RU2801121C2 - Production of alkaline phosphatases - Google Patents

Abstract

FIELD: biotechnology.

SUBSTANCE: following is described: a recombinant polypeptide obtained in a mammalian cell culture containing asphotase alpha (SEQ ID NO: 1) and having an average total sialic acid content of 0.9–3.5 mol sialic acid/mol protein monomer and another characteristic selected from the group consisting of the following: isoelectric focusing between 5.2 and 6.7, large glycan structure, as shown either on FIG. 41 or FIG. 42 is a 2-AB labeled oligosaccharide chromatogram profile as shown either on FIG. 38 or FIG. 39, MALDI-ToF fingerprint profile of the glycopeptide shown either on FIG. 40 or FIG. 44–49 and others. Also the following is provided: a pharmaceutical composition for treating a subject suffering from a condition associated with alkaline phosphatase deficiency containing the said recombinant polypeptide.

EFFECT: obtaining polypeptide with a high degree of efficiency.

25 cl, 53 dwg, 33 tbl, 10 ex

Description

Sequence list

[001] The amino acid sequences listed in the accompanying sequence listing are shown using the standard three letter code for amino acids as defined in 37 CFR 1.822. The sequence listing is provided as a Unicode text file created on August 16, 2016, named 0351 Sequence Listing.txt and 6604 bytes in size, which is incorporated herein by reference.

FIELD OF TECHNOLOGY

[002] The description of the invention relates to a method for producing a recombinant polypeptide, including: (a) providing a fed-batch bioreactor with a volume of from 100 to 25,000 liters containing (i) cells capable of expressing the recombinant asphotase alpha polypeptide (SEQ ID NO: 1) and (ii ) a culture medium suitable for such expression, wherein the culture medium contains from about 25 μM to about 300 μM zinc; (b) culturing the cells under conditions suitable for the expression of recombinant asphotase alpha; where the pH of the culture medium is from about 6.7 to about 7.1, and wherein zinc is added to said culture medium such that the concentration of zinc in the culture medium is maintained at a concentration of about 25 µM to about 300 µM zinc.

BACKGROUND OF THE INVENTION

[003] Hypophosphatasia (HPP) is a life-threatening, genetic and extremely rare metabolic disorder that results in an inability to produce functional tissue-specific alkaline phosphatase (TNSALP). This leads to an accumulation of non-mineralized bone matrix (eg, bone softening, osteomalacia) characterized by hypomineralization of bones and teeth. When growing bone is not properly mineralized, impaired growth is the result, which leads to distortion of the joints and bones. This, in turn, affects motor skills, respiratory function, and can even lead to death. Various forms of HPP have been found to include perinatal, infantile, juvenile, and adult HPP. Recently, six clinical forms have been identified, most based on age at onset of symptoms, including perinatal, benign prenatal, infantile, juvenile, adult, and odonto-HPP. Asfotase alfa is an approved, first-in-class targeted enzyme replacement therapy designed to address the problem of defective endogenous TNSALP levels. For the treatment of HPP with TNSALP, see Whyte et al., 2012 N Engl J Med. 366 :904-13.

[004] Asfotase alfa (STRENSIQ®, Alexion Pharmaceuticals, Inc.) is a soluble hybrid glycoprotein consisting of a human TNSALP catalytic domain, a human immunoglobulin G1 Fc domain, and a deca-aspartate peptide (i.e., D10) used as bone-oriented domain. In vitro, asphotase alfa binds with greater affinity for hydroxyapatite than soluble TNSALP lacking deca-aspartate peptide, allowing the TNSALP asfotase alfa fragment to effectively degrade excess local inorganic pyrophosphate (PPi) and restore normal mineralization. Pyrophosphate hydrolysis promotes bone mineralization and its effects are similar between species evaluated in non-clinical studies. Initial efficacy studies were conducted in a mouse model of HPP (Akp2 ^-/- mice). A mouse model of Akp2 ^-/- created by inactivating the TNSALP gene (Narisawa et al., 1997 Dev Dyn. 208 :432-46) shares many common features of the human condition, including the accumulation of unmineralized bone matrix.

SUMMARY

[005] This document describes compositions of alkaline phosphatases (e.g., asphotase alfa) that have specific characteristics (e.g., specific glycan structures, specific total sialic acid content (TSAC) values, etc.) and manufacturing processes used to obtain alkaline phosphatases (eg asphotase alfa) with specific characteristics. Such alkaline phosphatases (e.g., asphotase alfa) are suitable for use in therapy, e.g., for the treatment of conditions associated with decreased levels and/or functions of the alkaline phosphatase protein (e.g., insufficient digestion of inorganic pyrophosphate (PPi)) in a subject, e.g., a human.

[006] In one aspect, the present disclosure provides a method for producing a recombinant polypeptide having an alkaline phosphatase function. In various embodiments, the alkaline phosphatase function may include any alkaline phosphatase function known in the art, such as enzymatic activity towards natural substrates including phosphoethanolamine (PEA), inorganic pyrophosphate (PPi), and pyridoxal-5'-phosphate ( PLP). Such a recombinant polypeptide may contain asphotase alpha (SEQ ID NO: 1).

[007] In some embodiments of the invention, the method described herein further comprises adding zinc to said culture medium to obtain a recombinant polypeptide. Zinc can help improve the activity and/or stability of the recombinant polypeptide. In some embodiments of the invention, zinc may be added to provide a concentration of zinc from about 1 to about 300 μm in the specified culture medium. In one embodiment, zinc may be added to provide a zinc concentration of about 10 to about 150 µM (e.g., 10, 20, 30, 40, 50, 60, 70, 80, 90, 100, 110, 120, 130, 140 or 150 μM) in culture medium. In some embodiments, zinc is added to provide a zinc concentration in the culture medium of about 25 µM to about 150 µM, or about 60 µM to about 150 µM. In one particular embodiment of the invention, zinc is added to provide a zinc concentration in the culture medium of about 30, 60, or 90 μM zinc. In one particular embodiment of the invention, zinc is added to provide a zinc concentration in the culture medium of about 28 μM. In some embodiments of the invention, zinc is added to the specified culture medium in a bolus, continuously or semi-continuously.

[008] In some embodiments, the method described herein further comprises controlling the pH of said culture medium to obtain a recombinant polypeptide. For example, the pH can be set from about 6.8 to about 7.0. In one particular embodiment of the invention, the pH is set to about 6.9.

[009] In some embodiments, the method described herein further comprises adding at least one additional bolus feed to the original culture medium containing the cells during culture and/or polypeptide production. Such addition of fresh culture medium can improve the activity (eg, specific activity) of the resulting recombinant polypeptide. In one embodiment of the invention, at least one, two, three, or four bolus feeds are added to the culture medium during culture. In one particular embodiment, at least four bolus feeds are added. In some embodiments of the invention, bolus addition(s) feed improves the specific activity of the recombinant polypeptide. The cells described herein for the production of a recombinant polypeptide can be any cells (eg, mammalian cells) known in the art. In some embodiments, cells are selected from the group consisting of CHO, NSO/1, PER.C6, COS-7, human kidney embryonic line (293 or 293 cells subcloned for growth in suspension culture), BHK, TM4, CVl , VERO-76, HeLa, MDCK, BRL 3A, W138, Hep G2, MMT 060562, TRI, MRC 5, FS4 cells and Hep G2 cells. In some embodiments, the cells are CHO cells.

[0010] In some embodiments of the invention, the cells are grown at the first temperature for a certain time for cell growth, and then go to the second temperature for the expression of the polypeptide. For example, in some embodiments, a method is described that further comprises culturing the cells at a first temperature until a cell density of at least about 2.5 x 10 ⁶ viable cells is reached, and then moving to a second temperature that is lower than the first temperature to express the recombinant polypeptide. For example, in some embodiments of the invention, the first temperature is from about 35°C to about 37.5°C. In some embodiments of the invention, the second temperature is from about 29°C to about 35°C. In some embodiments of the invention, the first temperature is about 37°C and the second temperature is about 30°C. In some embodiments of the invention, the first temperature is about 36.5°C and the second temperature is about 33°C.

[0011] In another aspect, the present description provides a recombinant polypeptide obtained by any of the methods described herein. Such a produced recombinant polypeptide may have at least one of the specific characteristics resulting from the production methods described herein. Such characteristics may include at least one selected from the group consisting of: (a) a total sialic acid content (TSAC) between about 0.9 and 3.5 mol sialic acid/mol protein monomer; (b) isoelectric focus (IEF) between about 5.2 to about 6.7; (c) the basic structure of glycans as shown in FIG. 41 or FIG. 42; (d) chromatographic profile of labeled 2-AB oligosaccharide as shown in FIG. 38 or 39; (e) MALDI-ToF fingerprint profile of the glycopeptide as shown in FIG. 40 or 44-49; (f) a major band on SDS-PAGE under reducing conditions having a molecular weight of about 88-108 kDa and not less than about 85% of the total recombinant polypeptide produced; (g) a major band on SDS-PAGE under non-reducing conditions having a molecular weight of from about 194 to about 273 kDa and not less than about 85% of the total recombinant polypeptide produced; (h) not less than about 95.0% for recombinant polypeptide dimers, and not more than about 5.0% for aggregates by high pressure liquid chromatography (HPLC) except by size; (i) a purity of at least about 95.0% by reverse phase high pressure liquid chromatography (RP-HPLC); (j) not less than about 90.0% for the major peak, not greater than about 6.0% for the acidic peaks, and not greater than about 4.0% for the major anion exchange chromatography (AEX) peaks; (k) a percentage binding of hydroxyapatite (HA) from about 75% to about 125%; (l) specific activity of the product (pNPP) from about 620 to about 1250 units/mg; (m) K _m from about 13 to about 69 μM in an inorganic pyrophosphate (PPi) hydrolysis assay; (n) K _cat from about 65 to about 165 s ^-1 in an inorganic pyrophosphate (PPi) hydrolysis assay; (o) pI range from about 6.45 to about 6.95 for all peaks in capillary electrophoresis; (p) peaks in the MALDI-ToF mass spectrum as shown in FIG. 34A after deglycosylation; (q) Peaks in the MALDI-ToF mass spectrum as shown in FIG. 34B after reduction and deglycosylation; (r) peaks in the MALDI-ToF mass spectrum as shown in FIG. 35; (s) phosphorylation profile as shown in FIG. 36; (t) profile of the sialylated glycans in the MALDI-ToF negative mass spectrum as shown in FIG. 37A; (u) Profile of neutral glycans in a positive MALDI-ToF mass spectrum as shown in FIG. 37B; (v) a molar ratio of magnesium per mole of recombinant polypeptide from about 0.03 to about 0.15; (w) a molar ratio of calcium per mole of recombinant polypeptide from about 0.5 to about 1.5; and (x) a molar ratio of zinc per mole of recombinant polypeptide from about 0.5 to about 3.0.

[0012] In one embodiment of the invention, the resulting recombinant protein has from about 0.7 to about 1.19 free cysteine per half molecule.

[0013] In one embodiment of the invention, the resulting recombinant protein has a phosphorylation in Ser 93 in percent from about 13.5% to about 35.7%.

[0014] In one embodiment of the invention, the resulting recombinant protein has no less than 90.0% of the main peak, no more than 6.0% for acidic peaks, and no more than 4.0% for the main peaks on the AEX chromatogram. In one specific embodiment of the invention, the resulting recombinant protein has no less than 93.7% of the main peak, no more than 4.9% for acidic peaks, and no more than 3.4% for the main peaks on the AEX chromatogram.

[0015] In some embodiments, the resulting recombinant protein has a total sialic acid content (TSAC) of between about 0.9 to about 3.5 moles of sialic acid/mol of protein monomer. In one embodiment, the resulting recombinant protein has an average total sialic acid content (TSAC) of about 1.2 to about 3.0 mol sialic acid/mol per monomer. In one embodiment of the invention, the resulting recombinant protein has an average total sialic acid content of from about 1.9 to about 2.7 mol sialic acid/mol per monomer. In one specific embodiment of the invention, the resulting recombinant protein has an average total sialic acid content of from about 1.85 to about 2.28 mol sialic acid/mol per monomer.

[0016] In one embodiment, the resulting recombinant protein has a magnesium ion binding molar ratio of less than about 0.15. In some embodiments, the resulting recombinant protein has a magnesium molar ratio of about 0.05 to about 0.10. In one specific embodiment of the invention, the resulting recombinant protein has a magnesium ion binding molar ratio of about 0.12.

[0017] In some embodiments, the resulting recombinant protein contains at least about 95.0% recombinant polypeptide dimers, about 5.0% or less polypeptide aggregates, ie, as measured by size exclusion HPLC. In one embodiment, the resulting recombinant protein contains at least about 96.8% recombinant polypeptide dimers and about 3.2% or less polypeptide aggregates, ie, as measured by size exclusion HPLC. In one particular embodiment, the resulting recombinant protein contains at least about 97.6% recombinant polypeptide dimers and about 2.4% or less aggregates, ie, as measured by size exclusion HPLC.

[0018] In one embodiment, the resulting recombinant protein is at least 95.0% pure as measured by RP-HPLC. In one specific embodiment of the invention, the resulting recombinant protein is at least 97.6% pure as measured by RP-HPLC.

[0019] In some embodiments, the resulting recombinant protein has a hydroxyapatite (HA) binding percentage of about 75 to about 125%. In one embodiment of the invention, the resulting recombinant protein has an average % binding of hydroxyapatite from about 85% to about 97%. In one specific embodiment of the invention, the resulting recombinant protein has an average % binding of hydroxyapatite from about 90% to about 91%.

[0020] In one embodiment, the resulting recombinant protein has a specific activity (pNPP) of about 620 to about 1250 units/mg. In one particular embodiment, the resulting recombinant protein has a mean specific activity (pNPP) of about 904.0 to about 907.7 U/mg.

[0021] In some embodiments of the invention, the recombinant polypeptide is encoded by a polynucleotide encoding a polypeptide containing the sequence indicated in SEQ ID NO: 1, or a sequence fully complementary to SEQ ID NO: 1.

[0022] The recombinant polypeptide described herein can be obtained on an industrial or commercial scale. For example, in some embodiments of the invention, the fed batch reactor has a volume of 200 to 20,000 liters. In some embodiments, the fed-batch reactor has a volume of 2,000 to 20,000 liters.

[0023] In another aspect, the present disclosure is a pharmaceutical composition comprising a composition comprising the recombinant polypeptide described herein in combination with a pharmaceutically acceptable carrier, diluent, or excipient.

[0024] In another aspect, the present disclosure provides a method of using a recombinant polypeptide or pharmaceutical discussed herein to increase the breakdown of inorganic pyrophosphate (PPi) in a subject.

[0025] In another aspect, the present disclosure provides a method of treating a subject, comprising administering to a subject suffering from an alkaline phosphatase deficiency condition a therapeutically effective amount of a recombinant polypeptide or a pharmaceutical composition as discussed herein. Such an alkaline phosphatase deficiency condition includes, for example, hypophosphatasia (HPP) and neurofibromatosis type I (NF1). Such hypophosphatasia (HPP) can be any of perinatal, infantile, juvenile or adult HPP. Such a condition can be characterized by an unmineralized bone matrix and/or hypomineralization of the bones and teeth. For example, such unmineralized bone matrix can lead to rickets and/or osteomalacia.

[0026] In some embodiments, such a subject is a mammal. In some embodiments of the invention, such a subject is a human.

BRIEF DESCRIPTION OF GRAPHICS

[0027] In the accompanying drawings:

[0028] Figure 1 is a graph showing comparisons of cell growth (insert A, left) and viability (insert A, right) and total glucose (insert B, left) and lactate (insert B, right) concentrations between an exemplary manufacturing process ( No. 1) with or without temperature change for protein production. The control is the average of two temperature runs.

[0029] Figure 2 is a graph showing comparisons of titers bound to protein A (Inset A), volumetric activity (Inset B) and specific activity (Inset C) of asphotase alpha obtained using the exemplary manufacturing process (No. 1) with modification temperature or not. The control is the average of two temperature runs.

[0030] FIG. 3 are graphs showing comparisons of measurements of the acid peak (%) AEX (anion exchange chromatography) (group A) and SEC (size exclusion chromatography) aggregate (%) (panel B) of asphotase alpha obtained using the exemplary manufacturing process (#1) with temperature change or not.

[0031] In FIG. 4 is a graph showing comparisons of LoC under non-reducing conditions (Lab-on-Chip panel, inset A), LoC under reductive conditions (insert B), and TSAC (total sialic acid, inset C) of asphotase alfa prepared using an exemplary manufacturing process. (No. 1) with or without temperature change.

[0032] FIG. 5 is a graph showing the neural glycan profiles (by matrix-assisted laser desorption time-of-flight ionization or MALDI-TOF) of asphotase alpha produced by an exemplary manufacturing process (No. 1) with or without temperature change. The reference is a standard asphotase alpha obtained from the previous 20K process.

[0033] FIG. 6 is a graph showing the effect of production temperature bias on aggregated levels (Inset A) and TSAC (Inset B) of asphotase alfa produced. Error bars represent the standard deviation for each condition.

[0034] FIG. 7 is a graph showing the effect of culture pH on TSAC values of asphotase alpha produced by exemplary process #2 after 10 or 12 days of culture.

[0035] FIG. 8 is a graph showing the effect of medium supplementation (represented by different concentrations of glycose (glc), inset A) and the supplementation regimen (inset B) on asphotase alpha TSAC in exemplary process #2.

[0036] FIG. 9 is a graph showing the effect of zinc supplementation on asphotase alpha activity (Inset A, zinc added on day 9) and the effect of zinc concentration on cell growth (Inset B) and viability (Inset C) in exemplary process #2.

[0037] FIG. 10 are graphs showing the effect of pH and production temperature on TSAC of asphotase alpha (Inset A), protein A titer (Inset B) and volumetric activity (Inset C) in exemplary process #3.

[0038] FIG. 11 is a graph showing the effect of growth temperature on volumetric activity (Inset A) and fragmentation of asphotase alpha (Inset B, measured by SEC) in exemplary process #3.

[0039] FIG. 12 are graphs showing the effect of production temperature and pH on asphotase alpha aggregation (Inset A, measured by SEC), % acid peak (Inset B, measured by AEX), fragmentation of Asphotase alpha (Inset C, measured by SEC), and % main peak (Inset D). measured by AEX) in exemplary process #3.

[0040] FIG. 13 is a graph showing the effect of seeding density and temperature shift time on volumetric activity (Inset A) and fragmentation of asphotase alpha (Inset B, measured by SEC). Seeding is maintained for seeding density in exemplary process #3.

[0041] In FIG. 14 are graphs showing the effect of medium replenishment (amount and time) on asphotase alpha volumetric activity (Inset A), aggregation (Inset B, measured by SEC), % Acid Peak (Inset C, measured by AEX), % Main Peak (Inset D, measured AEX) and TSAC (panel E) in exemplary process #3.

[0042] FIG. 15 is a graph showing the effect of culture medium pH on cell growth (Inset A) and viability (Inset B) up to ten days in exemplary process #4.

[0043] FIG. 16 are graphs showing the effect of media production temperature on cell growth (Inset A) and viability (Inset B) up to ten days in exemplary process #4.

[0044] FIG. 17 are graphs showing the effect of culture medium pH on glucose (Inset A) and lactate (Inset B) concentrations in culture medium up to ten days in exemplary process #4.

[0045] FIG. 18 are graphs showing the effect of production temperature on glucose (Inset A) and lactate (Inset B) concentrations in culture media up to ten days in exemplary process #4.

[0046] FIG. 19 are graphs showing the effect of culture medium pH on protein A titer (Inset A), volume activity (Inset B), protein A specific productivity (Inset C), and specific activity (Inset D) up to ten days in exemplary process #4.

[0047] FIG. 20 are graphs showing the different effects of production temperature on protein A titer (Inset A), volumetric activity (Inset B), protein A specific productivity (Inset C), and specific activity (Inset D) in illustrative process #4.

[0048] FIG. 21 is a graph showing the effect of culture medium pH (Inset A) and production temperature (Inset B) on asphotase alpha TSAC in exemplary process #4.

[0049] FIG. 22 is a graph showing the effect of culture medium pH (Inset A) and production temperature (Inset B) on the acid peak AEX of the produced asphotase alpha in exemplary process #4.

[0050] FIG. 23 are graphs showing the effect of culture medium pH (Inset A) and production temperature (Inset B) on asphotase alpha aggregation (measured by SEC) in exemplary process #4.

[0051] FIG. 24 is a graph showing the different effects of pH (Inset A) of culture medium and production temperature (Inset B) on the measurement of the main peak LoC (% non-reducing conditions) of asphotase alpha produced in exemplary process #4.

[0052] FIG. 25 shows graphs showing the effect of culture medium pH on neutral glycan profiles of the resulting asphotase alpha (measured by MALDI-TOF) in exemplary process #4. The top panel shows all MALDI-ToF peaks for asphotase alfa obtained at 33°C for 10 days, but at different culture medium pH. The value of each peak was calculated and compared in the lower inset.

