In vitro derivation of macrophage from guinea pig bone marrow with human M-CSF

Journal of Immunological Methods 389 (2013) 88–94

Contents lists available at SciVerse ScienceDirect

Journal of Immunological Methods journal homepage: www.elsevier.com/locate/jim

Technical note

In vitro derivation of macrophage from guinea pig bone marrow with human M-CSF Karl O.A. Yu a,⁎, Steven A. Porcelli b, Howard A. Shuman c a b c

Section of Infectious Diseases, Department of Pediatrics, Comer Children's Hospital, University of Chicago Medical Center, Chicago, IL 60637, USA Department of Microbiology and Immunology and Department of Medicine (Rheumatology), Albert Einstein College of Medicine, Bronx, NY 10461, USA Department of Microbiology, University of Chicago, Chicago, IL 60637, USA

a r t i c l e

i n f o

Article history: Received 26 October 2012 Received in revised form 9 December 2012 Accepted 2 January 2013 Available online 16 January 2013 Keywords: Macrophage Bone marrow Guinea pig M-CSF CSF-1

a b s t r a c t The guinea pig has a storied history as a model in the study of infectious disease and immunology. Because of reproducibility of data and availability of various reagents, inbred mice have since supplanted the guinea pig as the animal model-of-choice in these fields. However, several clinically-significant microorganisms do not cause the same pathology in mice, or mice may not be susceptible to these infections. These demonstrate the utility of other animal models — either as the primary method to study a particular infection, or to confirm or refute findings in the mouse before translating basic science into clinical practice. The mononuclear phagocyte, or macrophage (Mφ), plays a key role in antigen presentation and the pathogenesis of intracellular bacteria, such as Mycobacterium tuberculosis and Legionella pneumophila. Because of variable yield and difficult extraction from tissue, the preferred method of producing Mφ for in vitro studies is to expand murine bone marrow (BM) precursors with mouse macrophage colony-stimulating factor (M-CSF). This has not been shown in the guinea pig. Here, we report the empiric observation that human M-CSF – but not mouse M-CSF, nor human granulocyte/macrophage colony-stimulating factor – can be used to induce BM precursor differentiation into bonafide Mφ. The differentiated cells appeared as enlarged adherent cells, capable of both pinocytosis and large particle phagocytosis. Furthermore, we showed that these guinea pig BM-derived Mφ, similar to human monocyte/Mφ lines but unlike most murine BM Mφ, support growth of wild type L. pneumophila. This method may prove useful for in vitro studies of Mφ in the guinea pig, as well as in the translation of results found using mouse BM-derived Mφ towards studies in human immunology and infectious disease. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Animal models remain the paradigm upon which to confirm experimental therapeutics prior to clinical use. While Abbreviations: BM, bone marrow; CFU, colony-forming units; h, hour(s); GM-CSF, granulocyte/macrophage colony-stimulating factor; LD50, median lethal dose; Mφ, macrophage; M-CSF, macrophage-colony stimulating factor; PBS, phosphate-buffered saline pH 7.4. ⁎ Corresponding author at: Section of Infectious Diseases, Department of Pediatrics, University of Chicago Medical Center, 5841 South Maryland Ave., MC 6054, Chicago, IL 60637, USA. Tel.: +1 773 702 9281; fax: +1 773 702 1196. E-mail address: [email protected] (K.O.A. Yu). 0022-1759/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jim.2013.01.005

inbred mice have dominated much non-human research, these are not the best models to use for the study of several infectious diseases (Padilla-Carlin et al., 2008). A classic example for this is infection with the intracellular gramnegative bacterium Legionella pneumophila, the etiologic agent for Legionnaires' disease and Pontiac fever. The guinea pig, Cavia porcellus, is the preferred and the more physiologic model of infection. Inhaled infection in the guinea pig leads to a diffuse, organizing pneumonia, with the animals experiencing fever, dyspnea, and weight loss. The guinea pig LD50 of 2 × 10 4 CFU is also more likely in line with epidemic aerosol infection in the human. On the other hand, most inbred mice are fairly resistant — with LD50's commonly 3 logs

