MacMARCKS is not essential for phagocytosis in macrophages.

MacMARCKS (also known as myristoylated alanine-rich protein kinase C substrate (MARCKS)-related protein) is a member of the MARCKS family of protein kinase C substrates. MacMARCKS contains within it a basic effector domain that contains the serine residues that are phosphorylated by protein kinase C, as well as a calcium/calmodulin and actin-binding site. Two previous reports demonstrated that a macrophage cell line expressing a mutant form of MacMARCKS that lacks the effector domain is defective in phagocytosis and cell adhesion (Zhu, Z., Bao, Z., and Li, J. (1995) J. Biol. Chem. 270, 17652-17655; Li, J., Zhu, Z., and Bao, Z. (1996) J. Biol. Chem. 271, 12985-12990). We report here that macrophages from MacMARCKS null mice phagocytose and spread normally. Thus, although MacMARCKS is recruited to phagosomes, it is not absolutely required for phagocytosis.

MacMARCKS, also known as myristoylated alanine-rich protein kinase C substrate (MARCKS)-related protein, 1 is a PKC substrate that binds calcium/calmodulin and actin (reviewed in Refs. [1][2][3]. MacMARCKS plays a role in coordinating the actin cytoskeleton during such diverse processes as neural tube closure and synaptic transmission and has been proposed to play an important role in phagocytosis by macrophages (4 -10). Mac-MARCKS and the closely related family member, MARCKS, are rod shaped proteins that can be divided into three domains based on sequence homology and function (4). The amino terminus consists of an amino-terminal glycine that is myristoylated followed by a short stretch of highly conserved amino acids with unknown function called the MH2 (MARCKS homology 2) domain (4). The MH2 domain is followed by a short stretch of basic amino acids, known as the effector domain, which binds calcium/calmodulin, actin, and acidic lipids and which contains the serines that are phosphorylated by PKC (4,5,11,12). PKC-dependent phosphorylation of the effector domain of MARCKS regulates the binding of calcium/calmodulin and F-actin to it and also regulates the association of MARCKS with membranes (13)(14)(15)(16). Much less is known about the effector domain of MacMARCKS; however, the domains are struc-turally very similar, and PKC also regulates the binding of calcium/calmodulin to the MacMARCKS effector domain (4,5). The biochemical data suggest that MacMARCKS, like MARCKS, plays a role in integrating the effects of PKC and calcium on actin dynamics.
MacMARCKS has recently been suggested to play a crucial role in regulating the actin cytoskeleton during phagocytosis in macrophages; it is rapidly recruited to the forming phagosome, and expression of a mutant form of the protein lacking an effector domain was reported to completely inhibit phagocytosis in a macrophage cell line (9). A subsequent report from the same group demonstrated that macrophages expressing this mutant protein were also incapable of spreading (10). In this report we demonstrate that although MacMARCKS associates with phagosomes, its presence is not required for phagocytosis; macrophages derived from MacMARCKS null mice phagocytose zymosan normally. Macrophages from MacMARCKS null mice also spread normally. In addition, we show that in our hands, the mutant form of MacMARCKS that lacks the effector domain does not associate with phagosomes, does not displace WT MacMARCKS from phagosomes, and does not inhibit phagocytosis when transfected into a macrophage cell line.

EXPERIMENTAL PROCEDURES
Materials-Unless indicated otherwise, all reagents were obtained from Sigma.
