Type 5 acid phosphatase is a lysosomal enzyme expressed in cells of monocyte/macrophage lineage frequently used as a marker of osteoclastic differentiation. Oligonucleotide primers for DNA amplification were designed following sequence alignment of rat bone and human macrophage type 5 acid phosphatases. DNA (330 bp in length) obtained using these primers and reverse transcribed total cell RNA from in vitro generated murine osteoclastic cells was cloned and sequenced. DNA sequence analysis of two clones demonstrates that the amplified material was 91% and 96% identical to rat bone type 5 acid phosphatase at the nucleotide and amino acid level, respectively. Northern blots of murine tissue RNA show the presence of 1.5‐kb transcripts that are most highly expressed in the long bones. Total cell RNA from the osteoclastic cells contain a marked level of type 5 acid phosphatase mRNA when compared to the levels seen in the tissue samples. Additionally, osteoclastic cell RNA contains two additional transcripts of 2.5 and 5 kb. Bone marrow macrophages grown in the presence of M‐CSF express low levels of the 1.5‐kb transcript with no signal observed for either of the two larger transcripts that were seen in the osteoclastic RNA samples. Importantly, bone marrow macrophage 1.5‐kb type 5 acid phosphatase transcript levels are increased by interleukin 4 treatment in both a time and concentration‐dependent manner. These findings indicate that type 5 acid phosphatase, while a cytochemical marker for osteoclasts, can be induced in macrophages by agents that block in vitro osteoclastic differentiation. Increased type 5 acid phosphatase may play a role in interleukin 4‐stimulated monocyte activities.

Osteoclasts degrade bone matrix, which is mainly type I collagen and hydroxyapatite, in an acidic extracellular compartment. Thus we reasoned that osteoclasts must produce an acid collagenase. We purified this enzyme, a 31 kDa protein, from avian osteoclast lysates (in 100 mM acetate/1 mM CHAPS/1 mM dithiothreitol, pH 4.4), fractionated by (NH2)2SO4 precipitation, gelatin-affinity, cation exchange, and gel filtration. Fraction activity was measured using diazotized collagen or 3H-labelled cross-linked collagen (decalcified and trypsin-treated metabolically L-[4,5-3H]proline-labelled bone) as substrates. Iodoacetate, leupeptin, antipain, pepstatin and mercurials inhibited collagenolysis by the isolated proteinase; mercurial derivatives could not be re-activated by dithiothreitol. Collagen degradation was maximal at pH 4.4; purified proteinase reproduced the collagenolytic activity of cell lysates. The N-terminal amino acid sequence from the isolated protein and its CNBr degradation fragments showed sequence similarity to mammalian cathepsin Bs, and near-identity with avian liver cathepsin B. Peptide substrate specificity of the osteoclastic enzyme resembled those of mammalian cathepsin B and its avian liver counterpart, but degradation of low-molecular-mass substrates by the osteoclastic enzyme was slower, reflecting generally lower kcat. values. Further, kcat/Km varied less between arginine-containing substrates than for previously reported cathepsin Bs, indicating different substrate specificity of the osteoclast enzyme. Polyclonal antibody raised to a 25 kDa fragment of the enzyme recognized a single 31 kDa band in SDS/PAGE of osteoclast lysates blotted to poly(vinylidene difluoride), adsorbed collagenolytic activity of osteoclast lysates, and stained avian osteoclasts in tissue sections. Degenerate sense- and antisense-oligonucleotide primers, predicted from segments of primary amino acid sequence, amplified a 486 bp DNA fragment; this was cloned and sequenced. Of 162 amino acids encoded, 77% are identical with those of human cathepsin B; hybridization identified a 2.4 kb RNA in osteoclast lysates. We conclude that the major avian osteoclast collagenolytic enzyme is a cathepsin B, whose activity varies from other enzymes of its class.

We prepared proton-transporting membrane vesicles from the avian osteoclast's ruffled membrane, a specialized region of the cell surface that acidifies the bone resorption space. We demonstrated a unique conductive Cl- permeability that is charge coupled to the vesicle H(+)-ATPase and is required for acidification. Ion replacement indicated an anion selectivity of Br- approximately Cl- greater than SO4(2-) greater than NO3- approximately SCN- in supporting acidification. The anion channel blocker 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (10 microM) was a competitive inhibitor of acidification and raised the Michaelis constant for ATP of the proton pump approximately 11-fold in 120 mM KCl. Inhibition was reversed by valinomycin, which provides an alternate path for charge neutralization. The Cl- dependence of acidification was nonlinear and yielded a Hill coefficient of 3-4, showing that it is distinct from a linear Cl- dependence reported for acidification of renal cortical endosomes. The K+ ionophore valinomycin augmented H+ transport in K2SO4, and not in KCl. Dependence of Cl- transport on membrane potential was confirmed by direct measurement of 36Cl- transport. We uncoupled charge transport from proton transport with a large excess of ammonia, which had no effect on 36Cl- accumulation in vesicles, and by measuring 36Cl- accumulation in response to a membrane diffusion potential, produced with a [K+] gradient and valinomycin in the absence of ATP. These experiments demonstrate that the electrogenic proton pump of the osteoclast ruffled membrane is charge coupled to a passive Cl- permeability in the same membrane.