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

higher. Furthermore, the two susceptible murine strains show pathology unlike that seen in humans. AKR/J mice develop hepatosplenic necrosis, whereas A/J mice develop acute pulmonic hemorrhage with little inflammation (Hedlund et al., 1979; Brieland et al., 1994). Consequently, animals other than the mouse remain important in the study of medicallyrelevant infections. With the dearth of reagents and protocols available for the guinea pig, tools relevant for this species have to be developed. The macrophage (Mφ) is the preferred host for several intracellular pathogens, including Mycobacterium tuberculosis and L. pneumophila. Because these cells are rare in peripheral blood and difficult to extract from tissue, protocols to derive Mφ from mouse BM are commonly used (Manzanero, 2012). Mouse macrophage colony-stimulating factor (M-CSF or CSF-1), commonly with conditioned medium from L929 murine fibroblasts being used as a crude source, is a reliable stimulus for Mφ expansion and differentiation for in vitro studies (Lee and Wong, 1980). Murine granulocyte/macrophage colony-stimulating factor (GM-CSF) has also been shown to generate granulocytes and dendritic cells as suspension cells in culture, and Mφ as the adherent cells (Inaba et al., 1992). No analogous protocols have been validated for use with the guinea pig. This study is an attempt to define such a method for this species. A search of the literature yielded descriptions of isolation of primary guinea pig BM Mφ (without expansion) by adherence onto glass slides in Leighton tubes, as well a report of expansion of fibroblasts from guinea pig BM when plated without exogenous stimulation (Friedenstein et al., 1970; Brade et al., 1982). Recombinant human GM-CSF with conditioned medium from the guinea pig lung fibroblast line JH4 has also been used to generate putative dendritic cells from guinea pig BM (Hiromatsu et al., 2002). The residual adherent cells were reported by the authors as Mφ, but this was not confirmed further. In this study, we have treated guinea pig BM with various cytokines, and have found human M-CSF, but not its murine analog, to be of promising use for the expansion and derivation of Mφ from guinea pig BM. As monoclonal antibodies against the guinea pig are not as well defined as they are for mice or humans, we have relied on morphologic and functional assays to characterize these as bonafide Mφ.

2. Materials and methods 2.1. Animals and reagents Female outbred Hartley guinea pigs (400–800 g) were obtained from Charles River Laboratories (Wilmington, MA). Female C57Bl/6 mice (6–15 weeks old) were obtained from Jackson Laboratories (Bar Harbor, ME). Animals were housed in specific pathogen-free facilities, and were handled as approved by the institutional animal care and use committee. Euthanasia was by CO2 asphyxiation. Recombinant human M-CSF and GM-CSF were obtained from Peprotech (Rocky Hill, NJ). Per the manufacturer, the human M-CSF used had a specific activity of ≥ 10 6 units/mg, as measured by proliferation by the cytokine-dependent murine myeloid leukemia line M-NSF-60.

