Characterization of a Leishmania stage-specific mitochondrial membrane protein that enhances the activity of cytochrome c oxidase and its role in virulence

Authors

  • Ranadhir Dey,

    1. Division of Emerging and Transfusion Transmitted Diseases, Center for Biologics Evaluation and Research, FDA, Bethesda, MD, USA.
    Search for more papers by this author
  • Claudio Meneses,

    1. Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, MD, USA.
    Search for more papers by this author
  • Poonam Salotra,

    1. Institute of Pathology (ICMR), Safdarjung Hospital Campus, New Delhi, India.
    Search for more papers by this author
  • Shaden Kamhawi,

    1. Laboratory of Malaria and Vector Research, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Rockville, MD, USA.
    Search for more papers by this author
  • Hira L. Nakhasi,

    1. Division of Emerging and Transfusion Transmitted Diseases, Center for Biologics Evaluation and Research, FDA, Bethesda, MD, USA.
    Search for more papers by this author
  • Robert Duncan

    Corresponding author
    1. Division of Emerging and Transfusion Transmitted Diseases, Center for Biologics Evaluation and Research, FDA, Bethesda, MD, USA.
      E-mail robert.duncan@fda.hhs.gov; Tel. (+1) 301 827 3160; Fax (+1) 301 480 7928.
    Search for more papers by this author

E-mail robert.duncan@fda.hhs.gov; Tel. (+1) 301 827 3160; Fax (+1) 301 480 7928.

Summary

Leishmaniasis is caused by the dimorphic protozoan parasite Leishmania. Differentiation of the insect form, promastigotes, to the vertebrate form, amastigotes, and survival inside the vertebrate host accompanies a drastic metabolic shift. We describe a gene first identified in amastigotes that is essential for survival inside the host. Gene expression analysis identified a 27 kDa protein-encoding gene (Ldp27) that was more abundantly expressed in amastigotes and metacyclic promastigotes than in procyclic promastigotes. Immunofluorescence and biochemical analysis revealed that Ldp27 is a mitochondrial membrane protein. Co-immunoprecipitation using antibodies to the cytochrome c oxidase (COX) complex, present in the inner mitochondrial membrane, placed the p27 protein in the COX complex. Ldp27 gene-deleted parasites (Ldp27−/−) showed significantly less COX activity and ATP synthesis than wild type in intracellular amastigotes. Moreover, the Ldp27−/− parasites were less virulent both in human macrophages and in BALB/c mice. These results demonstrate that Ldp27 is an important component of an active COX complex enhancing oxidative phosphorylation specifically in infectious metacyclics and amastigotes and promoting parasite survival in the host. Thus, Ldp27 can be explored as a potential drug target and parasites devoid of the p27 gene could be considered as a live attenuated vaccine candidate against visceral leishmaniasis.

Introduction

Leishmania is the causative agent of leishmaniasis, a spectrum of diseases affecting more than 12 million people worldwide (Desjeux, 2004). Visceral leishmaniasis, caused by Leishmania donovani, is fatal if untreated. Available drugs are toxic, expensive and there is the problem of emergence of drug-resistant parasites (Handman, 2001; Selvapandiyan et al., 2006; Santos et al., 2008). Dimorphic Leishmania cycle between the alimentary tract of the sandfly vector and phagolysomes of mammalian macrophages (Molyneux and Killick-Kendrick, 1987; McConville and Ralton, 1997). Promastigotes, the insect stage, normally survive and proliferate in the sugar-enriched environment of the alimentary canal of the vector (McConville et al., 2007; Volf and Myskova, 2007). Metacyclogenesis, the differentiation of non-infective procyclic promastigotes to virulent metacyclics, is an important stage in the leishmanial life cycle allowing transmission of the parasites from the sandfly vector to the mammalian host (Sacks and Perkins, 1984; Bates, 2007). Within the sandfly midgut, a percentage of procyclics, a midgut-bound stage, differentiate into free-swimming highly motile metacyclics that are resistant to lysis by complement, a pre-adaptation to infection of mammalian cells (Cunningham et al., 2001; Akopyants et al., 2004; Kamhawi, 2006; Bates, 2008). Sandflies are pool feeders and when the infective sandfly bites a mammal for its blood meal, it introduces metacyclic promastigotes into the skin where they are subsequently phagocytosed by macrophages. In the macrophage phagolysosomes, the parasites confront an acidic and nutrient scarce hostile environment where they differentiate to amastigotes (Burchmore and Barrett, 2001; Dey et al., 2007). To cope with the hostile environment inside the macrophages, these amastigotes change their gene expression levels to modify morphological and biochemical activity (Shapira et al., 1988; Joshi et al., 1993; McConville and Ralton, 1997; Alexander et al., 1999; McConville et al., 2007). These changes in gene expression levels may point to unique parasite genes that could be targeted to block new infection. In our previous studies using microarrays, we identified genes that may relate to Leishmania virulence during the process of differentiation from promastigotes to axenic amastigotes and from amastigotes isolated from patients (Duncan, 2004; Duncan et al., 2004; Srividya et al., 2007)

In search of functions that may be unique to amastigotes, we noted that the shift of metabolism from promastigotes to amastigotes leads to the expression of a spectrum of genes that could be targets to control Leishmania pathogenesis. Whereas promastigotes utilize glucose as their primary energy source, intracellular amastigotes depend primarily on amino acids and fatty acids as their carbon source (McConville and Handman, 2007; Naderer and McConville, 2008). Increased mitochondrial activity may play a crucial role in the survival of amastigotes inside host cells (McConville et al., 2007; Naderer and McConville, 2008). The mitochondrion harnesses the energy from numerous substrates through the electron transport chain. Electron transport depends on multi-protein complexes I, II, III and IV embedded in the inner mitochondrial membrane ultimately passing the electrons to oxygen. This oxygen consumption is referred to as respiration. The proton gradient produced by electron transport drives the F1/F0 ATPase (complex V) in a coupled process termed oxidative phosphorylation. Active respiration is required for survival of both promastigote and amastigote forms of Leishmania (Hart et al., 1981; Van Hellemond and Tielens, 1997). Investigations of the individual complexes of the respiratory chain suggest NADH dehydrogenase (complex I) (Cermakova et al., 2007; Panigrahi et al., 2008), succinate dehydrogenase (complex II), cytochrome c reductase (complex III) and cytochrome c oxidase (complex IV) are found in both Leishmania and Trypanosoma (Santhamma and Bhaduri, 1995; Hellemond et al., 2005).

