
>30-fold as Leishmania chagasi promastigotes develop
in vitro from a less infectious form during logarithmic
growth to a highly infectious form in the stationary phase of
cultivation. Nuclear run-on experiments demonstrate that this increase
in GP46 mRNA abundance is regulated post-transcriptionally.
Plasmids containing the 3-untranslated regions (UTRs) and downstream
intergenic regions (IRs) of two different GP46 genes fused immediately
downstream of the
-galactosidase coding region were transfected into
L. chagasi, and
-galactosidase activity and mRNA
levels were examined. The presence of the 3
-UTR + IR of one GP46 gene
(gp46A) resulted in a steady increase in
-galactosidase
activity and mRNA level as the transfected promastigotes developed
from logarithmic to stationary phase. This differential effect
parallels that of the 3
-UTRs + IRs of a family of genes for an
unrelated Leishmania surface glycoprotein, GP63. Thus, post-transcriptional regulation of the genes for two different surface
glycoproteins of Leishmania occurs via a similar
mechanism.
Protozoan parasites of the genus Leishmania cause a diverse group of diseases collectively called leishmaniasis, which range in severity from spontaneously healing cutaneous ulcers to potentially fatal visceral disease. These parasites have a digenetic life cycle, passing from the infected sandfly vector to the mammalian host as the female fly takes a blood meal. In the fly, Leishmania exist as extracellular flagellated promastigotes within the alimentary canal, and in mammals, they exist as intracellular aflagellate amastigotes within phagolysosomes of macrophages.
While multiplying in the sandfly gut, promastigotes progress through a series of morphologically distinct developmental stages culminating in the highly infectious metacyclic stage (1). Some aspects of this development are mirrored in vitro during the growth of promastigotes from logarithmic (less infectious) phase to stationary (highly infectious) phase in liquid culture medium (2). For example, in several Leishmania species, the glycosyl side chain of the abundant lipophosphoglycan on the surface of promastigotes elongates as parasites grow to their infectious metacyclic form in stationary phase (3, 4). This increase in metacyclic lipophosphoglycan size has been correlated with enhanced virulence of the parasite (5, 6). Another major surface constituent of promastigotes is a well characterized 63-kDa glycoprotein called GP63,1 a zinc protease with broad substrate specificity and a wide pH optimum (7, 8). In Leishmania chagasi, the cause of South American visceral leishmaniasis, the amount of GP63 increases 11-fold as the promastigotes grow from logarithmic to stationary phase (9). Biological functions ascribed to GP63 include evasion of complement-mediated lysis, attachment to macrophages, cleavage of various substrates including C3b, and elicitation of both humoral and cellular immune responses that are protective in several mouse models (10-14).
In earlier studies of GP63 expression, we found that L. chagasi has >18 tandemly arrayed GP63 genes divisible into three
classes on the basis of their developmental expression (9, 12). Genes of the class encoding 2.7-kb mRNAs occurring predominantly in logarithmic phase promastigotes are called mspL, where
msp and L refer to the genes encoding major
surface protease (GP63) and logarithmic phase, respectively. Genes of
the class encoding 3.0-kb mRNAs found predominantly in stationary
phase promastigotes are called mspS. Promastigotes at an
intermediate growth phase possess both RNAs. Class mspC is
comprised of a single gene that is constitutively expressed at a low
level throughout promastigote growth as 2.6- and 3.1-kb mRNAs. The
3-UTR and downstream intergenic region (IR) of the mspS
genes play an important role in the stationary phase expression of
these genes, whereas the corresponding 3
-UTR + IR sequence of the
mspL genes does not seem to be responsible for their
differential RNA expression (15).
Another abundant protein present on the surface of promastigotes is a 46-kDa glycoprotein called GP46 or promastigote surface antigen 2 (16, 17). Genes encoding GP46 have been detected in Crithidia fasciculata and in all Leishmania sp. examined except for members of the Leishmania braziliensis complex (18, 19). The organization of this gene family has not been fully characterized in any Leishmania species, but in those that have been investigated, multiple nonidentical copies of GP46 genes are arranged in clusters (18). Although the biological function(s) of GP46 are not known, immunization with GP46 partially protects experimental mice against challenge with Leishmania amazonensis (20, 21). Recent experiments have demonstrated that Leishmania major amastigotes and promastigotes express different GP46 mRNAs and proteins (22), indicating that these genes undergo developmental regulation.
