Down-regulation of MARCKS-related Protein (MRP) in Macrophages Infected with Leishmania *

Leishmania, a protozoan parasite of macrophages, has been shown to interfere with host cell signal transduction pathways including protein kinase C (PKC)-dependent signaling. Myristoylated alanine-rich C kinase substrate (MARCKS) and MARCKS-related protein (MRP, MacMARCKS) are PKC substrates in diverse cell types. MARCKS and MRP are thought to regulate the actin network and thereby participate in cellular responses involving cytoskeletal rearrangement. Because MRP is a major PKC substrate in macrophages, we examined its expression in response to infection by Leishmania. Activation of murine macrophages by cytokines increased MRP expression as determined by Western blot analysis. Infection with Leishmania promastigotes at the time of activation or up to 48 h postactivation strongly decreased MRP levels. Leishmania-dependent MRP depletion was confirmed by [3H]myristate labeling and by immunofluorescence microscopy. All species or strains ofLeishmania parasites tested, including lipophosphoglycan-deficient Leishmania majorL119, decreased MRP levels. MRP depletion was not obtained with other phagocytic stimuli including zymosan, latex beads, or heat-killedStreptococcus mitis, a Gram-positive bacterium. Experiments with [3H]myristate labeled proteins revealed the appearance of lower molecular weight fragments inLeishmania-infected cells suggesting that MRP depletion may be due to proteolytic degradation.

The ability of various intracellular pathogens including Leishmania to inhibit macrophage effector activities, also termed "deactivation", is well documented (1,2). Functional alterations in Leishmania-infected macrophages include decreases in cytokine production, oxidative burst activity, antigen presentation, and expression of major histocompatibility complex class II genes in response to interferon (IFN) 1 -␥. One mechanism of deactivation is indirect, involving induction of autoinhibitory molecules. In addition, there is evidence for direct interference of Leishmania with macrophage signal transduction pathways including inhibition of signaling through Janus kinases and Stat1 (3), or alterations in stimulus-induced intracellular calcium gradients related to decreased production of inositol 1,4,5-trisphosphate (4). Leishmania also inhibits protein kinase C (PKC)-dependent signaling in host macrophages as evidenced by alterations in PKC translocation and activity (5) and decreased expression of the transcriptional regulatory protein c-fos (6). Some of these effects may be ascribed to the properties of lipophosphoglycan (LPG), the major surface glycoconjugate of Leishmania, which has been shown to inhibit macrophage PKC-dependent signaling (7) as well as the activity of purified PKC in vitro (8). Thus, phagocytosis of LPG-coated beads inhibited phosphorylation of both a PKC-specific substrate peptide and myristoylated alanine-rich C kinase substrate (MARCKS), an endogenous PKC substrate in murine macrophages (9). Furthermore, depletion of PKC rendered macrophages more permissive for the proliferation of intracellular Leishmania suggesting that PKC-dependent events might contribute to parasite destruction (9).
MARCKS and MARCKS-related protein (MRP), also known as MacMARCKS (Macrophage-MARCKS), are members of a highly acidic myristoylated family of PKC substrates widely distributed in diverse cell types including macrophages (10,11). Phosphorylation of MARCKS proteins following activation of PKC has been observed in fibroblasts (12,13), macrophages (14) and neutrophils (15). Both proteins are essential for brain development and survival as shown by mice deficient in the genes macs or mrp (16,17).
MARCKS has been shown to cross-link actin filaments in vitro (18). In macrophages, MARCKS colocalizes with actin, vinculin, and talin at the site of attachment of the cytoskeleton to the plasma membrane (19,20). MRP colocalizes with paxillin at membrane ruffles at the leading edge of spreading macrophages, suggesting that it also associates with the actin cytoskeleton (21). Consequently, MARCKS and MRP are thought to regulate the actin cytoskeleton and thereby participate in major cellular responses such as phagocytosis, secretion, motility, mitogenesis, and membrane trafficking.