The trypanosomatid cytochrome c oxidase (COX) complex (complex IV) is a multicomponent complex composed of more than 14 subunits (Speijer et al., 1996; Horvath et al., 2000a). It has three mitochondrially encoded subunits and all the others are nuclear-encoded subunits. Most of the nuclear-encoded components have no apparent homologue outside of trypanosomatids (Speijer et al., 1996; Horvath et al., 2000b). Some of the nuclear-encoded subunits are essential for proper function of complex IV (Horvath et al., 2005).

Recently, a nuclear-encoded subunit, MIX, was identified as a subunit of the COX complex in Leishmania major and Trypanosoma brucei. However, its function in the COX complex has not been determined (Zikova et al., 2008). In this study, we characterize a gene encoding a 27 kDa mitochondrial membrane protein (Ldp27), a subunit of the active COX complex, specific to amastigotes and metacyclics, the infectious stages in Leishmania. We also demonstrate that Ldp27 is necessary for the high level of COX activity in amastigotes and that Ldp27 gene-deleted parasites (Ldp27−/−) show significantly less COX activity and reduced ATP synthesis in intracellular amastigotes compared with wild type. Moreover, the Ldp27−/− parasites are less virulent both in human macrophages and in BALB/c mice.

Results

Ldp27 is an amastigote/metacyclic specific protein in Leishmania and is only present in trypanosomatids

Ldp27 was identified previously by transcriptome analysis as a more abundantly expressed gene in L. donovani amastigotes (Srividya et al., 2007). This open reading frame encodes a protein 221 amino acids in length with a predicted molecular weight of 27 kDa (Ldp27). The alignment of L. donovani p27 protein with orthologues from Leishmania infantum, L. major, L. braziliensis, T. brucei and T. cruzi was performed (Fig. 1). The sequences of p27 are highly conserved at the amino acid level in trypanosomatids (Fig. 1). The similarity of p27 among all the Leishmania species is 80% or more whereas the similarity of Ldp27 with T. brucei and T. cruzi p27 sequences is 65% and 62% respectively. Ldp27 has a predicted N-terminal mitochondrial targeting sequence nine amino acids in length (http://wolfpsort.org) and according to InterPro Scan profile search there is a single predicted transmembrane domain (Fig. 1). blast searches of the GenBank database found no p27-related genes in other organisms.

Figure 1.

Multiple alignment of p27 sequences among trypanosomatids. L. donovani (GenBank submission in process), T. brucei (Tb11.0400) and T. cruzi (Tc00.1047053508823.70) p27 amino acid sequences. Dotted box indicating the predicted mitochondrial targeting signal sequence and solid box indicating the predicted transmembrane domain.

To confirm the differential expression of this gene, we isolated RNA from log- (24–36 h in culture) and stationary-phase (5 days in culture) promastigotes and amastigotes derived from promastigotes in vitro by culturing under conditions that induce differentiation, referred to as axenic amastigotes (Debrabant et al., 2004). Northern blot analysis with the p27ORF as a probe clearly showed that Ldp27 was highly expressed in both log- and stationary-phase axenic amastigotes and substantially less in promastigotes (compare lane 3 and 4, with lane 1 and 2, Fig. 2A). Western blot analysis (lane 3 and 4, Fig. 2B) and immunofluorescence staining (Fig. 2C) using anti-Ldp27 antibody revealed abundant expression of Ldp27 protein in the axenically cultured amastigote stage with only a background level in promastigotes. Interestingly, when we infected human macrophages with L. donovani, 12 h post infection, we were able to detect the Ldp27 protein by immunoprecipitation (lane 2, Fig. 2D) as well as by immunofluorescence using anti-Ldp27 antibody (Fig. 2E). These intracellular amastigote results suggest early onset of expression upon differentiation into amastigotes.

Figure 2.

Differential expression of the Ldp27 gene in promastigotes and amastigotes.
A. Northern blot of RNA from different stages of the L. donovani parasite using Ldp27 or Ld tubulin ORFs as probes.
B. Western blot of protein lysates from different stages of L. donovani parasites, using anti-p27 purified IgG antibody raised against p27 recombinant protein. Lower panel shows protein loading control, with anti-tubulin antibody.
C. Immunofluorescence analysis of expression of p27 protein in promastigotes and axenic amastigotes, using anti-p27 antibody, DAPI to stain the nucleus and kinetoplast. NRS, normal rabbit serum shows no cross reactivity with Leishmania proteins.
D. Ldp27 expressed in intracellular amastigotes. Human monocytes were cultured in the presence of MCSF, and infected with L. donovani stationary-phase promastigotes, after 6 h of infection, extracellular promastigotes were washed away and the infected macrophages lysed. The lysates (5 × 106 cells) were immunoprecipitated using either anti-p27 antibody or NRS, and Western blot probed with anti-p27 antibody. Mφ, macrophage.
E. Immunofluorescence analysis of expression of Ldp27 protein in intracellular amastigotes using anti-p27 antibody, DAPI to stain the nucleus and kinetoplast. Human monocytes were cultured and infected with L. donovani promastigotes for 6 h. Extracellular promastigotes were washed away, cells were fixed and stained.
F. Expression of p27 protein in L. infantum chagasi and L. donovani procyclics (day 4 midgut parasites) and metacyclics (50–80% of midgut population) recovered from sandflies, and hamster spleen-derived amastigotes. Western blots bound with anti-Ldp27 and anti-Ld calreticulin antibodies. UIG, Un-infected sandfly gut lysates used as negative control. The data presented here are representative of two separate experiments.