Since GP63 genes are differentially expressed during L. chagasi promastigote growth, we questioned whether GP46 gene
expression also varies during promastigote growth. We found that the
abundance of GP46 RNA in L. chagasi promastigotes increases
dramatically as they grow from logarithmic to stationary phase,
paralleling the increase in mspS mRNA. Stationary phase
promastigotes have approximately equivalent steady state levels of GP46
and mspS mRNAs. Two GP46 genes were isolated from the
L. chagasi genome, and a plasmid-borne -galactosidase
(
GAL) reporter gene was used to show that the 3
-UTR + IR of one
GP46 gene is responsible for its growth-regulated expression.
An isolate of L. chagasi derived from a Brazilian patient with visceral leishmaniasis was maintained in hamsters. Promastigotes were cultured in vitro at 26 °C as described (9) in a modified minimal essential medium (HOMEM) supplemented with 10% fetal bovine serum and 5.6 µg/ml hemin (23). Logarithmic and stationary phase promastigotes were defined by morphology and concentration criteria (24). Medium for stably transfected promastigotes contained 40 µg/ml G418 (Life Technologies, Inc.).
Genomic and cDNA LibrariesA L. chagasi genomic DNA library and a L. chagasi cDNA library made from RNA isolated from stationary phase promastigotes were described earlier (9, 12). Clones hybridizing to radiolabeled DNA probes were purified and processed as described (Stratagene, La Jolla, CA).
PCR Amplification of Partial GP46 cDNA ClonesA series
of primers designed from the coding sequence for L. amazonensis GP46 (accession number M38368) were used in attempts to PCR-amplify a partial-length GP46 coding region from L. chagasi and Leishmania mexicana genomic DNA. The
forward primer, 5 GGGGACGAGCGACTTCAC 3
, and reverse primer, 5
CCGGCACAGACCACGAGA 3
, successfully amplified a fragment from L. mexicana DNA that was subsequently used to screen the L. chagasi cDNA library.
To obtain a cDNA sequence corresponding to the 5 end of GP46
mRNA, reverse transcriptase-PCR was conducted using RNA isolated from stationary phase L. chagasi promastigotes and a forward
primer containing a partially spliced leader sequence (5
AACTAACGCTATATAAGTATCAGTT 3
). The reverse primer for this reversed
transcriptase-PCR, 5
CGACGTTGGTGTAGTCGA 3
, was designed from the
sequence of a partial length (2.4 kb) GP46 cDNA isolated from the
L. chagasi cDNA library.
Total L. chagasi DNA was isolated using DNAzolTM (Life Technologies), and Southern blots were conducted as described (25). For Northern blots, the L. chagasi RNA was isolated using a guanidinium-based method (26). RNA was separated by electrophoresis on a 1.2% agarose gel (6% formaldehyde buffered with 160 mM NaH2PO4, pH 6.8), transferred overnight in 6 × sodium saline citrate (25) to positively charged nylon membrane, and fixed by baking 30 min at 80 °C.
DNA probes used in Southern or Northern blot analysis were 1) a 0.6-kb
HpaI fragment from the 3.0-kb GAL coding sequence in
plasmid pX-
gal 2, 2) a 0.98-kb EcoRI-SacII
fragment from the 1.3-kb L. chagasi GP46 coding sequence, 3)
a 0.89-kb NdeI-ScaI fragment from the 1.8-kb
stationary GP63 coding sequence in plasmid pGP63S (9). The tubulin
probe used in initial experiments (see Fig. 2, panel C)
contained the sequence for Trypanosoma brucei rhodesiense
- and
-tubulin (27). The
-tubulin probe used in later
experiments was PCR-amplified from L. chagasi genomic DNA
using oligonucleotide primers 5
ATGCGTGAGGCTATCTGCAT 3
and 5
TTAGTACTCCTCGACGTCC 3
derived from the coding sequence for
-tubulin
of Leishmania donovani (accession number U09612).