89

The murine fibroblast line L929 (NCTC clone 929, originally from strain L), the guinea pig lung fibroblast line JH4, and the human monocytoid leukemia line THP-1 were obtained from the American Tissue Culture Collection (Manassas, VA). THP-1 monocytes were differentiated to a Mφ phenotype with 30 nM phorbol 12-myristate 13-acetate (Sigma-Aldrich, St. Louis, MO; stock dissolved in dimethylsulfoxide) for 48 h prior to use. Cells were generally maintained in advanced RPMI-1640 supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, 0.1 mM nonessential amino acids, 55 μM 2-mercaptoethanol, 100 U/ml penicillin, and 100 μg/ml streptomycin (Gibco, Grand Island, NY). Fibroblast lines were maintained in similarly-supplemented DMEM. Conditioned media from L929 or JH4 cells consisted of 7-day supernatants of confluent cultures; these were aliquoted and stored in −20 °C, and sterile-filtered prior to use. These were used at 30% final concentration. Antibiotics were omitted for infection experiments. Cells were grown at 37 °C in humidified air with 5% CO2. The mouse anti-guinea pig monocyte/Mφ monoclonal antibody MR-1 (IgG1) and mouse anti-human CD11b (clone ICRF44, IgG1) were obtained from Serotec (Raleigh, NC). Antibody MAC387, which is a mouse IgG1 that stains human Mφ, is specific for L-calprotectin, and cross reacts to guinea pig Mφ, was obtained from Pierce/Thermo Scientific (Rockford, IL). Ovalbumin-specific mouse IgG1 antibody from clone F2-3.58, used as isotype control, was obtained from the Fitch Monoclonal Antibody Facility at the University of Chicago. L. pneumophila strain JR32 is a wild type, sodium-sensitive variant derived from the serotype 1 strain Philadelphia-1, whereas the dotA mutant (strain LELA 3118) is a transposoninduced mutant that is incompetent for intracellular replication due to disruption of the bacteria's type IV secretion system (Sadosky et al., 1993). Strain AL4 is a JR32-derived mutant with a kanamycin-resistance cassette inserted into the flaA gene encoding for flagellin (Shuman and Levi, unpublished data). Bacteria were grown on CYE plates, which is comprised of 10 g/l N-(2-acetamido)-2-aminoethanesulfonic acid (Sigma-Aldrich; adjusted to pH 6.9 with KOH), 10 g/l yeast extract, 15 g/l agar, 2 g/l activated charcoal, 2.3 mM L-cysteine, and 0.62 mM Fe(NO3)3·9H2O. flaA was grown on CYE plates supplemented with 50 μg/ml kanamycin. Previously frozen stocks were grown to stationary phase in humidified room air at 37 °C for 48 h prior to use. Inoculum was estimated by turbidity, with an OD600 of 1 ml −1 cm −1 representing 1.4× 109 CFU/ml, as confirmed by colony counts. Saccharomyces cerevisiae strain W303 (clone DKB3213, Mat a haploid) was a gift from J. Grub and T. Bishop (University of Chicago). This strain was maintained in YPD plates (10 g/l yeast extract, 20 g/l each of peptone, D-glucose, and agar), and grown in room air at 30 °C. For preparation of yeast particles, 72 h cultures of yeast cells were fixed in 4% paraformaldehyde in PBS for 1 min at room temperature with intermittent vortexing. After washing in PBS (1000 g for 10 min), the particles were stored in PBS with 0.05% NaN3 at 4 °C. Aliquots were washed twice in PBS prior to re-suspension in culture medium prior to use. These particles were confirmed to be killed by plating. 2.2. Expansion of BM cells Muscle was removed from femora and tibia by abrasion with Kimwipes. The bones were then immersed for 1 min in

90

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

70% ethanol, and then air-dried. After the bones were cut open, the marrow was flushed out with 5–10 ml medium by syringe, and mechanically dispersed. After one wash (400 g, 5 min), and passing the cells through a 40 μm nylon filter, the cells were aliquoted and frozen in complete RPMI supplemented with 50% fetal bovine serum and 10% dimethylsulfoxide in − 80 °C until use. Frozen BM cells were quick-thawed in a 37 °C water bath, and after washing in PBS, were plated at various densities in 3 ml medium with cytokines per well in tissue culturetreated 6-well plates (Costar, Corning, NY). The cell yield post-thaw was about 4–10 × 10 7 intact marrowcytes per guinea pig by trypan-blue exclusion. The cytokines used included human M-CSF or GM-CSF (typically at 20 ng/ml), or L929 cell-conditioned medium used as a source of murine M-CSF. As previously described, human GM-CSF was used with 30% JH4 cell-conditioned medium. Every 3 days, half of the culture medium was removed, spun down, and replaced with fresh medium with the relevant cytokines together with the recovered cells. Most experiments were done with cells at 7–10 days of culture.

permeable by addition of saponin at 0.5% (w/v) final concentration for 20 min at 4 °C. Cells were then stained and washed as above, with all steps until the last wash conducted in the presence of 0.5% saponin.

2.3. Analysis

3. Results and discussion

In various days of culture, cells were collected for counting. Non-adherent cells were removed, and adherent cells were collected by incubation of the washed cell monolayer with a proprietary trypsin-like solution (TypLE Express, Gibco), used 1:1 with PBS, at 37 °C. Suspension and adherent cells were enumerated separately in an improved Neubauer chamber, and distinction was made in counting small and large leukocytes. Photographs were taken under phase-contrast microscopy (Eclipse TE200, Nikon Instruments, Melville, NY).