[0053] FIG. 26 is a graph showing the effect of manufacturing temperature on the neutral glycan profiles of the produced asphotase alpha (measured by MALDI-ToF) in exemplary process #4. The top inset shows all MALDI-ToF peaks for asphotase alfa obtained at different production temperatures at pH 6.9 for 10 days. The value of each peak was calculated and compared in the lower inset.

[0054] FIG. 27 are graphs showing the effect of culture medium pH (Inset A) and production temperature (Inset B) on peak percent capillary isoelectric focusing for the iCE280 asphotase alfa quality assay in exemplary process #4.

[0055] FIG. 28 shows graphs comparing cell growth (Inset A) and viability (Inset B) for various processes (blue line: control process (performed on two bioreactor tanks) with one bolus of media feed); red line: improved process (performed for six bioreactor tanks) with quadruple culture media; gray line: the previous 20K processes (performed on twenty bioreactor tanks) were used as a standard. The gray line represents the mean of 20K processes and the dotted lines represent the mean±1 x standard deviation.

[0056] FIG. 29 shows graphs comparing glucose utilization (Inset A) and lactate production (Inset B) for various processes (blue line: control process (performed on two bioreactor tanks) with one bolus feed of medium); red line: improved process (performed for six bioreactor tanks) with four times the medium make-up; gray line: the previous 20K processes (performed on twenty bioreactor tanks) were used as a standard. The gray line represents the mean of 20K processes and the dotted lines represent the mean±1 x standard deviation.

[0057] FIG. 30 shows graphs comparing protein A titer (Inset A) and specific activity (Inset B) of asphotase alfa prepared by various methods (blue line: control process (performed on two bioreactor tanks) with one bolus feed of medium); red line: improved process (performed for six bioreactor tanks) with four times the medium make-up; gray line: the previous 20K processes (performed on twenty bioreactor tanks) were used as a standard. The gray line represents the mean of 20K processes and the dotted lines represent the mean±1 x standard deviation.

[0058] FIG. 31 shows graphs comparing active titer (Inset A) and total volumetric activity (Inset B) for various processes (blue line: control process (performed on two bioreactor tanks) with one bolus feed of media); red line: improved process (performed for six bioreactor tanks) with four times the medium make-up; gray line: the previous 20K processes (performed on twenty bioreactor tanks) were used as a standard. The gray line represents the mean of 20K processes and the dotted lines represent the mean±1 x standard deviation.

[0059] FIG. 32 is a graph comparing the specific activity of asphotase alpha from the control processes (grey, one bolus medium feed) and the improved processes (dark grey, four medium feeds).

[0060] Figure 33 is a graph comparing the specific activity of asphotase alpha from the previous Y process (lower line without additional zinc supplementation) and another improved process (upper line showing the average of 30-90 μM zinc supplementation conditions and culture time extended up to 14 days).

[0061] In FIG. 34 is a graph showing the MALDI-ToF mass spectrum data for the produced asphotase alpha after deglycosylation (Inset A) and for the produced Asphotase alpha that has been reduced and deglycosylated (Inset B).

[0062] Figure 35 is a graph showing the MALDI-ToF mass spectrum of produced asphotase alpha. Figure 35 provides "TYNTNAQVPDSAGTATAYAYLCGVK" as residues 83-105 of SEQ ID NO: 1, "SAGTA" as residues 93-97 of SEQ ID NO: 1, and "AYLCGV" as residues 99-104 of SEQ ID NO: 1, respectively, in order of appearance.

[0063] Figure 36 is a graph showing the determination of the MC/MC phosphorylation site on asfotase alpha.

[0064] FIG. 37 are graphs showing the negative MALDI-ToF mass spectrum of sialylated glycans (Inset A) and the positive MALDI-TOF mass spectrum of neutral glycans (Inset B) on asphotase alpha produced.

[0065] Figure 38 is a graph showing a fluorescence chromatogram of asphotase alpha oligosaccharides.

[0066] Figure 39 is a graph showing a fluorescence chromatogram of the asphotase alpha reference standard.

[0067] FIG. 40 is a graph showing the mass spectrum of glycopeptides derived from asphotase alfa.

[0068] FIG. 41 shows exemplary proposed structures of asphotase alpha core glycans (C7108H11008O2206S56 (protein moiety dimer) or C3554H5506O1103S28 (monomer)). The number of NeuAc per glycan (FA2G2, FA2G1 and A2G2) is an estimated number.

[0069] FIG. 42 shows typical evaluated structures of glycans at asphotase alpha glycosylation sites.

[0070] Figure 43 is a graph showing a typical asphotase alpha electropherogram.

[0071] In FIG. 44 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N123 in T15-16). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0072] FIG. 45 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N213 in T26-27). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0073] FIG. 46 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N254 in T33). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0074] FIG. 47 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N286 in T35). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0075] FIG. 48 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N413 in T45-46). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0076] FIG. 49 are graphs showing glycopeptide mass fingerprints for asphotase alpha obtained from 20K and 2K reservoirs (N564 to T55). 2K tank numbers: #35, #36 and #38; 20K: #40, #42 and #34.

[0077] FIG. 50 is a graph comparing the ICP mole ratio of metal ions for zinc (Inset A), magnesium (Inset B), and calcium (Inset C) asphotase alpha obtained from batches 2K and 20K.

[0078] Figure 51A shows a graph comparing viable cell density (VCD) over culture time under different pH conditions. Figure 51B is a graph comparing cell viability over culture time under various pH conditions.

[0079] FIG. 52A is a graph showing the concentration of glucose in the bioreactors over elapsed culture time. In FIG. 52B is a graph showing lactate concentration in the bioreactors over elapsed culture time.

[0080] Figure 53 is a graph comparing specific activity profiles under various pH conditions.

DETAILED DESCRIPTION OF THE INVENTION

Definitions

[0081] "About", "About": As used herein, the terms "about" and "about", in relation to one or more specific cell culture conditions, refer to a range of values that are similar to the specified reference value for that condition. or culture conditions. In some embodiments of the invention, the term "about" refers to a range of values that are within 25, 20, 19, 18, 17, 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6 , 5, 4, 3, 2, 1 percent or less of the specified reference value for that culture condition or conditions.

[0082] "Amino acid": As used herein, the term "amino acid" refers to any of the twenty naturally occurring amino acids that are commonly used in the formation of polypeptides or analogues or derivatives of these amino acids. The amino acids of the present invention may be presented in cell culture media. The amino acids present in the medium may be in the form of salts or in the form of a hydrate.

[0083] "Batch culture": As used herein, the term "batch culture" refers to a method of culturing cells in which all components that will ultimately be used for culturing cells, including the medium (see definition of "medium" below), as well as the cells themselves, are provided at the beginning of the culturing process. The batch culture is usually stopped at some point and the cells and/or components in the medium are harvested and optionally purified.

[0084] "Bioreactor": As used herein, the term "bioreactor" refers to any vessel used to grow a cell culture (eg, mammalian cell culture). The bioreactor can be of any size as long as it is useful for cell culture. Typically, the bioreactor will be at least 1 liter and may be 10, 100, 250, 500, 1000, 2500, 5000, 8000, 10000, 120000, 20000 liters or more, or any volume in between. The internal conditions of the bioreactor, including but not limited to pH and temperature, are typically controlled during the culture period. The bioreactor may be comprised of any material suitable for holding mammalian or other cell cultures suspended in the media under the culture conditions of the present invention, including glass, plastic, or metal. As used herein, the term "production bioreactor" refers to the final bioreactor used in the production of a polypeptide or protein of interest. The volume of a large cell culture bioreactor is typically at least 500 liters and may be 1,000, 2,500, 5,000, 8,000, 10,000, 120,000, 20,000 liters or more, or any volume in between. One of skill in the art will know and be able to select suitable bioreactors for use in the practice of the present disclosure.

[0085] "Cell density": As used herein, the term "cell density" refers to the number of cells present in a given volume of medium.

[0086] "Cell viability": The term "cell viability" as used herein refers to the ability of cells in culture to survive under a given set of culture conditions or experimental variations. As used herein, the term also refers to that fraction of cells that are alive at a particular time relative to the total number of cells, alive and dead, in culture at the same time.

[0087] "Culture" and "cell culture": These terms, as used herein, refer to a population of cells that is suspended in a medium (see definition of "medium" below) under conditions suitable for the survival and/or growth of the cell population. . As will be appreciated by those skilled in the art, these terms used herein may refer to a combination containing a population of cells and the medium in which the population is suspended.

[0088] "Fed-batch culture": The term "fed-batch culture" as used herein refers to a cell culture method in which additional components are provided to the culture for some time after the culture process has started. The components provided typically contain nutritional supplements for cells that have been depleted during the culture process. The fed batch culture is usually stopped at some point and the cells and/or components in the medium are harvested and optionally purified. The fed-batch culture can be run in a suitable fed-batch bioreactor.

[0089] "Fragment": As used herein, the term "fragment" refers to a polypeptide and is defined as any discrete portion of a given polypeptide that is unique or characteristic of that polypeptide. As used herein, the term also refers to any discrete portion of a given polypeptide that retains at least a portion of the activity of the full-length polypeptide. In some embodiments of the invention, the proportion of retained activity is retained by at least 10% of the activity of the full-length polypeptide. In various embodiments of the invention, the proportion of retained activity is at least 20%, 30%, 40%, 50%, 60%, 70%, 80% or 90% of the activity of the full-length polypeptide. In other embodiments of the invention, the proportion of retained activity is at least 95%, 96%, 97%, 98% or 99% of the activity of the full-length polypeptide. In one embodiment of the invention, the percentage of activity retained is 100% of the activity of the full length polypeptide. As used herein, the term also refers to any portion of a given polypeptide that includes at least the identified sequence element found in the full-length polypeptide. In some embodiments, the sequence element spans at least 4-5 amino acids of the full length polypeptide. In some embodiments, the sequence element spans at least about 10, 15, 20, 25, 30, 35, 40, 45, 50 or more amino acids of the full length polypeptide.

[0090] "Integrated Viable Cell Density": The term "Integrated Viable Cell Density", as used herein, refers to the average viable cell density during a culture multiplied by the amount of time the culture has been cultured. Assuming that the amount of polypeptide and/or protein produced is proportional to the number of viable cells present during culture, the integrated viable cell density is a useful tool to estimate the amount of polypeptide and/or protein produced during culture.

[0091] "Media", "cell culture medium" and "culture medium": These terms, as used herein, refer to a solution containing nutrients that nourish growing mammalian cells. Typically, these solutions provide the essential and non-essential amino acids, vitamins, energy sources, lipids, and trace minerals required by the cell for minimal growth and/or survival. The solution may also contain components that enhance growth and/or survival above the minimum rate, including hormones and growth factors. The solution, for example, is formulated to a pH and salt concentration that is optimal for cell survival and proliferation. The medium can also be "defined medium" - serum-free medium that does not contain proteins, hydrolysates or components of unknown composition. Certain media do not contain animal ingredients and all ingredients have a known chemical structure.

[0092] "metabolic end product": As used herein, the term "metabolic end product" refers to a compound produced by a cell culture as a result of normal or abnormal metabolic processes that are detrimental to the cell culture in some way, especially with respect to expression or the activity of the desired recombinant polypeptide or protein. For example, metabolic end products may adversely affect cell culture growth or viability, may reduce the amount of recombinant polypeptide or protein produced, may alter the folding, stability, glycosylation, or other post-translational modification of the expressed polypeptide or protein, or may be detrimental to cells and/or expression or activity of the recombinant polypeptide or protein in any other way. Exemplary metabolic end products include lactate, which is formed from the metabolism of glucose, and ammonium, which is formed from the metabolism of glutamine. In one embodiment, methods are used to slow down the production, reduce, or even eliminate end products of metabolism in cell cultures.

[0093] "Osmolality" and "osmolarity": Osmolality is a measure of the osmotic pressure of dissolved particles in an aqueous solution. Solute particles include both ions and non-ionized molecules. Osmolality is expressed as the concentration of osmotically active particles (i.e. osmoles) dissolved in 1 kg of solution (1 mOsm/kg H ₂ O at 38° C. is equivalent to an osmotic pressure of 19 mmHg). "Osmolarity", in contrast, refers to the number of dissolved particles dissolved in 1 liter of solution. As used herein, the abbreviation "mOsm" means "milliosmoles/kg solution".

[0094] "Perfusion culture": The term "perfusion culture" as used herein refers to a cell culture method in which additional components are provided to the culture continuously or semi-continuously after the culture process has started. The components provided typically contain nutritional supplements for cells that have been depleted during the culture process. A portion of the cells and/or components in the medium is typically collected on a continuous or semi-continuous basis and optionally purified.

[0095] "Polypeptide": As used herein, the term "polypeptide" refers to a sequential chain of amino acids linked together via peptide bonds. The term is used to refer to an amino acid chain of any length, but one skilled in the art will appreciate that the term is not limited to long chains and may refer to a minimal chain containing two amino acids linked together via a peptide bond.

[0096] "Protein": As used herein, the term "protein" refers to one or more polypeptides that function as a discrete unit. Where one polypeptide is a discrete functional unit and does not require constant physical association with other polypeptides to form a discrete functional unit, the terms "polypeptide" and "protein" as used herein are used interchangeably.

[0097] "Recombinantly expressed polypeptide" and "recombinant polypeptide": These terms, as used herein, refer to a polypeptide expressed from a host cell that has been genetically engineered to express that polypeptide. The recombinantly expressed polypeptide may be identical or similar to a polypeptide that is normally expressed in a mammalian host cell. The recombinantly expressed polypeptide may also be foreign to the host cell, ie heterologous to peptides normally expressed in the host cell. Alternatively, a recombinantly expressed polypeptide may be chimeric in those parts of the polypeptide that contain amino acid sequences that are identical or similar to polypeptides normally expressed in a mammalian host cell, while other parts are foreign to the host cell.

[0098] "Inoculation": As used herein, the term "inoculation" refers to the process of loading a cell culture into a bioreactor or other vessel. The cells may have previously been grown in another bioreactor or vessel. Alternatively, the cells may be frozen and thawed immediately prior to being fed into the bioreactor or vessel. The term refers to any number of cells, including one cell.

[0099] "Title": As used herein, the term "titer" refers to the total amount of recombinantly expressed polypeptide or protein produced by a cell culture divided by a given amount of mean volume. The titer is usually expressed in units of milligrams of polypeptide or protein per milliliter of medium.

[00100] Acronyms used herein include, for example, VCD: Viable Cell Density; IVCC: Viable cell concentration integral; TSAC: Total Sialic Acid; HPAE-PAD: High Performance Anion Exchange Chromatography with Pulsed Amperometric Detection; SEC: Gel Permeation Chromatography; AEX: Anion Exchange Chromatography; LoC: Lab-on-Chip and MALDI-TOF: Matrix Laser Desorption/Ionization - Time of Flight.

[00101] The present disclosure provides a method for culturing cells (eg, mammalian cells, including but not limited to Chinese hamster ovary (CHO) cells) expressing the recombinant protein. The present disclosure provides production systems for producing alkaline phosphatase (eg, asphotase alfa) using cell culture. In some embodiments of the invention, systems are provided that minimize the production of one or more metabolic products that are detrimental to growth, cell viability and/or protein production or quality. In specific embodiments of the invention, the cell culture is a batch culture, fed-batch culture, culture or continuous culture. Other embodiments of the description are discussed in detail below. Those skilled in the art will appreciate, however, that various modifications of these embodiments are included within the scope of the description.

Squirrels

[00102] The present disclosure relates to the expression of an alkaline phosphatase protein, asphotase alpha, in cell culture. In some embodiments of the invention, such as asfotase alfa, when obtained by the methods described herein, can be used to treat or prevent diseases or disorders associated with alkaline phosphatase. For example, such asphotase alpha may be administered to a subject having reduced and/or inactive endogenous alkaline phosphatase, or having overexpressed (eg, above normal levels) alkaline phosphatase substrates. In some embodiments of the invention, asfotase alpha in this description is a recombinant protein. In some embodiments of the invention, asfotase alpha is a fusion protein. In some embodiments, asphotase alfa in this specification specifically targets a cell type, tissue (e.g., connective, muscle, nervous, or epithelial tissue), or organ (e.g., liver, heart, kidney, muscle, bone, cartilage, ligament, tendon). etc.). Asfotase alfa is a soluble Fc fusion protein consisting of two sTNALP-Fc-D10 polypeptides, each containing 726 amino acids, as shown in SEQ ID NO: 1. Underlined aspartic (N) residues correspond to potential glycosylation sites (i.e. 123, 213, 254, 286, 413 and 564). The underlined amino acid residues (L ₄₈₆ -K ₄₈₇ and D ₇₁₅ -I ₇₁₆ ) correspond to linkers between the sALP and Fc domains, Fc and D ₁₀ , respectively.

[00103] Each polypeptide or monomer consists of five parts. The first part (sALP), containing the amino acids L1-S485, is the soluble part of the non-specific human tissue alkaline phosphatase enzyme, which contains a catalytic function. The second part contains amino acids L486-K487 as a linker. The third part (Fc), containing the amino acids D488-K714, is the Fc part of human immunoglobulin gamma 1 (IgG1) containing the hinge, CH ₂ and CH ₃ domains. The fourth part contains D715-I716 as a linker. The fifth part contains the amino acids D717-D726 (D10), which are the target bone fragment that allows asphotase alpha to bind to the mineral phase of the bone. In addition, each polypeptide chain contains six potential glycosylation sites and eleven cysteine (Cys) residues. Cys102 exists as free cysteine. Each polypeptide chain contains four intrachain disulfide bonds between Cys122 and Cys184, Cys472 and Cys480, Cys528 and Cys588, Cys634 and Cys692. The two polypeptide chains are linked by two interchain disulfide bonds between Cys493 on both chains and between Cys496 on both chains. In addition to these covalent structural features, mammalian alkaline phosphatase is believed to have four metal binding sites on each polypeptide chain, including two zinc sites, one magnesium site, and one calcium site.

Asphotase alfa

[00104] In one embodiment, the alkaline phosphatase protein (eg, bone-targeting fusion protein sALP) is asphotase alpha (ie, sTNALP-Fc-D ₁₀ , SEQ ID NO: 1). In particular, asfotase alfa is a complex soluble glycoprotein with a polypeptide length of 726 amino acids. Asfotase alfa is an Fc fusion protein composed of 3 domains. From the N-terminus to the C-terminus, asphotase alfa contains: (1) human non-specific alkaline phosphatase (TNSALP) soluble catalytic domain (UniProtKB/Swiss-Prot Accession No. P05186), (2) human immunoglobulin G1 Fc domain (UniProtKB/Swiss -Prot Accession No. P01857) and (3) deca-aspartate peptide (D ₁₀ ) used as a bone targeting domain (Nishioka et al., 2006 Mol Genet Metab 88 : 244-255). The protein binds into a homodimer of two primary protein sequences. This fusion protein contains 6 confirmed complex N-glycosylation sites. Five of these N-glycosylation sites are located in the sALP domain and one in the Fc domain. Another important post-translational modification present in asfotase alpha is the presence of disulfide bridges that stabilize the enzyme and the Fc domain structure. A total of 4 intramolecular disulfide bridges are present per monomer, and 2 intermolecular disulfide bridges are present per dimer. One cysteine of the alkaline phosphatase domain is free.

[00105] Asfotase alfa can be used as an enzyme replacement therapy for the treatment of hypophosphatasia (HPP). In HPP patients, loss-of-function mutation(s) in the gene encoding TNSALP causes a deficiency in TNSALP enzymatic activity, resulting in elevated circulating levels of substrates such as inorganic pyrophosphate (PPi) and pyridoxal-5'-phosphate (PLP). Administration of asphotase alfa to HPP patients cleaves PPi, releasing inorganic phosphate for combination with calcium, thereby promoting hydroxyapatite crystal formation and bone mineralization and restoring a normal skeletal phenotype. For more information on asphotase alfa and its use in therapy, see PCT publications WO 2005103263 and WO2008138131, the principles of which are incorporated herein by reference in their entirety. In another embodiment of the invention, asfotase alfa can be used as an enzyme replacement therapy for the treatment of neurofibromatosis type I (NF1). For more information on asphotase alfa and its use (along with the use of other alkaline phosphatases) in the treatment of NF1, see PCT Publication No. WO 2013/058833, which is incorporated herein by reference in its entirety.

Manufacturing process

[00106] An alkaline phosphatase protein (eg, asphotase alfa) can be obtained from mammalian or other cells using standard methods known in the art. Such cells can be grown in culture dishes, glass flasks or bioreactors. Specific methods for culturing cells and obtaining recombinant proteins are known in the art, as described in Nelson and Geyer, 1991 Bioprocess Technol. 13 :112-143 and Rea et al., Supplement to BioPharm International March 2008, 20-25. Exemplary bioreactors include batch, fed batch, and continuous reactors. In some embodiments, the alkaline phosphatase protein is produced in a fed-batch bioreactor.

