W. Robert McMaster. E-mail firstname.lastname@example.org; Tel. (604) 875 4134; Fax (604) 875 4497.
The major surface glycoprotein of Leishmania major is a zinc metalloproteinase of 63 kDa referred to as leishmanolysin or GP63, which is encoded by a family of seven genes. Targeted gene replacement was used to delete gp63 genes 1–6 encoding the highly expressed promastigote and constitutively expressed GP63. In the L. major homozygous mutants deficient in gp63 genes 1–6, there was no expression of GP63 as detected by reverse transcription–polymerase chain reaction (RT–PCR) or fluorescent staining in promastigotes from the procyclic stage (logarithmic growth phase). The remaining L. major gP63 gene 7 was shown to be developmentally regulated, as it was expressed exclusively in infectious metacyclic stage (late stationary growth phase) promastigotes and in lesion amastigotes. The gp63 genes 1–6-deficient mutants showed increased sensitivity to complement-mediated lysis. The sensitivity to lysis was greater in procyclics than in metacyclics when compared with the equivalent wild-type stages. Increased resistance of the mutant metacyclic promastigotes correlated with the expression of gp63 gene 7 and was restored to the same levels as wild-type promastigotes by transfection with gp63 gene 1. Thus, expression of GP63 is clearly involved in conferring resistance to complement-mediated lysis. The L. major GP63 1–6 mutants were capable of infecting mouse macrophages and differentiating into amastigotes. Similar levels of infection and subsequent intracellular survival were observed when mouse macrophages were infected in vitro with wild type, GP63 1–6 mutants and mutants transfected with gp63 gene 1. The GP63 1–6 mutants were capable of lesion formation in BALB/c mice and, thus, gp63 genes 1–6 do not play a role in the survival of the parasite within mouse macrophages. The role of gp63 genes 1–6 in parasite development within the sandfly vector was studied. GP63 1–6 mutants grew normally in the blood-engorged midgut of both Phlebotomus argentipes and P. papatasi. However, both wild-type and mutant promastigotes were lost after 2 days' growth in P. papatasi. The complete developmental pathway in P. argentipes was observed for wild-type promastigotes, GP63 1–6 mutants and mutants transfected with gp63 gene 1. Normal stage differentiation from amastigotes to procyclics, to nectomonads, to haptomonads and to infectious metacyclics was observed. Thus, the highly expressed promastigote forms of GP63, encoded by gp63 genes 1–6, do not appear to be required for nutrient utilization in the bloodmeal during the early stages of development in the sandfly or for midgut attachment and further development. gp63 1–6 genes do, however, play a major protective role against complement-mediated lysis when promastigotes are introduced into the mammalian host.
Leishmania are pathogenic trypanosomatids that are the aetiological agents for leishmaniasis. Leishmania have a dimorphic life cycle consisting of extracellular promastigotes that multiply and develop within the midgut of the sandfly vector and intracellular amastigotes that reside and multiply within the phagolysosomal vacuoles of the host macrophages. Leishmania express two major surface molecules, a surface zinc metalloproteinase (Etges et al., 1986; Button and McMaster, 1988; Chaudhuri and Chang, 1988; Bouvier et al., 1989), referred to as leishmanolysin or GP63, and a lipophosphoglycan, referred to as LPG (King et al., 1987).
The L. major gp63 genes have previously been mapped to a single chromosome-sized band and are tandemly arranged as direct repeats (Button et al., 1989). In L. major, the gp63 genes have been shown to be expressed differentially; gp63 genes 1–5 were expressed at high levels in promastigotes only, whereas gp63 gene 6 was constitutively expressed at lower levels in both promastigotes and amastigotes (Voth et al., 1998).
Despite extensive information on the structure and expression of GP63 genes from several species, the exact role(s) of the GP63 metalloproteinase in the life cycle of Leishmania is not yet understood. Targeted gene replacement is a powerful approach for determining the role of specific genes in trypanosomatids (Cruz et al., 1991; Cooper et al., 1993, Souza et al., 1994; Webb and McMaster, 1994; Wiese et al., 1995). Using this approach, L. major strains have been derived that lack the highly expressed promastigote and the constitutively expressed GP63 by deletion of L. major gp63 genes 1–6. The resulting L. major gene 1–6 mutants were analysed for their sensitivity to complement-mediated lysis, interaction with macrophages, infectivity in mice and survival and development inside the sandfly vector.