For determining the relative abundance of GP46 and GP63 mRNAs, purified 0.98-kb GP46 and 0.89-kb GP63 DNA fragments were quantified by A260 and by the relative intensities of ethidium bromide stains of agarose gel-separated fragments compared with mass standards (Life Technologies). These fragments were serially diluted (in 0.4 N NaOH, 0.01 M EDTA, 0.08 µg of plasmid pAcUW21/ml as a carrier) and applied in parallel rows to positively charged nylon membranes via slot blot. The serial dilutions (1:2, 1:4, 1:8, etc.) spanned a range from 1.25 × 103 to 0.305 picograms of target DNA/slot. These membranes and Northern blot membranes containing RNA from stationary phase promastigotes were incubated in the same hybridization solution and probed with either the 0.98-kb GP46 or 0.89-kb GP63 DNA fragments. The relative abundance of the GP46 and GP63 RNAs was determined by comparing the intensity of the signals on the Northern blots containing the RNAs with the intensity of the signals on the slot blots containing the GP46- and GP63-DNA mass standards.
DNA ConstructionsThe starting plasmids were
pBluescript® SK (Stratagene) and pX-gal 2 (a generous
gift from S. Beverley). A fragment containing the 3
-UTR and IR of
gp46A was isolated after NotI digestion of recombinant phage DNA containing this genomic region of L. chagasi DNA and then ligated into the NotI site of
pX-
gal 2. NotI cleaves 76 bases upstream (5
) of the stop
codon in gp46A; therefore, this clone contains a small piece
of the coding region. A partial 3
-UTR and IR of gp46B was
similarly isolated from recombinant phage DNA by XhoI
digestion, blunt ended, and ligated into pX-
gal 2 that had
previously been digested with NotI, blunt ended, and dephosphorylated. The orientation of the inserts was determined by
sequencing across the insert boundaries. The pX-
gal 2 plasmid constructs containing the 3
-UTRs and IRs of the GP63 genes have been
described (15).
Promastigotes were harvested in logarithmic or stationary phase, pelleted by centrifugation for 5 min at 3000 × g, and washed twice in Hanks' balanced salt solution. Disruption of cells and labeling of nuclei were as described previously (28).
DNA Transfections andL. chagasi
promastigotes were transfected with plasmids and plated onto solid
medium containing 25 or 40 µg/ml G418 for selection and isolation of
clonal transfectants as described (29). Clonal isolates were grown in
liquid culture, and aliquots of 1.5 × 108
promastigotes were removed daily, washed three times (centrifuged for 5 min at 4000 × g then resuspended in 1.4 ml of
phosphate-buffered saline), resuspended in 100 µl of
phosphate-buffered saline, and then stored at 70 °C. For
GAL
assays, aliquots were thawed quickly with the addition of 2 volumes
(200 µl) of lysis buffer (100 mM KH2PO4, pH 7.8, 0.33% Triton X-100), lysed by
three freeze-thaw cycles in dry ice/ethanol and a 37 °C water bath,
and centrifuged (5 min at 10,000 × g). The supernatant
was assayed for protein concentration (bicinchoninic acid reagent and
assay, Pierce) and for
GAL activity (using
Galacton-StarTM chemiluminescent substrate in a
fluorometric assay, CLONTECH, Palo Alto, CA).
Fluorescence was measured in a Monolight® 2010 luminometer
(Analytical Luminescence Laboratory, San Diego, CA).
Standard techniques of molecular biology were used (25). DNA was sequenced using dye terminator cycle sequencing chemistry and analyzed on a 373A stretch fluorescent automated sequencer (Perkin-Elmer). The radioactivity of the blots was quantified using InstantImagerTM electronic autoradiography (Packard Instrument Co.). DNA and protein sequences were compared using the University of Wisconsin Genetic Computer Group program version 7.0.
The cDNA
sequences encoding GP46 have been reported for L. major and
L. amazonensis (30, 31) but not for L. chagasi.
Attempts to PCR-amplify a partial GP46 cDNA using degenerate and
nondegenerate primers designed from the L. amazonensis
sequence yielded a 1.3-kb product with L. mexicana genomic
DNA template but no product with L. chagasi genomic DNA.