Generation of MacMARCKS Null Macrophages-Heterozygous Mac-MARCKS Ϯ mice (7) were mated, and macrophages were prepared from fetal liver using a variation of a protocol developed by Moore and Williams (17). Briefly, fetuses were sacrificed at day 14.5 of gestation, and livers were harvested. Limbs and tails were collected at the same time for genotyping as described previously (7). The livers were dispersed by passage two times through a 26G needle, red blood cells were lysed, and the remaining cells were plated in 10-cm tissue culture dishes in Iscove's modified Dulbecco's medium (JRH Biosciences, Lenexa, KS) with 20% fetal calf serum (Hyclone, Logan, UT), 15% L cell conditioned medium, and supplemented with 1% L-glutamine, 100 units/ml penicillin G, and 100 g/ml streptomycin. Cells were cultured at 37°C in a 5% CO 2 atmosphere, and half of the medium was changed every 3 days. After 7-10 days nonadherent cells were washed away, and the remaining cells (macrophages) were scraped off the plate in ice-cold phosphate-buffered saline (PBS) and plated for subsequent experiments.
Immunofluorescence Microscopy-Macrophages were processed for immunofluorescence microscopy as described previously (18). Briefly, cells were grown on glass coverslips, stimulated with 50 ng/ml lipopolysacharide for 3 h, fixed in 10% neutral buffered formalin (Sigma), and permeabilized in acetone at Ϫ20°C. Nonspecific sites were blocked by incubation in PAB (PBS, 0.5% bovine serum albumin, and 0.05% sodium azide) containing 10% horse serum, and primary antibodies diluted in PAB with horse serum were applied to the cells for 1 h at room temperature. The cells were washed in PAB, and the appropriate sec-ondary antibodies diluted in PAB with horse serum were incubated with the cells for 1 h. After further washing in PAB, cells were mounted and observed by confocal microscopy using a MRC-1024 system (Bio-Rad) equipped with LaserSharp software and mounted on an Axiovert TV microscope (Carl Zeiss, Inc., Thornwood, NY) as described previously (18).
Phagocytosis Assays-Phagocytosis of zymosan by primary macrophages was measured as described previously (18). Zymosan particles (Sigma) in Hepes-buffered RPMI were centrifuged onto macrophage monolayers at 4°C, the cells were warmed by addition of RPMI at 37°C, and internalization was allowed to proceed. The cells were either fixed and processed for microscopy after 3 min of internalization as described above, or live cells were stained with 0.4% trypan blue in PBS after 90 min of internalization to facilitate counting of phagocytosed particles. Phagocytosis of zymosan by RAW cells was measured by centrifuging FITC-zymosan (Molecular Probes, Eugene, OR) onto macrophage monolayers as described above. After 1 h of internalization, the cells were washed, extracellular zymosan was dissolved by addition cDNA Constructs-The pcDNA 3 (Invitrogen, Carlsbad, CA) expression vectors encoding MacMARCKS and MacMARCKS ⌬-effector domain fused to HA at their carboxyl termini have been described before (20). The wild type construct encodes all 200 amino acids of Mac-MARCKS, whereas the effector domain deletion mutant excludes amino acids 86 -108. DNA for transfections was prepared using Qiagen Tip-500s according to the manufacturer's instructions (Qiagen, Santa Clarita, CA).
Macrophage Transfection-RAW 264 cells were transfected using a variation on the method described by Stacey et al. (21). Five million cells in 210 l of culture medium were placed in a 0.4-cm gap cuvette along with 5 g of plasmid DNA in 40 l of PBS. After 10 min at room temperature, the cells were electroporated using a Gene Pulser (Bio-Rad) set at 960 microfarad and 300 V. The cells were immediately transferred to 5 ml of medium, pelleted by centrifugation at 600 ϫ g, resuspended in fresh medium, and cultured for 24 h before addition of 400 g/ml G418 (Life Technologies, Inc.). After 10 days of selection, cells were cloned twice by limiting dilution, and cells expressing the desired proteins were analyzed by immunoblotting.