Deletion of L. major gp63 genes 1–6 by targeted gene replacement
The L. major GP63 genomic region was originally reported to contain six gp63 genes consisting of five homologous tandemly linked gp63 genes (gp63 genes 1–5; each consisting of 1.3 kb coding +1.8 intergenic region) and gp63 gene 6, located 8 kb 3′ to gp63 gene 5 (Button et al., 1989). Further Southern blot analyses of genomic DNA isolated from L. major A2 strain identified an additional gene, gp63 gene 7 (Voth et al., 1998), located 3′ to gp63 gene 6, as shown in Fig. 1A. L. major gp63 genes 1–5 have been shown to be expressed at high levels only in promastigotes, whereas gp63 gene 6 was constitutively expressed at lower levels in both promastigotes and amastigotes (Voth et al., 1998).
The following strategy was used to derive L. major mutants deficient in gp63 genes 1–6 to study their role in the Leishmania life cycle and to study the expression of the L. major gp63 gene 7. Transfecting DNA fragments containing either the sat or hyg gene positioned between the 5′ flanking region of gp63 gene 1 and the 3′ flanking region of gp63 gene 6 (Fig. 1A) were used to replace the entire 25 kb L. major gp63 1–6 region on each chromosome by homologous recombination. After the first round of transfection with the XbaI/ApaI fragment from the plasmid pLGT39, genomic DNA isolated from one nour-resistant clone, A2-4, was analysed by Southern blot hybridization (Fig. 1B). In wild-type L. major A2, the entire gp63 gene region is contained within a genomic 32 kb XhoI fragment (Button et al., 1989). Transfectants, in which homologous replacement had occurred, would have a smaller genomic XhoI fragment containing an extra SpeI site. Genomic DNA digested with XhoI and SpeI was hybridized with a 0.5 kb SalI/SacI fragment (Fig. 1A, probe i). With wild-type L. major DNA, this fragment hybridized to a 32 kb XhoI fragment and, with transfectant clone A2-4 DNA, this probe hybridized to two fragments: a 32 kb fragment and a 7.8 kb fragment (Fig. 1B, panel a). When the blot was hybridized with a probe containing the sat gene, a 7.8 kb SpeI/XhoI fragment was detected only in DNA from clone A2-4 but not from the wild-type strain A2 (Fig. 1B, panel d), indicating replacement of gp63 genes 1–6 by the sat gene.
Flow cytometric analysis showed that the heterozygous clone A2-4 expressed reduced levels of GP63 (Fig. 2B) compared with wild-type promastigotes (Fig. 2A). A2-4 promastigotes were subsequently transfected with the XbaI/ApaI fragment from pLGT40 (Fig. 1A). Twelve days after electroporation, flow cytometric analysis of the double drug-resistant transfectants showed a distribution of 60% negative and 40% positive staining with anti-GP63 monoclonal antibody 96 (Fig. 2C). After day 7, the fraction of promastigotes in the transfected culture that did not bind to mAb 96 diminished to less than 10% with a concomitant increase in promastigotes that bound mAb 96 (data not shown). The negative staining population (Fig. 2C) was sorted by fluorescence-activated cell sorting (FACS) to select for promastigotes that did not express GP63 (Fig. 2D). These cells were immediately cloned on semisolid JB199 medium containing 60 μg ml−1 and 70 μg ml−1 nour and hygro respectively. Eight colonies were isolated that were expanded in liquid culture and continued to be deficient in the expression of GP63. One clone, A2-4/40-5 was used for further analysis (Fig. 2E).
Southern blot analysis of genomic DNA from A2-4/40-5 digested with XhoI and SpeI and hybridized with probe i (Fig. 1) detected fragments at 8.3 and 7.8 kb (Fig. 1B, panel a). No hybridization to a 32 kb fragment was detected, indicating the absence of a wild-type gp63 gene region (Fig. 1B, band a). Hybridization of the blot with the hyg coding region (Fig. 1, probe v) detected a 8.3 kb fragment present only in the lane containing DNA from clone A2-4/40-5 (Fig. 1B, panel e).
Hybridization of the blot with a fragment specific for gp63 genes 1–5 and 7 (Fig. 1A, probe ii) identified the XhoI 32 kb fragment in wild-type L. major and clone A2-4 (Fig. 1B, panel b). Hybridization with this probe also detected a 7.8 kb hybridizing fragment in clone A2-4 and a 7.8 kb and a 8.3 kb fragment in clone A2-4/40-5 (Fig. 1B, panel b). Hybridization of the blot with a fragment unique to gp63 gene 6 (Fig. 1, probe iii) identified a 32 kb fragment in wild-type L. major and clone A2-4 (Fig. 1B, panel c). Probe iii did not hybridize to the 7.8 and 8.3 kb fragments in A2-4/40-5, demonstrating that the deletion of gp63 gene 6 from both chromosomes had occurred (Fig. 1B, panel c). Southern blot analysis of the other seven clones obtained from the sorted fraction of GP63-deficient promastigotes showed that they were similar to clone A2-4/40-5 (data not shown). In order to demonstrate that gp63 genes 1–5 were also deleted, Southern blot analysis was carried out using genomic DNA digested with XhoI and AflII and hybridized with probe ii (see Fig. 1A). For wild-type A2 and A2-4 DNA, this probe hybridized to two fragments (Fig. 1B, panel f): a 25 kb fragment containing gp63 genes 1–6 and a 6 kb fragment containing gp63 gene 7. For DNA from A2-4/40-5, probe ii hybridized to a 6 kb fragment containing gp63 gene 7 (Fig. 1B, panel f). Together, these results demonstrate that, in clone A2-4/40-5, gp63 genes 1–6 had been deleted on both chromosomes.