Expression of MARCKS and MRP is strongly up-regulated in macrophages stimulated with bacterial lipopolysaccharide (LPS) (22) or zymosan (23). LPS stimulation increases MRP steady state mRNA levels 30-fold in murine macrophages, and high levels persist for more than 8 h (22). MARCKS mRNA and protein expression can be decreased in fibroblasts through either PKC-dependent or -independent pathways by a post-transcriptional mechanism (24,25). MARCKS concentrations may also be regulated by specific proteolytic cleavage of the unphosphorylated protein by a cysteine protease (26,27), which has recently been identified as cathepsin B (28). To our knowledge, no reports concerning the down-regulation of MRP are avail-able. Inasmuch as MRP is a major PKC substrate in macrophages, we have examined the expression of MRP in response to infection with Leishmania promastigotes. Our finding that Leishmania infection markedly depresses MRP levels may provide an important mechanism for regulating PKC-dependent effector function in macrophages.

EXPERIMENTAL PROCEDURES
Mice-CBA/J mice were purchased from Harlan (Horst, The Netherlands) and were used between 8 and 16 weeks of age.
Macrophage Cultures and Activation-Bone marrow-derived macrophages were obtained by in vitro differentiation of bone marrow precursor cells as described previously (32). Briefly, cells flushed from mouse tibia and femurs were grown in DMEM with 20% horse serum (Life Technologies, Inc.) and 30% L cell-conditioned medium. Day 10 -11 macrophages were detached by pipetting, suspended in DMEM and 10% fetal bovine serum, and distributed in 35-mm tissue culture dishes (3 ϫ 10 6 macrophages/dish) or in 24-well cell culture plates (5 ϫ 10 5 macrophages/well), each well containing a round sterile glass coverslip. After 24 h, macrophages were washed and stimulated with IFN-␥ and/or TNF-␣ or LPS in the presence or absence of Leishmania (5 parasites/macrophage unless indicated otherwise). To quantitate phagocytosis of Leishmania, coverslips were removed 24 h after infection, rinsed with phosphate-buffered saline (PBS), fixed and stained with Diff-Quick (Mertz and Dade, Dü dingen, Switzerland) according to the manufacturer's instructions.
Nitrite Determination-After 24 h of macrophage activation, 100 l of supernatants were harvested for nitrite determination (33). Macrophage supernatants were mixed with an equal volume of Griess reagent and incubated for 10 min at room temperature. Absorbance was measured at 550 nm in a micro-enzyme-linked immunosorbent assay reader (Dynatech MR5000) using a 690-nm reference filter. NO 2 Ϫ concentration (M) was determined using NaNO 2 as a standard.
Radiolabeling of MRP-Aliquots of [9,10-3 H]myristic acid (Amersham, Zurich Switzerland, 53 Ci/mmol) in ethanol were dried under a stream of nitrogen gas, dissolved in dimethyl sulfoxide (Me 2 SO) and diluted in DMEM containing 10% fetal bovine serum. Macrophages cultured in 35-mm tissue culture plates (3 ϫ 10 6 cells/plate) as described above were washed and stimulated with IFN-␥ ϩ TNF-␣ for 4 h. Medium was then aspirated, and 1 ml of fresh medium containing IFN-␥ ϩ TNF-␣ and 50 Ci of [ 3 H]myristic acid in Me 2 SO (final concentration 0.4%) or Me 2 SO alone in the presence or absence of LV39 promastigotes (15 ϫ 10 6 /plate) was added. After 6 h, macrophages were washed 3 times with PBS, and cell lysates were prepared as described below.
Preparation of Macrophage Lysates-For routine Western blot analysis, macrophages were washed three times with PBS and detached with ice-cold PBS containing 5 g/ml pepstatin, 1 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin and 10 g/ml aprotinin. Samples were sonicated and total cellular protein was measured by the microbicinchoninic acid assay (Pierce). Laemmli sample buffer was then added, and samples were placed in a 100°C heat block for 5 min. For some experiments, parallel samples were prepared by adding heated SDS sample buffer directly to the tissue culture plates. Similar results were obtained for lysates prepared by these 2 protocols (not shown). For experiments involving radiolabeled macrophages, washed cells were lysed in PBS containing the same protease inhibitors and 0.5% (v/v) Triton X-100. After determination of total cellular protein, macrophage lysates were heated to 100°C for 5 min and centrifuged at 11,000 ϫ g for 5 min at 4°C to obtain a heat-stable protein fraction (22). Supernatants were collected, and aliquots for SDS-PAGE were prepared in Laemmli sample buffer as described above.