This 1.3-kb fragment was shown by DNA sequencing to encode GP46 and was
used to screen 2.7 × 104 phage in an L. chagasi cDNA library. Fourteen phage clones hybridized to the
probe (0.05%), and 10 were plaque-purified for further characterization. The complete 2.4-kb sequence of the largest of these
10 cDNA inserts was determined and found to be a partial-length cDNA lacking the 39-nucleotide-spliced leader found on the 5 ends
of all Leishmania mRNAs (32).
To isolate the missing 5 segment, a reverse transcriptase-PCR was
conducted on total L. chagasi promastigote RNA using a forward primer containing part of the spliced leader sequence and a
reverse primer based on the cDNA sequence. The resulting 500-bp PCR
product showed complete sequence identity in a 194-bp overlap with the
cDNA. The combined GP46 cDNA sequence of 2754 bp (Fig.
1A) is consistent with the 2.8-kb size of the
major GP46 mRNA species observed on a Northern blot (Fig.
2A). The deduced nascent protein sequence of
44 kDa (Fig. 1B) is similar to GP46 proteins of other
Leishmania species.
The Steady State Level of GP46 mRNA Is Developmentally Regulated by Post-transcriptional Events
Total RNA was isolated from promastigotes at various times during growth in culture and probed in Northern blots with the GP46 cDNA (Fig. 2). Promastigotes in these cultures entered logarithmic phase of growth within 1-2 days after passage and reached stationary phase by day 7, according to morphology and concentration as previously defined (24). Panel A of Fig. 2 shows that two GP46 RNA species that hybridize to the cDNA probe exist in stationary phase L. chagasi promastigotes, a major species of 2.8 kb and a minor one of 4.8 kb. Both species steadily increase in abundance as the cells grow from logarithmic to stationary phase. InstantImagerTM analysis of the blot indicated that the steady state level of the 2.8-kb GP46 RNA is >30-fold higher at day 7 than at days 3-4. The 4.8-kb RNA increases a similar amount. On the basis of size similarity, the cDNA shown in Fig. 1A is likely derived from the 2.8-kb GP46 RNA.
Panel B shows the same RNAs probed with the GP63 coding region. The hybridization pattern is very similar to the pattern reported previously (15). A GP63 RNA of 2.7 kb occurs in logarithmic phase promastigotes (days 3-4), and a GP63 RNA of 3.0 kb occurs in stationary phase promastigotes (days 6-7). Promastigotes at an intermediate phase of growth (day 5) have both GP63 RNA species. A comparison of panels A and B demonstrates that the abundance of GP46 RNA closely parallels that of the stationary 3.0-kb GP63 RNA. Both RNAs are very rare in days 3-4, begin to appear in day 5, and increase in abundance during days 6-7.
Panels C and D show a hybridization with a
tubulin probe and an ethidium bromide stain, respectively, to detect
variations in RNA loadings to the gel lanes. Both panels indicate that
in this particular experiment more RNA was added to lane 5 (day 5) and less to lane 7 (day 7) than to the other lanes.
When these differences are taken into account, the increases in GP46
and stationary GP63 RNA in stationary phase promastigotes (day 7) is
even more dramatic than those indicated by the relative band intensities in panels A and B. As shown in
panel C, the control hybridization with the T. brucei - and
-tubulin probe resulted in the typical one
-
and three
-tubulin band pattern seen in prior studies (28). To
eliminate this pattern of multiple bands, subsequent blots utilized a
probe from an
-tubulin gene of L. chagasi (see Fig.
6B).
The similar sizes of the GP46 and GP63 RNAs (2.8 kb and 3.0 kb) precluded a quantification of their relative abundance in a single Northern blot analysis. Thus, Northern blots hybridized separately with each probe were incubated in the hybridization solution with a second filter containing serial dilutions in slot blots of DNA fragments possessing the GP46 and GP63 coding regions. The relative amount of GP46 and GP63 RNA was estimated by comparing the intensities of their signals on the Northern blot filter to the signals of the serially diluted DNAs on the slot blot filter. Using this approach, the abundance of GP46 2.8-kb RNA was found to be about the same as that of GP63 RNA in stationary phase day 7 cells (data not shown).