GP63 1–6-deficient promastigotes express GP63 in late stationary phase cultures but not in log phase cultures
Expression of gp63 gene 7 in clone A2-4/40-5 was followed during in vitro development of promastigotes. Log phase A2-4/40-5 promastigotes did not react with anti-GP63 mAb 96 when analysed by FACS (Fig. 2E). Furthermore, reverse transcription–polymerase chain reaction (RT–PCR) using GP63-specific primers (that detect gp63 gene 7) did not detect the presence of GP63 RNA in log phase promastigotes (data not shown). Expression of GP63 in A2-4/40-5 promastigotes was also followed by Western blot analysis (Fig. 3). GP63 was not detected in early, late log and early stationary phase cultures of A2-4/40-5 promastigotes (Fig. 3, lanes 4 and 5), whereas wild-type A2 promastigotes expressed high levels of GP63 at all stages (Fig. 3, lanes 1–3). A band corresponding to GP63 was clearly detected in late stationary phase (9–12 days) A2-4/40-5 promastigotes (Fig. 3, lane 6). Thus, in A2-4/40-5 promastigotes, gp63 gene 7 was developmentally regulated and was expressed only in late stationary phase promastigotes. In wild-type promastigotes, the GP63 encoded by gp63 gene 7 could not be assayed for independently, as the available antibodies also recognized the products of gp63 genes 1 and 6.
GP63 expression is regained in log phase A2-4/40-5 promastigotes by transfection with a gp63 gene 1 expression plasmid
Clone A2-4/40-5 was transfected with the GP63 expression plasmid pLEXNEO-gp63 gene 1, and individual clones were isolated. Promastigotes of a representative G418-resistant clone were analysed by flow cytometry for the expression of surface GP63. Results show that the G418-resistant clones expressed levels of surface GP63 similar to those of the wild-type L. major clone A2 (Fig. 2, compare profiles A and F). The reintroduction of L. major gp63 gene 1 into A2-4/40-5 also restored proteinase activity in log phase cultures (data not shown).
Metacyclic A2-4/40-5 promastigotes are unaffected in their initial attachment, entry and intracellular survival within macrophages
Wild-type L. major, A2-4/40-5 (deficient in gp63 genes 1–6) and A2-4/40-5 transfected with gp63 gene 1 metacyclic promastigotes, obtained from late stationary phase cultures by concanavalin A (con A)-negative selection, were incubated with mouse peritoneal macrophages in a 1:1 ratio. All three strains were equally infective, as there were no differences in the percentage of infected macrophages and the number of parasites per infected cell after 1 h and 96 h (Fig. 4). These results suggest that GP63 encoded by gp63 genes 1–6 is not required for the attachment and uptake of metacyclic promastigotes. Furthermore, no differences were observed in the subsequent intracellular survival and growth of amastigotes after 96 h (Fig. 4B).
A2-4/40-5 promastigotes are more sensitive to complement-mediated lysis
Susceptibility to complement-mediated lysis was determined for log phase and con A-negative selected metacyclic promastigotes by incubation in increasing concentrations of fresh human serum. Figure 5A shows that A2-4/40-5 promastigotes from a log phase culture began to lyse at a serum concentration of 0.4%, and complete lysis was observed after incubation in 1% serum. Lysis of wild-type L. major A2 promastigotes occurred in 2% serum, and complete lysis occurred in 10% serum. Resistance to complement-mediated lysis in A2-4/40-5 promastigotes was restored to wild-type levels by transfection with gp63 gene 1 (Fig. 5A). When metacyclics were analysed, there was a shift towards greater complement resistance for each of the clones examined. The A2-4/40-5 metacyclics, nonetheless, remained significantly more sensitive to lysis than metacyclics of either the wild-type A2 or the mutant A2-4/40-5 transfected with gp63 gene 1 (Fig. 5B). The difference in sensitivity to lysis between the mutant A2-4/40-5 and the wild-type A2 strain was greater when log phase rather than metacyclic promastigotes were compared. Taken together, these results clearly demonstrate that the product encoded by L. major gp63 gene 1 plays an important role in conferring complement resistance in both log phase and metacyclic promastigotes of L. major.