SDS-PAGE and Western blot Analysis-For
Western blot analysis of total cell lysates, equal amounts of protein (30 g) were electrophoresed in a 12% polyacrylamide gel, electroblotted to nitrocellulose, and probed with a polyclonal rabbit antibody recognizing murine MRP. The anti-MRP antibody was raised against purified recombinant unmyristoylated MRP (34) by injection of 30 g of protein in complete Freund's adjuvant followed by four subsequent injections of 30 g in incomplete Freund's adjuvant. The anti-MRP antibody recognizes both myristoylated and unmyristoylated MRP as well as MRP phosphorylated in vitro by the catalytic subunit of PKC (35). 2 Anti-MRP was used as a 1:2000 dilution of serum followed by a 1:5000 dilution of horseradish peroxidase-conjugated goat anti-rabbit IgG. Immunoreactive MRP was detected with supersignal chemiluminescent substrate (Pierce). Films exposed to chemiluminescent blots were scanned on a ScanJet 4c/T densitometer (Hewlett Packard, Geneva, Switzerland) using the Adobe Photoshop software package (Adobe Systems, Inc., Mountain View, CA) and NIH image 1.60 software (NIH Division of Computer Research and Technology). For experiments with radiolabeled macrophages, aliquots containing the equivalent of 100 g of total cellular protein were electrophoresed in 12% polyacylamide gels. Gels containing radiolabeled protein were treated with 0.13 M salicylic acid in 10% v/v methanol, pH 7.0, dried, and exposed to x-ray film at Ϫ70°C. Fluorographs were scanned on the ScanJet 4c/T densitometer.
Immunofluorescence Microscopy-Immunofluorescence studies were performed using a polyclonal rabbit antibody (prepared by Eurogentec, Seraing, Belgium) directed against a synthetic peptide containing the 15 C-terminal amino acids of murine MRP preceded by the 21 amino acid tetanus toxoid P30 helper epitope (36). Immune serum recognizing murine MRP in enzyme-linked immunosorbent assay and Western blot analyses (not shown) was affinity purified by HiTrap N-hydroxysuccinimide-activated affinity column chromatography (Pharmacia LKB Biotechnology, Uppsala, Sweden). Control or Leishmania-infected macrophages were cultured on glass coverslips with or without IFN-␥ ϩ TNF-␣ as described above. After 24 h, macrophages were washed with medium without serum, fixed, and permeabilized with ice-cold methanol for 1 min, dried, and frozen at Ϫ20°C. Cells were rehydrated with cold PBS and incubated for 30 min with 1% bovine serum albumin in PBS at room temperaure before staining. Coverslips were then incubated with affinity purified rabbit anti-MRP antibody for 1 h at room temperature followed by a 50 min incubation with fluorescein-conjugated AffiniPure donkey anti-rabbit IgG (HϩL) (Jackson ImmunoResearch Laboratories, Westgrove, PA). Antibodies were diluted in PBS containing 1% bovine serum albumin. Coverslips were mounted in Citifluor (Kent Scientific and Industrial Projects, UK) and stored at 4°C. Microscopy was performed using a Zeiss Axioskop microscope fitted with a 100x Plan Neofluar objective.

Down-regulation of MRP Expression by Leishmania-Mu-
rine macrophages activated with IFN-␥ ϩ TNF-␣ produce high levels of nitric oxide (NO) and are capable of killing intracellular Leishmania (37). We examined the expression of MRP in normal and activated macrophages by Western blot analysis of total cell lysates. Because of its acidic amino acid composition, MRP, whose calculated molecular mass is 20 kDa, exhibits anomalous migration on SDS gels and is recognized as a 42-kDa doublet in Western blots (38,39). As shown in panels A and B of Fig. 1, IFN-␥ ϩ TNF-␣ increased the level of immunoreactive MRP protein after 4 h of culture (lane 2) though much stronger induction was observed after 24 h (lane 6). A comparison with known amounts of recombinant murine MRP (lanes 9 and 10) indicates that MRP is present at a concentration of approximately 1 ng/g total protein in macrophages activated with IFN-␥ ϩ TNF-␣. To determine whether infection by Leishmania modulates MRP expression, macrophages were challenged with LV39 promastigotes at the same time as stim-ulation with IFN-␥ ϩ TNF-␣. Under these conditions, a strong decrease in MRP levels was consistently observed either 4 or 24 h after infection (Fig. 1, A and B, lanes 4 and 8). In many experiments, Leishmania also decreased the level of MRP in control unstimulated macrophages (Fig. 1B, lane 3; Fig. 4, lane 3, below and data not shown). As shown in Fig. 1C, a strong increase in MRP was also observed when macrophages were stimulated with TNF-␣ (lane 4) or LPS (lane 5) alone, and LV39 inhibited such induction (lanes 7 and 8). LV39 also inhibited the weak induction of MRP obtained with IFN-␥ alone (lanes 3 and 6).