3.1. Human M-CSF, but not murine M-CSF, promotes differentiation of Mφ-like cells from guinea pig BM in vitro

2.4. Pinocytosis and phagocytosis Cells were plated at 10 5 cells in 1 ml of medium in 24-well tissue culture treated plates overnight. To this was added India ink to a final concentration of 1:1000 (BD Biosciences, Franklin Lakes, NJ; the ink preparation was previously diluted 10-fold with and dialyzed against PBS to remove the phenol preservative), or prepared S. cerevisiae at 10 6 particles per ml. Cells were incubated for 2–20 h, then the supernatants were carefully removed and the cell monolayer washed with medium prior to phase contrast microscopy. 2.5. Flow cytometry Cells were stained by standard methods. Cells were blocked in PBS with 0.1% bovine albumin, 1% normal goat serum (Southern Biotech, Birmingham, AL) and 0.05% NaN3. This was followed by staining with primary antibody (either at 10 μg/ml in blocking buffer or half-strength supernatant) for 45 min at 4 °C. Cells were then washed with PBS (1000 g for 5 min). Phycoerythrin-labeled goat anti-mouse Ig(H + L) (Southern Biotech; used at 1:200 in blocking buffer) was added for 30 min at 4 °C. After a final wash in PBS, flow cytometry data were collected on a FACSCalibur (BD Biosciences), and analyzed using FlowJo (Tree Star, Ashland, OR). For staining of intracellular antigens, cells were first fixed in 4% paraformaldehyde in PBS (10 min at 4 °C), after which cells were made

2.6. Infection with L. pneumophila Cells (approximately 5 × 10 5) were plated overnight in 1 ml of supplemented, antibiotic-free RPMI per well in 24-well tissue culture-treated plates. For derived cells, the appropriate cytokine stimuli were included, as these were found to be needed for extended cell survival. Cells were then infected with different strains of L. pneumophila at 100 CFU/ well. Plates were centrifuged 800 ×g for 10 min to synchronize infection. Supernatants (~ 25 μl aliquots) were sampled periodically and serially diluted 1:5 in PBS. Five microliter drops were plated and dried onto CYE plates for quantitation. Plates were scored after 72–96 h of growth in humidified air at 37 °C, when individual Legionella colonies are observed. The minimum quantifiable bacterial burden in this setup, assuming 100% recovery, is hence 200 CFU/ml.

In attempting to adapt the standard murine protocols for derivation of BM Mφ to the guinea pig, there were a few issues to consider. As the commonly-used strains of guinea pigs are all outbred, experiments would likely necessitate the use of a number of animals to ensure generalizability of data. Hence, we preferred that the use of cryopreserved BM be validated, as well as development of straightforward expansion of BM-derived cells, without the need for precursor enrichment or depletion methods. Pilot experiments with human GM-CSF showed that small numbers of Mφ-like cells and fibroblasts can be grown from guinea pig BM. While murine BM Mφ are conventionally grown on non-treated plastic Petri dishes because of their strong adherence properties, we opted to continue the experiments with tissue culturetreated dishes, as the cells elicited from guinea pig BM were fairly easy to detach by enzymatic treatment, and the cell morphology was difficult to interpret on untreated plastic dishes — especially in differentiating fibroblasts from adherent monocytoid cells. We compared various cytokines in their ability to promote differentiation of guinea pig BM into Mφ. As previously reported, L929 cell supernatant – commonly used as a source of murine M-CSF – led to a several-log expansion from mouse BM of strongly adherent Mφ from day 3 of culture onwards, with cells easily growing beyond confluence. Human M-CSF induced differentiation of a smaller number of murine BM Mφ, whereas human GM-CSF induced a mixture of adherent and non-adherent cells (data not shown). The latter appears similar to the reported generation of granulocytes, dendritic cells, and Mφ when mouse BM is treated with murine GM-CSF (Inaba et al., 1992). In contrast, guinea pig BM yielded small numbers of Mφ-like cells in response to mouse M-CSF or human GM-CSF (Fig. 1A–C). These cells appeared monocytoid with spread-out