Since procyclic promastigotes undergo differentiation to metacyclic promastigotes in the sandfly gut, a stage preceding differentiation to amastigotes, we investigated the expression of p27 in metacyclic promastigotes isolated from infected sandflies. Laboratory-reared Lutzomyia longipalpis sandflies, susceptible to experimental infection with several Leishmania species (Kamhawi, 2006), were fed on blood meals containing hamster-derived L. infantum chagasi and L. donovani amastigotes. L. longipalpis is found naturally infected with L. infantum chagasi which also causes visceral disease similar to L. donovani. Four days post infection, the sandflies were dissected and the promastigotes (designated as procyclics) were collected. Another group of flies were dissected 9 days post infection, which allowed enough time for differentiation into metacyclics that represented 50–80% of the midgut parasite populations. Immunoblot analysis using Ldp27 antibody revealed abundant expression of p27 protein in sandfly-derived metacyclics of L. infantum chagasi and L. donovani as well as in hamster tissue-derived amastigotes of both species. In contrast, no expression was seen in sandfly-derived procyclic promastigotes (compare lanes 3 and 4 with lane 2 in Fig. 2F). Interestingly, sandfly-derived metacyclic promastigote-expressed p27 protein in both Leishmania species has a slower and more diffuse electrophoretic mobility than p27 from tissue-derived amastigotes. Thus the above experiments indicate that Ldp27 is a unique trypanosomatid protein differentially expressed during metacyclogenesis inside the vector gut and in intracellular amastigotes.

Ldp27 is a mitochondrial membrane protein in amastigotes

Since Ldp27 protein has a predicted mitochondrial targeting signal and a transmembrane domain, we used the mitochondrion-specific marker, MitoTracker Red and biochemical analysis to confirm its localization in mitochondria. Immunofluorescence analysis using axenic amastigotes showed that the Ldp27 signal had good correlation with the MitoTracker staining of the mitochondria (Fig. 3A and inset). To verify the localization of Ldp27 biochemically, subcellular fractionation of mitochondria was performed by sequential lysis and differential centrifugation as described in Experimental procedures. Each mitochondrial fraction was characterized by Western blot using anti-p27 antibodies and antibodies to proteins that are markers for each fraction. The results showed that Ldp27 is an inner-mitochondrial membrane protein (Fig. 3B, lane 5). Solubilization of the inner mitochondrial membrane fraction with Na2CO3, resulted in the majority of Ldp27 protein remaining in the pellet fraction (PL) (lower panel, Fig. 3C, lane 8), suggesting that this protein is membrane bound similar to COX complex subunit VI (COVI) possibly through the trans-membrane domain. These results suggest that Ldp27 is an inner mitochondrial membrane protein.

Figure 3.

Ldp27 is a mitochondrial inner membrane protein.
A. Immunofluorescence analysis of subcellular localization of p27 protein in axenic amastigotes, using MitoTracker to stain the mitochondria (red), anti-Ldp27 antibody plus FITC conjugated secondary antibody to stain p27 (green) and DAPI to stain the nucleus and kinetoplast (blue). Arrows indicate the colocalization (yellow) of Ldp27 and MitoTracker.
B. Subcellular localization of Ldp27 in axenic amastigote fractionation was determined using standard markers for each fraction: anti-Ld APRT (adenine phosphoribosyltransferase) for the cytoplasmic fraction (CF), anti-CytC (cytochrome c) for intermembrane space (IMS), anti-COVI for inner membrane (IM), and HSP70 for mitochondrial matrix (MX). WCL, whole cell lysates; OM, outer mitochondrial membrane.
C. Ldp27 is a membrane-bound protein, and it does not solubilize with 0.1 M Na2CO3. SL, soluble fraction; PL, pellet fraction. The data presented here are representative of more than three separate experiments.

Ldp27 is a component of the active COX complex in amastigotes

The orthologue of Ldp27 in T. brucei (Tb11.0400) is a part of the COX complex in the mitochondrial inner membrane (Zikova et al., 2008). In the current study, Ldp27 has been localized to the inner mitochondrial membrane, suggesting Ldp27 could be a part of the COX complex. To establish the association of Ldp27 with this COX complex, co-immunoprecipitation followed by immunoblotting was performed with anti-Ldp27 antibody and antibodies to two of the most well-characterized trypanosomatid-specific subunits of the COX complex, COIV (Maslov et al., 2002) and COVI (Horvath et al., 2005). Immunoprecipitation with anti-p27 pulled down the COIV and COVI subunits of the COX complex in the amastigote stage, demonstrating the interaction of p27 with these two components of the COX complex (Fig. 4A). Consistent with the absence of p27 in Western blots of promastigote lysates, the COX components were not immunoprecipitated by p27 from promastigotes (Fig. 4A). Interestingly, immunoprecipitation with anti-COVI pulled down the COIV and COVI subunits both in promastigotes and amastigotes, suggesting the presence of a COX complex in promastigotes (Fig. 4A). To further confirm the expression of COX components in both the life cycle stages, we performed Western blot analysis of whole cell lysates of promastigotes and amastigotes, which clearly revealed the abundant expression of COIV and COVI subunits in both promastigotes and axenic amastigotes; as expected, p27 was only expressed in amastigotes (Fig. 4B). To investigate the role of p27 in COX activity, we isolated the mitochondria from both promastigotes and amastigotes. The activity assay showed that amastigote mitochondria have significantly (P < 0.01) higher COX activity in comparison with promastigote mitochondria (Fig. 4C). Altogether these results suggest an important role for Ldp27 in the increased activity of the COX complex in the amastigote stage.

Figure 4.

Ldp27 is a component of the cytochrome c oxidase complex.
A. Promastigotes and axenic amastigotes were lysed and immunoprecipitated with anti-p27 or anti-COVI antibodies and Western blots are probed with anti-COIV or anti-COVI antibodies.
B. Western blot analysis of whole cell lysates of promastigotes (Pro) and axenic amastigotes (AxAm) using anti-COIV, anti-COVI, anti-p27 and anti-tubulin antibody.
C. COX activity of mitochondria isolated from promastigotes and axenic amastigotes measured as described in Experimental procedures. The data presented here are representative of more than three separate experiments, *P < 0.01.