We then examined whether it was possible to reduce MRP levels by challenging macrophages with LV39 at various times after addition of IFN-␥ ϩ TNF-␣. As shown in Fig. 2 As an alternative proof that MRP levels were decreased in Leishmania-infected cells, the incorporation of [ 3 H]myristic acid was examined. MRP expression was first induced for 4 h with IFN-␥ ϩ TNF-␣, followed by the addition of fresh medium containing [ 3 H]myristic acid in the presence or absence of LV39 promastigotes. After an additional 6 h, heat-stable fractions of total cell lysates were prepared and subjected to SDS-PAGE. Fluorography revealed three major proteins, a 74 -78-kDa protein, most probably MARCKS, an uncharacterized protein of approximately 48 -50 kDa (designated p50), and a broad 42-46-kDa doublet corresponding to MRP (Fig. 3). Western blot analyses, performed in parallel on the myristic acid-labeled lysates, confirmed the identity of MARCKS and MRP (data not shown). In agreement with data presented above, myristoylated MRP levels were increased upon cytokine activation and decreased in Leishmania-infected macrophages. Interestingly, MARCKS levels were also strongly decreased in Leishmaniainfected cells. Although little or no induction of MARCKS expression was observed in macrophages stimulated with IFN-␥ ϩ TNF-␣, it should be pointed out that constitutive levels of MARCKS are generally higher and induction of MARCKS mRNA and protein is both less pronounced and occurs with more rapid kinetics when compared with MRP (22,39). Expression of the third heat-stable protein, p50, was increased by cytokine stimulation but, unlike MRP and MARCKS, was unaffected by Leishmania. In addition to the three major bands discussed above, additional lower molecular weight bands were observed for the samples from infected macrophages (lanes 3 and 4) possibly representing degradation products of MRP and/or MARCKS (see "Discussion").
Comparison of Different Species of Leishmania-Several additional species of Leishmania were then compared with LV39 for their effects on MRP levels. Because LPG may be responsi-

FIG. 3. Leishmania infection decreases the levels of [ 3 H]myristate-labeled MRP and MARCKS in murine macrophages.
Macrophages were cultured with IFN-␥ (50 units/ml) plus TNF-␣ (250 ng/ml) or medium alone for 4 h. Medium were then aspirated and fresh medium containing [ 3 H]myristic acid (50 Ci/ml) plus or minus the initial concentrations of IFN-␥ ϩ TNF-␣ added in the presence or absence of LV39 promastigotes (5 per macrophage). Incorporation of labeled myristate was demonstrated by SDS-PAGE and fluorography of heat-stable macrophage proteins. Panel A, fluorograph of SDS-PAGE exposed for 3 days; panel B, the fluorograph was scanned and integrated optical density is presented as relative changes in MARCKS, p50, or MRP with the DMEM controls considered as 100%.
ble for certain inhibitory effects of Leishmania, the LPG-deficient L. major strain L119 (30) was also tested. All parasites decreased MRP levels in macrophages stimulated with IFN-␥ ϩ TNF-␣ albeit to somewhat different degrees. L119 strongly decreased MRP levels (Fig. 4B) as did L. mexicana and another L. major strain IR75 (not shown). However, L. donovani was consistently less potent in down-regulating MRP in either 4-or 24-h lysates (Fig. 4B). Because we and others (40 -42) have shown that Leishmania strongly up-regulates NO production by murine macrophages, NO 2 Ϫ release was determined in parallel. A very similar enhancement of NO production was observed regardless of which Leishmania was used (Fig. 4A).