A2-4/40-5 mutants cause a slightly delayed onset of disease in mice and have little effect on the progression of disease
The ability of A2-4/40-5 to cause lesions in BALB/c mice was investigated by infecting the left hind footpad with conA-negative selected metacyclic promastigotes. Figure 6 shows the increase in mean footpad thickness at various time points after infection. In mice infected with the wild-type L. major, lesions developed between 4 and 6 weeks after infection. With A2-4/40-5, lesions developed between 6 and 8 weeks after infection. After the disease was established (≈8 weeks after infection), the rate of lesion progression for both strains did not differ significantly (0.34 mm week−1 for the wild type and 0.28 mm week−1 for A2-4/40-5). When A2-4/40-5 transfected with gp63 gene 1 was used to infect mice, lesions appeared at between 4 and 6 weeks with a similar rate of development after 8 weeks (0.38 mm week−1). These results suggest that the deletion of gp63 genes 1–6 in the A2-4/40-5 strain results in a delay of 2–3 weeks in their ability to produce progressive disease when compared with the wild-type A2 strain. This delay was not observed in the A2-4/40-5 strain transfected with gp63 gene 1.
Amastigotes were purified from mouse lesions produced by infection with either the wild-type L. major or the A2-4/40-5 mutant strain. Western blot analysis of amastigote lysates clearly showed that amastigotes from both strains expressed GP63 (Fig. 3, lanes 7 and 8 respectively). Thus, in the A2-4/40-5 clone, gp63 gene 7 was expressed in both stationary phase promastigotes and lesion amastigotes.
Growth and development of the mutant A2-4/40-5 clone within sandfly vectors is unaffected by the loss of gp63 genes 1–6 gene products
The ability of A2-4/40-5 promastigotes to survive and undergo development within P. papatasi, the natural sandfly vector for L. major, was investigated. Sandflies that had been membrane fed with log phase promastigotes were dissected, and the number and morphology of the promastigotes within the midgut was determined. 7Figure 7A shows that there was no difference in growth between wild-type and A2-4/40-5 promastigotes on day 2 and day 3 after infection. The early differentiation during growth from dividing procyclic forms at 48 h to elongated nectomonads at 72 h was unchanged in the mutant strain when compared with the wild-type strain (not shown). However, parasites of both strains were lost between days 4 and 6. This loss was indicative of excretion of the parasites along with the digested bloodmeal (Pimenta et al., 1994) and was presumed to occur with the wild-type L. major A2 strain, because its LPG lacks the side-chain oligosaccharides necessary to mediate attachment to the P. papatasi midgut. P. argentipes has been shown experimentally to be a permissive vector for numerous Leishmania species, including those expressing relatively unsubstituted LPGs (e.g. L. donovani ). P argentipes sandflies, therefore, were membrane fed on heparinized mouse blood containing tissue amastigotes purified from BALB/c mouse footpad lesions. 7Figure 7B shows the change in the number of promastigotes of the wild-type A2 and the mutant A2-4/40-5 parasites within the midgut of P. argentipes. No significant differences were observed between the two strains during the transformation to promastigotes and subsequent growth during the first 2 days within the digesting bloodmeal. At day 5 and day 8, well after the time of bloodmeal excretion, no significant differences were observed in the number of promastigotes (Fig. 7B). By 8 days, the infections had moved anterior into the thoracic midgut for each of the clones. The differentiation of A2-4/40-5 promastigotes, as determined by the changes in morphology, was not significantly different from that of the wild-type L. major (Table 1). In each case, complete development from procyclic to nectomonad forms occurred between day 2 and day 5, and the appearance of metacyclic promastigotes commenced at day 5. By day 8, approximately 10% of the midgut promastigotes of the wild type, A2-4/40-5 and A2-4/40-5 transfected with gp63 gene 1 had transformed into metacyclics. These data indicate that the A2-4/40-5 mutant survives, replicates and differentiates normally within a permissive sandfly vector.
Table 1. . Stage differentiation of L. major promastigotes in P. argentipes. The percentage of each morphological form was determined by scoring Giemsa-stained parasites from preparations of individual flies at different time intervals after infective feed. Numbers are expressed as mean percentage + SD of 5–7 flies per group. Wild type L. major (wild type); A2-4/40-5 gp63 genes 1–6-deficient mutants (null); A2-4/40-5 transfected with gp63 gene 1 (null + GP63).