[00107] Potential variability in the physicochemical environment of a cell culture process includes, for example, changes in pH, temperature, composition of the cell culture medium, change in raw material, media filter media, bioreactor scale differences, gassing strategy (air, oxygen, and carbon dioxide ), etc. As described herein, the glycosylation profiles of the produced alkaline phosphatase protein can be affected by changes in one or more parameters.

Development of cell culture processes

[00108] To obtain a recombinant protein in cell culture, the recombinant gene with the necessary transcriptional regulatory elements is first transferred into a host cell. As a rule, a second gene is transferred, which gives the recipient cells a selective advantage. In the presence of a selection agent, which is usually applied a few days after gene transfer, only those cells that express the selector gene survive. Two popular genes for selection are dihydrofolate reductase (DHFR), an enzyme involved in nucleotide metabolism, and glutamine synthetase (GS). In both cases, selection occurs in the absence of the appropriate metabolite (hypoxanthine and thymidine in the case of DHFR, glutamine in the case of GS), preventing the growth of untransformed cells. In general, it does not matter for efficient expression of a recombinant protein whether the biopharmaceutical coding gene and the selector genes are on the same plasmid or not.

[00109] After selection, surviving cells can be transferred as single cells to a second culture vessel and cultures are expanded to obtain clonal populations. Ultimately, individual clones are evaluated for expression of the recombinant protein, with the highest producers being retained for further cultivation and analysis. From these candidates, one cell line with appropriate growth and performance characteristics is selected for the production of the recombinant protein. Then the cultivation process is started, which is determined by the needs of production.

Cells

[00110] Any mammalian cell or non-mammalian cell type that can be cultured to produce a polypeptide can be used in accordance with the present invention. Non-limiting examples of mammalian cells that can be used include, for example, Chinese hamster ovary +/- DHFR cells (CHO, Urlaub and Chasin, 1980 Proc. Natl. Acad. Sci. USA, 77 :4216); mouse myeloma line BALB/c (NSO/1, ECACC accession no: 85110503); human retinoblasts (PER.C6 (CruCell, Leiden, The Netherlands)); monkey kidney CVl line transformed with SV40 (COS-7, ATCC CRL 1651); human kidney embryonic line (293 or 293 cells subcloned for growth in suspension culture, Graham et al., 1977 J. Gen Virol., 36:59 ); baby hamster kidney cells (BHK, ATCC CCL 10); mouse Sertoli cells (TM4, Mather, Biol. Reprod., 23:243-251 (1980)); monkey kidney cells (CVl ATCC CCL 70); African green monkey kidney cells (VERO-76, ATCC CRL-I 587); human cervical carcinoma cells (HeLa, ATCC CCL 2); dog kidney cells (MDCK, ATCC CCL 34); buffalo rat liver cells (BRL 3A, ATCC CRL 1442); human lung cells (W138, ATCC CCL 75); human liver cells (Hep G2, HB 8065); mouse mammary tumor (MMT 060562, ATCC CCL51); TRI cells (Mather et al., 1982, Annals NY Acad. Sci. 383 : 44-68); MRC 5 cells; FS4 cells; and human hepatoma lines (Hep G2). In a particular embodiment of the invention, the cultivation and expression of the polypeptides and proteins is from a Chinese hamster ovary (CHO) cell line.

[00111] In addition, any number of commercially available and non-commercially available recombinant cell lines that express polypeptides or proteins can be used in accordance with the present invention. One skilled in the art will appreciate that recombinant cell lines may have different nutritional requirements and/or may require different culture conditions for optimal polypeptide or protein growth and expression, and will be able to modify the conditions as needed.

[00112] As noted above, in many cases cells will be selected or engineered to produce high levels of a protein or polypeptide. Often, cells are genetically engineered to produce high levels of protein, for example, by introducing a gene encoding a protein or polypeptide of interest and/or by introducing control elements that regulate the expression of a gene (whether endogenous or introduced) encoding a polypeptide of interest.

Temperature change

[00113] The running time of cell culture processes, especially non-continuous processes (eg, fed-batch processes in bioreactors), is typically limited by remaining cell viability, which typically declines during culture. Therefore, it is desirable to increase the cell viability to improve the production of the recombinant protein. Product quality issues are also motivated to minimize the decline in viable cell density and maintain high cell viability, since cell death can release sialidases in the culture supernatant, which can reduce the sialic acid content of the expressed protein. The problems of protein purification offer another motivation for minimizing the decline in viable cell density and maintaining high cell viability. Cell debris and dead cell content in culture can adversely affect the ability to isolate and/or purify the protein product at the end of the culture run. Thus, keeping the cells viable for a longer period of time in culture reduces contamination of the culture medium with cellular proteins and enzymes (eg, cellular proteases and sialidases) that can cause degradation and eventual degradation of the desired glycoprotein produced by the cells.

[00114] Many methods can be used to achieve high cell viability in cell cultures. One of them involves lowering the temperature of the culture after initial cultivation at normal temperature. For example, see Ressler et al., 1996, Enzyme and Microbial Technology 18 : 423-427). Typically, mammalian cells or other cell types capable of expressing the protein of interest are first grown at normal temperature to increase cell numbers. Such "normal" temperatures for each cell type are typically about 37°C (e.g., from about 35°C to about 39°C, including, for example, 35.0°C, 35.5°C, 36.0°C , 36.5°C, 37.0°C, 37.5°C, 38.0°C, 38.5°C and/or 39.0°C). In one specific embodiment of the invention, the temperature for obtaining asfotase alpha is first set to about 37°C. When a sufficiently high cell density is reached, the culture temperature for the entire cell culture is then shifted (eg, reduced) to promote protein production. In most cases, lowering the temperature shifts cells to the non-growth G1 phase of the cell cycle, which can increase cell density and viability compared to the previous higher temperature medium. In addition, lower temperature may also promote recombinant protein production by increasing the rate of cellular protein formation, facilitating post-translational modification of the protein (e.g., glycosylation), reducing fragmentation or aggregation of newly produced proteins, facilitating protein folding, and forming a three-dimensional structure (thus preserving activity) and/or reducing the degradation of newly produced proteins. In some embodiments of the invention, the lower temperature is from about 30°C to about 35°C (for example, 30.0°C, 30.5°C, 31.0°C, 31.5°C, 32.0° C, 32.5°C, 33.0°C, 33.5°C, 34.0°C, 34.5°C and/or 35.0°C). In other embodiments of the invention, the temperature for obtaining asfotase alpha is first set from about 35.0°C to about 39.0°C, and then shifted from about 30.0°C to about 35.0°C. In one embodiment of the invention, the temperature for producing asphotase alpha is first set to about 37.0°C and then shifted to about 30°C. In another embodiment of the invention, the temperature for obtaining asfotase alpha is first set to about 36.5°C, and then shifted to about 33°C. In another embodiment of the invention, the temperature for obtaining asfotase alpha is first set to about 37.0°C, and then shifted to about 33°C. In yet another additional embodiment of the invention, the temperature for obtaining asfotase alpha is first set to about 36.5°C, and then shifted to about 30°C. In other embodiments of the invention, multiple (eg, more than one) temperature steps may be used. For example, the temperature may be reduced from 37°C to 33°C and then further to 30°C.

[00115] The time to maintain the culture at a certain temperature before shifting to another temperature can be determined to achieve sufficient (or desired) cell density while maintaining cell viability and the ability to produce the protein of interest. In some embodiments, the cell culture is grown at the first temperature until the density of viable cells reaches about 10 ⁵ cells/ml to about 10 ⁷ cells/ml (for example, 1×10 ⁵ , 1.5×10 ⁵ , 2.0×10 ⁵ , 2.5×10 ⁵ , 3.0×10 ⁵ , 3.5×10 ⁵ , 4.0×10 ⁵ , 4.5×10 ⁵ , 5.0×10 ⁵ , 5 .5×10 ⁵ , 6.0×10 ⁵ , 6.5×10 ⁵ , 7.0×10 ⁵ , 7.5×10 ⁵ , 8.0×10 ⁵ , 8.5×10 ⁵ , 9, 0×10 ⁵ , 9.5×10 ⁵ , 1.0×10 ⁶ , 1.5×10 ⁶ , 2.0×10 ⁶ , 2.5×10 ⁶ , 3.0×10 ⁶ , 3.5 ×10 ⁶ , 4.0×10 ⁶ , 4.5×10 ⁶ , 5.0×10 ⁶ , 5.5×10 ⁶ , 6.0×10 ^{6 , 6.5×10 6 ,} 7.0× 10 ⁶ , 7.5×10 ⁶ , 8.0×10 ⁶ , 8.5×10 ⁶ , 9.0×10 ⁶ , 9.5×10 ⁶ , 1×10 ⁷ cells/ml and more) before transition to a different temperature. In one embodiment of the invention, the cell culture is grown at the first temperature until the viable cell density reaches from about 2.5 to about 3.4×10 ⁶ cells/ml before switching to another temperature. In another embodiment of the invention, the cell culture is grown at the first temperature until the viable cell density reaches about 2.5 to about 3.2×10 ⁶ cells/ml before changing to another temperature. In yet another embodiment of the invention, the cell culture is grown at the first temperature until the viable cell density reaches about 2.5 to about 2.8×10 ⁶ cells/ml before changing to another temperature.

[00116] In some embodiments of the invention, the cell culture is grown at 37°C until the density of viable cells reaches about 2.5-2.8×10 ⁶ cells/ml before shifting to 30°C for protein production. In other embodiments of the invention, the cell culture is grown at 37°C until the density of viable cells reaches about 2.5-3.4×10 ⁶ cells/ml before shifting to 30°C for protein production.

pH

[00117] Changing the pH of the growth medium in cell culture can affect cellular proteolytic activity, secretion, and levels of protein production. Most cell lines grow well at around pH 7-8. Although the optimal pH for cell growth varies relatively slightly between different cell strains, some normal fibroblast cell lines work best at pH 7.0-7.7, and transformed cells generally work best at pH 7.0-7.4 (Eagle , 1973 The effect of environmental pH on the growth of normal and malignant cells J Cell Physiol 82 :1-8). In some embodiments, the pH of the asphotase alpha culture medium is about pH 6.5-7.7 (e.g., 6.50, 6.55, 6.60, 6.65, 6.70, 6.75, 6.80, 6.85, 6.90, 6.95, 7.00, 7.05, 7.10, 7.15, 7.20, 7.25, 7.30, 7.35, 7, 39, 7.40, 7.45, 7.50, 7.55, 7.60, 7.65 and 7.70). In some embodiments of the invention, the pH of the culture medium for the production of asphotase alpha is about pH 7.20-7.60. In other embodiments of the invention, the pH of the culture medium for the production of asphotase alpha is about pH 6.9-7.1. In one specific embodiment of the invention, the pH of the culture medium for producing asphotase alpha is about pH 6.9. In another embodiment of the invention, the pH of the culture medium for producing asphotase alpha is about pH 7.30. In another embodiment of the invention, the pH of the culture medium for the production of asphotase alpha is about pH 7.39.

[00118] All references cited in this document are incorporated by reference in their entirety.

[00119] Although this description has been described in more detail by way of illustration and example for clarity of understanding, it will be apparent to those skilled in the art that some minor changes and modifications will be made. Therefore, the description and examples should not be construed as limiting the scope of the description.

Examples

Example 1: General process for the production of asphotase alfa

[00120] As described herein, a method has been developed for the production of alkaline phosphatases (eg, asphotase alpha (sTNALP-Fc-D10)).

Table 1. Differences between illustrative manufacturing processes (previous)

Process X Process Y Process Z Parent cell line CHO CHO CHO Work environment HyClone SFM4CHO HyClone SFM4CHO HyClone SFM4CHO+BD selection make-up CHO Makeup (0.5%) CHO Replenishment (2% i.e. 4 x medium nutrient) CHO Replenishment (2%) + cell growth accelerator 2+5 (9%) pH set point 6.90 6.90 6.90 Temperature 37.0°C, then shifted to 30°C 37.0°C, then shifted to 30°C 37.0°C, then shifted to 30°C DO 40% 40% 40% Additive N/A N/A ZnSO4 collection time 240±12 h 240±12 h 240±12 h

[00121] Stable CHO cell lines expressing asphotase alpha have been developed using the GS gene expression system. Secondary clones were generated from high yielding primary clones in one round of limited serial dilution cloning and the final cell line was selected.

[00122] An exemplary manufacturing process, Process X, is described herein. The Master Cell Bankl vial was thawed and the entire volume of the vial was resuspended. The entire volume was transferred to a 250 ml shake flask for growth. Samples were taken daily to check the number and viability (also for all subsequent expansion steps). Cells were expanded in several steps and inoculated into a 1000 L seed bioreactor (N-3 low level), a 1000 L seed bioreactor (N-2 high level) and a 4000 L seed bioreactor (N-1), followed by a 20,000 L bioreactor. After the production of asphotase alfa, a harvest cleaning process was used to remove intact cells and cellular debris using sterile filtration. The cell harvest was then ultrafiltered (Post Harvest UF) for concentration and buffer dilution. Further processes include, for example, viral inactivation (for chemical inactivation of viral particles), MabSelect SuRe chromatography, hydrophobic interaction chromatography (HIC), post-HIC UF/DF (UF/DF2), capto-adhesion in mixed chromatography, virus filtration (with using size exclusion), composition (UF/DF3) and bulk filler. Numerous production processes have been carried out, including, for example, a 2000 l-scale process and the next increase to a production scale of 20,000 liters. Illustrative differences between the 2000 L (2K) and 20000 L (20K) processes are shown in Tables 2 and 3. The 2000 L (2K) scale process had a more pronounced lag phase and more variable late viability (data not shown).

Table 2. Comparative Analysis of Inoculum Parameters Between an Exemplary 2000 L Process and a 20,000 L Process

Parameter Process 2000 l Process 20000 l culture vessel Shaking flasks, cell bags Shaking Flasks, Spinning Bottles, Cell Bags, Expansion Bioreactors Temperature 37°C 36.5-37.5°C Wednesday HyQSFM4CHO with 4 mM glutamine and 100 nM methotrexate HyQSFM4CHO with 4 mM glutamine and 100 nM methotrexate Target population Thaw: 0.50×10 ⁶ viable cells/ml
Passage: 0.40×10 ⁶ viable cells/ml Thaw: 0.50×10 ⁶ viable cells/ml
Passage: 0.35×10 ⁶ viable cells/ml Passage length 3-4 days 3-4 days Bioreactor production time Fixed (7 passages, 21-28 days) Varies (11 passages, with the possibility of maintaining a high volume stage per 100 L) Cell density during passage ≥1.00×10 ⁶ viable cells/ml ≥1.40×10 ⁶ viable cells/ml

Table 3. Comparison of production bioreactor parameters between an exemplary 2000 L process and a 20,000 L process

Parameter Process 2000 l Process 20000 l Temperature 37°C at inoculation;
30°C when the cell density is 2.5-2.8×10 ⁶ viable cells/ml 37°C at inoculation;
30°C when the cell density is 2.5-3.4×10 ⁶ viable cells/ml pH The pH was adjusted to 7.10 on day 1, to 7.00 on day 2, and then controlled at 6.90 from day 3 to harvest 6.90±0.10 Dissolved oxygen 40% 40% Seeding density 0.50×10 ⁶ viable cells/ml 0.55×10 ⁶ viable cells/ml Serving time Single bolus refill to temperature shift Single bolus refill to temperature shift Foam control Defoamer Antifoam-C as needed Duration of cultivation Day 10 or when viability <50% Day 10 or when viability <60%

[0123] The overall yields of 2K processes and 20K processes were compared, as well as the quality of the corresponding asphotase alpha produced. Analytical methods used to compare product characteristics include, for example, SEC-HPLC, RP-HPLC, and other methods to measure the specific activity, protein concentration, pH, and total sialic acid content (TSAC) of asphotase alpha produced. In addition, contamination and safety tests were performed to measure, for example, residual DNA, residual protein A, host cell proteins, bioburden, and endotoxin in asphotase alpha derived from different processes. Three additional tests were also performed, i.e., isoelectric focusing (IEF), dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE), and oligosaccharide mapping to compare drugs from the 2K and 20K batch scales. The results of these three tests showed comparable profiles for these batches. In conclusion, the quality of asfotase alfa between the 2000 L and 20,000 L scales was comparable across all lots.

Example 2: Effect of Temperature Shift on Productivity and Quality of Asphotase Alpha

[0124] Among the various manufacturing processes adopted and practiced for the production of sTNALP-Fc-D ₁₀ , temperature variation generally has an impact on the performance and final quality of asphotase alpha. The temperature shift from growth temperature (comparatively high temperature) to production temperature (comparatively low temperature) was implemented in all processes.

[0125] To compare the effects of temperature shift versus no shift, repeated runs of the sTNALP-Fc-D ₁₀ bioreactor were performed with or without temperature shift. Sartorius 2L and 10L bioreactors have been used for various industrial processes. In this exemplary process, the feedstock used in the production bioreactors included, for example, production medium, 10% sodium carbonate, CHO feed, and glutamine stock solution as shown in Table 4. SFM4CHO refers to serum-free CHO medium.

Table 4. Process parameter in an exemplary process

Process parameter Process #1 cell line Cell line X (derived from CHO) Work environment CM70 (SFM4CHO+Pluronic+Sodium Bicarbonate)+Glutamine Seeding density 5.5e ⁵ cells/ml pH 6.90 without (w/o) dead zone DO 40% (air 0.0016 VVM+O ₂ if required) Temperature 37.0°C ->30°C (when VCD is 25-32e ⁵ cells/ml) make-up 5 ml/l CHO refill when VCD is 25-32e ⁵ cells/ml Other additives If [gln] < 0.18 g/L (1.2 mM), increase [gln] to 0.29 g/L (2.0 mM) before temperature shift. Collection Criteria 240±12 hours (within 30 hours viability <60%)

Ways

Design of Cell Culture Experiments

[0126] Two sets of experiments were implemented to evaluate the effect of temperature shift on the performance and quality characteristics of asphotase alpha for an exemplary process. As shown in Table 5, two bioreactor runs (i.e., #1 and #2) were run with a temperature shift (37°C to 30°C), and the other two runs (i.e., #3 and #4) were carried out without a temperature shift from 37°C.

Table 5. Bioreactor Conditions for Process No. 1

Bioreactor ID Condition #1 Pace. shift 1 #2 Pace. shift 2 #3 No temp. shift 1 #4 No temp. shift 2

[0127] Cell density and viability were counted using a cell counter (ie ViCell VR, Beckman Coulter). pH and autonomic gas were measured using pHOx, and major metabolites, including glucose and lactate, were measured using a sensor (Nova Profile 100, Nova Biomedical, Waltham, MA). Enzymatic activity was measured in a standard manner with the modification that each sample was diluted only once, rather than three times, prior to measuring enzymatic activity.

Harvesting and cleaning methods

[0128] Fifty microliters of day 10 (240±4 hours) samples were collected from all bioreactors using syringes. After removal of cells by centrifugation (3000×g, 5 min), supernatants were clarified using 0.22 μm vial neck filters and stored at -80°C until purification. In this study, a single high throughput protein A purification step was applied. Samples were transferred to a buffer (Buffer Exchange) with low salt content (5 mm Na ₃ PO ₄ , pH 7.4) prior to analytical analysis and analysis of protein characterization.

Analytical methods and protein characterization methods

[0129] Quality criteria analyzed in this study included asphotase alpha aggregation, fragmentation, charge distribution, total sialic acid content (TSAC), and levels of neutral glycan species. Aggregate level was assessed as the percentage of cumulative peaks to total protein as determined by SEC. Fragment level was assessed as the percentage of fragment to total protein measured in LoC. The charge distribution was assessed by the percentage of main peaks, main peak and acid peak to total protein, respectively, quantified in AEX. Total sialic acid content (TSAC) was calculated using HPAE-PAD. Detection of neutral glycan species was performed using MALDI-TOF mass spectrometry.

results

Cell culture performance

[0130] In process #1, the temperature setpoint was shifted from 37°C to 30°C within 5 hours after the viable cell density (VCD) reached 25-32×10 ⁵ cells/ml. Without the temperature shift, the cell culture achieved a higher peak VCD but experienced a faster decline in viability (Figure 1A). Higher total glucose uptake was also observed under conditions without temperature shift (FIG. 1B). Glucose bolus addition was applied on day 5 and day 8 under no temperature shift conditions as described in the process. Interestingly, lactate was still consumed, but at a significantly lower rate under no temperature shift conditions than with temperature shift conditions (FIG. 1B).

[0131] Higher peak VCD under no temperature shift conditions resulted in early production of protein A binding titer (FIG. 2A). With a faster decline in viability under no temperature shift conditions, similar binding protein A titers were achieved on day 10 under both conditions. The total specific productivity capable of binding to protein A was 6.4 pg/cells/day for the temperature shift condition and 6.7 pg/cells/day for the non-temperature shift condition. Interestingly, although a similar binding protein A titer was observed under both conditions, the bulk activity under no temperature shift conditions was significantly lower than the temperature shift condition (FIG. 2B). In addition, the specific activity under no temperature shift conditions was significantly lower than the temperature shift condition throughout the duration of the culture (FIG. 2C). These data show that a temperature shift from 37°C to 33°C is optimal to maintain specific activity for Process #1.

