Tartrate-Resistant Acid Phosphatase (TRACP): A Personal Perspective

As a young hematologist in the late 1960s at the New England Medical Center in Boston, one of us (LTY) was assigned to take charge of the Cytology Section of the Hematology Laboratory. It was the section's responsibility to render a diagnosis from cytologic examination of bone marrow aspirations. In addition, research specimens from humans and animals were submitted for cytological interpretation. Although our diagnostic accuracy was very good, we soon recognized the need for more objective means of diagnosis through accurate cell identification. The section staff embarked on investigations to develop specific cell markers for cells in the hemopoietic system. We reasoned that the different lineages and cell types perform different biological functions; therefore, their biochemical constituents may reflect these specific functions. After examining and testing a wide array of special cytochemical reactions, we found a handful useful for the differentiation of various blood cell types. Among these were α-naphthyl butyrate esterase, chloroacetate esterase, and acid phosphatase.1, 2 At the time, serum acid phosphatase was recognized as a potential disease marker. Patients with prostatic cancer were known to have high levels of acid phosphatase activity in both blood and urine.3 Soon, however, it was realized that serum contained other acid phosphatases that were of nonprostatic origin. In a search to increase specificity of the biochemical tests for prostatic acid phosphatase, it was found that tartaric acid was a strong and effective inhibitor of prostatic acid phosphatase. Thus, to increase the specificity for prostate disease, acid phosphatase was measured as the tartrate-sensitive fraction. The tartrate-resistant fraction was believed to come primarily from red blood cells and was considered simply an interference.4 However, it was also noted that patients with bone diseases, particularly cancer patients with lytic bone lesions, also had increased acid phosphatase activity in the serum, and this acid phosphatase was resistant to inhibition by tartaric acid.5 Histochemically, TRACP activity was very strong in osteoclasts, and therefore, it was believed that the origin of the high TRACP in serum of patients with bone diseases may come from osteoclasts.6, 7 However, because other acid phosphatase activities, particularly from red blood cells, were also resistant to tartrate, it was difficult to distinguish osteoclastic from non-osteoclastic TRACP activity by biochemical assay. This lack of absolute specificity for bone has held back progress on the use of biochemical assay of TRACP activity for bone diseases. While investigating cytochemical methods for hemopoietic cell identification, we noted that acid phosphatase activity was strong in the leukemic cells of hairy cell leukemia and weak in normal lymphocytes or those from chronic lymphocytic leukemia.2 Therefore, acid phosphatase cytochemistry appeared to be useful as a test for differentiating hairy cell leukemia from other lymphoproliferative disorders which, at the time, was difficult. To optimize the cytochemical test for hairy cell leukemia, we investigated various substrates for acid phosphatase and discovered that when using α-naphthyl phosphate as substrate, the polymorphonuclear neutrophils had strong activity and the hairy cells had weak activity. On the other hand, when naphthol ASBI phosphate was used as substrate, the neutrophils were weak, and the hairy cells were very strong. This observation led to our hypothesis that different isoenzymes may be demonstrable by their substrate specificity and perhaps have specific cellular expression. Using non-denaturing polyacylamide gel electrophoresis and histochemical staining, Li et al.8 showed that there were at least seven identifiable acid phosphatases, some of which were selectively expressed in certain cell types. For example, band 2 was present in prostate and semen, band 4 was present in neutrophils, band 3 was present in lymphocytes and platelets, and the most cationic, band 5, was strongly expressed in hairy cells. The rapidly migrating band 5 was the only one that was resistant to tartaric acid inhibition. Other tartrate-resistant acid phosphatases present in platelets and red blood cells were not demonstrable by this gel system; hence, the nomenclature for TRACP as type-5 acid phosphatase is derived from the rapid electrophoretic mobility of this hairy cell enzyme in this gel system. Recognizing the potential of TRACP as a highly useful diagnostic marker for disease, we worked to perfect the cytochemical technique using naphthol ASBI phosphate as substrate, and it has subsequently proved to be a robust method over the years.9 Today, cytochemical testing has been largely replaced by the more objective and quantitative flow cytometry methods for cytodiagnosis. The diagnostic potential of the TRACP enzyme also stimulated us to investigate some of its fundamental biochemical properties. Lam et al.10 showed that type 5 TRACP was unique immunologically and catalytically from the other acid phosphatases. Further work disclosed that TRACP actually consisted of two related structural forms: 5a and 5b. Polyclonal antisera derived from rabbits immunized with hairy cell TRACP reacted equally with both 5a and 5b, but TRACP 5a was found to contain N-acetyl neuraminic acid. The 5a and 5b isoforms separated from serum differed too in their pH