Targeted gene deletion of the highly expressed promastigote and constitutively expressed L. major gp63 genes 1–6 was accomplished by replacement of the entire 24 kb region of each chromosome with the drug-selectable genes conferring resistance to nour and hygro. In A2-4/40-5, deficient in gp63 genes 1–6, there was no expression of GP63 as detected by RT–PCR or fluorescent staining in promastigotes from the procyclic stage (logarithmic growth phase). The remaining L. major gp63 gene 7 was shown to be developmentally regulated, as it was expressed exclusively in the infectious metacyclic stage (late stationary growth phase) promastigotes and lesion amastigotes. In wild-type L. major, gp63 genes 1–5 were highly expressed in all stages of promastigote development. Developmental expression of certain gp63 genes in promastigotes has also been reported for L. mexicana (Medina-Acosta et al., 1993) and L. chagasi (Ratmmamoorthy et al., 1992). The L. major gp63 gene 7 pattern of expression correlated to that of L. chagasi mspS genes, while the expression of L. major gp63 1–5 genes correlated to that of L. chagasi mspC and L. mexicana C2 genes.
It has been reported that resistance to complement-mediated lysis increases as L. major promastigotes develop from the procyclic to the metacyclic stage (Franke et al., 1985; Puentes et al., 1988). Although at each stage A2-4/40-5 promastigotes showed an increased sensitivity to complement-mediated lysis, the sensitivity to lysis was greater in procyclics than in metacyclics when compared with the equivalent wild-type stages. The increased resistance of A2-4/40-5 metacyclic promastigotes correlates with the expression of gp63 gene 7. Resistance to complement-mediated lysis of both procyclic and metacyclic promastigotes was restored to the same levels as wild-type A2 strain in A2-4/40-5 transfected with gp63 gene 1. These results clearly demonstrate an important role for GP63 in conferring resistance to complement-mediated lysis. Previous studies have also implicated GP63 in complement resistance using both transfected CHO cells and L. amazonensis promastigotes expressing the cloned L. major gp63 gene 1 (Brittingham et al., 1995). These studies have shown that the L. amazonensis promastigotes, which were deficient in high-level expression of GP63, were more sensitive to complement lysis than wild-type L. amazonensis and that complement resistance was restored by transfection and expression of the cloned L. major gp63 gene 1 but not a proteolytically inactive active site mutant of GP63. The complement-resistant transfectants bound higher levels of the iC3b neoantigen, suggesting a role for proteolytically active GP63 in the inactivation of C3b to a form resembling iC3b. This inactivation prevented the formation of C5 convertase and the subsequent formation of the membrane attack complex 1 (Brittingham et al., 1995).
Despite lacking gp63 genes 1–6, A2-4/40 mutants were capable of infecting mouse macrophages and differentiating into amastigotes. Similar levels of infection and subsequent intracellular survival were observed when mouse macrophages were infected in vitro with wild type, A2-4/40-5 and A2-4/40-5 transfected with gp63 gene 1 (Fig. 6). A2-4/40-5 was also capable of lesion formation in BALB/c mice as measured by footpad swelling. These mutants showed a 2–3 week delay in initial lesion formation, after which a rate of disease progression similar to wild-type L. major was observed. The initial delay in lesion development could be attributed to either a decrease in infectivity or a reduction in viability of the initial inoculum. As the progression of disease after the initial delay in lesion appearance was unaltered, this further supports the conclusion from the in vitro infection studies that the gp63 genes 1–6 products do not play a role in the survival of the parasite within mouse macrophages. Therefore, the delay in the onset of disease observed in mice infected with A2-4/40-5 was most probably a consequence of reduced viability of the initial parasite inoculum owing to increased sensitivity to complement-mediated lysis. No delay in the establishment of disease was observed in A2-4/40-5 transfected with gp63 gene 1, strongly suggesting a role for GP63 in maintaining inoculum viability.
The presence of GP63-like proteinase and GP63 gene homologues in monoxenous trypanosomatids has led to speculation that GP63 may play a role in the parasite survival within the insect gut (Etges, 1992; Inverso et al., 1993). A2-4/40-5 promastigotes and amastigotes, however, were capable of development within the phlebotomine vectors. Similar development within P. papatasi was observed for L. major A2 wild-type and A2-4/40-5 promastigotes. However, promastigotes of both types were lost between days 4 and 6 and were probably excreted along with the bloodmeal. The inability of the parasites to remain anchored to the midgut epithelium of P. papatasi could be caused by the lack of typical galactose-substituted side-chains on the LPG of the wild-type L. major strain A2 (McConville et al., 1990). The role of gp63 genes 1–6 in parasite development in the later stages of midgut infection was therefore investigated in another sandfly species, P. argentipes, previously shown to be broadly permissive to Leishmania strains expressing unsubstituted phosphoglycan chains. The A2-4/40-5 promastigotes were retained in the gut as efficiently as the wild type, suggesting that GP63 encoded by gp63 genes 1–6 have little or no role in midgut attachment. Furthermore, the stage differentiation of these parasites in the gut, from amastigotes to procyclics, to nectomonads, to haptomonads and to infectious metacyclics, was not altered by the absence of gp63 genes 1–6. The complete developmental pathway from lesion amastigotes to metacyclic promastigotes was observed for both strains. It is interesting to note that A2-4/40-5 procyclics grew normally in the blood-engorged midgut of both P. argentipes and P. papatasi when the parasite was exposed to proteolytic enzymes in the gut. Susceptibility to these digestive enzymes has been shown to be a major barrier to vector competence (Schlein and Romano, 1986; Schlein et al., 1990). Thus, GP63 does not appear to confer proteinase resistance to competent strains, nor does it appear to be required for nutrient utilization in the bloodmeal during the early stages of development in the fly. gp63 1–6 genes do, however, play a major protective role against complement-mediated lysis when promastigotes are introduced into the mammalian host.