All quantitative quality indicators determined in process No. 1 are shown in table 6.

Table 6. Quantitative indicators of quality for process No. 1

Sample ID Condition AEX (%Main) AEX (% Main) AEX (% Sour) LoC (% Main, NR) LoC (% Main, R) SEC (% Agr.) SEC (% dimer) SEC (% Fragment) TSAC #1 Pace. shift 0.6 98.5 0.9 95.9 100.0 10.3 89.6 0.0 1.9 #2 Pace. shift 0.2 98.8 1.1 97.1 100.0 10.9 89.1 0.0 2.0 #3 No temp. shear 0.6 99.1 0.2 100.0 96.8 24.0 76.0 0.0 3.9 #4 No temp. shear 0.6 99.2 0.2 100.0 98.3 25.6 74.4 0.0 3.9

[0132] AEX results showed that fewer acidic species were produced under the no temperature shift condition compared to the temperature shift condition (FIG. 3A). However, significantly higher aggregates were quantified using SEC results under no temperature shift conditions (Fig. 3B). Under conditions without temperature shift, the average LoC peak percentage under non-reducing conditions was higher (FIG. 4A, also in the "LoC (% Principal, NR)" column in Table 6), while the recovered average LoC peak percentage was lower (FIG. 4B, also in the column "LoC (% Principal, R)" in Table 6).

[0133] Sialylation was approximately doubled under conditions without temperature shift (FIG. 4C).

[0134] MALDI-TOF analysis of neutral glycans did not detect any higher order mannose species or atypical glycan species found under non-temperature shift conditions (FIG. 5). However, the no temperature shift condition resulted in a decrease in A2 (the predominant afucosylated glycan species under temperature controlled conditions), higher fucosylation (higher ratio of FA2 to A2), and an increase in higher order glycans including FA3G3, FA4G3, FA4G1L1 and FA4G4L1 ( Fig. 5).

[0135] The effect of temperature without shift on the productivity and quality of asphotase alpha in process No. 1 is shown in Table 7. Without temperature shift, the resulting asphotase alpha had a significantly lower specific activity. Meanwhile, higher production temperatures under no temperature shift conditions resulted in higher sialylation and higher fucosylation.

Table 7. Effect of constant temperature on the productivity and quality of asphotase alfa

Options Process #1 - No temp. shear Protein A titer no influence Volume activity Significantly reduced volumetric activity Specific activity Significantly less specific activity AEX sour peak Less acidic peak SEC unit Further evaluation needed LoC Fragment Further evaluation needed TSAC Significantly increased TSAC Neutral glycan Higher fucosylation and higher order glycans

Example 3: Initial Evaluation of Upstream Process Parameters for Exemplary Process #2

[0136] This example summarizes an initial assessment of the prior process parameters for an exemplary asphotase alfa #2 manufacturing process and their potential to affect the critical quality characteristics of the produced asphotase alfa. Some of the upstream process parameters include % aggregation, % fragmentation, sialylation, glycosylation, charge distribution, and specific activity.

Cell culture

[0137] All manufacturing processes mentioned in this study were carried out either in shake flasks or in bioreactors. After thawing, cells were grown using a series of shake flasks and spinner flasks prior to inoculation of the production bioreactor. The production bioreactor (2 L, 5 L, 10 L, or 200 L) was scaled using matched power per volume and bubbling gas volume per liquid volume per minute (VVM) unless otherwise noted. The temperature of the production bioreactor was controlled at 36.5°C from 0 to about 120 hours and shifted to 33.0°C after about 120 hours, unless otherwise noted. The dissolved oxygen setpoint was maintained at 30%. Approximately 2.7% (v/v) of CPN nutrition per bolus was added at 96 hours (hour), 144 hours and 192 hours when the culture was harvested on day 10 (240 h±6 h) and an additional bolus nutrition was given at 240 hours when the culture was collected after the 10th day.

Collection and cleaning

[0138] Samples collected from production bioreactors were clarified by 0.22 µm filtration after centrifugation at 3000×g for 5 minutes. Purification of the clarified filtered crop was carried out, and the degree of purification was dictated by the requirements for the study and testing of the samples. For most of the samples tested, either a single chromatographic step (Protein A) or two consecutive chromatography steps (Protein A followed by Hydrophobic Interaction Chromatography) (HIC) was performed. In addition, prior to analytical testing, the samples were transferred (Buffer Exchange) to a low salt buffer (5 mM Na ₃ PO ₄ , pH 7.4).

Analytical characterization

[0139] The quality characteristics mentioned in this study were identified as subsets of broader quality measures such as purity, potency, and structure. Quality subtypes described include % Aggregation, % Fragmentation, Charge Distribution, Total Sialic Acid Content (TSAC), Neutral Glycan Profile, and Specific Activity. Both the aggregate level and the fragment level were determined using relative peak areas quantified using gel permeation HPLC (GP-HPLC). In addition, fragment level was determined using either SDS-PAGE or Lab-on-Chip (LoC) capillary electrophoresis. The charge distribution was evaluated using anion exchange (AEX) chromatography. TSAC was quantified by high performance anion exchange chromatography with pulsed amperometric detection (HPAE-PAD). Detection of neutral glycans was performed using matrix-assisted laser desorption/ionization (MALDI) mass spectrometry. Specific activity was measured after a para-nitrophenyl phosphate (pNPP) alkaline phosphatase (pNPP) enzymatic assay.

results

1. Growth and production temperature

[0140] The effect of temperature on product fragmentation is discussed. For asphotase alpha produced by a selected cell clone grown at 36.5°C, the main asphotase alpha band was calculated as 95.9%, with approximately 4% as fragments measured by SDS-PAGE. When the temperature was shifted from 36.5°C to 33.0°C on day 5, the same clone produced almost no asphotase alpha fragments (basic band calculated as >99%). Other secondary clones showed very similar trends in fragment recovery by changing the temperature from 36.5°C to 33.0°C, although not all secondary clones were able to reduce fragment levels to less than 1%. Therefore, growth temperature and production temperature can affect fragmentation. In addition, the time associated with conducting a temperature shift can also affect the quality of asphotase alpha.

[0141] Higher aggregated levels detected with GP-HPLC were observed at lower production temperatures during clone secondary screening in 2 L bioreactors. Figure 6A shows the aggregate level of asphotase alpha produced by 6 secondary clones when the temperature shifted to 33°C compared to material obtained with 10 secondary clones when the temperature was not shifted (i.e. maintained at 36.5°C ). These data indicate that a change in temperature from 36.5°C to 33.0°C can lead to an increase in the aggregate level.

[0142] In addition, the effect of temperature change on TSAC was investigated during initial clone selection and initial process development (FIG. 6B). The primary clone was used to investigate the effect of production temperature (day 5 onwards) on sialylation or TSAC in 2 L bioreactors. In this study, the production temperature was shifted from 36.5°C to 30.0°C on day 5 under condition 1, from 36.5°C to 33.0°C on day 5 under condition 2, and control was carried out without temperature shift. Samples were collected when viability was between 60% to 80% and these cultures were run for 11-18 days prior to harvest depending on viability. As shown in FIG. 6B, the introduction of a temperature shift resulted in a decrease in TSAC compared to the control at constant temperature at the same viability, and the transition to a lower temperature corresponded to a lower TSAC. The required minimum TSAC value in this document is 1.8. Because sialylation (or TSAC) changes the charge profile of the product and converts neutral glycans to negatively charged glycans, these three quality attributes (production temperature, aggregation, and TSAC) are considered to be closely related.

2.pH

[0143] Preliminary results showed that changing the pH set point from 6.90 to 7.00 resulted in an increase in TSAC (FIG. 7). Using a selected secondary clone, two 2 L bioreactors set to pH 6.90 and 7.00, respectively, were used to study the effect of pH on asphotase alpha sialylation. Samples were collected from the bioreactors on day 10 and day 12, respectively. It was noted that TSAC decreased from 1.9 on day 10 to 1.4 on day 12 under the condition of pH 6.90. Although the same downward trend in TSAC values was observed at pH 7.00, TSAC values at pH 7.00 on both days (2.6 on day 10 and 2.1 on day 12) were higher than those at lower pH level. These data indicate that TSAC decreases with culture time, and setting a lower pH results in a lower TSAC value.

3. Adding medium make-up

[0144] Preliminary data show that delaying the addition of the first bolus or increasing the glucose concentration in the medium can lead to a significant increase in TSAC compared to the control condition (Fig. 8A).

[0145] In addition, how the medium is fed (either continuously or by adding a bolus) also affects TSAC. Data from both the 2 L and 10 L sizes show that significantly higher TSACs were observed when the feed was added in the bolus mode compared to the continuous addition mode (FIG. 8B). Therefore, CPN recharge is defined as being added as a bolus. In addition, no effect on TSAC was observed when the make-up was delivered by adding through the top hole or by subsurface addition (immersion).

4. Addition of zinc sulfate

[0146] Zinc ions are known to be essential for the stability of alkaline phosphatase (eg, asphotase alfa) as it helps maintain their structure and activity. For example, two zinc atoms are associated with one placental alkaline phosphatase molecule (Helene Le Du et al., 2001 J. Biol. Chem. 276 :9158-9165). Based on this ratio, for a titer of 1 g/L asphotase alpha, obtained from an illustrative manufacturing process developed in small scale models, approximately 20 μM zinc is required for asphotase alpha activity.

[0147] It has been noted that as the cell culture progresses, protein A-binding asfotase alpha is produced continuously, but the corresponding volumetric activity is not synchronized with protein A titer, indicating that the protein produced is inactive (FIG. 9A). In addition, although the volumetric activity decreased from day 5 to day 9, protein activity was immediately restored after the addition of zinc sulfate on day 9. These data indicate the need to add zinc to the production environment. In this study, the zinc sulfate concentration was increased by 25 µM by adding a bolus on day 9. Given that the protein A titer was approximately 1000 mg/l on day 9, this addition of a bolus of zinc increased the ratio of zinc ion to asphotase alpha to at least 2:1.

[0148] In the following study, the level of cytotoxicity of zinc sulfate on cell growth and viability was investigated using a mini-bioreactor system in batch mode, in which zinc sulfate at various concentrations was added to the culture medium on day 0. When the zinc sulfate concentration was below 300 μM, there was little effect on growth or viability (FIGS. 9B and 17C). However, under conditions where the zinc concentration was greater than or equal to 500 μM, both growth and viability deteriorated sharply. Therefore, the optimal concentration of total zinc was below 500 μM. Thus, it has been estimated that the optimal concentration of zinc to be applied on day 0 may be between about 25 μM to about 300 μM (taking into account both cell growth/viability and problems with protein function). In fact, the zinc concentration at 150 μM may even be better than 300 μM, as the former results in less inhibition of cell growth after day 5 (FIG. 9A). It is also surprising that while 20 µM zinc is theoretically sufficient to produce functional asphotase alpha (i.e. 2 zinc ions per active enzyme), actual zinc supplementation may require significantly higher concentrations of zinc (e.g. 150 µM) in alkaline phosphates (e.g., asphotase alfa, TNALP, PALP, GCALP, IAP or their fusion proteins/protein variants) manufacturing processes.

5. Population density

[0149] Typically, a higher seeding density results in a higher peak viable cell density, which is primarily due to certain nutrient limitations in the production environment. Because feed and temperature shift are based on culture time, seed density, in combination with other upstream process parameters, especially temperature shift time, can potentially affect the quality of asphotase alpha.

6. Collection time

[0150] The collection time in this exemplary process was 240±6 hours. The effect of collection time on TSAC has been demonstrated (see, for example, Figure 7). In addition, harvest time is also considered to be related to viability (see, for example, Figure 14), and the decrease in viability towards the end of the culture indicates that harvest time can potentially affect the quality of asphotase alpha.

[0151] Various upstream process parameters were evaluated, including control of seed population, temperature, pH, DO, gassing strategies, agitation, CPN feed, glucose addition, galactose addition, zinc addition, and harvest time. Additional parameters of the prior process were evaluated, including % aggregation, % fragmentation, sialylation, glycosylation, charge distribution and specific activity, for example, as described below.

[0152] Temperature has a strong influence on aggregation, fragmentation and sialylation. pH also influences the sialylation of asphotase alfa. The gassing and/or agitation strategy appears to affect pot life. The addition of CPN recharge appears to affect sialylation. Zinc supplementation was shown to be necessary to maintain enzymatic activity and the amount of zinc supplementation (150 μM) was determined based on the studies presented. Harvest time has been shown to be related to viability, with TSAC decreasing towards the end of culture.

Example 4: Further Evaluation of Preceding Process Parameters for Exemplary Process #3

[0153] This example summarizes a further characterization of prior processes that were performed to screen bioreactor process parameters. For initial evaluations in this exemplary process, a list of prior process parameters was tested including culture production pH, temperature, seed density, gassing strategy, agitation, CPN feed, glucose addition, galactose addition, and harvest time.

ABBREVIATIONS

[0154] CPP: critical process parameter, a process parameter (process input) that affects the critical quality characteristic and therefore will be monitored and controlled to ensure that the process produces the desired quality.

CQA: critical quality characteristic

HPLC: High Performance Liquid Chromatography

LoQ: Limit of quantification

P/V: Power per volume

Materials and methods

[0155] Were further investigated the parameters of the process, which were shown in the initial experiments with potential impact. Six blocks of bioreactors were carried out in the following study characterization process (Table 8). Three levels for each process parameter were tested with a median level as the current setpoint (control), taking into account the control capabilities in manufacturing and testing a range of process parameters. In blocks 1 to 5, a three-level full factorial design with two replication controls is implemented. In block 6, four combinations of extreme P/V and VVM charges were run, plus two controls.

Table 8. Process parameters and levels studied

Block Process parameter Researched levels Short Average High 1 Glucose Addition Threshold (g/L) 1.5 2.0 3.0 Addition of galactose (g/l) 7 10 15 2 Starting point pH 6.75 6.90 7.10 Production temperature ( ^o C) 30.0 33.0 35.0 3 growth temperature 35.0 36.5 37.5 Starting point DO (%) 15 thirty 60 4 Feed Amount (%) 67 100 133 Feed start (day) 3 4 5 5 Seeding density
(105 cells/ml) 4.0 5.5 8.0 Temperature shift time (h) 108 120 132 6 P/V (W/m3) 9->41 17->81 26->121 Maximum bubbling 0.015 0.020 0.025

Cell culture

[0156] All manufacturing processes mentioned in this study were carried out in either 2 L or 10 L bioreactors. After thawing, cells were grown using a series of shake flasks and spinner flasks prior to inoculation of the production bioreactor. The production bioreactor (2 L or 10 L) was scaled up using P/V and VVM sparging unless otherwise noted. Under control conditions, the cell culture process parameters were as follows. The initial pH of the culture was 6.90 with a 0.05 dead zone. The temperature was controlled at 36.5°C from 0 to 120 hours and shifted to 33°C after 120 hours. The dissolved oxygen setpoint was maintained at 30%. Initial P/V was 17 W/m ³ (0-96 hours) and shifted to 81 W/m 3 before the first addition of the bolus feed at 96 hours. 2.67% (v/v) CPN feed per bolus was added at 96 hours, 168 hours and 192 hours. Glucose was maintained at ≤2 g/L. If the glucose was below 2 g/L, a daily bolus addition was used to increase the glucose concentration in the range of 1.8 to 2.0 g/L. Galactose was added throughout the culture: 2.0 g/l on day 0 on day 4 and 1.0 g/l daily from day 5 to day 9. Part of the culture was collected on the 10th day (240±4 hours) and on the 11th day (264±4 hours).

Collection and cleaning

[0157] Samples collected from production bioreactors were clarified with a 0.22 µm filter after centrifugation at 3000×g for 5 minutes. The clarified filtered crop was cleaned. The degree of purification was dictated by the requirements for the study and testing of samples. Either one step (Protein A) or two consecutive steps (Protein A and HIC) were implemented. In addition, prior to analytical testing, the samples were transferred to a low salt buffer (5 mM Na ₃ PO ₄ , pH 7.4).

Analytical characterization

[0158] Quality criteria specified in this study include % aggregation, % fragmentation, charge distribution, total sialic acid content (TSAC), and neutral glycan profile. Both the aggregate level and the fragment level were determined using relative peak areas quantified using size exclusion chromatography (SEC). In addition, the purity of asphotase alpha (main peak) was also determined using Lab-on-Chip (LoC) capillary electrophoresis. The charge distribution was evaluated using anion exchange chromatography (AEX). TSAC was quantified by high performance anion exchange chromatography with pulsed amperometric detection (HPAE-PAD). Detection of neutral glycan species was performed using time-of-flight laser desorption ionization using matrix (MALDI-TOF) mass spectrometry. Volumetric activity was measured after a para-nitrophenyl phosphate (pNPP) alkaline phosphatase (pNPP) enzymatic assay.

Statistical analysis and data visualization

[0159] The potential of the process variables being tested to affect performance and each quality attribute was evaluated using the statistical software package JMP (Carey, NC). Three continuous parameters were analyzed (three levels for the two process parameters tested, respectively, and two levels for collection). Individual p-values calculated with a two-tailed t-test were reported and 0.05 was chosen as the threshold of significance. Contour plots generated with JMP were used to visualize the potential impact of test parameters on each quality attribute.

results

[0160] Because the acceptable range for each CQA in this exemplary process was still under investigation, an exemplary experimental range was created and used for comparison purposes (Table 9). Baseline data was derived from all day 10 results obtained in this study. For impurity measurements, including fragment and aggregate with SEC, and base and acid peak with AEX, the mean value plus two times the standard deviation was used as the upper threshold. Since the calculated upper threshold for the fragment measured by SEC and the base peak measured by AEX are lower than the LoQ (1%) of the SEC and AEX analysis, the upper threshold was set to 1% for both qualities. For TSAC, the mean, minus, and plus two times the standard deviation were used as the experimental range.

Table 9. Small scale experimental ranges for CQA

Critical quality characteristics TSAC SEC aggregation (%) Fragment SEC (%) Main Peak AEX (%) Acid peak AEX (%) Average (average) 2.2 8.0 0.5 0.4 8.6 Standard deviation (stotcl) 0.4 0.8 0.2 0.1 0.8 Average - 1 x stocl 1.8 7.3 0.3 0.3 7.8 Average - 2 x stocl 1.5 6.5 0.1 0.2 7.0 Average+1 x stdcl 2.5 8.8 0.7 0.5 9.4 Average+1 x stdcl 2.9 9.6 0.8 0.6 10.2 Experimental Range 1.5 to 2.9 ≤9.6 ≤1.0 ≤1.0 ≤10.2

[0161] The significance of the direct impact of each process variable on each CQA is summarized in Table 10. The p -value of P/V and the bubbling effect of VVM were not calculated due to lack of statistical power. p -values less than or equal to 0.05 are considered statistically significant.

Table 10. The value of the direct impact of each process parameter on CQA, analyzed using JMP

p-value TSAC SEC AEX Unit Fragment Basic Acid Glucose 0.41 0.63 - 0.98 0.67 Galactose 0.04 0.37 - 0.14 0.44 pH 0.00 0.63 0.00 0.05 0.29 Production temperature 0.00 0.00 0.00 0.64 0.00 DO 0.72 0.73 0.72 0.64 0.79 growth temperature 0.16 0.47 0.01 0.26 0.04 Make-up amount 0.03 0.00 0.14 0.70 0.00 Make-up time 0.37 0.00 0.71 0.03 0.00 Seeding density 0.81 0.34 0.00 0.31 0.49 Temperature time
shear 0.06 0.64 0.00 0.70 0.26

[0162] Table 11 summarizes the average value of the quality characteristic obtained at the level of the minimum and maximum test parameters of the process. For example, in a glucose and galactose supplementation study, three bioreactors used 1.5 g/L as the glucose addition threshold, and each had a different level of galactose addition. The CQA value at the 1.5 g/L glucose threshold was then calculated using the average of these three bioreactors. In the following sections, day 10 data is used to discuss quality unless otherwise noted.