As shown in Table I, infection measured by microscopic examination of coverslips 24 h after parasite challenge was lower for both L119 and L. donovani than for LV39. However, 4 h after infection, parasite loads for L119 and L. donovani were equal to or greater than for LV39. Both L119 and L. donovani were rapidly destroyed by host macrophages even in the absence of cytokine activation (Table I). LV39 persists in the presence or absence of cytokines for up to 24 h (Table I) (37) but is subsequently eliminated by 48 -72 h of culture (37).
Effect of Other Phagocytic Stimuli-The effect of other phagocytic stimuli on MRP levels was also examined. Both zymosan and latex beads were previously shown to augment cytokinedependent NO production similar to Leishmania (40), and these results were confirmed as shown in Fig. 5A. As shown in Fig. 5B, neither latex beads (lane 4) nor zymosan (lane 9) markedly reduced the levels of MRP observed in activated macrophages (lanes 2 and 7) unlike the strong inhibition obtained with LV39 ( lanes 5 and 13). Indeed, zymosan increased MRP levels when added alone (lane 8) in agreement with a previous report by Aderem et al. (23). Addition of LV39 together with zymosan (lane 10) resulted in lower levels of MRP induction than obtained with zymosan alone (lane 8). Like zymosan, the Gram-positive bacterium S. mitis up-regulated NO production and MRP expression (lanes 14 and 15).
Immunofluorescence Microscopy Studies of MRP Expression-MRP was localized in murine macrophages by indirect immunofluorescence using an affinity-purified anti-C-terminal peptide antibody, which recognizes a single 42-kDa doublet in Western blots of macrophage lysates (data not shown). Strong punctate staining of MRP was observed in the cytosol of acti-vated macrophages (Fig. 6). In agreement with Western blot analyses, staining was much less intense in nonactivated macrophages or in macrophages infected with Leishmania. No staining was observed with a control rabbit IgG or in the absence of primary antibody (not shown). DISCUSSION Recently, in vitro studies have demonstrated that Leishmania is capable of interfering with host macrophage signal transduction machinery (1, 2) thereby modifying the capacity of this cell to combat infection. One well studied effect of Leishmania involves inhibition of macrophage PKC activity and consequently PKC-dependent cell function. Results presented here suggest that Leishmania might also regulate PKC-dependent cell function in a more selective fashion by decreasing levels of MRP, a major PKC substrate in macrophages. Addition of Leishmania promastigotes to macrophages strongly reduced levels of cytokine-induced MRP as early as 4 h after infection. To date, all species or strains of Leishmania promastigotes tested were capable of down-regulating MRP levels in response to IFN-␥ ϩ TNF-␣. This effect did not require viable parasites as heat-killed (15 min, 56°C) promastigotes exhibited comparable activity (data not shown). Other phagocytic stimuli including yeast cell wall zymosan, latex beads, or heat-killed S. mitis had either no effect or increased MRP levels by themselves. Interestingly, the LPG-deficient strain L119 was as efficient as LV39 suggesting that LPG is not responsible for the effect of Leishmania infection on MRP. Moreover 10 or 25 M purified LPG from L. donovani (kind gift of S. Turco) had no Ϫ release in 24-h supernatants is shown in panel A. MRP levels at 24 or 4 h determined by Western blot analysis of total cell lysates are shown in panels B and C, respectively. Note because MRP is found at lower levels 4 h after activation compared with 24 h, the blot in panel C was exposed for a longer period as evident from the stronger rMRP signal. inhibitory effect on macrophage MRP expression in two independent experiments (data not shown). The reason for the less pronounced down-regulation of MRP observed with L. donovani is unknown. However, it appears unlikely that decreased inhibition is due entirely to a more rapid parasite clearance from macrophage cultures because L119, which was as effective as LV39 in depleting MRP, was also efficiently killed by nonactivated macrophages.
It is highly unlikely that the observed down-regulation of MRP in Leishmania-infected macrophages reflects an overall inhibition of protein synthesis or cell function for several reasons. Nitrocellulose blots stained with Ponceau red showed no significant differences in lanes containing lysates from infected versus noninfected macrophages (not shown). Secondly, the expression of an uncharacterized 50-kDa heat-stable myristoylated protein was unaffected by Leishmania infection. Third, we have previously shown that Leishmania increases bone marrow macrophage synthesis of TNF-␣ and prostaglandin E2 in an identical experimental system (40). Finally, as shown previously (40 -42) and confirmed here, phagocytosis of Leishmania strongly up-regulates the synthesis of NO.