The wild-type L. major strain NIH S (MHOM/SN/74/Seidman) clone A2 (Wallis and McMaster, 1987) was used for all experiments. All parasites were routinely cultivated in M199 medium (BRL) containing 10% heat-inactivated fetal calf serum (FCS) at 26°C. During selection for GP63-deficient parasites after the second round of gene replacement, the transfectants were recovered, selected, cloned and initially expanded in M199 supplemented with 30% FCS and 5 mg ml−1 trypticase (BBL/Becton Dickinson). This enriched M199 medium is referred to as JB199.
Plasmid construction for gp63 gene deletion and complementation studies
For targeted gene replacement in L. major promastigotes, targeting DNA fragments were constructed in which either the sat gene (Heim et al., 1989; Tietz and Brevet, 1990), which confers resistance to nourseothricin (nour; Joshi et al., 1995), or the hyg gene, which confers resistance to hygromycin B (hygro; Gritz and Davies, 1983), were flanked at the 5′ end by the unique upstream region of gp63 gene 1 and at the 3′ flanking region of gp63 gene 6 to generate fragments that would ensure specific replacement of gp63 genes 1–6 (Fig. 1). To ensure high-level expression in promastigotes (Souza et al., 1994), the 3′ flanking region (1.3 kb) of the L. major dhfr-ts gene was inserted directly after the stop codon of both the sat and hyg gene before ligation to the 3′ region of L. major gp63 gene 6. The plasmids pLGT39 and pLGT40 were constructed for the first and second rounds of gp63 gene 1–6 replacement respectively. pLGT39 was derived from the plasmid pLGT37, which contained the entire 549 bp coding region of the sat gene flanked at the 5′ end by a 1.365 kb region from immediately upstream of L. major gp63 gene 1 and at the 3′ end by a 0.946 kb fragment from immediately downstream of gp63 gene 6. pLGT40 was derived from plasmid pLGT35 and is identical to pLGT37, except that the sat gene was replaced by the 1.0 kb coding region of the hyg gene. The 5′ flanking region of gp63 gene 1 was obtained by PCR amplification from plasmid pLMS10.1.3 (Button and McMaster, 1988) using the oligonucleotide primers o5′gp63 gene 1 XbaI [sequence: 5′-GTGGtctagaTGACGCTCCACATGGGA-3′ containing an XbaI site (lower case)] and 5′gp63 gene 1 SpeI [sequence: 5′-GGTGGactagtGGCTCTGCAGGCGCGGGGGCGCTGT-3′ containing an SpeI site (lower case)] corresponding to positions −1348 to −1365 and −1 to −25, respectively, upstream of the ATG translation initiation codon. The sat gene was obtained by PCR amplification from the plasmid pLEXSat (Joshi et al., 1995) using the oligonucleotide primers oSatSpe [sequence: 5′-GTGactagtATGAAGATTTCGGTGATCC-3′ containing a SpeI linker (lower case)] and oSatEV [sequence: 5′-GGTCgatatcTTAGGCGTCATCCTGTGC-3′ containing an EcoRV linker (lower case)] corresponding to the first 19 bp and last 18 bp, respectively, of the sat-1 coding sequence (Heim et al., 1989). The hyg gene was obtained by PCR amplification from pX63-HYG (Cruz et al., 1991) using the primers oHygSpe [sequence: 5′-GTGGactagtATGAAAAAGCCTGAACTC-3′ containing an SpeI site (lower case)] and oHygEcV [sequence: 5′-GTGGgatatcCTATTCCTTTGCCCTCGG-3′ containing an EcoRV site (lowercase)] corresponding to the first 18 bp and last 24 bp of the hyg coding sequence (Gritz and Davies, 1983). The 3′ flanking region of gp63 gene 6 fragment was obtained by PCR amplification from the plasmid pLGS6.1 (Voth et al., 1998) using the primers o6 5′ EV [sequence: ATCgatatcGGGACGGCGGCCTG containing an EcoRV site (lower case)] and o6 3′ Apa [sequence: GGCGgggcccGTTGATGCACGATGACG containing an ApaI site (lower case)] corresponding to positions 1–12 and 942–915 bp downstream from the translation stop codon of gp63 gene 6 (Voth et al., 1998). The PCR products were digested with the appropriate enzymes and the XbaI/SpeI 5′ flanking, the SpeI/EcoRV sat- or hyg-containing and the EcoRV/ApaI 3′ flanking fragments were ligated sequentially into the corresponding sites of the Bluescript vector KS (II)– (Stratagene) to give plasmids pLGT37 (sat containing) and pLGT35 (hyg containing). To obtain pLGT39 and pLGT40, a 1.3 kb BamHI fragment containing the 3′ flanking region of the dhfr-ts gene from L. major was isolated from the plasmid pX63-HYG (Cruz et al., 1991), the ends filled in with the Klenow fragment of DNA polymerase I and ligated into either pLGT37 or pLGT35, which had been linearized with EcoRV. This strategy placed the 3′dhfr-ts flanking region between the drug resistance gene and the 3′ flanking region of gp63 gene 6. For transfection, CsCl-purified plasmids pLGT39 and pLGT40 were digested with ApaI and XbaI, the DNA precipitated with ethanol, redissolved in TE and used for transfections without further purification. For gene complementation studies, the entire coding region of gp63 gene 1 was amplified by PCR and blunt end cloned into the SmaI site of the Leishmania expression vector pLEXNeo (Joshi et al., 1995) to give the plasmid pLEXNeo-gp63 gene 1.