Absence of Ldp27 in the intracellular amastigote affects the COX activity as well as ATP synthesis by oxidative phosphorylation

An L. donovani p27 null mutant was generated by homologous recombination (Fig. 5A). The Southern blot analysis showed complete loss of the coding region of the Ldp27 in double knockout mutants (Fig. 5B, lanes 2, 4, 6). Loss of Ldp27 expression in the Ldp27−/− parasite was confirmed by Western blot analysis using anti-Ldp27 antibody (Fig. 5C, lane 2). Ldp27 expression was restored by transfecting the null mutant cells with the pXG-vector containing the coding sequence of Ldp27 (Ldp27−/− AB, Fig. 5C, lane 3). Previously it has been shown that deletion of other components of the COX complex led to destabilization of the COX complex, eventually affecting the COX activity (Horvath et al., 2005; Zikova et al., 2008). To evaluate the stability of the COX complex in the Ldp27−/− cell line, we performed co-immunoprecipitation using anti-Ldp27 or anti-COVI antibody and then immunoblotted with anti-COIV, anti-COVI or anti-Ldp27 antibodies. The analysis showed that anti-Ldp27 antibody did not immunoprecipitate COVI or COIV molecules from the Ldp27−/− lysates (Fig. 6A, lane 3). However, anti-COVI antibody immunoprecipitated COIV and COVI subunits in both the wild-type as well as Ldp27−/− parasites (Fig. 6B, lanes 2 and 3). These results suggest that even in the absence of Ldp27 a COX complex remains intact. The Western blot of whole cell lysates, using COIV, COVI, Ldp27 and tubulin antibodies, showed that in the Ldp27−/− cell line, the expression of the other COX components was as abundant as in the wild-type cells (Fig. 6C). Although we have only evaluated the presence of three out of more than 14 subunits in a COX complex, these results suggest that even in the absence of Ldp27 protein the expression and assembly of at least some of the COX subunits are not affected.

Figure 5.

p27 gene disruption in the L. donovani genome.
A. Schematic diagram showing design and use of constructs for Ldp27 gene disruption in the L. donovani genome. Neo construct is flanked on the 5′ and 3′ sides with Ldp27 5′-UTR and 3′-UTR respectively. The position of PvuI, HindIII and SphI restriction sites are indicated.
B. Southern blot analysis of the genomic DNA of Leishmania wild-type (+/+) and Ldp27 gene-deleted (−/−) parasites with Neo and Ldp27 ORFs as 32P-labelled probes. The genomic DNA from the parasites cut with the restriction enzymes HindIII, SphI or PvuI were used in the analysis.
C. Western blot analysis of the lysates of wild-type (+/+), Ldp27 gene-deleted (−/−) and the null mutant parasites episomally expressing Ldp27 (−/− AB) parasites. The data presented here are representative of more than three separate experiments.

Figure 6.

In Ldp27−/− parasites, the integrity of the cytochrome c oxidase complex and expression of other components is not affected.
A and B. Lysates of the wild-type and Ld p27−/− axenic amastigote parasites, immunoprecipitated with either anti-p27 (A) or anti-COVI antibodies (B). Western blots of immunoprecipitated products were probed with anti-COIV, anti-COVI or anti-p27 antibodies. White asterisk (*) indicates Ldp27 band immunoprecipitated from wild type (B, lane 2).
C. Western blot analysis of whole cell lysates of wild-type and Ldp27−/− axenic amastigote using anti-p27, anti-COIV, anti-COVI or anti-tubulin antibodies. The data presented here are representative of more than three separate experiments. WT, wild type.

To evaluate the effect of p27 on COX activity under more physiological conditions, we infected human macrophages with Ldp27−/−, Ldp27−/− AB or wild-type promastigotes and isolated the amastigotes from these infected cells. We measured the COX activity of the mitochondria of isolated wild-type, Ldp27−/− and Ldp27−/− AB amastigotes. The COX activity was significantly (P < 0.05) reduced in Ldp27−/− cells (Fig. 7A). This impaired COX activity was significantly (P < 0.05) restored in the Ldp27−/− AB cell line (Fig. 7A), indicating Ldp27 is specifically required for normal COX activity in intracellular amastigotes. The Western blot analysis of these mitochondrial preparations showed the same level of COIV and COVI in all three cell types. The striking difference was the lack of p27 in Ldp27−/− cells which had reduced COX activity (Fig. 7B). However, further investigation will be required to determine whether some other subunit(s) besides COIV and COVI are affected.

Figure 7.

COX activity and oxidative phosphorylation are impaired in intracellular Ldp27−/− amastigote mitochondria.
A. Human elutriated monocytes were cultured in the presence of MCSF and infected with wild-type (Wt), p27−/− or p27−/− AB promastigotes. Twelve hours post infection the intracellular amastigotes were isolated from these infected cells. COX activity was measured as described in Experimental procedures from the isolated mitochondria of these amastigotes. *P < 0.05.
B. Western blot analysis of the mitochondrial lysates of indicated intracellular parasites isolated as described in (A). Cell type is shown at the top, antibody staining on the right.
C. ATP production by crude mitochondria prepared by digitonin extraction from the indicated types of intracellular amastigotes was evaluated for three substrates, succinate, α-ketoglutarate and pyruvate each indicating a separate mitochondrial pathway. The tested substrate is indicated at the top. Malonate, a specific inhibitor of succinate dehydrogenase, was used to inhibit ATP production by oxidative phosphorylation and attractyloside was used to inhibit the import of ADP into mitochondria, which inhibits all mitochondrial ATP production. Addition of these compounds to the sample or no inhibitor (None) is indicated by the bar graph shading according to the legend. ATP production in mitochondria isolated from wild-type cells without addition of malonate or attractyloside was set to 100% for each substrate. The data presented here are representative of more than three separate experiments. WT, wild type; *P < 0.05.

Since COX activity results in ATP production in the mitochondria by the oxidative phosphorylation pathway, we investigated the production of ATP in the presence of substrates that distinguish substrate-level phosphorylation from oxidative phosphorylation. In the closely related trypanosomatid, T. brucei, procyclic stage, mitochondrial ATP is produced via three different pathways that can be assayed in the isolated intact mitochondria (Bochud-Allemann and Schneider, 2002). Succinate is the main substrate for the oxidative phosphorylation pathway, α-ketoglutarate induces ATP production by substrate-level phosphorylation occurring in the citric acid cycle, and pyruvate induces ATP production by substrate-level phosphorylation during the acetate-succinate coenzyme A (CoA) transferase cycle. Malonate, a specific inhibitor of succinate dehydrogenase, was used to inhibit ATP production by oxidative phosphorylation. All three forms of ATP production are sensitive to attractyloside, an inhibitor that prevents mitochondrial import of the added ADP in the reaction buffer (Bochud-Allemann and Schneider, 2002).