Table 11. Average value of the quality characteristic under the tested conditions

Process parameters and tested ranges critical quality characteristics TSAC Unit, SEC (%) Fragment, SEC (%) Main peak, AEX (%) Acid peak, AEX (%) Glucose threshold (g/l) Min 1.5 2.5 5.9 0.0 0.3 5.9 Max 3.0 2.4 5.5 0.0 0.4 5.5 Total galactose addition (g/l) Min 7 2.3 5.5 0.0 0.3 5.6 Max 15 2.4 6.3 0.0 0.5* 5.7* pH Min 6.75 1.9 5.7 0.0 0.5 5.7 Max 7.10 2.7 5.7 0.1 0.3 6.1 Production temperature ( ^o C) Min 30.0 1.6 8.2 0.0 0.3 8.5 Max 35.0 2.7 3.8 0.1 0.6 3.7 Dissolved oxygen (%) Min 15 2.3 7.8 0.5 0.4 8.7 Max 60 2.3 7.6 0.6 0.4 8.9 Growth temperature ( ^o C) Min 35.0 2.2 7.7 0.4 0.3 8.5 Max 37.5 2.4 7.6 0.8 0.4 9.2 Feed Amount (%) Min 67 2.1 8.5 0.4 0.3 8.9 Max 133 1.9 8.8 0.4 0.3 9.4 Start of recharge (days) Min 3 2.1 8.9 0.4 0.4 9.3 Max 5 2.0 8.2 0.4 0.3 8.6 Seeding (10 ⁵ cells/ml) Min 4.0 2.4 9.0 0.6 0.5 8.3 Max 8.0 2.4 8.3 0.8 0.6 8.9 Temperature shift time (h) Min 108 2.5 8.6 0.6 0.5 8.5 Max 132 2.3 8.8 0.8 0.5 9.2 Collection (days) Min 10 2.1 8.2 0.4 0.3 8.4 Max eleven 1.8 8.1 0.5 0.4 8.9 Generation No. Min 38 1.9 8.8 0.4 0.4 8.6 Max 48 2.1 7.7 0.5 0.4 8.7 P/V (W/m ³ ) Min 9
(0-96 h);
41 (96-240 h); 2.2 6.5 0.3 0.5 6.8 Max 26
(0-96 h);
121
(96-240 hours); 1.9 6.9 0.3 0.2 7.3 Maximum load VVM Min 0.015 2.0 6.7 0.3 0.4 7.1 Max 0.026 2.1 6.8 0.3 0.3 7.0

Note: When the average percentage of main peak and acid peak measured by AEX was calculated for a state of 15 g/l total galactose addition, the outlier sample was excluded due to its low peak response in the assay. His biological re-sample from another bioreactor was left for calculation purposes.

1.pH

[0163] Previous results have shown that pH affects the TSAC of asphotase alpha produced. The potential effect of culture pH and production temperature on the production and quality of asphotase alfa was further investigated in 2 L bioreactors. A full factorial construct was implemented containing three culture pH levels (6.75, 6.90 and 7.10) and three production temperature levels (30.0°C, 33.0°C and 35.0°C), and samples from all conditions were collected at 240±4 hours and 264±4 hours, respectively. Prior to analytical analysis, a two-step purification (ProA+HIC) was used.

[0164] Setting the culture pH in the tested range of 6.75 to 7.10 was not considered a critical process parameter (CPP) in that all CQAs tested were within the small experimental range shown in Table 9. However, it was shown that pH has a significant effect on several quality characteristics tested. Increasing the pH setpoint from 6.75 to 7.10 resulted in an increase in TSAC from 1.9 to 2.7 at a post-shear production temperature of 33° C. (FIG. 10A). pH also had a significant effect on fragmentation measured by SEC and main peak measured by AEX (Table 10), but the effect was less than or equal to 0.2% (Table 11), which was below the HPLC-based analysis variance of 0 ,5%. In addition, at the control condition (pH 6.90 and product temperature 33.0° C.), maximum performance was achieved in both protein A titer and volumetric activity (FIGS. 10B and 10C).

[0165] The data show that the effect of the pH setpoint in the range of 6.75 to 7.10 on CQA was within the experimental range, suggesting that the production culture pH is unlikely to be CPP. In various embodiments of the invention, an alkaline phosphatase (e.g., asphotase alfa) is obtained at a given pH value of 6.65, 6.70, 6.75, 6.80, 6.85, 6.90, 6.95, 7.00 , 7.05, 7.10, 7.15, 7.2 or higher.

2. Temperature

[0166] Preliminary results have shown that temperature has a strong influence on several parameters, including aggregation, fragmentation, and TSAC. As with previous processes using a temperature shift from 36.5°C to 33.0°C after 120 hours, the study of the effect of temperature was divided into three parameters: growth temperature (from 0 to 120 hours), production temperature (from 120 hours to harvest ) and temperature shift time. These three temperature-related parameters were further investigated separately and summarized below.

2.1 Growth Temperature

[0167] The effect of growth temperature was investigated along with the DO set point in 2 L bioreactors. Three growth temperature levels (35.0°C, 36.5°C and 37.5°C) were maintained from 0 to 120 hours under different test conditions followed by a temperature shift to 33.0°C after 120 hours. Samples from all conditions were collected at 240±4 hours and 264±4 hours. A one-step purification (ProA) was used prior to analytical analysis. Growth temperature affects growth rate and viability. A higher growth temperature correlated with faster growth rates in the early stages, followed by a rapid decrease in temperature after the temperature shift, resulting in a lower VCD peak. In addition, cultures with a higher growth temperature also had a 2-4% lower terminal viability on the day of harvest. Interestingly, growth temperature has a significant effect on performance, with volumetric activity peaking at 36.5°C growth temperature (FIG. 11A).

[0168] With regard to the effect on the quality of asphotase alpha, growth temperature has been shown to have a significant effect only on fragmentation measured by SEC (FIG. 11B). Increasing the growth temperature from 35.0° C. to 37.5° C. resulted in the fragment level being increased from 0.3% to 0.8%. It is noted that these measured values were below the LoQ (1%) of the SEC analysis.

[0169] Since the effect on CQAs with growth temperature was within the experimental range (Table 9 and Table 11), growth temperatures ranging from 35.0°C to 37.5°C are unlikely to be CPP. In various embodiments of the invention, alkaline phosphatase (for example, asfotase alfa) is obtained at a growth temperature of 33.0°C, 33.5°C, 34.0°C, 34.5°C, 35.0°C, 35, 5°C, 36.0°C, 36.5°C, 37.0°C, 37.5°C, 38.0°C or 38.5°C. In one specific embodiment of the invention, alkaline phosphatase (eg, asfotase alfa) is produced at a growth temperature of 35.0°C, 36.5°C, 37.0°C, or 37.5°C.

2.2 Production temperature

[0170] The effect of production temperature (30.0° C., 33.0° C., and 35.0° C.) on CQA was investigated along with the set culture pH in 2 L bioreactors. A one-step purification (ProA) was used prior to analytical analysis.

[0171] Manufacturing temperature significantly affects CQA, especially TSAC, aggregation and fragmentation measured by SEC, and acid peak measured by AEX (Table 10 and Table 11). Increasing the production temperature from 30.0°C to 35.0°C resulted in an average increase in TSAC from 1.6 to 2.7 (FIG. 10A) and decreasing the production temperature from 35.0°C to 30.0°C resulted in a significant increase in aggregates as measured by SEC from 3.8% to 8.2% (FIG. 12A). In addition, the acid peak measured by AEX increased from 3.7% to 8.5% as the production temperature was lowered (FIG. 12B). These data suggest that a production temperature higher than 30.0°C is likely beneficial in reducing the risk of low TSAC and preventing an increase in aggregation/acid peak. Meanwhile, there appears to be a very fine positive correlation between production temperature and fragmentation as measured by SEC (FIG. 12C) and the main peak as measured by AEX (FIG. 12D). These data suggest that a production temperature lower than 35°C may be preferable to reduce the risk of fragmentation of asphotase alfa.

[0172] In terms of performance impact, production temperature has the greatest impact on protein A titer and volumetric activity (FIGS. 10B and 10C). In the range from 30.0°C to 35.0°C, an increase in production temperature correlates with an increase in productivity. However, below pH 7.10, the production temperature interacts with the pH to increase productivity, with the maximum volumetric efficiency occurring at a production temperature of 33.0°C. It has been shown that peak productivity in terms of both volumetric activity and protein A titer occurs at pH 6.90 at a production temperature greater than or equal to 33.0°C. Although production temperature was shown to have a significant effect on several CQAs (Table 10), the effect was still within the experimental range (Table 9 and Table 11). Therefore, a production temperature in the range of 30.0°C to 35.0°C is unlikely to be CPP.

In various embodiments of the invention, alkaline phosphatase (for example, asfotase alfa) is produced at a production temperature of 29.0°C, 29.5°C, 30.0°C, 30.5°C, 31.0°C, 31, 5°C, 32.0°C, 32.5°C, 33.0°C, 33.5°C, 34.0°C, 34.5°C, 35.0°C, 35.5° C, 36°C or higher. In one particular embodiment of the invention, alkaline phosphatase (eg, asphotase alfa) is produced at a production temperature of 30.0°C, 30.5°C, or 35.0°C.

3. Temperature shift time

[0173] The effect of temperature shift time on CQA was investigated along with seeding density in 2 L bioreactors. Three temperature shift times were tested (108 hours, 120 hours and 132 hours) and samples from all conditions were collected at 240±4 hours and 264±4 hours. A one-step purification (Protein A) was used prior to analytical analysis.

[0174] The temperature shift time has little effect on volumetric activity (FIG. 13A). At high seeding densities, delaying the temperature shift time from 108 to 132 hours appears to reduce volumetric activity. However, at a lower seeding density, the maximum volumetric activity is observed under the condition of a later temperature shift.

[0175] In terms of asphotase alpha quality, temperature shift time has been shown to have a significant effect on fragmentation as measured by SEC (FIG. 13B). Assuming the last temperature shift (132 hours), the average fragment level was 0.8%, close to the small scale experimental range for fragmentation (<1.0%). Therefore, a narrower temperature shift time window is likely to be useful in reducing the risk of fragmentation.

[0176] Because the effect on CQA by temperature shift time was within the experimental range (Table 9 and Table 11), a temperature shift time in the range of 108 to 132 hours is unlikely to be CPP.

[0177] In various embodiments, alkaline phosphatase (e.g., asphotase alfa) is produced by a process in which there is a temperature shift from about 100 hours, 105 hours, 108 hours, 110 hours, 115 hours, 120 hours, 125 hours, 130 hours, 132 hours, 135 hours, 140 hours, 145 hours or 150 hours after the starting point of growth (eg inoculation). In one particular embodiment of the invention, alkaline phosphatase (eg, asphotase alfa) is produced by a process in which a temperature shift occurs at about 108 hours, 120 hours, or 132 hours. In various embodiments, alkaline phosphatase (e.g., asphotase alfa) is collected at about 200 hours, 210 hours, 220 hours, 230 hours, 240 hours, 250 hours, 260 hours, 264 hours, 270 hours, 280 hours, 288 hours (i.e. 12 days) or more than 12 days. In one particular embodiment of the invention, alkaline phosphatase (eg, asphotase alfa) is collected at about 240, 264, or 288 hours.

4. Feeding the environment

[0178] Previous results have shown that the addition of feed affects sialylation. As a control for this exemplary process, three boluses of medium were added to the production bioreactor on day 4, day 6 and day 8, and the total amount was equivalent to 8% (v/v) of the original culture volume. With the same number of boluses, three replenishment amounts were tested: 67%, 100% (control), and 133%. In terms of timing of bolus addition, three strategies were investigated: (1) day 3, day 5, and day 7; (2) day 4, day 6 and day 8 (control); (3) 5th day, 7th day and 9th day. All bioreactions were carried out in 2 L bioreactors and samples from all conditions were collected within 240±4 hours and 264±4 hours. A one-step purification (Protein A) was used prior to analytical analysis.

[0179] Although the average amount of feed and initiation time did not directly affect performance, their interaction appeared to have little effect on volumetric activity (FIG. 14A). Under control conditions, the maximum volumetric activity was achieved. When 33% less feed was provided, delaying the start of the feed resulted in lower volumetric activity.

[0180] The increased acid peak (10.1%) and aggregation (9.5%) observed in the higher (133%) and earlier (first bolus on day 3) fed state is close to the small-scale experimental range (10, 2% for acid peak and 9.6% for aggregation, Table 9), indicating that at least one parameter requires tighter scaling control, e.g., 15% change in average recharge amount per bolus and/or relatively short addition time window (96±3 hours, 144±3 hours and 192±3 hours). Since the effect on CQA by the addition of medium fed was within the experimental range (Table 9 and Table 11), the average amount of medium fed was in the range of 67% to 133% and the feeding time (96±24 hours, 144±24 hours and 192±24 hours) is unlikely to be CPP.

[0181] In various embodiments, alkaline phosphatase (eg, asphotase alfa) is produced by a process in which additional boluses of culture medium are added to the production bioreactor. For example, one, two, three, four, five, six or more culture medium boluses may be added. In one particular embodiment of the invention, three boluses of culture medium are added. In various embodiments of the invention, such additional boluses of the culture medium can be added in various amounts. For example, such culture medium boluses can be added in amounts of about 20%, 25%, 30%, 33%, 40%, 45%, 50%, 60%, 67%, 70%, 75%, 80%, 90% , 100%, 110%, 120%, 125%, 130%, 133%, 140%, 150%, 160%, 167%, 170%, 175%, 180%, 190%, 200% or more of original volume culture medium in the production bioreactor. In one specific embodiment of the invention, such culture medium boluses can be added in an amount of about 33%, 67%, 100%, or 133% of the original volume. In various embodiments of the invention, such addition of additional boluses may occur at different times during cell growth or during protein production. For example, boluses can be added on Day 1, Day 2, Day 3, Day 4, Day 5, Day 6, Day 7, Day 8, Day 9 day, day 10, day 11, day 12 or later in the process. In one particular embodiment of the invention, such culture medium boluses can be added every other day (e.g., on (1) Day 3, Day 5, and Day 7; (2) Day 4, Day 6 and Day 8, or (3) Day 5, Day 7, and Day 9. In practice, the frequency, amount, time point, and other parameters of culture medium bolus additions can be freely combined according to the above limitation and determined experimental practice.

5. Population density

[0182] In this study, the effect of seed density on CQA was investigated along with temperature shift time in 2 L bioreactors. Three seeding densities were tested (4.0×10 ⁵ cells/ml, 5.5×10 ⁵ cells/ml and 8.0×10 ⁵ cells/ml) and samples from all conditions were collected at 240±4 hours and 264 ±4 hours. A one-step purification (Protein A) was used prior to analytical analysis.

[0183] The temperature shift time has a very strong impact on performance. As shown in FIG. 13A, increasing the seed density from 4.0×10 ⁵ cells/mL to 8.0×10 ⁵ cells/mL resulted in an increase in volumetric activity from 806 U/mL to 1012 U/mL on average.

[0184] Regarding the quality of asphotase alpha, seeding density was shown to significantly affect fragmentation as measured by SEC (FIG. 13B). At the highest seeding density (8.0×10 ⁵ cells/mL), the average fragment level was 0.8%, close to the small scale experimental range for fragmentation (<1.0%). Therefore, a denser seeding density range may be useful to reduce the risk of fragmentation at the scale.

[0185] Since the effect on CQA by seeding density was within the experimental range (Table 9 and Table 11), seeding density ranging from 4.0×10 ⁵ cells/mL to 8.0×10 ⁵ cells/mL is unlikely will be CPP. In various embodiments of the invention, alkaline phosphatase (for example, asfotase alfa) is produced by a method in which cells are seeded at a density of about 1.0×10 ⁵ cells/ml, 1.5×10 ⁵ cells/ml, 2.0×10 ⁵ cells/ml, 2.5×10 ⁵ cells/ml, 3.0×10 5 cells/ml, 3.5×10 ^{5 cells} /ml, 4.0×10 ⁵ cells/ml, 4.5×10 ⁵ cells/ml, 5.0×10 ⁵ cells/ml, 5.5×10 ⁵ cells/ml, 6.0×10 ⁵ cells/ml, 6.5×10 ⁵ cells/ml, 7.0×10 ⁵ cells/ml, 7.5×10 ⁵ cells/ml, 8.0×10 5 cells/ml, 8.5×10 ⁵ cells/ml, 9.0× ^{10 5 cells} /ml, 9.5×10 ⁵ cells/ml, 1.0×10 ⁶ cells/ml, 1.5×10 ⁶ cells/ml, 2.0×10 ⁶ cells/ml or higher density. In one specific embodiment of the invention, in such processes cells are seeded at a density of about 4.0×10 ⁵ cells/ml, 5.5×10 ⁵ cells/ml, or 8.0×10 ⁵ cells/ml.

6. Collection time

[0186] Preliminary data indicate that collection delay was associated with viability and TSAC reduction, so collection time may have a potential impact on other CQAs. In all process performance studies, samples were collected from day 10 (240±4 hours) and day 11 (264±4 hours) from all bioreactors. Ten 2 L bioreactors served as control conditions in multiple process characterization units. Since all were performed under the same conditions with the current process parameter settings, they allowed for a study of collection time on productivity and quality of asphotase alfa. The analytical result of these ten 2 L bioreactors is shown in Table 12. It is noted that samples from two bioreactors (No. 18 and No. 19) were purified using a two-step purification (ProA + HIC), and samples from the other eight bioreactors (No. 10-No. 17) purified only with ProA. It should be noted that HIC is able to increase sample purity without affecting TSAC.

Table 12 Analytical Result from Lots of Exemplary 2 L Bioreactors

# Collection
(Day) AEX (%) SEC (%) TSAC cleaning Basic Main Sour Unit Main Fragment 10 10 0.2 90.8 9.1 8.0 91.6 0.4 2.1 ProA 10 eleven 0.4 90.1 9.5 8.4 91.2 0.5 1.7 eleven 10 0.3 90.7 9.0 7.5 92.0 0.6 2.3 eleven eleven 0.3 90.4 9.3 7.7 91.5 0.8 1.8 12 10 0.2 90.9 8.9 8.5 91.2 0.3 1.7 12 eleven 0.4 90.4 9.3 8.8 90.8 0.4 1.4 13 10 0.3 90.9 8.8 8.3 91.2 0.5 1.5 13 eleven 0.4 90.3 9.3 8.6 90.8 0.6 1.4 14 10 0.5 91.0 8.6 8.0 91.5 0.5 2.2 14 eleven 0.4 91.6 8.0 7.2 92.3 0.5 2.1 15 10 0.5 91.0 8.5 7.9 91.6 0.5 2.5 15 eleven 0.5 92.1 7.4 6.6 92.9 0.5 2.4 16 10 0.3 94.3 5.4 9.1 90.7 0.2 2.1 16 eleven 0.4 90.3 9.3 9.2 90.6 0.3 2.0 17 10 0.4 91.0 8.7 8.9 90.8 0.3 2.3 17 eleven 0.4 90.3 9.4 8.9 90.8 0.4 2.4 18 10 0.5 93.4 6.1 5.9 94.1 0.0 2.5 ProA+HIC 18 eleven 0.5 92.8 6.6 6.3 93.6 0.0 1.5 19 10 0.4 93.7 5.9 5.8 94.2 0.0 1.8 19 eleven 0.5 92.5 7.1 6.7 93.3 0.0 1.6

[0187] In each individual study, collection time was considered as the third parameter for statistical analysis. As shown in each individual study, delaying collection from day 10 to day 11 resulted in an increase in protein A titer and volumetric activity of approximately 10-15%.

[0188] A significant effect of collection time on TSAC was observed in all five separate studies. One day delay in harvesting resulted in a decrease in TSAC of about 0.3-0.4 moles of sialic acid per mole of protein on average, although TSAC under both conditions was still within the small-scale experimental range (1.5-2.9). as indicated in Table 9. Due to the significant impact of collection time on TSAC, early collection may be performed to increase TSAC.

[0189] With respect to aggregation as measured by SEC and acid peak as measured by AEX, collection time was shown to be statistically significant in three of the five studies (Table 13).

[0190] However, the magnitude of the effect on CQA was negligible, less than 1% compared to control conditions in each individual study. In addition, all values fell within the small scale experimental ranges shown in Table 9 (≤9.6% for aggregates and ≤10.2% for acid peak).

[0191] In one of the five studies, the levels of fragments were affected by the timing of collection (see Table 13). However, a one day delay in collection resulted in only a 0.1-0.2% increase in fragment under two control conditions, and the fragment level at day 11 was still below the LoQ (1%) of the SEC assay. Similarly, the level of the main peak was increased from 0.2-0.3% on day 10 to 0.4% on day 11 with a p-value of 0.03 (Table 13). However, both values were still well below the LoQ (1%) of the AEX analysis. Therefore, collection times from day 10 (240±4 hours) to day 11 (264±4 hours) are unlikely to be CPP.

Table 13. Effect of timing of collection on CQA from individual studies

p-value for collection time TSAC SEC AEX Unit Fragment Basic Sour Link 1 0.00 0.84 --- 0.55 0.61 Link 2 0.02 0.00 0.37 0.13 0.01 Link 3 0.02 0.21 0.34 0.64 0.00 Link 4 0.00 0.00 0.21 0.03 0.00 Link 5 0.01 0.01 0.00 0.33 0.43

Conclusion

[0192] The process parameters tested in this study include culture pH set point, temperature, seed density, amount and time of feeding, glucose addition, galactose addition, and harvest time. Table 14 lists all process parameters used.