Immunoblotting-Whole cell extracts of RAW cells were prepared by resuspending cells in solubilization buffer (1% Triton X-100, 150 mM NaCl, 2 mM EDTA, 50 mM Tris, pH 7.4, plus 0.5 mg/ml leupeptin, 0.09 trypsin inhibitory units aprotinin, and 1 mM phenylmethylsulfonyl fluoride) and incubating the mixture on ice for 20 min. Insoluble material was pelleted by centrifugation at 12,000 ϫ g for 10 min in a microfuge, and the protein content of the supernatant was analyzed by the BCA method (Pierce). 20 g of protein from each sample were resolved by 10% SDS-polyacrylamide gel electrophoresis and transferred to Immobilon-P (Millipore, Bedford, MA). The membranes were blocked overnight at 4°C in blotto (PBS with 5% milk, 0.01% Tween 20, and 5 mM azide). Primary antibodies were diluted in blotto and applied to the membrane for 2 h at room temperature. The membranes were washed thoroughly with PBS ϩ 0.01% Tween 20 and incubated for 30 min with secondary horseradish peroxidase-conjugated antibodies (Amersham Pharmacia Biotech) diluted in blotto without azide. After further washing, specific antibody binding was detected by chemiluminescence using ECL reagents (Amersham Pharmacia Biotech). The hemagglutinin tag was detected using a polyclonal rabbit anti-HA antibody (Babco, Berkeley, CA), and MacMARCKS was detected using a MacMARCKS-specific polyclonal antiserum (6).

RESULTS AND DISCUSSION
MacMARCKS null mice die at birth due to neural tube closure defects, so macrophages from adult MacMARCKS knockout mice could not be assessed for phagocytic competence (7). To study MacMARCKS null macrophages, we cultured fetal liver-derived macrophages from 14 1 ⁄2 day MacMARCKS null embryos. Macrophages were recovered identically from wild type and null (MMKO) fetuses; cells were recovered in similar numbers and had similar morphologies (Fig. 1). Immunofluorescence microscopy and immunoblotting studies with antibodies to MARCKS and MacMARCKS confirmed that the Mac-MARCKS Ϫ/Ϫ macrophages had no detectable MacMARCKS, whereas the protein was easily detected in macrophages from wild type litter mates (7) (Fig. 1 and data not shown). The distribution and expression of MARCKS in the MacMARCKS null macrophages was unchanged; MARCKS was found diffusely distributed throughout the cells and on plasma membranes with occasional accumulation on plasma membrane ruffles ( Fig. 1 and data not shown). MacMARCKS null macrophages spread normally in response to bacterial lipopolysacharide and in response to phorbol esters ( Fig. 1 and data not  shown).
When wild type and MacMARCKS null macrophages were incubated with zymosan particles, no differences in phagocytic ability were observed. Wild type cells had a phagocytic index (particles/100 cells) of 444 Ϯ 35, whereas MacMARCKS null macrophages had a phagocytic index of 397 Ϯ 45. Because MacMARCKS was not required for internalization of zymosan, we examined more carefully the formation and maturation of the nascent phagosome. One of the earliest steps in phagocytosis is the polymerization of actin beneath a surface-bound particle. Localization of F-actin with FITC-phalloidin revealed that actin polymerized identically around nascent phagosomes in wild type and MacMARCKS null macrophages (Fig. 2) and  6, respectively). Cell extracts were subjected to SDS-polyacrylamide gel electrophoresis and transferred to Immobilon, and HA-tagged proteins (HA, upper panels) and MacMARCKS proteins (MM, lower panels) were detected by immunoblotting with specific antibodies. B, localization of HA-tagged proteins to phagosomes. RAW 264 cells expressing either the HA-tagged wild type MacMARCKS (upper panels) or the HA-tagged MM-⌬ED (lower panels) were allowed to ingest zymosan particles for 15 min, fixed, and processed for immunofluorescence microscopy as described under "Experimental Procedures." Actin was localized with rhodamine-phalloidin (left panels), and HA-tagged MacMARCKS proteins were localized with an antibody to the HA epitope (right panels).
