br Materials and methods br

Materials and methods
Results
Discussion We confirmed the published cDNA and protein sequence of feline CSF-1R and stably cloned the cat receptor into the Ba/F3 cell line. Activation of feline CSF-1R by recombinant human CSF-1 was previously tested by Woolford et al. [14]. Expression of oncogenic forms of feline CSF-1R in Rat-2 cells, which ordinarily are unable to grow in soft agar [51], permitted the formation of colonies in soft agar upon the addition of human CSF-1 [14]. However, the native receptor was insufficient to cause transformation even in the presence of human CSF-1. This study also gave no insight into relative efficacy. In the current study, the addition of either human or mouse CSF-1 to Ba/F3 cells expressing full length feline CSF-1R produced survival and proliferation of cells that are dependent on CSF-1 for survival. Mouse CSF-1 was substantially less active on the feline receptor than human CSF-1. In this respect, the cat CSF-1R is idiosyncratic compared to the other species tested. Mouse and pig CSF-1R respond equally to mouse, pig and human CSF-1, human CSF-1R responds to human and pig, but not to mouse CSF-1 [2]. Publication of the mouse CSF-1: CSF-1R contact WZ4003 [48] permits a structure-based analysis of receptor specificity. There are 6 non-conserved contact amino acids (His6, Asn13, Phe55, Glu78, Arg79, and Asn85 in mouse) between mouse and human CSF-1 [2]. His6 is conserved in cat, and is Asn in pigs, but is the bulky aromatic amino acid, Tyr, in humans. This difference could explain why mouse CSF-1 can bind the pig and cat CSF-1R, but not the human receptor. Amino acids 78 and 79 of CSF-1 are Glu and Arg (positive) in mice, Val and Thr in cat (neutral), and Val and Gln (neutral) in both human and porcine CSF-1. These differences could explain the relatively lower efficacy of mouse CSF-1 on the feline CSF-1R. Both human and mouse IL-34 were also able to activate the feline CSF-1R in vitro; again the mouse ligand was less active. IL-34 is more species-specific than CSF-1 as identified by the reduced activity of human IL-34 on the mouse CSF-1R [8]. Human IL-34 and CSF-1 were similarly active on the pig CSF-1R [2] but with the cat receptor the EC50 for human IL-34 was substantially higher than for CSF-1. Again, the differences between the species probably reflect differences in CSF-1R contact amino acids for IL-34 that may account for the reduced activity of mouse IL-34 on the cat CSF-1R. The difference that is most likely to disrupt function is Gln258 (human and cat) which is substituted with a charged Lys in the mouse CSF-1R. The growth of macrophages from bone marrow cells using recombinant CSF-1 has been described previously for human, mouse, chicken and pig bone marrow cells [7], [17], [50], [52], [53], [54]. Culture of feline macrophages is a valuable tool to allow the in vitro study of the effects of drugs, e.g. chemotherapeutic agents, which may cause immune-suppression or immune-modulation in cats. Equally, the mechanisms of feline infectious diseases that can infect macrophages, e.g. FIV or mycobacteria, may be further investigated. It will be of interest to compare the response of feline BMDMs to LPS in a similar fashion to the analysis performed for mice, human and porcine BMDMs [50], [55]. We have demonstrated previously that both recombinant human and porcine CSF-1 proteins can activate feline and canine CSF-1Rs using bone marrow aspirates [2]. The primary marrow cells used in the present study were harvested post-mortem, by repeated flushing from the femur. An earlier study Daniel et al. [56] noted that feline primary bone marrow cultures could be successfully maintained without the addition of exogenous CSF-1 due to the production of CSF-1 by the adherent bone-marrow cell population. This is a rather less-reproducible and inefficient approach. Bone-marrow cells can be frozen and thawed, so that cells preserved from a small number of animals can provide a long term resource for studies of macrophage biology.