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

projections — then beyond 7 days some fibroblast contamination was observed. This yield of putative Mφ was only minimally increased over BM incubated in medium without exogenous cytokines. Human M-CSF, on the other hand, reproducibly induced production of Mφ-like cells (Fig. 1D). In situ, these were enlarged, adherent cells typically comprising of a main cell body of 15–20 μm wide, with a few pseudopodia-like projections ranging 10–30 μm in length. When grown on non-tissue culture-treated dishes, the cells have a more spread-out phenotype. On suspension after trypsinization, they appeared as mononuclear cells about 10 μm in diameter. They appeared by 3–5 days of culture; although for one animal, these cells did not appear until day 8. There was substantial inter-animal variation in the yield of these cells. This may be explained by the outbred genetic background of the animals. For most cases, an optimal yield appeared at days 7–9, with observable fibroblast contamination by day 12 onwards in a minority of cultures. Fibroblasts appeared as spread-out cells > 100 μm in width when grown in tissue-culture plastic. We observed an apparent optimal precursor plating density. The highest yields were about 5 × 10 5 Mφ derived from 2 × 10 5− 1 × 10 6 BM cells per 3 ml well. Of note, the number of Mφ per primary BM cell varied inversely with plating density (Fig. 2). This phenomenon suggests inhibition of Mφ differentiation by other endogenous cell products, a limiting amount of human M-CSF, or a requirement for sufficient “floor area” for Mφ precursor adherence. This is not the case for conventional expansion of Mφ from inbred mouse BM with the common protocols.

A

B

D

E

91

Adherent cells were detachable from the tissue culture plate with 10–15 min of incubation with half-strength TrypLEExpress. This stands in contrast with the BM Mφ derived from C57Bl/6, which in our hands could be enzymatically detached only when grown on bacteriologic plastic Petri dishes, and required mechanical dislodgement when grown in tissue culture-treated plates. 3.2. Human M-CSF-derived guinea pig BM Mφ exhibit phagocytosis and pinocytosis To confirm that the guinea pig BM-derived adherent cells were indeed Mφ, we performed functional studies demonstrating their phagocytic potential. We compared guinea pig BM-derived adherent cells with murine BM Mφ, with Mφ derived by differentiating THP-1 cells (a human monocytoid leukemia line), and with the guinea pig fibroblast line JH4 as a negative control. India ink (colloidal carbon) is efficiently pinocytosed by Mφ, while larger particles such as yeast or bacteria are phagocytosed by a number of mechanisms, including by receptors for Fc, complement, or mannose (Treves et al., 1976). We found that guinea pig BM-derived adherent cells efficiently uptake colloidal carbon by the 2 h time point in all cells (Fig. 3B). Similar pinocytosis was seen with Mφ derived from murine BM or from THP-1 cells. Phagocytosis of yeast particles was also demonstrated of guinea pig BM-derived Mφ (Fig. 3C). This was similar to phagocytosis seen with murine BM-derived Mφ, with most Mφ from either species having multiple visible intracellular S. cerevisiae particles within 2 h

C

Mϕ - like other cells

medium Hu GM-CSF Mo M-CSF Hu M-CSF 0

1

2 3 4 5 x 105 live cells / well

100 μ m 10 μm

Fig. 1. Phase contrast photomicrography of 6-day cultures of guinea pig BM cells incubated with (A) medium, (B) 20 ng/ml human GM-CSF with 30% JH4 fibroblast conditioned medium, (C) mouse M-CSF in the form of 30% L929 cell supernatant, and (D) 20 ng/ml human M-CSF. In this experiment, marrowcytes were plated at 106 per well in a 6-well tissue culture plate. (E) Quantitation of Mφ-like cells (large, adherent cells) versus cells of other morphology from these cultures. Representative results from at least 3 experiments.