Figure 7C shows that the mitochondria isolated from p27−/− amastigotes produced about 50% (P < 0.05) of the ATP produced by wild-type mitochondria when we used succinate as a substrate. Additionally, ATP production approached the wild-type level in Ldp27−/− AB mitochondria where p27 expression was restored. Malonate reduced the succinate-derived ATP production by 50% (P < 0.05) in wild-type and Ldp27−/− AB mitochondria, but had little effect on Ldp27−/−, indicating that oxidative phosphorylation through succinate was eliminated in these mitochondria in the absence of p27. When we used α-ketoglutarate or pyruvate as substrates, the ATP production was not significantly affected by the absence of p27 indicating that substrate-level phosphorylation does not require this molecule. The attractyloside inhibition of ATP production with all three different substrate types reveals the background level of ATP measurement in this assay. These results suggest that p27 protein is an essential component of an active COX complex that acts specifically through the oxidative phosphorylation pathway to generate ATP.

Amastigotes of the Ldp27 null mutant are unable to survive in macrophages or mice

Since p27 protein was shown to be important for mitochondrial oxidative phosphorylation in amastigotes, we explored whether such an activity has a role to play in leishmanial pathogenesis in human macrophages ex vivo and in mice in vivo. Human monocytes differentiated by MCSF were infected with stationary-phase wild-type, Ldp27−/− and Ldp27−/− AB promastigotes. All three types of parasites infected the macrophages at similar levels at early hours of infection. However, Ldp27−/− parasites failed to replicate normally inside the macrophages as the days progressed, while wild-type and p27-complemented, Ldp27−/− AB, parasites underwent normal replication and survived in host cells (Fig. 8). Interestingly, there were a few viable amastigotes among the Ldp27−/−-infected macrophages even after 7 days of infection suggesting the lower level of COX activity observed in the absence of p27 may be sufficient to maintain a low number of parasites viable inside host cells.

Figure 8.

Ldp27−/− parasite does not survive well inside human macrophages ex vivo. Human macrophages infected with wild-type (WT), p27−/− or p27−/− AB, stationary-phase promastigotes and macrophages cultured for the indicated time periods. Cells were stained with Wright stain and counted to measure the intracellular parasite load in each type. Infection efficiency (% of infected cells, top), intracellular growth (parasites per infected cell, middle) and parasite survival (parasites per 100 cells, bottom) were recorded. The data presented here are representative of more than three separate experiments. *P < 0.01.

Susceptible BALB/c mice were infected with 3 × 106 metacyclic promastigotes of wild-type, Ldp27−/− and Ldp27−/− AB strains and the parasite load was determined at 5 and 13 weeks after infection from both liver and spleen either by the serial dilution method or by real-time PCR. Wild-type and Ldp27−/− AB parasites survived well at 5 and 13 weeks post infection in both liver and spleen (Fig. 9 and Table S1). However, at 5 weeks post infection, Ldp27−/− parasites failed to survive inside the animals showing a significant (P < 0.01) reduction of parasite burden in both liver and spleen compared with the wild-type parasite. After 13 weeks, there were a few surviving parasites in the liver, and none were detectable in the spleen. These data suggest that Ldp27 is essential for parasite survival and growth in the mammalian host.

Figure 9.

Ldp27−/− parasites are less virulent in BALB/c mice. Survival of wild-type (WT), Ldp27−/− and Ldp27−/− AB parasites in BALB/c mice. Parasite burdens were analysed at 5 weeks or 13 weeks after infection by serial dilution from the organs of mice infected intravenously with metacyclic parasites. *P < 0.01.

Discussion

In this study, we show that Ldp27 is a protein specifically expressed in amastigotes and metacyclics, the infectious stages of Leishmania parasites, is part of the active COX complex involved in oxidative phosphorylation and is essential for survival of the intracellular parasite. Infection of a mammalian host by the Leishmania parasite requires differentiation from promastigote to amastigote to survive the acidic and nutrient-restricted phagolysosome of the macrophage cell. This differentiation has been associated with dramatic changes in gene expression (Joshi et al., 1993; Charest and Matlashewski, 1994; Duncan et al., 2001; 2004; Saxena et al., 2007; Srividya et al., 2007; Rochette et al., 2009).

Previous studies documenting changes in gene expression during the differentiation to amastigotes have been most meaningful when these genes were characterized and shown to be required for survival. Some of these genes are directly related to defending against host immunity, some are involved in self replication, and some are metabolically essential enzymes (Cruz et al., 1991; Alexander et al., 1998; Streit et al., 2001; Zhang and Matlashewski, 2001; Papadopoulou et al., 2002; Spath et al., 2003a,b; Selvapandiyan et al., 2004; Stewart et al., 2005; Vergnes et al., 2005; Boitz et al., 2009; Gaur et al., 2009). Our characterization of Ldp27 is similarly thorough, demonstrating stage-specific expression of its encoded protein, its biochemical function and its impact on parasite survival and virulence.

Identification of the function of the amastigote-specific Ldp27 protein was greatly aided by the availability of the annotated genomes of Leishmania and Trypanosoma species (http://www.genedb.org). A recent study of the components of the T. brucei COX complex identified Tb11.0400 among more than 14 proteins that appear to be core components of this complex (Zikova et al., 2008). Tb11.0400 is the T. brucei orthologue of Ldp27. The results of co-immunoprecipitation experiments demonstrate that Ldp27 is also a member of the COX complex in Leishmania and is important for its activity. COX activity is substantially reduced in p27 gene-deleted intracellular amastigotes and is restored by the episomal expression of Ldp27 in these gene-deleted cells. A functional role for Ldp27 is also suggested by the lower level of COX activity in the wild-type procyclic promastigote stage that does not express Ldp27. It has been established that the respiratory chain is active in Leishmania promastigotes (Santhamma and Bhaduri, 1995), and the inhibition of promastigote proliferation by cyanide indicates the requirement for an active COX in this stage (Van Hellemond and Tielens, 1997). In this study, COX activity was also detected in the promastigote form, although significantly less than the amastigote form. Unlike some of the other COX components examined (Horvath et al., 2005; Zikova et al., 2008), depletion of Ldp27 does not lead to complete loss of the complex. Other components, COIV and COVI, can be immunoprecipitated and are as abundant in promastigotes and Ldp27−/− amastigotes where Ldp27 is absent as they are in wild-type amastigotes, where Ldp27 is present. Thus Ldp27 may play a role in increasing the enzymatic activity of the COX complex, but not in the abundance or assembly of at least some of its components.