To determine whether the increase in the steady state level of GP46
mRNA that occurs as promastigotes grow from logarithmic to
stationary phase is due to enhanced transcription initiation or to
post-transcriptional events, nuclear run-on experiments were conducted
using nuclei isolated from logarithmic and stationary phase
promastigotes (Fig. 3). The radioactive RNA isolated
from logarithmic nuclei consistently had a higher specific activity than that obtained from stationary nuclei, probably indicating that
logarithmic phase cells are transcriptionally more active. Thus, the
amount of radioactive RNA hybridizing to each of the test DNAs in Fig.
3 was normalized to the amount hybridizing to DNA encoding -tubulin.
The normalized ratio of GP46 run-on RNA in stationary versus
logarithmic nuclei was found to be about 0.51, approximating the
corresponding ratio of GP63 run-on RNA. This result indicates that
there is no difference in the transcription rate of the GP46 genes in
logarithmic and stationary phase promastigotes relative to
-tubulin
gene transcription despite the 30-fold increase in the steady state
level of GP46 mRNA in stationary phase cells. Thus, this increase
in GP46 mRNA must be regulated primarily by post-transcriptional
events.
The 3
To examine which sequences of the GP46
genes might contribute to the increased level of their mRNAs in
stationary promastigotes, we first screened a bacteriophage library
of L. chagasi genomic DNA for genomic DNA clones that
contain GP46 genes. Two such clones were isolated and characterized by
restriction mapping, and several of their restriction fragments were
subcloned for DNA sequencing. The genomic DNA segments in these two
clones were found to overlap in a region of the genome containing two
GP46 genes (gp46A and gp46B) that are separated
by 7.2 kb (Fig. 4). The end of one of the genomic clones
occurs within codon 163 of gp46A, so the sequence preceding
that codon could not be determined. However, the sequence of the rest
of the gp46A coding region and its entire 3
-UTR of 1.3 kb
was determined and found to be identical to the GP46 cDNA sequence
(Fig. 1A), indicating that the corresponding mRNA could have arisen from this gene. The complete sequence of gp46B
was determined (not shown), and its coding region was found to have only 78% nucleotide identity with the corresponding region of the
cDNA sequence (Fig. 4), largely due to short stretches of DNA
sequence in gp46B that were not present in gp46A.
Most of its 3
-UTR has 93% identity with the cDNA, but it also has
an internal 1.3-kb segment with no similarity (shaded
rectangle in Fig. 4). In Northern analysis (Fig. 2E),
this 1.3-kb segment of gp46B predominately hybridized to
4.8-kb RNAs that increased in abundance during promastigote development
in a pattern equivalent to that seen for the 2.8-kb RNAs of Fig.
2A. Thus, gp46B may be a source of the 4.8-kb
GP46 RNAs.
Fragments containing all of the gp46A 3-UTR and about 65%
of the gp46B 3
-UTR plus their downstream IRs were cloned
immediately downstream of the
GAL coding region in pX-
gal 2 (Fig.
4). This parent plasmid is maintained episomally in
Leishmania and contains a neomycin resistance gene that
allows for the selection of cells that contain the plasmid by growth in
the presence of the drug G418. In earlier studies, we had already
constructed derivatives of pX-
gal 2 in which the corresponding
3
-UTRs and IR regions of the three GP63 gene classes, i.e.
mspS (stationary), mspL (logarithmic), and
mspC (constitutive), were cloned at the same site downstream of the
GAL gene (15). These five plasmids containing the various Leishmania 3
-UTRs + IRs and the parent plasmid pX-
gal 2 were introduced into promastigotes, and clones of stable transfectants were selected (listed in Table I).
|
Fig. 5 shows the GAL enzymatic activities that were
determined during growth of the six transfectants from logarithmic to stationary phase and corresponds to experiment 1 of Table I. No change
in
GAL activity with time occurred in the transfectants containing
the parent plasmid and the plasmids with the 3
-UTRs + IRs of
mspL and mspC, as demonstrated previously (15).