92

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

Mϕ yield per 106 BM cells

Mϕ yield per well 10

7

106

105

104

105

106

107

105

106

107

BM cells plated per well Fig. 2. Quantitation of Mφ derived from guinea pig BM cells incubated with human M-CSF. Left, Mφ yield per well graphed against initial well inoculum. Right, Mφ yield per well, normalized against 106 inoculated BM cells per well. Linear regression of log/log transformed data drawn on the graph on the right. Slope −0.86, Pearson r2 = 0.54, p b 0.0001. Identical symbols indicate separate cultures from BM from the same animal. These data represent results from separate cultures from five guinea pigs.

of co-incubation. Phagocytosed particles were difficult to visualize with THP-1 Mφ; these were absent in JH4 fibroblasts. Interestingly, effective yeast phagocytosis by THP-1-derived Mφ is seen in the presence of non-heat-inactivated human serum (data not shown). This may suggest a requirement for complement- or antibody-mediated facilitation of ingestion of these particles by this particular cell line, in variance to BM-derived Mφ. For further confirmation of the phenotype of these putative guinea pig BM Mφ, we tested these cells using various monoclonal antibodies that may bind guinea pig Mφ. In our experiments, antibodies ICRF44 (mouse anti-human CD11b/ Mac-1) and MAC387 (anti-human Mφ, against the S100 family member calprotectin/MRP14) did not exhibit staining of these cells (data not shown). Of interest is clone MR-1. This monoclonal antibody was described for use in the immunohistochemistry against an as-yet unidentified intracellular antigen of blood monocytes and Mφ from lymph node, spleen, liver, Peyer's patches, and corticomedullary thymus in this animal, but not of Langerhans cells (Kraal et al., 1988). This antibody showed intracellular staining of our cells with a unimodal distribution (Fig. 3D). This staining was consistently above the isotype control, and was absent from murine BM Mφ and THP-1 cells (data not shown).

Fig. 3. Phenotype of guinea pig BM-derived Mφ. Cells were incubated with medium, colloidal carbon, or killed S. cerevisiae particles for 2 h, then washed prior to photomicrography (A–C, respectively). (D) Histogram of intracellular staining for flow cytometry of guinea pig BM Mφ by the Mφ-specific monoclonal antibody MR-1 (bold line) versus isotype control (thin line).

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

These data, together with the morphologic descriptions in Section 3.1, suggest that the cells derived from guinea pig BM using human M-CSF were bonafide Mφ.

3.3. Human M-CSF-derived guinea pig BM Mφ support growth of L. pneumophila To test if guinea pig BM-derived Mφ may be useful for the study of infectious disease, we examined if these cells support growth of the intracellular pathogen L. pneumophila. The guinea pig is particularly important in the study of this bacterium, since this species represents a susceptible pulmonic infection model for Legionnaires' disease, whereas most mice are fairly resistant to infection. It has been previously reported that L. pneumophila infects guinea pig and hamster peritoneal Mφ in vitro, exhibits delayed infection of rat Mφ, and does not easily infect Mφ from most mouse strains (Yoshida and Mizuguchi, 1986). Furthermore, this pathogen has been demonstrated to infect human Mφ cell lines and primary alveolar Mφ (Nash et al., 1984; Marra et al., 1990).

93

We infected guinea pig BM-derived Mφ with L. pneumophila with a multiplicity of infection of ~0.0002. As with THP-1 Mφ, there was a 4- to 5-log increase in bacterial burden and gravid cytopathic effects — supporting evidence for Mφ-tropic infection and replication by this pathogen. The intracellular growthdeficient mutant, dotA, which has a defect in the dot/icm type IV secretion system, does not grow in either cell type (Fig. 4). As expected, the guinea pig fibroblast cell line JH4 (data not shown), and Mφ derived from C57Bl/6 murine BM did not sustain robust L. pneumophila growth. The apparent doubling time of L. pneumophila in the susceptible Mφ – presumably a composite of intracellular infection, multiplication, cytolysis, and then re-infection – is 3–4 h. The bacteria die within 3 days in RPMI medium without antibiotics in the absence of relevant cell hosts (data not shown). Of interest are results with infection with the non-motile flagellin mutant flaA. Genetic studies have shown that the susceptibility of A/J mice and derived Mφ to Legionella infection, in contrast to the inherent resistance of other strains, such as C57Bl/6, is due to the gene Naip5/Birc1e (neuronal apoptosis inhibitory protein 5/baculoviral IAP repeat-containing 1e), a member of the nucleotide-binding domain/C-terminal leucine

Fig. 4. In vitro support of growth of Legionella pneumophila. THP-1 Mφ, or BM-derived Mφ from either guinea pig or C57Bl/6 mice (5 × 105/well) were infected with 100 CFU wild type L. pneumophila strain JR32, or mutants dotA or flaA. Supernatants were then serially tested for bacterial content at various times post-infection. Cell types as listed. Cultures with no demonstrable bacterial colonies were plotted at 100 CFU/ml. Representative results of at least 2 experiments for THP-1- and guinea pig BM-derived Mφ.