Since promastigote stages are only present in the sandfly gut in nature, we wanted to explore the expression of Ldp27 in infected sandflies. Leishmania parasites differentiate into several distinct developmental stages in the sandfly gut beginning with procyclic parasites and culminating with infectious metacyclics (Kamhawi, 2006). On day 4, the majority of the parasites are represented by leptomonads, a dividing stage that binds to the gut and is adapted for survival in the sandfly (Bates, 2007). Metacyclics, which begin to appear on day 5 onwards, are non-dividing, rapidly motile forms that resist complement lysis and are highly specialized towards successful transmission to a mammalian host and for survival until phagocytosed by a macrophage (Bates, 2007). Not much is known about the metabolic differences between these two distinct forms inside the insect gut. The expression of Ldp27 in metacyclics may reflect a metabolic shift aimed at generating more energy to sustain the increased motility of this stage. Alternately, metacyclics may express Ldp27 to prepare the parasite to rapidly differentiate into the amastigote stage. In either case, we have observed an alternate form of Ldp27 in the metacyclic stage parasites, indicated by a slower migration in SDS-PAGE compared with tissue-derived amastigotes in both L. donovani and L. infantum chagasi. At this time, we do not know what produces the migration difference of Ldp27 in metacyclics and it is the subject of future studies.

The expression of Ldp27 in amastigotes, correlated with an increase in COX activity, is essential for Leishmania survival in the human macrophage and for virulence in the mouse. The increase in COX activity in amastigotes is likely to be part of the metabolic shift adapting the parasite to the intracellular environment. Intracellular amastigotes are more dependent on the tricarboxylic acid cycle and mitochondrial respiration than on glycolysis for energy production (Naderer et al., 2006). Intracellular Ldp27-deficient parasites do not proliferate in human macrophages. This may be due to the reduction in ATP production by oxidative phosphorylation observed in mitochondria isolated from these parasites. The failure to proliferate in macrophages is reflected by the drastic reduction in parasite load in mice infected with the Ldp27−/− strain. Both replication in macrophages and virulence in mice were restored to near wild-type virulence by episomal expression of p27 in these cells, specifically implicating Ldp27 as a virulence factor for amastigotes.

The functional characterization of a protein specifically expressed in infectious stages of Leishmania and its requirement for survival and intracellular proliferation of amastigotes in the mammalian host is an important step towards understanding this human pathogen and developing means to control it. Drugs aimed at disrupting the function of Ldp27, which has no orthologue in the human cell, may be a means of clearing the parasites. In addition, the Ldp27−/− cell line, which proliferates well as promastigotes and shows little or no pathogenesis in mice, could be further investigated as a possible genetically modified live attenuated Leishmania vaccine. The Ldp27−/− cell line offers various attributes of a genetically targeted vaccine such as genetic stability, testable genetic identity and possibly an improved safety based on better control and monitoring of reversion to virulence (Selvapandiyan et al., 2006). Studies are underway to evaluate protective immunity conferred by Ldp27 gene-deleted parasites in animal models against challenge by virulent Leishmania parasites.

Experimental procedures

Parasites and animals

The L. donovani cloned line designated by the World Health Organization as MHOM/SD/62/1S-C12D (Ld1S2D) was used in all the experiments (Debrabant et al., 2004; Selvapandiyan et al., 2004). Promastigotes and the axenic amastigotes were grown and harvested as described previously (Debrabant et al., 2004). L. infantum chagasi strain (MHOM/BR/00/1669), originally isolated from a patient with visceral leishmaniasis in northeast Brazil, was used in this study. Ldp27-deleted (Ldp27−/−) and Ldp27 episomal add back (Ldp27−/− AB) parasites were derived from the L. donovani Ld1S2D wild type. Five- to six-week-old female BALB/c mice from the National Cancer Institute were used in the experiments. Tissue-derived amastigotes of both parasite species were harvested from spleens of Syrian hamsters 3 months following intracardial injection of parasites. Procedures used were reviewed and approved by the Animal Care and Use Committee, Center for Biologics Evaluation and Research, Food and Drug Administration.

Sandfly infection

Lutzomyia longipalpis sandflies (Jacobina strain) were maintained at the Laboratory of Malaria and Vector Research at the National Institute of Allergy and Infectious Diseases. Three to four days post eclosion sandflies were membrane fed on a blood meal containing 5 × 106 tissue-derived L. donovani or L. infantum chagasi amastigotes per ml of heparinized mouse blood. Blood fed females were separated the day after, maintained at 26°C, and given a 30% sucrose solution. At days 4 and 9 post infection, sandflies infected with either L. donovani or L. infantum chagasi were dissected in phosphate-buffered saline (PBS). For each time point, up to 50 midguts were pooled in a microfuge tube containing 50 µl of PBS and macerated using a Teflon-coated micro-tissue grinder. The number of promastigotes and per cent metacyclics were counted using a haemocytometer.

Isolation of RNA and Northern blot analysis

Total RNA was isolated from promastigote and axenic amastigote cultures of L. donovani using RNA STAT-60 according to the manufacturer's instructions (Tel-Test, Friendswood, TX). Total RNA (10 µg) was analysed by Northern blot as described (Selvapandiyan et al., 2004). Both Northern and Southern blots were hybridized with a 32P-labelled Ldp27 coding region probe, stripped and rehybridized with a 32P-labelled Ld tubulin probe. The membranes were exposed on X-ray film (Amersham Hyperfilm) and developed as described (Selvapandiyan et al., 2004).

Antibody generation and Western blot analysis

Production of polyclonal rabbit antibodies to Ldp27 was described previously (Duncan et al., 2009). Total IgG was purified from the rabbit anti-serum using an IgG purification kit (Pierce Biotech, USA) and used at 1:2000 dilution for Western blots. The proteins were visualized using the SuperSignal Chemi-luminescent substrate system (Pierce) (Selvapandiyan et al., 2001; Duncan et al., 2009). Other antibodies for COX subunits, anti-COIV and anti-COVI, we received as a gift from Julius Lukes, University of South Bohemia, Czech Republic. Rabbit polyclonal anti-Ld calreticulin was a gift of Alain Debrabant (Debrabant et al., 2002). For cultured parasites, the volume of lysate corresponding to 106 parasites was used in each lane; for sandfly-derived procyclics, metacyclics (representing 50–80% of the total gut population) and tissue-derived amastigotes, the volume of lysate corresponding to 105 parasites was used in each lane.