optimum; type 5a enzyme had an optimum of pH 5.2, whereas type 5b TRACP had a pH optimum of 5.8. Gel analysis showed that type 5b was elevated in normal children's serum. Type 5b TRACP was also abundant in the spleen from hairy cell leukemia; however, type 5a enzyme was abundant in the spleen from Gaucher's disease. The significance of the existence and nature of two TRACP isoforms was obscure to us at the time and remained undeveloped for years. Meanwhile, we found that TRACP activity in serum was also elevated in a high percentage of patients with breast cancer, especially those with bone metastasis. That this TRACP was probably derived from osteoclasts was well supported by the previous histochemical studies showing strong TRACP activity in osteoclasts. The potential for TRACP as a biomarker of bone resorption was long recognized, however, because RBC, platelets, and other sources of TRACP existed and contributed to total activity in serum, the specificity of the biochemical TRACP assay for bone disease was compromised and the assay was not widely accepted. Clearly, a more specific immunoassay was needed for TRACP, both for histochemical identification of hairy cells in paraffin sections of bone marrow biopsy, and for quantitative immunoassay of bone specific TRACP in serum. We originally had difficulty immunizing mice to respond with high titer antibodies to natural TRACP purified from hairy cell leukemia spleen. This was probably because of the phylogenetically conserved structure of TRACP among species and its lack of foreignness as an immunogen. After heat denaturation of pure TRACP in the presence of SDS, human TRACP became a potent antigen in mice and we were able to clone an antibody-producing hybridoma, 9C5. Antibody 9C5 was shown to be highly useful for staining TRACP in paraffin sections, provided sections were subjected to heat or enzymatic epitope retrieval.11 In addition to hairy cells, osteoclasts and some macrophages were strongly stained consistent with the known presence of TRACP activity in these cells. 9C5, however, was not useful for enzyme immunoassay techniques as it recognizes an epitope present only in denatured TRACP protein. Repeated efforts to immunize mice with increasingly large doses of TRACP finally yielded a suitable immune response and successful cloning of an additional hybridoma, 14G6, secreting antibody to native TRACP. While we were optimizing 14G6 for use in immunoassay, a number of papers appeared describing similar immunoassays for TRACP using a variety of antibodies and assay strategies. Some measured total circulating TRACP protein, while others measured immunoreactive TRACP activity.12-14 All methods showed significantly increased TRACP in patients with a variety of metabolic bone diseases; however, the reported levels, both normal and pathological, varied widely among studies. It seemed that not all assays were measuring the same thing. Discrepancies could be due to different antibody specificities, technical differences, or the presence of circulating inactive enzyme that would be detected with some assays, but not others. It was not known if any assays were selective for TRACP 5a or TRACP 5b, or whether such isoform specificities would be clinically significant. In collaboration with Halleen et al. who supplied a second anti-TRACP antibody, J1B, we developed two assays for TRACP, one for activity and one for protein.13 Using these assays simultaneously to measure TRACP in serum from patients with rheumatoid arthritis (RA) and end-stage renal disease (ESRD), we were able to show that each assay may measure different forms of TRACP and may have their own clinical utility in specific diseases. TRACP protein was elevated in about one-third of the patients with RA, whereas TRACP activity was elevated in a similar proportion of patients with ESRD. Refinement of the activity assay to measure primarily TRACP 5b at pH 6.1 has improved specificity for osteoclast activity and bone resorption rate.15 The clinical significance of the TRACP protein assay and its potential use to measure circulating TRACP 5a is the topic of another contribution in this issue.16 Our efforts to develop cell markers for blood cell identification were originally met with skepticism. Now, the use of cell and disease markers for diagnosis is often a requirement. Much current research effort is spent in the development of such markers, and TRACP 5b is no exception. With continuous refinements, the TRACP 5b activity assay is showing increasing promise as a bone resorption marker. A new kit for TRACP 5b activity assay is now available for research use. Although the TRACP enzyme is a single protein derived from a single gene, tissue specific post-translational processing may give rise to at least two isoforms; 5a and 5b. Preliminary indications are that TRACP 5a may be related to inflammation and have its own clinical and biological implications. It is gratifying to witness that our humble beginnings with a simple cytochemical stain have blossomed into a highly relevant and exciting field of research. Funding for TRACP research now supports many laboratories in their explorations of the molecular regulation, cell biology, biochemistry, and clinical applications of this key enzyme in bone physiology and immune responses. This study was supported by grants from the Research Service of the Department of Veterans Affairs, Washington, DC, and the Research Foundation, Veterans Affairs Medical Center, Louisville, KY.