Transfection, selection and cloning of L. major promastigotes
Log phase L. major promastigotes were transfected with 5 μg of linearized DNA by electroporation as described previously (2.25 kV cm−1, 500 μF; Kapler et al., 1990) In the first round of gp63 gene 1–6 deletion, electroporated promastigotes were recovered in 10 ml of drug-free M199 media for 2 days, after which the parasites were centrifuged (1000 × g for 10 min), resuspended in 100–200 μl of media and plated onto semisolid M199 containing 50 μg ml−1 nour. In the second round of gp63 gene 1–6 deletion, electroporated promastigotes were grown in 10 ml of JB199 media containing 60 μg ml−1 nour and, at 24 h, hygro (Sigma) was added to 70 μg ml−1. After 96 h, the promastigotes were centrifuged (1000 × g for 10 min) and resuspended in fresh JB199 medium containing 60 μg ml−1 and 70 μg ml−1 of nour and hygro respectively. The double drug-resistant parasites were grown for 12 days after electroporation before cell sorting was performed. The cell-sorted parasites (in PBS) were centrifuged (1000 × g for 10 min) and resuspended in 10-fold increments of cell numbers ranging from 1 × 104 to 1 × 106 promastigotes in 200 μl of conditioned media (supernatant of a mid-log culture of the wild-type L. major strain NIH S) diluted 1:5 with JB199 and spread onto semisolid JB199 containing 60 μg ml−1 and 70 μg ml−1 nour and hygro respectively. Double drug-resistant parasite colonies appeared after 1 week and were expanded in JB199 containing 60 μg ml−1 and 70 μg ml−1 nour and hygro respectively.
Transfection with the plasmid pLEXNeo-gp63 gene 1 and selection of G418-resistant promastigotes was performed as described previously (Joshi et al., 1995), except that transfectants were initially selected in liquid M199 containing 8 μg ml−1 G418 followed by cloning on semisolid M199 containing 16 μg ml−1 G418. Isolated clones were expanded and grown routinely in M199 containing 32 μg ml−1 G418.
DNA and RNA isolation and Southern blot analysis
Genomic DNA from L. major was isolated as described previously (Medina-Acosta and Cross, 1993). Total L. major RNA was isolated using TRIzol (BRL) according to the manufacturer's instructions. RNA was treated routinely with DNAse I before use. Southern analysis of genomic DNA was performed as described previously (Button et al., 1989). Briefly, restriction enzyme-digested DNA was separated on 0.8% agarose gels, transferred onto Hybond-N (Amersham) and hybridized according to the manufacturer's instructions with probes that had been uniformly labelled with [α-32P]-dCTP by the random-priming method. After hybridization, the filters were washed in 0.1% SSPE (18 mM NaCl, 1 mM Na3PO4 and 0.1 mM EDTA, pH 7.7)/0.1% SDS at 65°C for 30 min.
Promastigote cellular lysates preparation and Western blot analysis
Whole-parasite lysate preparation, SDS–PAGE and Western blots were prepared as described previously (Wallis and McMaster, 1987). Immunodetection on Western blots was performed with anti-GP63 monoclonal antibody CP3.235 (Button et al., 1991) using an ECL detection kit (Amersham) according to the manufacturer's instructions.