Immunoprecipitation

A total of 5 × 106 cells were washed in PBS and lysed in immunoprecipitation (IP) buffer [Tris 25 mM pH 8.0, NaCl 100 mM, Glycerol 10%, NP40 0.5%, MgCl2 5 mM, DTT 1 mM, PMSF 1 mM and freshly added protease inhibitor cocktail (Roche Applied Science)] for 1 h on ice. The respective antibodies added, incubated at 4°C overnight with gentle rotation. The immunoprecipitation reactions were mixed with Protein A Sepharose (Amersham) and incubated for 1 h at 4°C and washed three times with IP buffer. After removing the supernatant, SDS sample buffer was added to the Sepharose pellet, heated at 95°C for 5 min and the eluted solution was loaded on the SDS-PAGE gel.

Immunofluorescence analysis

Leishmania donovani promastigotes and amastigotes were fixed in suspension in 4% (w/v) paraformaldehyde in PBS (50 mm Na2HPO4, 150 mm NaCl, pH 7.4) for 20 min at room temperature, washed three times in PBS, and allowed to attach to glass slides. After air drying, the slides were first immersed in cold methanol (−20°C) for 5 min, blocked for 30 min in 1% (w/v) bovine serum albumin (United States Biochemical, Cleveland, OH) in PBS, and incubated 1 h with the anti-p27 antibody (1:200 dilution) diluted in 1% bovine serum albumin in PBS. After three washes in PBS, slides were incubated for 1 h with affinity-purified fluorescein-conjugated anti-rabbit IgG (H+L) as a secondary antibody. The secondary antibody (Vector Laboratories, Burlingame, CA) was diluted 1:200-fold in PBS containing 1% bovine serum albumin. Cells were subsequently washed three times with PBS and mounted in Vectashield containing 4′6-diamidino-2-phenylindole (DAPI, Vector Lab) to stain both nucleus and kinetoplast. Cells were examined for fluorescence under the microscope (Nikon Eclipse TE2000-U) and 0.3-micron-thick optical sections were captured and identical slices from MitoTracker and FITC channels were processed with Open lab 5.2 software (Perkin Elmer, Waltham, MT) to generate deconvoluted images. The focal plane chosen in all the cells was in the middle of the cells. The images were further processed using Adobe Photoshop 5.5 (Adobe Systems, Mountain View, CA) (Selvapandiyan et al., 2004).

Macrophage infection

Human elutriated monocytes were resuspended at 1.8 × 105 cells ml−1 in RPMI medium containing 10% FBS, macrophage colony-stimulating factor (20 ng ml−1, ProSpec, Israel), plated in 0.5 ml on eight-chamber Lab-Tek tissue-culture slides (Miles Laboratories) and incubated for 9 days for differentiation into macrophages. The differentiated macrophages were infected with L. donovani stationary-phase promastigotes (10:1 parasite-to-macrophage ratio). After incubation for 6 h at 37°C in 5% CO2, the free extracellular parasites were removed by RPMI washes and the cultures were incubated in macrophage culture medium for 7 days. At the indicated days the culture medium was removed, and the slides were air-dried, fixed by immersion in absolute methanol for 5 min at room temperature and stained using Diff-Quick Stain set (Baxter Healthcare Corporation, Miami, FL). To measure parasite load in these cultures, a minimum of 300 macrophages were counted.

Isolation of the intracellular amastigotes

For isolation of intracellular amastigotes, 12 h post infection, human macrophages were suspended in chilled PBS containing 1 mM EDTA and 11 mM Glucose and passed five times through a 27-gauge needle. Cellular debris was removed by centrifugation (60 g for 5 min at 4°C) and the supernatant passed through a 3 µm pore filter. Amastigotes were recovered by centrifugation of the filtrate (800 g for 10 min) (Naderer and McConville, 2002).

Isolation and fractionation of mitochondria from axenic or intracellular amastigotes

Cell fractionation was performed as described previously (David et al., 2006; Gannavaram et al., 2008). Briefly, amastigotes are washed three times in 15 ml of MES buffer (20 mM MOPS, pH 7.0, 250 mM sucrose, 3 mM EDTA). The cell pellet was resuspended in 0.2 ml of MES buffer containing 1 mg ml−1 digitonin and protease inhibitor cocktail (Roche Applied Science). The suspension was incubated at room temperature for 5 min and centrifuged at 10 000 g for 5 min. The resulting supernatant was collected as a cytosolic fraction (CF). The mitochondrial pellet was resuspended in buffer A (10 mM KH2PO4, pH 7.4), incubated for 15 min at 4°C with gentle rocking. An equal volume of buffer A containing 32% (w/v) sucrose, 30% glycerol (v/v), 10 mM MgCl2 was added to the suspension. After 15 min incubation, the suspension was centrifuged for 10 min at 12 000 g, and the pellet (P1) and supernatant (S1) were separated. The pellet was resuspended in buffer A and incubated for 30 min at 4°C with rocking. The supernatant was centrifuged at 16 000 g for 30 min and the resulting supernatant (S2) and pellet (P2) were designated matrix (MX) and inner membrane (IM) respectively. S1 was centrifuged at 16 000 g, and the resulting pellet (P3) and supernatant (S3) were designated as outer membrane (OM) and inter membrane space (IMS) respectively. Protein concentration was determined and equal amount of protein was loaded for Western blot analysis with various antibodies. To extract protein from the inner mitochondrial membrane, the P2 fraction was incubated with 0.1 M Na2CO3 in ice for 20 min. Separation of the pellet from the released proteins was achieved by centrifugation at 16 000 g for 20 min.

Measurement of cytochrome c oxidase activity

The mitochondrial vesicles from 5 × 108 cells were isolated by hypotonic lysis as described elsewhere (Horvath et al., 2005) and lysed with 1% dodecyl maltoside. Cytochrome c oxidase (Complex IV) activity was measured as described previously with little modifications (Zikova et al., 2006). Briefly, cytochrome c oxidase activity was measured in a cuvette containing the COX buffer [40 mM sodium phosphate buffer, pH 7.4; 0.5 mM EDTA; 30 µM sodium ascorbate; 20 µM horse heart cytochrome c (Sigma); 0.005% dodecylmaltoside] to which 5 µl of mitochondrial lysates were added. The change in absorbance was measured every 10 s over a period of 2 min. Antimycin was added to the final concentration of 300 ng ml−1 as an inhibitor of the interfering reductase activity. The unit activity was defined as the amount of enzyme that catalyses the oxidation of 1 µmol of cytochrome c per min, assuming an extinction coefficient of 21.1 mM−1 cm−1. Total protein concentration of mitochondrial lysates was determined by the Bradford assay and activity was expressed as mU per mg protein.