Table 14. Summary of process parameters

Process parameter Tested ranges Starting point pH 6.75-7.10 Growth temperature (°C) 35.0-37.5 Production temperature (°C) 30.0-35.0 Temperature shift time (h) 108-132 Average recharge amount (%) 67-133 Average start of recharge day 3 - day 5 Seeding density (cells/ml) 4.0-8.0×10 ⁵ Collection time (day) day 10 - day 11

Example 5 High Resolution Study of the Effect of pH and Manufacturing Temperature on Asphotase Alpha Quality Characterization for Exemplary Process #4

[0193] In addition to previous estimates, this study provides a higher resolution effect of culture pH and production temperature (≥120 hours) on asphotase alpha performance and quality in an exemplary manufacturing process. Scenarios tested include: 1) different pH levels (ie 6.70, 6.80, 6.90, 7.00 and 7.10) and the same production temperature of 33.0°C; and 2) various production temperature levels (ie 31.0°C, 32.0°C, 33.0°C, 34.0°C and 35.0°C) and a pH setpoint of 6.90. Samples from all conditions were collected at 240±4 hours, and protein A-purified materials were tested for several quality characteristics (e.g., percent fragmentation, percent aggregation, charge distribution, total sialic acid content (TSAC), and neutral glycan coverage. This study confirmed that that the process proceeding at a given pH value in the range from 6.70 to 7.10 and a production temperature from 31.0°C to 35.0°C, supports all CQA.

Cell culture

[0194] All experiments performed in this study used disposable Mobius 3L bioreactors (CR0003L200, EMD Millipore). A list of the process parameters used in this study is presented in Table 15. In addition, four additional culture pH conditions with a control production temperature (33.0°C) and four additional temperatures with a control culture pH (6.90) were investigated (Table 16).

Table 15 Cell Culture Process Parameters for Exemplary Process #4

Process condition starting point Initial working volume 1.6 l Temperature 36.5°C (0-120 h); 33.0°C (120-240 h) pH 6.90±0.05 DO thirty% Adding glucose If [glc] ≤ 1.8 g/L, add a bolus daily to increase glucose to 1.8-2.0 g/L Medium with galactose 2.7% (v/v) bolus supplement at 96, 144, 192 hours Addition of _CuSO4 20 μM after 24 hours

Table 16. Conditions tested

no. culture pH Production temperature 1 6.70 33.0°C 2 6.80 33.0°C 3 6.90 33.0°C 4 7.00 33.0°C 5 7.10 33.0°C 6 6.90 31.0°C 7 6.90 32.0°C 8 6.90 35.0°C 9 6.90 34.0°C 10 6.90 33.0°C

[0195] Cell density and viability were calculated using ViCell VR. pH and autonomic gases were measured using pHOx, and major metabolites, including glucose and lactate, were measured using Nova Profile 100 (Nova Biomedical, Waltham, MA). Enzymatic activity was measured by standard methods with the modification that each sample was diluted only once, rather than three times, prior to measuring enzymatic activity.

Collection and cleaning

[0196] Fifty microliters of day 10 (240±4 hours) samples were collected from all bioreactors using syringes. After removal of cells by centrifugation (3000×g, 5 min), supernatants were clarified using 0.22 μm vial neck filters and stored at -80°C until purification. In this study, a single high-throughput wafer-based purification step (Protein A) was applied. Samples were transferred to low salt buffer (5 mM Na ₃ PO ₄ , pH 7.4) prior to analytical and protein characterization analysis.

Analytical characterization and protein characterization

[0197] Critical quality characteristics analyzed in this study included aggregate percentage, fragment percentage, charge distribution, TSAC, and distribution of neutral glycans. Aggregate level was assessed by the percentage of cumulative peaks to total protein as determined by SEC. Fragment level was assessed by the percentage of fragment to total protein determined in SEC as well as in LoC. The charge distribution was assessed by the percentage of main peaks, main peak and acid peak to total protein, respectively, quantified by AEX. The level of sialylation (or TSAC) was determined using HPAE-PAD. Detection of neutral glycan species was performed using MALDI-TOF mass spectrometry. Isoelectric focusing was performed using the iCE280 system.

results

Cell culture performance

[0198] As noted, both culture pH and production temperature affect cell performance. The low pH (pH=6.70) resulted in the lowest growth rate and the lowest maximum VCD, reaching a maximum density of 11.1×10 ⁶ cells/ml (FIG. 15A) on the day of collection (day 10). Higher pH values (pH=7.00 and 7.10) resulted in a 1-2% reduction in viability compared to other conditions on day 10 (FIG. 15B).

[0199] Higher production temperatures were also associated with enhanced growth (FIG. 16A) and a faster decline in viability towards the end of culture (FIG. 16B). The lower production temperature resulted in a lower VCD peak but provided comparable pot life to the 33.0°C breakpoint.

[0200] Increasing culture pH also caused much faster glucose uptake (FIG. 17A) and caused both more lactate production early in the culture (before the temperature shift) and more overall lactate accumulation throughout culture (FIG. 17B).

[0201] The culture cells were grown for five days (120 hours) until the growth temperature shifted to the production temperature. Both glucose and lactate profiles were very comparable between day 0 and day 5 (Fig. 18A and Fig. 18B). After the temperature shift, higher production temperatures were associated with higher levels of glucose intake. A slight decrease in maximum lactate concentration was observed for production temperature conditions of 33.0°C compared to other conditions. In addition, the production temperature of 35.0°C showed a significant reduction in lactate consumption during 9 and 10 days in culture compared to all other conditions.

[0202] It has previously been observed that production temperature had a more dramatic effect than culture pH on both protein A titer and specific activity for a production temperature range of between 30.0°C to 35.0°C and culture pH between 6 .75 and 7.10. However, the low pH condition in this study (6.70) had a significant effect on protein A production titer, decreasing by 35% compared to the control condition (pH=6.90) (Fig. 19A) and significantly lower volumetric activity compared to all other pH conditions tested (FIG. 19B). Culture pH did not appear to affect the specific performance of Protein A (Fig. 19C). However, the low pH condition induced the most specific activity (FIG. 19D) and a general trend was observed between decreasing culture pH and increasing specific activity.

[0203] Low temperature manufacturing conditions (31.0°C and 32.0°C) caused a significant decrease (26% and 21%, respectively) in protein A titer compared to control conditions (FIG. 20A). In addition, increased production temperature was associated with higher volumetric activity (Fig. 20B) and a slight increase in protein A specific performance (Fig. 20C). There was no effect on specific activity from changing the production temperature (FIG. 20D).

[0204] All quality characteristics of this study and previous work studying the effect of culture pH and production temperature are summarized in Table 17.

Table 17. Quantitative indicators of quality for various pH and various production temperatures

# pH Pace. production (°C) AEX (%) LoC (% Main, NR) LoC (% Main, R) SEC (%) TSAC Basic Main Sour Unit Main Fragm. 20 6.70 33.0 0.2 93.2 6.5 98.6 100.0 6.2 93.8 0 1.5 21 6.80 33.0 0.1 93.4 6.4 98.7 100.0 6.1 93.9 0 1.9 22 6.90 33.0 0.1 93.1 6.7 97.2 100.0 6.2 93.8 0 1.9 23 7.00 33.0 0.1 94.4 5.4 96.8 100.0 5.6 94.4 0 2.4 24 7.10 33.0 0.1 94.4 5.5 97.1 100.0 5.8 94.2 0 2.6 25 6.90 31.0 0.1 92.3 7.6 97.0 100.0 7.0 93.0 0 1.8 26 6.90 32.0 0.1 92.4 7.5 97.2 100.0 6.7 93.3 0 1.9 27 6.90 35.0 0.1 95.2 4.6 98.5 100.0 5.0 94.8 0.2 2.8 28 6.90 34.0 0.1 93.2 6.6 98.2 100.0 6.6 93.2 0.2 2.6 29 6.75 33.0 0.4 93.8 5.8 97.0 100.0 5.6 94.4 0 1.9 thirty 6.90 30.0 0.4 90.8 8.8 97.1 100.0 8.5 91.4 0 1.4 31 6.90 33.0 0.5 93.4 6.1 100.0 100.0 5.9 94.1 0 2.5 32 6.90 35.0 0.5 95.4 4.0 100.0 100.0 4.2 95.7 0.1 2.4 33 7.10 33.0 0.2 94.0 5.8 100.0 100.0 5.3 94.7 0 2.8

[0205] As in previous studies, TSAC was affected by both culture pH and production temperature (FIG. 21). The small scale experiment range for TSAC was 1.5-2.9 and higher TSAC was more desirable in this range. Increasing either culture pH or production temperature has been shown to increase the TSAC of the product. On the one hand, increasing the pH of the culture from 6.70 to 7.10 resulted in an increase in TSAC from about 1.5 to 2.6-2.8, which was in the range of the small scale experiment. On the other hand, increasing the production temperature from 30.0°C to 35.0°C resulted in an increase in TSAC from about 1.4 to 2.6-2.8. Because the TSAC value (1.4) at a production temperature of 30.0°C was lower than the lower limit (1.5) of the small scale experiment range, a production bioreactor should probably avoid a production temperature of 30.0°C. These data indicate that TSAC can be increased by increasing culture pH or production temperature, or both.

[0206] The level of impurity, which is currently measured by the percentage of acidic peak using AEX, is shown in FIG. 22. Increasing the production temperature reduced the percentage of acid peak in the product. Culture pH does not appear to correlate with impurity as measured by the AEX acid peak. Although the production temperature showed an effect on the AEX acid peak, all values were within acceptable limits for cleaning (≤10.2%). The main AEX peak was consistently below 1% LoQ for analysis for all conditions considered.

[0207] A previous study showed a strong correlation between AEX acid peak and aggregate percentage measured by SEC. This trend continued for this study, as shown in FIG. 23. Higher production temperature resulted in lower aggregate percentage. All aggregate percentages were within the range of the small scale experiment (≤9.6%). Cultural pH does not have a strong effect on the percentage of aggregate. Asphotase alpha fragments measured by SEC were below 1% LoQ for analysis for all conditions.

[0208] The percentage of the main peak, measured using LoC, as previously identified, is significantly affected by the production temperature. A positive correlation was found between the increase in production temperature and the percentage of the main peak under non-reducing conditions (FIG. 24B). There was no apparent effect of culture pH on the percentage of the main peak LoC (under non-reducing conditions) (FIG. 24A). LoC analysis under reducing conditions determined a 100% major peak for all conditions.

[0209] No new neutral species were observed for any of the conditions compared to the BDS materials described in the exemplary process of Example 2. However, a profile shift was observed between the low pH condition and the higher pH at which higher order glycoforms were observed in material obtained at a higher culture pH (FIG. 25). A similar trend was observed for production temperature with higher order glycoforms present in material produced at higher production temperatures (34.0° C. and 35.0° C.) (FIG. 26).

[0210] The determination of the peak from the iCE280 chromatograms was carried out by determining the most basic peak as peak 6 and determining the next four peaks as peak 5 through 2 in reverse order. Peak 1 then consisted of all peak percentages below the pI of the next peak after peak 2. The resulting pI values for the identified peaks are shown in Table 18. Both the pH of the medium and the production temperature affect the results of capillary isoelectric focusing (Fig. 27). An increase in either culture pH or production temperature resulted in an increase in the percentage of low pI peaks identified as peak 1. In addition, the associated percentage of high pI peaks, peak 5 and 6, decreased with increasing culture pH or production temperature. These factors indicate a general shift towards more acidic species associated with an increase in pH or production temperature. This shift may be due to an increase in sialylation associated with both an increase in pH and production temperature, which will lead to an increase in the percentage of acidic species.

Table 18. pI ranges for identified iCE280 peaks

Peak ID pi range Peak 1 ≤6.63 Peak 2 6.62-6.67 Peak 3 6.67-6.72 Peak 4 6.71-6.77 Peak 5 6.77-6.82 Peak 6 ≥6.83

conclusions

[0211] This study confirms previous results indicating that culture pH (6.70-7.10) and production temperature (31.0-35.0°C) are unlikely to be CPPs in that the measured CQAs were within the small scale, as indicated earlier. Both culture pH and production temperature affect many cell culture performance parameters and quality characteristics. Notably, an increase in either culture pH or production temperature results in an increase in maximum VCD, but also leads to a rapid decline in viability in late culture stages.

[0212] Both culture pH and production temperature can affect performance in terms of protein A titer and volumetric activity. Cultural pH 6.70 led to a decrease in protein A titer and volumetric activity. Increasing the pH value in the culture led to a decrease in specific activity. Therefore, although culture pH has a positive effect on protein A titer, there is little to no difference in volumetric activity between pH 6.80 and 7.10. In addition, increasing the production temperature resulted in an increase in protein A titer and volumetric activity, but did not affect specific activity.

[0213] Increasing culture pH or production temperature also resulted in higher TSAC values. Increasing the latter parameter also reduced the formation of aggregates and acid peaks. Finally, culture pH and production temperature appear to shift the isoelectric focusing profile of the product. These results indicate that a pH in the range of 6.70 to 7.10 and a production temperature in the range of 31.0°C to 35.0°C are acceptable, and tight control of both process parameters is recommended to maintain process stability and consistency in scaling (for example, to increase the scale of the process up to 10,000 liters).

Example 6: Improvement of active asphotase alpha titer with supplemental feeding

[0214] The previous illustrative methods for producing asphotase alpha (sTNALP-Fc-D ₁₀ ), as described in the previous examples, were low-throughput processes with active asphotase alpha titers of approximately 0.1 g/l at harvest. An alternative manufacturing process is illustrated herein to improve the process. Both upstream and downstream processes were evaluated.

[0215] The potential influence of the parameters of the previous cell culture process on the quality of asphotase alpha and its productivity was evaluated in a previous study. Cell culture parameters, including addition of metals (Zn ²⁺ /Mg ²⁺ /Ca ²⁺ ), temperature shift time, CHO feed amount, and harvest time were evaluated to improve titer and/or volumetric activity without affecting other critical quality characteristics ( CQA). Previous study data indicated that metal addition resulted in a significantly lower TSAC than control conditions and was therefore excluded from further evaluation. Delaying the temperature shift time from 54 hours to 78 hours resulted in a decrease in specific activity and a higher protein A titer. As a result, the total volumetric activity at later temperature shift conditions remained unchanged. Increasing the amount of CHO refill bolus from 1X (5 ml/L on day 2) to 4X (5 ml/L on days 2, 4, 6, and 8) resulted in a significant increase in volumetric activity without adversely affecting another quality characteristic tested (e.g., TSAC, charge distribution and distribution of neutral glycans). Delaying collection day from day 10 to day 12 was associated with higher productivity (eg, volumetric activity, protein A titer, and specific activity) and slightly lower TSAC. Based on these data, an increase in the amount of CHO refeed was chosen for further evaluation at the 10 L scale in this study.

Materials and methods

[0216] The previous process for the production of asphotase alpha (sTNALP-Fc-D ₁₀ ) was carried out using the following small scale seed fermenter system and cell culture process: thawing cell inoculums at target seeding density at 5×10 ⁵ cells/ml; increase in inoculum (flask/cylinder stage) with target seeding density at 3.5×10 ⁵ cells/ml; increase in inoculum (cell bag stage) with target seeding density at 3.5×10 ⁵ cells/ml; increase in inoculum (100 l - low level) with a target seeding density of 3.5×10 ⁵ cells/ml; increase in inoculum (100 l - high level) with a target seeding density of 3.5×10 ⁵ cells/ml; seeding 1000 l bioreactor (N3 - low level) with a target seeding density at 3.5×10 ⁵ cells/ml; seeding 1000 l bioreactor (N2 - high level) with a target seeding density at 3.5×10 ⁵ cells/ml; seeding 4000 l bioreactor (N1) with a target seeding density of 3.5×10 ⁵ cells/ml; production in a 20,000 L bioreactor at a target seeding density of 5.5 x 10 ⁵ cells/ml. For this study, the small scale seed fermenter system used as bioreactor inoculum for both control and enhanced bioreactor processes included: inoculum thawing; passage 1 (1×100 ml shake flask, target seeding 5×10 ⁵ cells/ml); passage 2 (2×100 ml shake flask, target seeding 3.5×10 ⁵ cells/ml); passage 3 (2×1 L rotator, target seeding 3.5×10 ⁵ cells/ml); passage 4 (2×15 L rotator, target seeding 3.5×10 ⁵ cells/ml); 8×10 L bioreactor, target seeding at 5.5×10 ⁵ cells/ml. The 10 L bioreactors used in this study were the Sartorius ^BIOSTAT® B-DCU II bioreactor system (Sartorius Stedim Biotech) and the data acquisition software MFCS/win 3.0 Module Shell 3.0 (level 32) (Sartorius Stedim Biotech). The list of raw materials used in bioreactor manufacturing includes: CM69 growth medium, 100 µM MTX (methotrexate), 200 mM L-glutamine, CM70 growth medium, Sigma CHO feed, 10% sodium carbonate, and FoamAway AOF (Gibco).

[0217] A cell flask from the asphotase alpha clone development cell bank (sTNALP-Fc-D ₁₀ ) was thawed and passaged through the seed fermenter system through the N-1 stage (equivalent to a 4000 liter seed bioreactor in the 20,000 liter process used previously). The cultivation of N-1 was carried out in two 15-liter rotary mechanisms with a working volume of 15 liters. Both N-1 cultures were used for inoculation. Agitation was set at 200 rpm with a total gas flow of 20 l/h. The gassing of oxygen and carbon dioxide was adjusted according to the measured dissolved oxygen and pH. Eight 10 liter bioreactors were inoculated and harvested on day 10. Two bioreactors were fed using the control fed strategy (single feeding), while six others were fed using the improved feeding strategy (multiple feedings on days 2, 4, 6 and 8). Table 19 summarizes the cell culture process parameters used for these bioreactors.

Table 19 Cell Culture Process Parameters

Process parameter Control Improved Process Initial working volume (medium and inoculum) 10 l (Sartorius bioreactor 10 l) Target population 5.5×10 ⁵ cells/ml pH 6.90 (without dead zone) Temperature 37.0°C, then shifted to 30.0°C when viable cell density is 25-32×10 ⁵ cells/ml (58 hours post inoculation) DO 40% bubbling 309 rpm (P/V= 81 W/m ³ ) Carbonation strategy Air flow 20ml/min+Oxygen as needed Upper layer 450 ml/min air CHO recharge 5 ml/l bolus (50 ml) on day 2 based on initial working volume 50 ml bolus *(50 ml) on days 2, 4, 6 and 8 Adding glucose If [glucose] is below 2 g/l, increase to 3 g/l by adding a bolus Glutamine addition If [glutamine] is below 0.18 g/l (1.2 mM), increase to 0.29 g/l (2.0 mM) by adding a bolus until temperature shift Base (10% w/w Na ₂ CO ₃ ) Add if needed Defoamer 5 ml on day 0, then 1 ml as needed during cultivation

*: Each bolus feed contains 1.2 mM zinc chloride, compared to basal medium containing only about 9 µM zinc chloride. Thus, each bolus feed increases about 6 μM zinc concentration (0.5 v/v).

[0218] The 20,000 L scale production process uses a VCD-based feeding strategy (one bolus per VCD about 25.0-32.0×10 ⁵ cells/mL, typically about 2 days of culture). In this study, the temperature shift time was programmed to 58 hours post-inoculation to meet the VCD criteria (25.0-32.0×10 ⁵ cells/mL) based on an average doubling time of 24.5 hours calculated from production data in a 20K bioreactor. . This change was made to simplify the process and provide tighter control over the timing of the recharge.

results

[0219] Cell culture values of all bioreactors are shown in FIGS. 28 in FIG. 31.

[0220] The growth rate and viability of the currently improved process were compared to the control and previous 20,000 L processes. The VCD during culture was similar to the control and 20,000 L runs (FIG. 28A, error bars are ±1X standard deviation). Cell viability in the control processes and the improved process #4 declined at the same rates, which was faster than the rate in the previous 20,000 l-scale processes (FIG. 28B). The final pot life was 73% and 70% for the control and improved processes, respectively, while the 20,000 L process typically remained around 80%. This appears to be a scaling issue for the 10 L system, possibly due to the speed of agitation. This was tested in the second material generation block, and the data from this run showed that the drop in viability was likely due to a combination of agitation speed and that the ViCell cell counter parameters are set to the standard CHO cell type, which has a wider acceptable cell diameter range ( 5-50 µM) than for cell type (11-30 µM) in previous processes.

[0221] Glucose utilization and lactate consumption are shown in FIG. 29. The improved process showed similar glucose utilization and lactate profiles compared to control.

[0222] Asphotase alpha protein A titer and specific activity across the whole culture is shown in FIG. 30. Protein A titer for both control and improved processes was comparable. However, they were lower than the titers normally observed in the 20,000 L process. This may be partly explained by the scale differences observed with viability, as the power to volume ratio used in this study was found to be 30% higher than that used in production (confirmed by a subsequent P/V ratio used in production scale ). Specific activity (units/mg protein) for improved process was on average 17% higher (532 U/mg vs 456 U/mg) than control. These data, taken with the protein A titer data, indicate that while obtaining the same amount of protein in each process, the resulting protein was more active when the culture was fed more.