We considered the possibility that down-regulation of MRP resulted from an effect of Leishmania on TNF-␣ receptor expression. However, similar results were obtained with other stimuli capable of up-regulating MRP levels including LPS and zymosan. Moreover, other markers of macrophage activation such as NO production or TNF-␣ synthesis (40) are enhanced under the same conditions. MRP levels in activated macrophages were also dramatically decreased when parasites were added 24 or 48 h after stimulation.
As mentioned above, examination of myristic acid incorporation revealed the presence of a 48 -50-kDa protein (designated as p50 in our studies) in macrophages stimulated by IFN-␥ ϩ TNF-␣. The identity of this protein remains unknown though at least two groups (39,43) have previously described myristoylated macrophage proteins of comparable size. Although p50 levels were similar in normal and infected macrophages, the same studies suggested a profound effect of Leishmania on the levels of MARCKS, a PKC substrate closely related to MRP. Further studies are now in progress to examine Leishmaniadependent modulation of MARCKS expression.
Although MRP has been shown to be induced at the transcriptional level by LPS (22), there are no reports concerning factors capable of down-regulating its expression. Down-regulation of MARCKS in fibroblasts can occur through a posttranscriptional decrease in MARCKS mRNA upon incubation with bombesin or platelet-derived growth factor (24). Downregulation could be mimicked by short term treatment with phorbol esters and was inhibited by PKC depletion. Somewhat paradoxically, Spizz and Blackshear (28) showed that PKC-dependent phosphorylation of MARCKS protects the protein from another down-regulatory pathway involving proteolysis by lysosomal cathepsin B. They speculated that targeting of MARCKS to the lysosomal membrane via a putative LAMP1specific sequence might permit the interaction of cytosolic MARCKS and the lysosomal enzyme. That similar mechanisms might be involved in the regulation of MRP levels is suggested by our observations that the disappearance of radiolabeled MARCKS proteins in Leishmania-infected macrophages correlates with the appearance of lower molecular weight species. Moreover, we recently demonstrated that rMRP is rapidly cleaved by LV39 lysates or by purified Leishmania surface metalloprotease, leishmanolysin, in a cell-free in vitro assay. 3 It remains to be determined if this proteolytic event occurs within the macrophage and, if so, how a Leishmania enzyme, which is presumably restricted to the phagosomal/phagolysosomal compartment might interact with a cytosolic protein such as MRP. In this regard, a recent report by Rittig et al. (44) provided intriguing evidence that some intracellular promastigotes of L. major are localized in the cytosol of infected macrophages.
The implications of MRP down-regulation during Leishmania infection are purely speculative for the time being. It has been proposed that down-regulation of PKC might favor parasite survival (9). Decreasing the expression of a given PKC substrate could represent an important mechanism for inhibiting specific PKC-dependent effector functions in the macrophage. Evidence of functional alterations in fetal cells from animals lacking MARCKS family proteins or from cell lines expressing incomplete or dominant-negative mutants of MRP or MARCKS is somewhat contradictory (45). In a recent investigation, Underhill et al. (46) reported that MRP is not essential for phagocytosis by macrophages. However, the authors speculated that due to their high effector domain homology, MRP and MARCKS might play overlapping roles explaining the normal phagocytic phenotype of MRP-deficient cells. It is, thus, particularly interesting that Leishmania infection appears to decrease levels of both MARCKS proteins in macrophages. Our data, taken together with the previously documented inhibitory effect of LPG on PKC activity, further establish the ability of Leishmania parasites to circumvent normal PKC-dependent function in macrophages.
Finally, we recently showed that peptides corresponding to the effector domain of MARCKS and MRP induce polymerization of monomeric actin and bundling of filamentous actin 4 in contrast to comparatively moderate effects found with the intact MARCKS and MRP proteins (18). 5 We postulated that in vivo proteolysis might facilitate the interaction between MARCKS proteins and actin by exposing their effector domain. Thus it is interesting to speculate that Leishmania-dependent degradation of MRP might in some way modulate the structure and function of the actin cytoskeleton in infected macrophages.