Measurement of ATP synthesis

ATP production was measured as described previously (Allemann and Schneider, 2000). Briefly, a crude mitochondria preparation from amastigotes isolated from infected macrophages was obtained by digitonin extraction (Tan et al., 2002). ATP production was measured in the presence of a 5 mM solution of the indicated substrates (Succinate, Pyruvate, α-ketoglutarate) and 67 µM ADP. Inhibitors were pre-incubated with mitochondria on ice for 10 min and used at the following concentration: 6.7 mM malonate, 33 µg ml−1 attractyloside. The concentration of ATP was determined by a luminometer using the ATP Bioluminescence kit CLS II (Roche Applied Science, IN).

Homologous recombination construct for the Ldp27 gene

The final construct contained a fragment from the 5′ region flanking the Ldp27 open reading frame, followed by the coding sequence for the neomycin resistance gene, followed by a fragment of the 3′ flanking region of the Ldp27 gene. The plasmid was assembled starting with a gene targeting plasmid containing neo previously described for Centrin (Selvapandiyan et al., 2004). The 5′ flanking fragment was amplified by PCR from L. donovani genomic DNA with a forward primer, 5′-ACTGGATCCAGGTGGAGGTCACGGGT-3′, that adds a BamHI restriction site, and a reverse primer, 5′-TCAGTCGACTGGTTCAGTGATGGATTC-3′, that adds a SalI restriction site. The targeting plasmid was digested with BamHI/SalI, gel purified and the 5′ flanking fragment of Ldp27 ligated into it. The 3′ flanking fragment was amplified by PCR from L. donovani genomic DNA with a forward primer, 5′-TTGACTAGTGCTTCGGCGACACAGACAGG-3′, that adds a SpeI restriction site, and a reverse primer, 5′-ACTGGTACCAGAGACGGAGACAGAGAGCACTAC-3′, that adds a KpnI restriction site. The plasmid containing the 5′ flanking fragment was digested with SpeI/KpnI, gel purified and the 3′ flanking fragment ligated into it. The final plasmid was confirmed by DNA sequencing.

The targeting construct was prepared for transfection by digestion with BamHI and HindIII which cuts out a linear fragment containing the p27 5′ sequence, the neo gene, the p27 3′ sequence and 6 bp beyond the KpnI site at the 3′ side of the p27 sequence. The fragment was gel purified and transfected.

Transfection and selection of Ldp27 null mutant parasites

Mid-log-phase amastigotes (2–3 × 107 cells ml−1) were harvested by centrifugation at 3000 g for 10 min at 4°C. The cell pellets were washed in ice-cold PBS and electroporated with the DNA construct using conditions described previously (Goyard et al., 2003). For clonal selection the transfected amastigotes were incubated overnight in Leishmania growth medium. The cultures are centrifuged as above and the pellet was resuspended in 5 ml of growth medium containing G418 (Geneticin; Invitrogen) 20 µg ml−1. Drug-resistant amastigotes were transformed into promastigotes as described (Debrabant et al., 2004). The drug-resistant promastigotes were cloned by limiting dilution and with 100 µg ml−1 of G418 concentration. A homozygous mutant was recovered from the heterozygote by a loss of heterozygosity approach (Goyard et al., 2003). We confirmed the deletion of both the alleles of p27 by Southern blot.

Complementation of Ldp27 in Ldp27−/− parasites

To restore p27 expression in the Ldp27−/− parasites, the Ldp27 ORF was first PCR amplified using a Ldp27 containing plasmid as template and the following forward primer: 5′-GATGGATCCATGTCTCGTTGCACGAAC-3′, and reverse primer: 5′-GATGGATCCTTACGCGTAGTCCGGCAC-3′, each containing a BamHI restriction site (bold). The amplified product was initially cloned at the T/A cloning site of pCRII-TOPO cloning vector. The fidelity of the cloned sequence was verified by nucleotide sequencing. The BamHI insert was ligated into the BamHI site of the pXG-PHLEO vector (Freedman and Beverley, 1993) and the recombinant plasmid, pXG-PHLEO-Ldp27, was transfected into the Ldp27−/− promastigotes as described previously (Selvapandiyan et al., 2001). Transfected promastigotes were selected with minimal doses of phleomycin (Sigma) (10 µg ml−1). The selected cell pool was designated ‘Ldp27−/− AB’ for p27 deleted, added back.

Isolation of genomic DNA and Southern blot analysis

Total genomic DNA was isolated from either promastigotes or axenic amastigotes according to the methods described in the manual for GENOME DNA isolation kit from Promega Biosciences (San Luis Obispo, CA). The DNA was digested with restriction endonucleases PvuI, SphI and HindIII and separated on 1% agarose gels. Southern blot analysis of electrophoresed DNA was done as described previously using 32P-labelled Ldp27 coding sequence or Ld tubulin coding sequence as probes (Selvapandiyan et al., 2004).

Mouse infection

In independent experiments, the mice were inoculated via tail vein with 3 × 106 metacyclic cells of either L. donovani wild-type, Ldp27−/− or Ldp27−/− AB parasites. Infective-stage metacyclic promastigotes of L. donovani were isolated from stationary cultures by density gradient centrifugation as described (Selvapandiyan et al., 2009). Five weeks and 13 weeks post infection, all the mice were sacrificed and parasite burdens of liver and spleen were measured by the serial dilution method as previously described (Selvapandiyan et al., 2009). As an additional confirmation of the presence of parasites in tissues, total DNA samples obtained from infected mouse livers and spleens were used as templates in a Taqman-based real-time PCR. The amplification target was the kinetoplast minicircle DNA of the parasite. The primers and methods were as previously described (Selvapandiyan et al., 2008), with the addition of a fluorescent probe for detection. The probe had the sequence 5′-RAAARKKVRTRCAGAAAYCCCGT-3′. A Black Hole Quencher moiety is coupled to the 3′ end and Calfluor Red is coupled to a C6 linker at the 5′ end. The degenerate letter code is according to the Nomenclature Committee of the International Union of Biochemistry (http://www.chem.qmul.ac.uk/iubmb/misc/naseq.html) (Selvapandiyan et al., 2009).

Statistical analysis

Student's t-test was employed to assess the significance of the differences between the mean values of control and experimental groups. A P-value of < 0.05 was considered significant.

Ancillary