The arrows indicate internalized particles. Talin, an actin-binding protein that colocalizes with actin during phagocytosis of zymosan (18), was also found associated with phagosomes in MacMARCKS null macrophages. MARCKS also accumulated normally around nascent phagosomes, raising the possibility that MARCKS and MacMARCKS may have interchangeable functions during phagocytosis. 90 min after particle internalization, the lysosomal membrane marker, lamp-1, was found on zymosan-containing phagosomes in wild type and MacMARCKS null macrophages, suggesting that after particle internalization and subsequent actin depolymerization, phagosome maturation proceeded normally (data not shown).
A requirement for MacMARCKS in phagocytosis and macrophage spreading has been proposed by Li and co-workers (9, 10) based on experiments using the mouse macrophage cell line, J774. In their hands, expression of a mutant form of Mac-MARCKS lacking the effector domain in J774 cells inhibited phagocytosis of zymosan and prevented cell spreading (9,10). Because these observations were at variance with our data using MacMARCKS null macrophages, we attempted to reproduce the inhibition of macrophage functions observed by Li et al. (9,10). RAW 264 cells were stably transfected with cDNAs encoding HA-tagged wild type MacMARCKS or HA-tagged MacMARCKS lacking the effector domain. The RAW 264 macrophage cell line was chosen instead of J774 cells because they have a much higher transfection frequency, thereby giving us confidence that we were not selecting for rare cells that harbored other potential defects. Immunoblots of cell extracts demonstrated that a variety of independently derived cell lines expressed HA-tagged MacMARCKS and mutant Mac-MARCKS. Two independent lines expressing each of the HAtagged proteins at approximately 2-5-fold higher levels than endogenous MacMARCKS were chosen for further study (Fig.  3A). No differences were seen in the phagocytic capacity of RAW cells expressing wild type or mutant MacMARCKS; the phagocytic indexes derived from 10,000 cells exposed to FITCzymosan are shown in Table I. Immunolocalization of actin and HA-tagged MacMARCKS proteins to nascent phagosomes revealed that the wild type HA-tagged protein (MM-WT) accumulated together with actin around early phagosomes (Fig.  3B). In cells expressing HA-tagged MacMARCKS lacking the effector domain (MM-⌬ED), actin accumulated normally around nascent phagosomes, but the HA-tagged protein did not colocalize with the actin (Fig. 3B). In addition, endogenous, WT MacMARCKS was recruited normally to the forming phagocytic cup in cells expressing an excess of MM-⌬ED (Fig. 4), suggesting that the mutant protein does not affect the localization or function of endogenous MacMARCKS. Also, none of the cell lines expressing MM-⌬ED showed any defects in spreading (data not shown). The failure of this data to recapitulate the observations of Li and co-workers (9, 10) is puzzling. It is possible that the J774 cells expressing the mutant Mac-MARCKS protein have some other defect that arose during selection and cloning. Alternatively, it is possible that there are redundant mechanisms in RAW 264 cells that are not present in J774 cells. However, given what is known about the functional domains of MARCKS and MacMARCKS (2, 3), it is not clear how MM-⌬ED acts as a dominant negative mutant; our data suggest that it does not.
Taken together, the above data demonstrate that although MacMARCKS localizes to phagosomes, it is not absolutely required for phagocytosis in macrophages. It is possible that MacMARCKS and MARCKS have identical or overlapping functions during phagocytosis and that in the absence of Mac- FIG. 4. MM-⌬ED does not displace endogenous MacMARCKS from phagsosomes. RAW cells expressing MM-⌬ED (clone BF5.1) were allowed to ingest zymosan particles for 5 min, after which MM-⌬ED was localized using the anti-HA antibody (upper panels), and endogenous MacMARCKS (as well as the mutant) was localized using a rabbit polyclonal antiserum to MacMARCKS (lower panels) (6). F-actin was localized using rhodamine-phalloidin as described under "Experimental Procedures." MARCKS, MARCKS activity is sufficient. This possibility is currently under investigation.