94

K.O.A. Yu et al. / Journal of Immunological Methods 389 (2013) 88–94

rich repeats (NBD/LRR) or NOD-like receptor (NLR) protein superfamily. A/J mice express lower levels of Naip5 protein, and have several amino acid polymorphisms when compared to C57Bl/6 mice. Flagellin-deficient Legionella mutants replicate well in C57Bl/6 BM Mφ — suggesting that recognition of this protein is required for the innate immunity of most mouse stains against L. pneumophila (Ren et al., 2006). Indeed, mouse Naip5 has been demonstrated as an inflammasome-associated receptor for the C-terminal region of flagellin (Lightfield et al., 2008). This, however, was not the case for Mφ derived from human THP-1 cells, nor from guinea pig BM. The growth curves of wild type JR32 and mutant flaA coincided (Fig. 4). These data stand in contrast with the published results with C57Bl/6, as well as data on our hands, which show that flagellin mutants replicate > 2 logs more efficiently in C57Bl/6 Mφ than do wild type bacteria. The relative defect of flagellin signaling through human, guinea pig, and A/J mouse Naip5 may be a strong contributor to the pathogenesis of L. pneumophila infection. This may be worth exploring as target for immunomodulatory chemotherapy in the setting of certain bacterial infections. 4. Concluding remarks The data presented here propose a straightforward method for moderate-yield derivation of Mφ from guinea pig BM using recombinant human M-CSF. With optimal plating conditions – even with use of cryopreserved BM – the yield would be about 1–2.5 × 10 8 Mφ per animal. This is on-par with expected yields with exudative Mφ elicited with intraperitoneal mineral oil-induced inflammation, but affords flexibility to the investigator — with the possibility of being able to work with frozen BM from multiple animals in parallel. In the mouse, peritoneal exudate Mφ exhibit proteomic patterns distinct from BM-derived Mφ or their further differentiated forms — and, in fact, appear closer to that of dendritic cells (Becker et al., 2012). If these findings extend to the guinea pig, then our BM-derived Mφ may represent an opportunity to investigate the phenotype of more naïve Mφ, as opposed to the more activated Mφ derived by peritoneal inflammation. It is difficult to speculate as to the difference between BM-derived Mφ and primary alveolar Mφ, as direct comparisons of these two populations in the mouse are sparse in the literature. While alveolar Mφ may represent the more physiologic cell host for pulmonic pathogens, the higher yield and more flexible setup of BM-derived Mφ may prove more helpful for in vitro experiments. We have also shown that these BM-derived Mφ may work as a model for intracellular infection by a clinically-relevant pathogen. For Legionella, these cells recapitulate the expected phenotype with human Mφ, based on experiments with human Mφ cell lines and from alveolar Mφ. This method should aid in the study of host/pathogen interactions in the guinea pig, with a better potential of translation to the human. Acknowledgments We thank L. Zitzow (Department of Surgery, University of Chicago) and L. Degenstein (Transgenic Core, University of