Fluorescence flow cytometry
L. major promastigotes, 5 × 106, were washed twice in phosphate-buffered saline (PBS) containing 0.5% BSA and centrifuged. The pellets were incubated on ice for 1 h with 100 μl of azide-free hybridoma culture supernatant containing mAb 96 to L. major GP63 (Macdonald et al., 1995). After extensive washing with PBS/BSA, the parasites were incubated on ice for 1 h in 50 μl of fluorescein isothiocyanate (FITC)-conjugated goat anti-mouse IgG (Southern Biotechnology) diluted to 10 μg ml−1 in PBS/BSA. The promastigotes were washed twice in PBS/BSA, resuspended in 0.5 ml of PBS and analysed by flow cytometry on a FACScan analyser (Becton Dickinson Immunocytometry Systems) using LYSIS II software. For preparative flow cytometry (cell sorting), an identical protocol for immunolabelling was followed, except that 2 × 107 promastigotes were used and aseptic conditions were maintained throughout the procedure. The immunolabelled parasites in 4 ml of PBS were sorted on a FACStar Plus cell sorter (Becton Dickinson Immunocytometry Systems) using CELLQUEST software.
Selection of metacyclic promastigotes
The L. major strain used in these studies was atypical insofar as it was not recognized by PNA lectin even during logarithmic phase growth (Sacks and de Silva, 1987; Sacks et al., 1990). The LPG expressed by this strain lacked terminally exposed galactose residues either in its side-chain substitutions or in its neutral capping oligosaccharides (D. L. Sacks, unpublished). Nonetheless, metacyclic promastigotes were isolated using a protocol described previously for L. donovani, which takes advantage of the fact that the mannose-containing, concanavalin A-binding capping sugars were no longer exposed on metacyclic LPG (Sacks et al., 1995). Briefly, equal volumes of promastigotes (2 × 107 parasites ml−1) from late stationary phase (7–8 days) cultures were mixed with 100 μg ml−1 concanavalin A in Hanks' balanced salt solution (HBSS) containing 1% BSA and incubated at room temperature for 30 min. After centrifugation at 150 × g for 5 min, the unagglutinated metacyclics remaining in the supernatant were aspirated and washed twice with HBSS.
In vitro infectivity studies
Peritoneal macrophages from BALB/c mice were adhered overnight on eight-chamber tissue culture plastic slides (Labtech) and infected at approximately 1 promastigote per macrophage with purified con A-negative metacyclics that had been preopsonized with serum from C5-deficient mice. After incubation for 1 h at 35°C, non-phagocytosed promastigotes were removed by gentle washing with HBSS. The initial attachment and uptake of promastigotes by macrophages and subsequent intracellular survival and growth of amastigotes was determined by counting the number of attached or intracellular parasites in stained slides 1 h and 96 h after infection. Slides were fixed in methanol and stained with Diff-Quick solution (American Scientific Products). The number of Leishmania per 100 macrophages was determined by examination of approximately 400 macrophages per assay.
Complement lysis assay
The sensitivity of procyclic and metacyclic promastigotes to complement lysis was measured as described previously (Franke et al., 1985). Briefly, promastigotes (1 × 108 ml−1) were incubated at 37°C in HBSS/1% BSA with an equal volume of diluted fresh pooled normal human serum. After 1 h, the samples were diluted in ice-cold PBS and placed on ice. Percentage lysis, assessed by loss of promastigote motility and morphological changes, was calculated relative to control samples incubated in heat-inactivated serum.
BALB/c mice (five mice per group) were inoculated subcutaneously in the left hind footpad with 1000 con A-negative selected metacyclic promastigotes in HBSS. Footpads were measured at various times after inoculation using a direct-reading vernier caliper.
Phlebotomus papatasi and P. argentipes sandflies were reared and maintained in the Department of Entomology, Walter Reed Army Institute of Research. The sandflies were infected as described previously (Pimenta et al., 1994). Briefly, flies were fed through chick skin membrane with heparinized mouse blood containing 2 × 106 ml−1 logarithmic phase promastigotes or containing 1 × 106 ml−1 tissue amastigotes purified from footpad lesions of BALB/c mice. For infections initially using promastigotes, mouse blood cells were washed twice in 0.86% saline and added back to the plasma that been heat inactivated at 56°C for 45 min. Blood-engorged sandflies were separated and maintained at 28°C with 30% sucrose solution. At various times after feeding, the flies were anaesthetized with CO2; then their midguts were dissected and examined microscopically for the presence and anatomic localization of midgut promastigotes. Dissected midguts were placed individually into microfuge tubes containing 30 μl of PBS. Midguts were homogenized with a Teflon pellet pestle (Kontes Scientific Glassware Instruments), and released parasites were counted in a haemocytometer, fixed on slides for Giemsa staining and examined microscopically for morphological changes associated with stage differentiation.
We are grateful to Dr Ben Kelly for help with Southern blot analysis and for critical reading of the manuscript. This work was supported by grant MT-7399 (W.R.M.) from the Medical Research Council of Canada.