[0223] The active titer and total volumetric activity (shown in FIG. 31) indicate that the improved process generates about 22-24% more activity than the control process at the 10 L scale. The improved process generated about the same active titer as at the 20,000 L scale, while the control process was about 19% lower.

Conclusion

[0224] This study confirms the effectiveness of a cell culture improved manufacturing process (scale 10 L) that uses an increased amount of feed, resulting in an increase in active titer of approximately 22-24% while maintaining the protein A titer of the product. A comparison of the specific activity of asphotase alpha between control and improved processes is shown in FIG. 32. It appears that all six replicates of the improved process conditions (each with three additional make-up additions) showed higher specific activity than the controls.

[0225] Each bolus feed contained 1.2 mM zinc chloride compared to the main medium containing only about 9 μM zinc chloride. Thus, each bolus feed resulted in an increase of about 6 μM zinc concentration (0.5 v/v) in the entire culture. Without wishing to be bound by any particular theory, it is postulated that the increase in specific activity by the addition of additional feed is due at least in part to the addition of zinc.

Example 7 Further Improvement of Asphotase Alpha Activity by Adding Zinc

[0226] Based on the manufacturing processes described in the previous examples, new studies were conducted to further optimize the parameters of the previous process. Exemplary Z and Z' processes are summarized and compared in Table 20.

Table 20. Differences between illustrative manufacturing processes (previous)

Process X Process Y Process Z Process Z' Parent cell line CHO CHO CHO CHO Work environment HyClone SFM4CHO HyClone SFM4CHO HyClone SFM4CHO+BD selection HyClone SFM4CHO+BD selection make-up CHO Makeup (0.5%) CHO Make-up (2%, i.e. 4 x medium make-up) CHO Replenishment (2%) + cell growth accelerator 2+5 (9%) CHO Replenishment (2%) + cell growth accelerator 2+5 (9%) Starting point pH 6.90 6.90 6.90 6.90 Temperature 37.0°C, then shifted to 30°C 37.0°C, then shifted to 30°C 37.0°C, then shifted to 30°C 37.0°C, then shifted to 30°C DO 40% 40% 40% 40% Additive N/A N/A _ZnSO4 ZnSO ₄ (30-90 µM) collection time 240±12 h 240±12 h 240±12 h 288±48 h

[0227] Further optimization was performed based on the previous Y process. For example, the new Z' process included adding about 30-90 μM zinc sulfate to the culture medium and increasing the total culture time from about 240 hours (i.e., 10 days) to about 288 hours (i.e. 12 days). A small-scale shake flask study was performed to test three variants of the Z' process (addition of 30, 60, or 90 μM zinc, respectively). 30 μM additives were added by front loading into the culture medium. Supplementation at 60 and 90 µM included the addition of zinc in two equal boluses on days 2 and 6 (assuming 60 µM) or three equal boluses on days 2, 6 and 10 (for 90 µM condition) . As shown in FIG. 33, zinc supplementation (red line showing the average of three experiments with different zinc concentrations) effectively improved the specific activity of asphotase alpha compared to the control process Y (blue line). This increase peaked around the 288-hour (i.e. 12 days) time frame and subsequently decreased.

Example 8 Characterization of Asphotase Alpha Produced

[0228] This document illustrates an exemplary characterization of manufactured asphotase alpha with several orthogonal analytical methods. These methods were used to evaluate the identity, purity, size, structure, glycosylation, and charge profiles of the produced asphotase alpha. Some methods have also been used to ensure consistency between batches. In addition, product-related substances and product-related impurities have also been characterized.

Structural analysis of asphotase alfa

Matrix-assisted laser desorption/ionization (MALDI-TOF) time-of-flight analysis of molecular weights after removal of oligosaccharides

[0229] MALDI-TOF mass spectrometric analysis of asphotase alpha after removal of oligosaccharides was used to establish the identity of asphotase alpha. In addition, this method will also detect the presence of any significant levels of post-translational modifications with molecular weights significantly different from the theoretical molecular weight of asphotase alpha. Removal of oligosaccharides was achieved by digestion of asphotase alpha using PNGase F with or without reduction of disulfide bonds. The samples were desalted and mixed with a 90:10 mixture of 2,5-dihydroxybenzoic acid and 2-hydroxy-5-methoxybenzoic acid in acetonitrile and trifluoroacetic acid. The samples were air dried and then analyzed using a MALDI-TOF mass spectrometer. Mass spectra were taken in positive mode. The instrument was externally calibrated using bovine serum albumin (BSA). The mass spectrum of asphotase alpha after deglycosylation is shown in FIG. 34A. Peaks at m/z 161140, 80571, and 53748 correspond to a molecule that is singly, doubly, or triply charged, respectively. The measured molecular weight of 161140 Da agrees with the calculated molecular weight (161135.2 Da) within the instrument's accuracy limit of 1%. The mass spectrum of asfotase alpha after reduction and removal of the oligosaccharide is shown in FIG. 34b. Peaks at m/z 80545, 40267, and 26808 correspond to a molecule that is singly, doubly, or triply charged, respectively. The measured molecular weight of 80545 Da agrees with the calculated molecular weight (80577.7 Da) within the instrument's accuracy limit of 1%. The results confirmed the identity of asphotase alfa and demonstrated the absence of significant levels of post-translational modifications other than glycosylation.

Analytical ultracentrifugation (AUC)

[0230] AUC was used to determine the percentage of aggregates, dimer and fragments, if present, and hence the purity of asphotase alpha. In addition, it defines the molecular weight of the asphotase alpha dimer, which is characteristic of the molecule due to the unique amino acid sequence and post-translational modifications. It is considered an orthogonal method for gel permeation chromatography (SEC). Asphotase alpha samples were diluted to approximately 1 mg/ml using 25 mM phosphate, 150 mM NaCl, pH 7.4. An analytical ultracentrifuge was used to analyze the sedimentation rate. Double sector cages equipped with quartz windows were used. The rotor was balanced in vacuum at 20°C and after about 1 hour the rotor accelerated to 36,000 rpm. Absorption scans at 280 nm were obtained at 4.5 minute intervals over a period of approximately 6 hours. The data was first analyzed using DCDT to determine the observed molecular weight, and then using the method (SEDFIT software) to determine the percentage of dimers, high molecular weight and low molecular weight species of compounds. The molecular weight of asphotase alpha was 211 kDa and the purity was 96.8% for an exemplary lot.

MALDI-TOF intact molecular weight

[0231] The MALDI-TOF asphotase alpha molecular weight measurement determines the size of the molecule, including all post-translational modifications, primarily glycosylation. Asphotase alpha samples were desalted and mixed with a matrix prepared from a 90:10 mixture of 2,5-dihydroxybenzoic acid and 2-hydroxy-5-methoxybenzoic acid. The samples were air dried and then analyzed using a MALDI-TOF mass spectrometer. Mass spectra were taken in positive mode. The instrument was externally calibrated using bovine serum albumin (BSA). The mass spectrum of asphotase alpha is shown in FIG. 35. Peaks at m/z 180157, 90223, and 60540 correspond to a molecule that is singly, doubly, or triply charged, respectively. Other low enriched peaks were formed due to fragmentation caused by hard instrument settings. The measured molecular weight of 180157 Da is significantly higher than the calculated molecular weight of 161135.2 Da based on the known amino acid sequence. This result indicates extensive post-translational modifications of asphotase alpha, mainly glycosylation, as MALDI-TOF and ESI-TOF analyzes of deglycosylated asphotase alpha do not reveal significant levels of modification.

Multiple angle size exclusion chromatography (SEC-MALS)

[0232] SEC-MALS is used to separate proteins based on size and then determine the molecular weight of each peak. Asphotase alpha was analyzed by SEC-MALS (data not shown). The calculated molecular weight of the peak in the retention time window of 7.0-8.2 minutes is 194.1 Da, which is in good agreement with the calculated molecular weight of asphotase alpha based on the known amino acid sequence and extensive glycosylation, and also consistent with the molecular weight observed with using MALDI-TOF.

Metal analysis

[0233] Inductively coupled plasma mass spectrometry (ICP-MS) was used to determine zinc and magnesium, and inductively coupled plasma atomic emission spectrometry (ICP-AES) was used to determine calcium. Asphotase alpha samples were digested using microwave digestion with heating with the addition of concentrated nitric acid first and then with the addition of hydrogen peroxide. The samples were then analyzed using the ICP-MS instrument for zinc and magnesium and the ICP-AES instrument for calcium. The molar ratios determined (mol ion/mol asphotase alpha monomer) were 2.27 for zinc, 0.97 for calcium, and 0.12 for magnesium. The ratios of zinc and calcium were in good agreement with the reported ratios in the literature (Stec et al., 2000 J Mol Biol 299 : 1303-1311 and Mornet et al., 2001 J Biol Chem 276 : 31171-31178). Mornet et al. (2001) states that calcium may occupy the magnesium binding site on TNALP and thus suggests that the ratio of magnesium in the asphotase alpha molecule may be less than one (1). However, the magnesium ratio tested as 0.12 is still unexpectedly low.

Phosphorylation Site Determination

[0234] UPLC-MS ^e was used to determine the site and percentage of phosphorylation. Asphotase alfa at a final concentration of 0.5 mg/ml was denatured and reduced in the presence of 6 M guanidine hydrochloride in 0.2 M Tris, pH 8.6 using 40 mM DTT at 37° C. for 1 hour. The reconstituted sample was alkylated by adding iodoacetic acid to a final concentration of 50 mM and incubating at room temperature for 30 minutes. The sample was then transferred to 50 mM ammonium bicarbonate buffer using a 10 kDa molecular weight cutoff dialysis tube. The sample was digested using trypsin at a final ratio (wt:wt) of 1:50 trypsin; protein at 37°C for 24 hours. A UPLC system equipped with a reverse phase C18 column and a Q-TOF mass spectrometer was used to analyze the digested samples. The tryptic peptide containing Ser93 was identified as both phosphorylated and non-phosphorylated. As shown in FIG. 36, the MS/MS spectrum showed that S93 was shifted by +80 Da of phosphate mass added by phosphorylation, as seen with y12 and y13 ions. The relative percentage of phosphorylation was determined using the peak areas of the extracted ion chromatograms (EIC) as 33.9%.

MALDI analysis of free glycans

[0235] MALDI-TOF analysis of free glycans was used to determine the types of oligosaccharides and their relative percentage. The oligosaccharides were released from the alphatase alpha reference standard by cleavage with PNGase F. The isolated glycans were purified using top-tipped reverse phase columns. The isolated glycans were mixed 1:1 with the sDHB matrix and then analyzed by MALDI-TOF. Analyzes were run in negative mode for sialic acid glycans and then in positive mode for neutral glycans. The mass spectrum obtained using the negative mode is shown in FIG. 37A. The mass spectrum obtained using the positive mode is shown in FIG. 37b. The types of oligosaccharides are determined by comparing the observed molecular weights with the theoretical molecular weights of commonly known oligosaccharides. The relative percentage was determined by dividing the individual peak intensity by the total peak intensity of all glycans. The main oligosaccharides are shown in Table 21 and Table 22. Based on molecular weights, the structure of sialic acid was defined as N-acetylneuraminic acid (Neu5Ac or NANA).

Table 21. Theoretical and observed molecular weights of sialylated glycans (M-H) ^-

Theoretical m/z Glycoforms Observed m/z Error (ppm) Intensity Relative intensity 1768.628 A2G1+NANA 1768.581 46.923 619,000 44 1914.685 FA2G1+NANA 1914.682 3.816 646.122 46 1930.680 A2G2+NANA 1930.677 3.596 774.051 55 1971.707 A3G1+NANA 1971.649 57.426 195,000 14 2076.738 FA2G2+NANA 2076.733 5.538 1402.453 100 2117.765 FA3G1+NANA 2117.755 9.504 359.057 26 2133.760 A3G2+NANA 2133.729 31.127 245,000 17 2279.818 FA3G2+NANA 2279.823 -5.668 421.152 thirty 2295.813 A3G3+NANA 2295.825 -12.668 333.253 24 2320.844 FA4G1+NANA 2320.75 94.536 155,000 eleven 2336.839 A4G2+NANA 2336.793 46.579 103,000 7 2441.870 FA3G3+NANA 2441.863 7.559 995.507 71 2482.897 FA4G2+NANA 2482.904 -7,393 296.679 21 2498.899 A4G3+NANA 2498.827 71.854 158.291 eleven 2644.950 FA4G3+NANA 2644.952 -2.229 289.645 21 2660.945 A4G4+NANA 2660.97 -25.771 165.363 12 2685.976 FA4G2+GN+NANA 2685.915 61.386 52.974 4 2807.003 FA4G4+NANA 2807.003 -0.322 849.240 61 2848.029 FA4G3+GN+NANA 2848.034 -4.806 67.026 5 3010.082 FA4G4+GN+NANA 3010.022 59.538 85,000 6 3026.077 A4G4L1+NANA 3025.985 91.732 80,000 6 3172.137 FA4G4L1+NANA 3172.11 26.443 277.854 20 3213.161 FA4G3L1+GN+NANA 3213.175 -13.611 15,000 1 3537.267 FA4G4L2+NANA 3537.181 86.085 72.238 5 3902.399 FA4G4L3+NANA 3902.451 -51.951 21.992 2

Table 22. Theoretical and observed molecular weights of neutral glycans (M+Na) ⁺

Theoretical (m/z) Glycoforms m/z Error (ppm) Intensity Relative intensity 933.317 M3 933.301 16 383.81 4 1136.401 A1 1136.4 0.6 2765.27 31 1257.427 M5 1257.435 -8.2 792.57 9 1282.454 FA1 1282.465 -10.7 1580.16 18 1298.453 A1G1 1298.454 -0.6 840.8 9 1339.48 A2 1339.484 -4 6704.09 75 1444.507 FA1G1 1444.522 -14.9 386.39 4 1485.538 FA2 1485.545 -7.1 8939.36 100 1501.533 A2G1 1501.53 2.8 4852.75 54 1542.559 A3 1542.556 3.4 1570.78 18 1647.591 FA2G1 1647.599 -8.3 3217.43 36 1663.586 A2G2 1663.582 3.6 2328 26 1688.617 FA3 1688.624 -6.7 2752.37 31 1704.612 A3G1 1704.609 3.2 1998.15 22 1745.639 A4 1745.624 14.8 803 9 1809.644 FA2G2 1809.647 -3.5 1704.05 19 1850.67 FA3G1 1850.675 -4.9 2024.14 23 1866.665 A3G2 1866.656 9 1263.32 14 1891.697 FA4 1891.7 -3.3 1384.76 15 1907.692 A4G1 1907.692 -0.4 937.19 10 2012.723 FA3G2 2012.728 -5.1 1352.46 15 2028.718 A3G3 2028.719 -1.2 992.03 eleven 2053.75 FA4G1 2053.761 -11.5 1404.04 16 2069.744 A4G2 2069.742 2.4 771.77 9 2174.776 FA3G3 2174.779 -3.3 1463.26 16 2215.793 FA4G2 2215.814 -20.7 1180.8 13 2231.797 A4G3 2231.791 6.2 572.1 6 2377.855 FA4G3 2377.864 -8.9 853.1 10 2393.85 A4G4 2393.849 1 529.86 6 2418.877 FA4G1L1 2418.954 -76.6 202 2 2539.904 FA4G4 2539.918 -14.3 1009.29 eleven 2580.93 FA4G2L1 2580.949 -18.8 188.9 2 2742.983 FA4G3L1 2742.985 -1.9 255 3 2905.036 FA4G4L1 2905.048 -12.1 301.29 3 3270.168 FA4G4L2 3270.102 66.1 204.96 2

Profiling of oligosaccharides labeled with 2-aminobenzamide (2-AB)

[0236] Asphotase-alpha-isolated oligosaccharides were labeled with 2-aminobenzamide (2-AB) and then analyzed by normal phase HPLC with fluorescence detection to determine types and relative percentages of various oligosaccharides. Asphotase alpha samples were reconstituted using dithiothreitol (DTT) at 37°C for 1 hour and then cleaved using PNGase F at 37°C overnight to release the N-linked oligosaccharides. The protein from the digested sample was precipitated and separated from the released glycans by cold ethanol and centrifugation. The supernatants containing the released glycans were dried using a vacuum centrifuge, reconstituted using 1% formic acid and incubated at room temperature for 40 minutes and then dried again. Samples were then labeled using 2-AB labeling reagents following the manufacturer's instructions (Ludger, Oxfordshire, UK). Labeled glycans were purified using S glycan cartridges (Prozyme, Hayward, CA) and then analyzed using HPLC with fluorescence detection at 330 nm excitation wavelength and 420 nm emission wavelength. The fluorescence chromatogram from the asphotase alpha assay is shown in FIG. 38. The accuracy of the peaks was determined by collection of fractions, mass spectrometric analysis of molecular weights and comparison of molecular weights with molecular weights of well-known oligosaccharide structures. The oligosaccharides identified by 2-AB labeling are summarized in Table 23. The chromatogram resulting from the analysis of the alphatase alpha reference standard is shown in FIG. 39.

Table 23. Glycoforms and their relative percentage using 2-AB labeling

Glycoforms Relative % M3
A1
FA1
A2
FA2
M5/A3
A2G1
FA3
FA2G1
FA2G1
FA4
A2G2
FA3G1
FA2G2
FA4G1
FA3G3
A3G3/A4G2
FA3G3
FA3G3
FA4G3
FA4G3
FA4G4
FA4G4L1
FA4G4L2
UNKNOWN 0.02
1.90
11.58
0.26
25.55
5.18
1.25
2.67
4.58
4.59
2.68
2.46
2.25
7.03
3.42
1.71
3.00
2.49
3.01
2.13
0.89
5.35
3.54
1.87
0.61

Glycopeptide profiling

[0237] A glycopeptide profile was used to determine the site-specific distribution of the oligosaccharide. Asphotase alpha samples at a final concentration of 0.5 mg/ml were denatured and reconstituted in the presence of 6 M guanidine hydrochloride in 0.2 M Tris, pH 8.6 using 40 mM DTT at 37° C. for 1 hour. The recovered samples were alkylated by adding iodoacetic acid to a final concentration of 50 mM and incubating at room temperature for 30 minutes. Samples were then transferred to 50 mM ammonium bicarbonate buffer using a dialysis tube with a 10 kDa molecular weight cutoff. Samples were digested using trypsin at a final ratio (wt:wt) of 1:50 trypsin; protein at 37°C for 24 hours. A UPLC system equipped with a reverse phase C18 column and a Q-TOF mass spectrometer was used to analyze the digested samples. The data was received in positive mode. Mass spectra over windows of elution time were averaged for each glycopeptide. Six glycopeptides were obtained by trypsin digestion of asphotase alfa, corresponding to six glycosylation sites per asphotase alfa monomer. Since the peptide fragment is the main determinant of retention time, glycopeptides are separated based on their unique amino acid sequence; therefore, a site-specific distribution of oligosaccharides can be obtained. The mass spectra of the glycopeptides are shown in FIG. 40. The types of oligosaccharides associated with each glycosylation site are summarized in Table 24.

Table 24 Oligosaccharide species associated with each glycopeptide

Tryptic peptides retention time Glycoforms Theoretical molecular weights Observed molecular weights T15-16 39.7-40.7 FA1
FA2
FA2G1
FA3
FA2G2
FA3G1
FA4
FA3G2
FA4G1
FA3G3 3078.34
3281.42
3443.47
3484.50
3605.53
3646.55
3687.58
3808.61
3849.63
3970.66 3078.32
3281.40
3443.45
3484.48
3605.51
3646.54
3687.57
3808.60
3849.62
3970.65 T26-27 17.4-17.9 A2
FA2
FA2G1
FA3
FA2G2
FA3G1
FA4
FA4G1
FA4G2
FA4G3 2755.12
2901.18
3063.23
3104.26
3225.28
3266.31
3307.34
3469.39
3631.43
3793.49 2755.11
2901.16
3063.21
3104.25
3225.27
3266.29
3307.32
3469.38
3631.43
3793.47 T33 28.6-29.9 A1
A2
A2G1
A3
A2G2
A3G1
A4 2191.92
2394.99
2557.05
2598.07
2719.10
2760.13
2801.15 2191.93
2395.01
2557.06
2598.07
2719.11
2760.14
2801.16 T35 99.5-102.8 A2
A3
FA3
A4
FA3G1
FA4
FA4G1 3496.66
3699.74
3845.80
3902.82
4007.85
4048.88
4210.93 3496.65
3699.73
3845.79
3902.80
4007.85
4048.87
4210.93 T45-46 46.3-47.0 FA2
FA2G1
FA3
FA2G2
FA3G1
FA4
FA3G2
FA4G1
FA4G2 4548.04
4710.09
4751.11
4872.14
4913.17
4954.19
5075.22
5116.25
5278.29 4547.99
4710.05
4751.08
4872.10
4913.13
4954.21
5075.18
5116.21
5278.26 T55 19.6-20.0 A1
A2
FA2
FA2G1 2284.91
2487.99
2634.05
2796.10 2284.92
2488.00
2634.06
2796.12

[0238] Based on the observed molecular weights and other analysis results (e.g., mass spectrum data and oligosaccharide profiling data with 2-aminobenzamides (2-AB)), the structures of the major glycans (≥4% relative abundance) of asphotase alpha were determined in FIG. . 41 and FIG. 42.