Likewise, the presence of the partial 3
-UTR + IR of gp46B
did not change the
GAL activity from that of the parent plasmid,
possibly due to the lack of a full-length 3
-UTR. However, the 3
-UTR + IR of gp46A stimulated a dramatic increase in
GAL
activity with time in culture, similar to that of the 3
-UTR + IR of
mspS. In both cases, the
GAL activity in logarithmic
phase promastigotes was lower on day 3 than in the other transfectants,
increased to the level observed in the other transfectants by day 4, and subsequently increased steadily until by day 8 it was about
20-30-fold higher than on day 3. In two other experiments similar to
that shown in Fig. 5, the ratio of
GAL activity on day 8 versus day 3 was
12 in transfectants containing the
3
-UTRs + IRs of gp46A and mspS, whereas no
substantive difference was observed in transfectants bearing the other
plasmids (Table I). Thus, the gp46A and mspS 3
-UTRs + IRs exert parallel effects on
GAL activity when placed downstream of the
GAL gene.
The
To ensure that the increase in GAL
enzymatic activity observed in the transfectants bearing plasmids with
the 3
-UTRs + IRs of gp46A and mspS was due to
increased
GAL mRNA, Northern blots were conducted on RNAs
extracted from the transfectants. Fig. 6A
shows that in transfectants containing the 3
-UTRs + IRs of gp46A and mspS, the steady state level of
GAL
mRNA is much higher in stationary phase (S, day 8) than
in logarithmic phase (L, day 3). In addition, the
GAL
mRNA containing the 3
-UTR of gp46A (1.3 kb) is slightly
larger than the
GAL mRNA with the 3
-UTR of mspS (1.1 kb), as expected from the known sizes of these 3
-UTRs. In contrast, in
the transformant bearing only the parent plasmid (called
-gal in Fig. 6), the level of
GAL RNA does not vary as
much with respect to the phase of growth. In addition, the major
GAL
RNA species in this transformant is about 7.0 kb, suggesting that its
primary polyadenylation site occurs about 4 kb downstream of the 3-kb
GAL coding region. The membrane shown in Fig. 6A was
stripped and reprobed with the
-tubulin gene to determine the
relative amounts of RNA added to each lane (Fig. 6B). The relative increase in
GAL mRNA in stationary versus
logarithmic phase transfectants was 13-fold for the clone containing
gp46A sequence, 12-fold for the clone containing
mspS sequence, and <2-fold for the parent plasmid. These
increases in steady state mRNA levels are consistent with the
increases observed in the
GAL enzymatic activity.
To verify that the plasmid copy number does not change during growth
from logarithmic to stationary phase, DNA was extracted from the
transfectants at days 3 and 8 and probed with the GAL coding
sequence (Fig. 7A). The same blot was
stripped and reprobed with GP46 coding sequence to adjust for DNA
loading (Fig. 7B). The relative amount of the plasmid in
each transformant was found to vary by <10% during growth of the
transfectants.
In these Southern blots, the GAL probe hybridizes only to a single
3.6-kb AccI fragment of the plasmid, since the probe is contained within this fragment. Conversely, because the plasmid constructs do not contain GP46 coding sequence, the GP46 coding region
probe hybridizes only to GP46 genes in the L. chagasi
genome. Thus, in panel B, all of the fragments are derived
from the genome. The 2.0-kb fragment and one of the
7.5-kb doublet
fragments correspond to expected fragments based on the locations of
AccI sites in the genomic and cDNA sequences shown in
Fig. 3. The additional 1.6, 2.7, and
7.5-kb fragments to which the
GP46 probe hybridizes with varying intensities indicate the existence
of other genes besides gp46A and gp46B. These
results are consistent with similar studies of GP46 gene organization
in other Leishmania species that showed that GP46 is encoded
by a family of nonidentical genes that are organized as a cluster(s)
within the genome (18). The weak signals of fragments whose sizes are
not indicated in Fig. 7B may suggest either the presence of
additional sequences in the genome that have partial identity to the
GP46 coding region probe or incomplete digestion of genomic DNA.
Previous work with in vitro cultured L. chagasi has documented differential expression of RNAs from distinct msp genes encoding the surface glycoprotein GP63 (9). The current study was based on the hypothesis that the genes coding for other surface proteins might be similarly regulated. We tested this hypothesis with another Leishmania gene family encoding GP46, which, like GP63, is a glycoprotein that is expressed on the parasite surface membrane.
