Enhancement of drug delivery to bone: characterization of human tissue-nonspecific alkaline phosphatase tagged with an acidic oligopeptide

Abstract

Hypophosphatasia is caused by deficiency of activity of the tissue-nonspecific alkaline phosphatase (TNSALP), resulting in a defect of bone mineralization. Enzyme replacement therapy (ERT) with partially purified plasma enzyme was attempted but with little clinical improvement. Attaining clinical effectiveness with ERT for hypophosphatasia may require delivering functional TNSALP enzyme to bone. We tagged the C-terminal-anchorless TNSALP enzyme with an acidic oligopeptide (a six or eight residue stretch of L-Asp), and compared the biochemical properties of the purified tagged and untagged enzymes derived from Chinese hamster ovary cell lines. The specific activities of the purified enzymes tagged with the acidic oligopeptide were the same as the untagged enzyme. In vitro affinity experiments showed the tagged enzymes had 30-fold higher affinity for hydroxyapatite than the untagged enzyme. Lectin affinity chromatography for carbohydrate structure showed little difference among the three enzymes. Biodistribution pattern from single infusion of the fluorescence-labeled enzymes into mice showed delayed clearance from the plasma up to 18 h post infusion and the amount of tagged enzyme retained in bone was 4-fold greater than that of the untagged enzyme. In vitro mineralization assays with the bone marrow from a hypophosphatasia patient using each of the three enzymes in the presence of high concentrations of pyrophosphate provided evidence of bone mineralization. These results show the anchorless enzymes tagged with an acidic oligopeptide are delivered efficiently to bone and function bioactively in bone mineralization, at least in vitro. They suggest potential advantages for use of these tagged enzymes in ERT for hypophosphatasia, which should be explored.

Fig. 1

Transient expression of rhTNSALP, CD6-TNSALP, and CD8-TNSALP and SDS–PAGE of the purified enzymes. (A) The cDNA of each enzyme subcloned in pCXN vector was transfected into CHO-K1 cells with lipofectamine. At 24 and 48h after transfection, the culture medium and cells were separately collected, and the enzyme activity was determined. (B) The purified enzymes (0.2 μg) were subjected to SDS–PAGE under reducing condition and stained with silver. A single band appeared in all the three enzymes. The molecular mass of the untagged rhTNSALP (lane 1) was approximately 80 kDa, while those of CD6- and CD8-TNSALP were larger (lanes 2 and 3, respectively).

Fig. 2

Concentration-dependent binding curves of the three enzymes to hydroxyapatite. The purified enzyme was mixed with the hydroxyapatite suspension at a final concentration of 1.0, 2.5, 5.0, and 10.0μg/ml. The mixture was mixed at 37°C for 1 h, and centrifuged at 14,000× rpm for 10 min to separate the unbound and bound enzymes. To determine the amount of the unbound enzyme, the enzyme activity in supernatant was measured, and the amount of the bound enzyme was determined from the amount of both total and unbound enzymes. The affinity to hydroxyapatite for the tagged enzymes was 30-fold higher than that for the untagged enzyme and the binding to hydroxyapatite was seen even at low concentration of the tagged enzyme.

Fig. 3

ConA affinity chromatography of the three enzymes. The rhTNSALP (A), CD6-TNSALP (B), and CD8 TNSALP (C) enzymes were applied to the ConA column. After washing the column, two fractions were eluted by the two different concentrations, 0.01 M (arrow; a) and 0.5 M (arrow; b) of αMM. There was no difference in the elution profile among the three enzymes.

Fig. 4

WGA affinity chromatography and SDS –PAGE of three enzymes before and after neuraminidasc digestion. (A) Three enzymes before (a –c) and after (d–f) neuraminidase digestion were applied to the WGA affinity chromatography. The rhTNSALP (a and d), CD6-TNSALP (b and e), and CD8-TNSALP (c and f) enzymes were applied to the WGA column. After washing the column, two fractions were eluted by the two different concentrations. 0.1 M (arrow; a) and 0.5 M (arrow; b) of GlcNAc. (B) The enzymes (0.3 μg) were subjected to SDS –PAGE under reducing condition and stained with silver. A single band was observed at all the lanes. After the treatment with neuraminidase, the molecular mass of the three enzymes decreased in a similar proportion.

Fig. 5

Biodistribution of fluorescence-labeled enzyme to bone. The three fluorescence-labeled enzymes were infused into mice by tail vein injections at a dose of 1 mg/kg of body weight. At the indicated time points, the legs were dissected and sectioned. The sections of legs were observed under a fluorescent microscopy to evaluate the enzyme distribution at the epiphyseal region. Three enzymes were distributed to the mineralized region, but not to the growth plate, gp, growth plate; m, mineralized region (A). The average of the relative areas of fluorescence from three fields of the fluorescent images at epiphyseal region was quantitated. *p < 0.05 in comparison with the untagged enzyme. **p < 0.01 in comparison with the untagged enzyme (B).

Fig. 6

Biodistribution of fluorescence-labeled enzyme to liver. The fluorescence-labeled three enzymes were infused to mice from tail vein at the dose of 1 mg/kg of body weight. At 6 h after the infusion, the livers were dissected and sectioned. The sections of legs were observed under a fluorescent microscopy. The distribution patterns in these tissues were comparable among three enzymes.

Fig. 7

Blood clearance of rhTNSALP, CD6-TNSALP, and CD8-TNSALP. The three purified enzymes were intravenously administered to 1-month-old mice at a dose of 2 U/g body weight. Plasma was obtained from infraorbital vein at the indicated time points, and the remaining enzyme activities in plasma were determined.