Chicago) for generous provision of guinea pigs and C57Bl/6 mice. S. cerevisiae was a kind gift of T. Bishop and J. Grub. S. Matushek (University of Chicago Medical Center Clinical Microbiology Laboratory) kindly provided India ink. We appreciate extended helpful discussions with several members of the laboratory, and thank A. Punnoose and A. Jongco for critique of the manuscript. K.Y. is supported by the Infectious Diseases Fellowship Program of the Department of Pediatrics at Comer Children's Hospital. This work was supported by the Department of Microbiology at the University of Chicago. References Becker, L., Liu, N.C., Averill, M.M., Yuan, W., Pamir, N., Peng, Y., Irwin, A.D., Fu, X., Bornfeldt, K.E., Heinecke, J.W., 2012. Unique proteomic signatures distinguish macrophages and dendritic cells. PLoS One 7, e33297. Brade, V., Diesselhoff-dem Dulk, M.M., van Furth, R., 1982. Isolation and characterisation of mononuclear phagocytes from the bone marrow, blood, and peritoneal cavity of the guinea-pig. J. Pathol. 137, 139. Brieland, J., Freeman, P., Kunkel, R., Chrisp, C., Hurley, M., Fantone, J., Engleberg, C., 1994. Replicative Legionella pneumophila lung infection in intratracheally inoculated A/J mice. A murine model of human Legionnaires' disease. Am. J. Pathol. 145, 1537. Friedenstein, A.J., Chailakhjan, R.K., Lalykina, K.S., 1970. The development of fibroblast colonies in monolayer cultures of guinea-pig bone marrow and spleen cells. Cell Tissue Kinet. 3, 393. Hedlund, K.W., McGann, V.G., Copeland, D.S., Little, S.F., Allen, R.G., 1979. Immunologic protection against the Legionnaires' disease bacterium in the AKR/J mouse. Ann. Intern. Med. 90, 676. Hiromatsu, K., Dascher, C.C., Sugita, M., Gingrich-Baker, C., Behar, S.M., LeClair, K.P., Brenner, M.B., Porcelli, S.A., 2002. Characterization of guinea-pig group 1 CD1 proteins. Immunology 106, 159. Inaba, K., Inaba, M., Romani, N., Aya, H., Deguchi, M., Ikehara, S., Muramatsu, S., Steinman, R.M., 1992. Generation of large numbers of dendritic cells from mouse bone marrow cultures supplemented with granulocyte/ macrophage colony-stimulating factor. J. Exp. Med. 176, 1693. Kraal, G., Shiamatey-Koolma, R., Hoffer, M., Baker, D., Scheper, R., 1988. Histochemical identification of guinea-pig macrophages by monoclonal antibody MR-1. Immunology 65, 523. Lee, K.C., Wong, M., 1980. Functional heterogeneity of culture-grown bone marrow-derived macrophages. I. Antigen presenting function. J. Immunol. 125, 86. Lightfield, K.L., Persson, J., Brubaker, S.W., Witte, C.E., von Moltke, J., Dunipace, E.A., Henry, T., Sun, Y.H., Cado, D., Dietrich, W.F., Monack, D.M., Tsolis, R.M., Vance, R.E., 2008. Critical function for Naip5 in inflammasome activation by a conserved carboxy-terminal domain of flagellin. Nat. Immunol. 9, 1171. Manzanero, S., 2012. Generation of mouse bone marrow-derived macrophages. Methods Mol. Biol. 844, 177. Marra, A., Horwitz, M.A., Shuman, H.A., 1990. The HL-60 model for the interaction of human macrophages with the Legionnaires' disease bacterium. J. Immunol. 144, 2738. Nash, T.W., Libby, D.M., Horwitz, M.A., 1984. Interaction between the Legionnaires' disease bacterium (Legionella pneumophila) and human alveolar macrophages. Influence of antibody, lymphokines, and hydrocortisone. J. Clin. Invest. 74, 771. Padilla-Carlin, D.J., McMurray, D.N., Hickey, A.J., 2008. The guinea pig as a model of infectious diseases. Comp. Med. 58, 324. Ren, T., Zamboni, D.S., Roy, C.R., Dietrich, W.F., Vance, R.E., 2006. Flagellindeficient Legionella mutants evade caspase-1- and Naip5-mediated macrophage immunity. PLoS Pathog. 2, e18. Sadosky, A.B., Wiater, L.A., Shuman, H.A., 1993. Identification of Legionella pneumophila genes required for growth within and killing of human macrophages. Infect. Immun. 61, 5361. Treves, A.J., Feldman, M., Kaplan, H.S., 1976. Primary cultures of human spleen macrophages in vitro. J. Immunol. Methods 13, 279. Yoshida, S., Mizuguchi, Y., 1986. Multiplication of Legionella pneumophila Philadelphia-1 in cultured peritoneal macrophages and its correlation to susceptibility of animals. Can. J. Microbiol. 32, 438.