The deduced amino acid sequence of L. chagasi GP46 shown in Fig. 1B displays 61% identity to GP46 of L. major (30) and 65% identity to GP46 of L. amazonensis (31). All of these GP46 sequences contain hydrophobic amino and carboxyl termini that are probably post-translationally cleaved during the translocation of the protein across the endoplasmic reticulum and its linkage to membrane-anchored glycosylphosphatidylinositol, respectively. Additionally, all contain 3 (L. major) to 7 (L. chagasi) leucine-rich repeats of 24 residues. Leucine-rich repeats of 24 residues are components of many other proteins and are thought to be sites of specific protein-to-protein interactions (33, 34), suggesting a similar role for these repeats in GP46. Protozoan parasites of the genus Giardia also have a surface glycoprotein that contains leucine-rich repeats of 24 residues (35), raising the possibility that the surface glycoproteins of these two distantly related genera may have analogous function.
Based upon our work on GP46 gene transcription and mRNA stability, we would predict that GP46 levels vary during promastigote development, as has been shown for GP63. Recent analysis in L. major showed that different GP46 RNAs are expressed in amastigotes and promastigotes, and that the glycosylphosphatidylinositol-anchored GP46 in amastigotes, but not promastigotes, is resistant to hydrolysis by phosphatidylinositol-specific phospholipase C (22). The work presented here establishes the need for a similar study of GP46 expression throughout promastigote development.
Northern blot and nuclear run-on experiments showed that the varied
abundance of GP46 mRNA in promastigotes is due to
post-transcriptional events (Fig. 2 and 3). The steady state level of a
number of RNAs in the order Kinetoplastidae are also regulated
post-transcriptionally (36-38). Such regulation may be linked to the
observation that in this order, many highly expressed proteins are
encoded by tandemly repeated genes present in a linked cluster. Some,
if not all, of these gene clusters are transcribed as polycistronic
precursor RNAs that subsequently undergo processing by the addition of
a 39-nucleotide-spliced leader at the 5 end and polyadenylation at the
3
end of each mRNA. 5
processing of a downstream gene is coupled
to polyadenylation of the upstream gene product (39). Thus,
post-transcriptional regulation would provide the means for controlling
levels of specific mRNAs derived from a polycistronic transcript.
The parallel expression profiles of GP46 and mspS RNA in
promastigotes as well as the equivalent effects of the 3-UTRs + IRs of
gp46A and mspS on
GAL activity and mRNA
levels suggest that both genes are post-transcriptionally regulated by
a similar mechanism. From this observation, one would hypothesize that
homologous regions within the 3
-UTRs of these two genes may be
responsible for the increased abundance of these mRNAs during
stationary phase growth. Comparison of the 3
-UTR + IR sequences of
gp46A and mspS with BESTFIT indicates several
regions of similarity within their 3
-UTRs, the highest scoring of
which is depicted in Fig. 8. In contrast, there is
little identity within their IRs. A test of whether these homologous
DNA segments in the 3
-UTRs are responsible for the stationary-specific
expression pattern will require further transfection experiments with
these putative regulatory regions either deleted or inserted downstream
of a reporter gene. The 3
-UTR of gp46B also contains this
region of similarity, leading us to hypothesize that the reason for the
lack of growth-regulated expression of the Bgal construct
containing the partial length gp46B 3
-UTR + IR is that all
sequences essential for its growth-regulated expression (Fig.
2E) were not contained in this construct.
The actual molecular mechanism by which the GP46 and mspS
genes are regulated remains unclear. Our prior work documented
differences in the mechanisms regulating expression of the
mspL and mspS genes within the same cluster (15,
28). In experiments comparable to those reported here, the 3-UTR + IR
of mspS, but not of mspL, was responsible for
augmenting both the steady state RNA and protein expression levels of
an upstream reporter gene during growth of L. chagasi from
logarithmic to stationary phase (15). In other experiments, the levels
and half-lives of mspL RNAs, but not of mspS
RNAs, were dramatically up-regulated by the addition of protein synthesis inhibitors that did not influence the rate of transcription, implying the involvement of a negative regulatory protein factor in
targeting mspL RNAs for rapid degradation (28). Similar
studies with the GP46 genes hold the promise of revealing further
intricacies about gene regulation in this class of protozoan
parasites.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBankTM/EMBL Data Bank with accession number(s) AF006588.