Capillary isoelectric focusing (cIEF)

[0239] cIEF separates proteins based on their isoelectric points and is used to control the charge of alpha alphatase variants. Asphotase alpha samples at a final concentration of 0.5 mg/ml in buffer containing urea, methylcellulose, sucrose, pharmacolytes and pI markers were analyzed using a capillary electrophoresis system. Samples were focused for 0.1 minute at 1500V followed by 14 minutes at 3000V. Separated protein variants were detected using UV absorption at 280 nm. An electropherogram from a representative batch of asphotase alpha is shown in FIG. 43. Six peaks were observed using cIEF. As a rule, peak 1 has a wide range of pI. Additional peaks were observed for peaks 4 and 5. This result demonstrated that asfotase alpha is highly heterogeneous with respect to charge, as expected due to the presence of multiple glycosylation sites with different amounts and levels of sialic acids at each site. The pI of each peak and their relative percent asphotase alpha of a representative lot are shown in Table 25.

Table 25. pI and relative percentage of peaks observed with cIEF

Peaks pi % 1
2
3
4
5
6 6.60
6.65
6.68
6.73
6.80
6.85 36.64
8.66
11.78
16.71
11.24
15.27

Product Related Substance

[0240] Structural heterogeneity of asphotase alpha, which is expressed in the CHO cell line, is expected due to post-translational modification. As described above, the heterogeneity of asphotase alpha is reflected in the charge variants, where six peaks are observed when analyzed by cIEF. A consistent pattern of charge profiles for asphotase alfa has been demonstrated in clinical and PV batches. As determined by AUC, SDS-PAGE, SEC and AEX using representative material, asphotase alfa is 96% pure, indicating that the multiple peaks observed by cIEF are entirely produced asfotase alfa and that all peaks are products associated with this substance.

Product related impurities

[0241] Product-related impurities are molecular variants that do not have properties comparable to the desired asphotase alpha in relation to activity, efficacy, and safety (Guidance for Industry, Q6B specifications: test procedures and acceptance criteria for biotechnological/biological products, August 1999, ICH). Product-related impurities were observed using three assays, and the characteristics of each impurity are shown in Table 26.

Table 26 Summary of Impurities Associated with Asphotase Alfa Production

Ways impurities SDS-PAGE Molecular weight groups other than the main band SEC Aggregates AEX truncated molecule

Characterization of SDS-PAGE bands

[0242] Asfotase alpha was analyzed by SDS-PAGE under non-reducing and reducing conditions. The results of SDS-PAGE analysis show that lane 1 corresponds to intact asphotase alpha, lane 2 corresponds to intact asphotase alpha with only one enzyme line, and lane 3 corresponds to the reduced form (monomer) of asphotase alpha (data not shown).

[0243] As expected, peptides from both the enzyme portion and the IgG1-Fc region were observed from the SDS-PAGE band under reducing conditions. This result, in addition to the apparent molecular weight of the SDS-PAGE band under reducing conditions, indicates that the main band corresponds to asphotase alpha in its reduced form. The band performance is summarized in Table 27.

Table 27. SDS-PAGE Band Performance Overview

SDS-PAGE Strip Description Identified molecular weights MAIDI-TOF and LC-MS band analysis after gel digestion View Under non-reducing conditions Band 1 (main band) Approximately 200-220 kDa Confirmed lane 1 as asfotase alfa (61.9% sequence coverage) Asfotase alfa, based on the identified molecular weight and sequence coverage. Stripe 2 Approximately 125 kDa Confirmed lane 2 is associated with asphotase alpha (13.8% sequence coverage) The putatively truncated asphotase alpha is based on the identified molecular weight that lies between the asphotase alpha homodimer and its single polypeptide chain. Stripe 3 Approximately 95-105 kDa Confirmed lane 3 is associated with asphotase alpha (7.4% sequence coverage) Probably one polypeptide chain of asphotase alpha. In restorative conditions Main strip Approximately 95-105 kDa Main band confirmed as asfotase alfa (65.4% sequence coverage) Reduced asphotase alpha (one polypeptide chain)
The putative truncated form detected by SDS-PAGE under non-reducing conditions was not detected by SDS-PAGE under reducing conditions due to its low concentration.

Characterization of SEC peaks

[0244] SEC-HPLC was used as a purity assay by separating high molecular weight species (aggregates), asphotase alpha dimer, and low molecular weight species. In a typical SEC chromatogram, the most abundant peak corresponds to the asphotase alpha homodimer. The peak that elutes before the main peak corresponds to high molecular weight species. The peak that elutes after the main peak corresponds to the low molecular weight species. Fractions corresponding to both high molecular weight species and major peaks were collected and analyzed by SDS-PAGE followed by gel digestion and mass spectrometry analysis. Matching the observed molecular weights of the peptides to the theoretical molecular weights derived from the known amino acid sequence showed a coverage rate of 86.6% for the main peak and 70.7% for high molecular weight species. Therefore, it is concluded that high molecular weight species are product-associated impurities.

AEX Peak Characterization

[0245] AEX was used to control the charge variants of asphotase alpha. Five AEX fractions along with starting material were analyzed by GP-HPLC and SDS-PAGE under non-reducing and reducing conditions (data not shown). The data suggest that the main AEX peaks contain mainly asphotase alpha and possibly a small percentage of high molecular weight species. The acid peak of AEX contains high molecular weight asphotase alpha species. Basic and acidic species are associated with the product, as confirmed by peptide fingerprinting using MALDI-TOF analysis.

[0246] Asfotase alfa has been well characterized using a variety of analytical methods. Its identity was confirmed by N-terminal sequencing, amino acid analysis and molecular weight analysis using MALDI-TOF and ESI-MS. The purity of asphotase alpha produced was monitored by AUC, SDS-PAGE, GP-HPLC and AEX. Asphotase alpha size was assessed by MALDI-TOF and SEC-MALS analysis of the oligosaccharide molecule. In addition, asphotase alpha molecular weight analyzes were also performed by AUC and SDS-PAGE. Consistently higher molecular weights of asphotase alpha were detected by these orthogonal methods, indicating extensive glycosylation of the molecule at multiple sites, which was confirmed by glycosylation analysis. The higher order structures of asphotase alpha were determined using both far-UV and ultra-UV-CD. The hydrodynamic size of the molecule can also be determined from AUC and SEC-MALS analysis, which also confirmed that asphotase alpha exists as a covalent dimer. A low number of aggregates was detected using SEC, AUC and SEC-MALS. The structure of the native disulfide bond and the location of the free cysteine at 102 was confirmed by LC-MS analysis of peptide maps under non-reducing conditions. The presence of only one basic free cysteine was further confirmed by assessing the total amount of free sulfhydryl. LC-MS analysis showed that the serine residue of the active site was partially phosphorylated. Three metals, including zinc, calcium and magnesium, have been reported in the literature in the asphotase alfa preparation. Asphotase alpha glycosylation was extensively analyzed by free glycan analysis using MALDI-TOF, normal phase HPLC analysis of labeled 2-AB oligosaccharides, LC-MS analysis of glycopeptides, and total sialic acid measurement. All six glycosylation sites per asphotase alpha monomer were found to be associated with a heterogeneous population of oligosaccharides. The charge profile, a heterogeneity caused mainly by the presence of sialic acid, was analyzed by cIEF where six peaks were observed. Product-bound material is reflected by the observed six cIEF peaks. Low levels of product related contaminants were detected using SEC, AEX and SDS-PAGE.

Example 9 Comparability of Asphotase Alpha Produced at 2000 L (2K) and 20000 L (20K) Scales

[0247] In this example, the asphotase alpha drug substance was made using the 2000 L (2K) and 20000 L (20K) processes. To demonstrate comparability between asphotase alfa produced at different scales, three batches of asphotase alfa produced using the 2K process (#35, #36 and #38) and three batches of asphotase alfa produced using the 20K process (#40, #42 and No. 34) were analyzed for their physicochemical properties in parallel, where possible, using the methods described in Table 28. Specific lots of 2K and 20K were selected based on clinical use, and previously tested lots proved lot-to-lot consistency for all 2K lots. and 20K.

[0248] The physicochemical properties tested were established as asphotase alpha purity, charge variants, size, identity, structure, glycosylation, and activity. The test results are shown in Table 29. The comparability of production was further assessed using a side-by-side forced degradation study comparing the degradation pathways and kinetics of the aforementioned 2K and 20K batches. Samples were incubated at 40° C. for 0, 14, 24, and 48 hours and analyzed by the methods listed in Table 30. Test results for forced thermal degradation studies are shown in Table 31.

[0249] The overall results show that asfotase alpha made using the 2K and 20K processes is comparable.

Table 28. Physical and chemical tests of Asfotase Alpha

Test category Test Purity Analytical ultracentrifugation Purity SDS-PAGE/LoC (Lab on Chip) Purity SDS-PAGE Purity GP-HPLC Purity Anion exchange chromatography Charge Profile Capillary isoelectric focusing (cIEF) Size Multi-angle light scattering gel filtration chromatography (SEC-MALS) Size Intact molecular weight analysis (MALDI-ToF-MS) Identity Deglycosylated/Reductive and Deglycosylated Molecular Weight Analysis (MALDI-ToF-MS) Identity Deglycosylated/Reductive and Deglycosylated Molecular Weight Analysis (ESI-ToF-MS) Structure Circular dichroism spectrometry Structure Disulfide bonding and free thiol by LC/MS Structure Metal Analysis (ICP-MS/ICP-AES) Structure Identification of phosphorylation site and amount via UPLC-MSe Glycosylation MALDI-Free Glycan Glycosylation Profiled 2-AB-labeled N-linked oligosaccharides by HPLC Glycosylation Glycopeptide Profile and Site Occupation (UPLC-QToF-MS) Glycosylation Total Sialic Acid Activity pNPP activity Activity HA binding Activity Kinetic analysis of the PPi enzyme

Table 29. Physical and chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Concentration (mg/ml) 93.5 99.9 99.8 NA NA 107.6 94.5 101.6 NA NA Analytical ultracentrifugation 96.5 95.9 94.5 95.6 1.0 96.0 97.6 96.8 96.8 0.8 % Asphotase alpha monomer * % Asphotase alpha dimer 2.9 3.3 3.9 3.37 0.5 3.4 2.3 2.9 2.9 0.6 % Asphotase alpha trimer 0.3 0.8 1.6 0.9 0.7 0.6 0.1 0.3 0.3 0.3 SDS-PAGE/LoC 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 Under non-reducing conditions (%) Under reducing conditions (%) 99.8 99.8 99.8 99.8 0.0 99.8 99.8 99.9 99.8 0.1 SDS-PAGE 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 Under non-reducing conditions (%) Under reducing conditions (%) 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 GP-HPLC 96.1 97.1 96.2 96.5 0.6 98.2 98.9 98.6 98.6 0.4 Dimer (%) * Aggregate (%) 2.4 2.9 3.8 3.0 0.7 1.8 1.1 1.4 1.4 0.4 Fragment (%) 1.5 0.0 0.0 0.5 0.9 0.0 0.0 0.0 0.0 0.0 AEX 94.29 94.60 93.62 94.17 0.50 95.68 96.84 96.63 96.38 0.62 % Main peak area

* Asphotase alpha monomer reported by analytical ultracentrifugation is equivalent/synonymous with "dimer" by GP-HPLC and SEC-MALS.

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD cIEF Peak 1 pi 6.61 6.59 6.62 6.61 0.02 6.59 6.59 6.60 6.59 0.01 % 33.55 29.01 33.09 31.88 2.50 24.46 24.40 36.64 28.50 7.05 Peak 2 pi 6.65 6.64 6.66 6.65 0.01 6.63 6.63 6.65 6.64 0.01 % 10.79 9.54 10.24 10.19 0.63 8.53 8.96 8.66 8.72 0.22 Peak 3 pi 6.69 6.68 6.70 6.69 0.01 6.67 6.67 6.68 6.67 0.01 % 12.36 13.49 13.62 13.16 0.69 12.97 10.60 11.78 11.78 1.19 Peak 4 pi 6.73 6.73 6.76 6.74 0.02 6.72 6.72 6.73 6.72 0.01 % 17.21 18.71 17.09 17.67 0.90 18.79 20.25 16.71 18.58 1.78 Peak 5 pi 6.80 6.80 6.80 6.80 0.00 6.79 6.79 6.80 6.79 0.01 % 10.29 11.78 11.21 11.09 0.75 14.62 13.27 11.24 13.04 1.70 Peak 6 pi 6.84 6.85 6.85 6.85 0.01 6.84 6.84 6.85 6.84 0.01 % 15.81 17.47 14.74 16.01 1.38 20.62 22.53 15.27 19.47 3.76 SEC-MALS
Mw (kDa) Dimer
Mw (kDa) high molecular weight species 185.5 188.5 193.6 189.2 4.1 187.0 189.4 194.1 190.2 3.6 789.6 1054.0 1047.0 963.5 150.7 673.4 579.1 722.1 658.2 72.7 MALDI-ToFMW. (Yes)
Glycosylated
Deglycosylated
Deglico and rebuilt 180984 180853 180869 180902 71 179649 180252 180157 180019 324 161122 160980 161048 161050 71 160778 160725 161140 160881 226 80534 80493 80562 80530 35 80563 80509 80545 80539 27 ESI-ToF MW (Yes) 161137.80 161136.05 161138.33 161137.39 1.19 161136.69 161136.46 161138.69 161137.28 1.23 Deglico
Deglico and rebuilt 80575.23 80575.16 80576.18 80575.52 0.57 80574.65 80574.86 80574.77 80574.76 0.11

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Circular dichroism About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) About UV (nm) Features near-UV (nm): λmax 1 291.0 290.5 290.5 290.7 0.3 290.5 290.5 290.5 290.5 0.0 λmax 2 285.5 285.5 285.0 285.3 0.3 285.5 286.0 285.0 285.5 0.5 λmax 3 268.0 268.5 268.0 268.2 0.3 268.0 268.5 268.5 268.3 0.3 λmax 4 261.0 261.5 261.0 261.2 0.3 261.5 260.5 261.0 261.0 0.5 λmax 5 256.5 256.5 257.0 256.7 0.3 257.0 257.0 257.0 257.0 0.0 λmin 1 293.0 293.0 293.0 293.0 0.0 292.5 292.5 293.0 292.7 0.3 λmin 2 287.0 287.0 286.5 286.8 0.3 286.5 286.5 286.5 286.5 0.0 λmin 3 281.0 281.0 280.0 280.7 0.6 281.0 280.5 280.0 280.5 0.5 CD features far-UV (nm): Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) Far UV (nm) signal intensity λ210 -7550 -7210 -6830 -7012 365 -7520 -7340 -7320 -7288 274 signal intensity λ220 -8410 -8180 -7680 -7934 355 -8450 -8270 -8400 -8250 263 Ratio λ210/λ220 0.9 0.88 0.89 0.88 0.01 0.89 0.89 0.87 0.88 0.01 Deconvolution Decon Decon Decon Decon Decon Decon Decon Decon Decon Decon α-helix 0.197 0.193 0.181 0.190 0.008 0.206 0.198 0.201 0.202 0.004 Spiral 3/10 0.049 0.046 0.046 0.047 0.002 0.047 0.045 0.046 0.046 0.001 β-sheet 0.192 0.196 0.199 0.196 0.004 0.182 0.188 0.193 0.188 0.006 turns 0.146 0.144 0.152 0.147 0.004 0.153 0.150 0.141 0.148 0.006 Structure Poly(Pro) II 0.060 0.064 0.061 0.062 0.002 0.060 0.066 0.059 0.062 0.004 disordered protein 0.351 0.357 0.358 0.355 0.004 0.350 0.350 0.358 0.353 0.005 General 0.995 1,000 0.997 0.997 0.003 0.998 0.997 0.998 0.998 0.001

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Disulfide bonding and free thiol analysis (LC/MS) T16ssT21 15479 14499 14976 14985 490 11128 14424 15068 13540 2114 T48ss 1222 1261 442 975 462 728 1141 1345 1071 314 T50ssT50 17743 15660 15268 16224 1330 14527 15217 17669 15804 1651 T52ssT58 17922 17324 18066 17771 393 17530 16264 15593 16462 984 T66ssT71 4205 3576 3482 3754 393 2852 3611 4077 3513 618 T13* 13768 11442 13817 13009 1357 10623 13261 12453 12112 1352 T16* 4792 4107 4602 4500 354 4110 4428 3729 4089 350 T21* 7332 6144 7307 6928 679 3332 5123 5423 4626 1131 T52* 1317 1096 1366 1260 144 820 1181 1140 1047 198 T71* 1408 966 1463 1279 272 795 1070 1167 1011 193 T13ssT13 90 100 92 94 5 17 33 70 40 27 T13ssT16 5870 4884 5636 5463 515 3538 5356 4365 4420 910 T13ssT21 4060 3893 3661 3871 200 1161 1931 3321 2138 1095

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Total free cysteine # Free Cys per half molecule 0.7 0.6 0.6 0.6 0.06 0.7 0.7 0.7 0.7 0 MALDI free glycan m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ View M3 933.45 933.36 933.39 933.40 0.05 933.4 933.38 933.38 933.39 0.01 A1 1136.52 1136.50 1136.52 1136.51 0.01 1136.48 1136.52 1136.42 1136.47 0.05 M5 1257.61 1257.51 1257.54 1257.55 0.05 1257.49 1257.53 1257.43 1257.48 0.05 FA1 1282.66 1282.56 1282.58 1282.60 0.05 1282.54 1282.58 1282.48 1282.53 0.05 A1G1 1298.61 1298.54 1298.56 1298.57 0.03 1298.51 1298.63 1298.48 1298.55 0.08 A2 1339.60 1339.59 1339.60 1339.60 0.01 1339.57 1339.61 1339.50 1339.56 0.06 FA1G1 1444.71 1444.69 1444.64 1444.68 0.03 1444.58 1444.65 1444.52 1444.58 0.07 FA2 1485.66 1485.65 1485.66 1485.66 0.01 1485.64 1485.68 1485.56 1485.62 0.06 A2G1 1501.65 1501.65 1501.67 1501.66 0.01 1501.61 1501.67 1501.56 1501.61 0.06 A3 1542.70 1542.68 1542.69 1542.69 0.01 1542.60 1542.70 1542.58 1542.63 0.06 FA2G1 1647.73 1647.71 1647.73 1647.72 0.01 1647.66 1647.74 1647.61 1647.67 0.07 A2G2 1663.72 1663.69 1663.73 1663.71 0.02 1663.62 1663.74 1663.60 1663.66 0.07 FA3 1688.76 1688.75 1688.76 1688.76 0.01 1688.69 1688.76 1688.64 1688.70 0.06 A3G1 1704.75 1704.74 1704.77 1704.75 0.01 1704.67 1704.75 1704.63 1704.69 0.06 A4 1745.77 1745.76 1745.79 1745.77 0.01 1745.70 1745.75 1745.64 1745.70 0.06 FA2G2 1809.79 1809.77 1809.82 1809.79 0.02 1809.69 1809.79 1809.65 1809.71 0.07 FA3G1 1850.82 1850.80 1850.83 1850.82 0.01 1850.72 1850.81 1850.69 1850.74 0.06 A3G2 1866.79 1866.79 1866.83 1866.80 0.02 1866.70 1866.79 1866.67 1866.72 0.06 FA4 1891.84 1891.83 1891.86 1891.84 0.01 1891.74 1891.84 1891.71 1891.77 0.07 A4G1 1907.85 1907.82 1907.87 1907.85 0.02 1907.71 1907.83 1907.69 1907.75 0.07 FA3G2 2012.88 2012.86 2012.89 2012.88 0.01 2012.73 2012.86 2012.72 2012.78 0.07 A3G3 2028.86 2028.85 2028.89 2028.87 0.02 2028.71 2028.84 2028.72 2028.77 0.07 FA4G1 2053.89 2053.89 2053.91 2053.90 0.01 2053.77 2053.88 2053.75 2053.81 0.07 A4G2 2069.96 2069.90 2069.94 2069.93 0.03 2069.77 2069.90 2069.74 2069.81 0.08 FA3G3 2174.91 2174.90 2174.94 2174.92 0.02 2174.78 2174.92 2174.77 2174.83 0.08 FA4G2 2215.93 2215.92 2215.95 2215.93 0.01 2215.77 2215.92 2215.80 2215.84 0.07 A4G3 2231.93 2231.87 2231.96 2231.92 0.04 2231.80 2231.90 2231.80 2231.84 0.05 FA4G3 2377.96 2378.01 2377.98 2377.98 0.02 2377.76 2377.96 2377.84 2377.87 0.09

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD MALDI Free Glycan (Cont.) m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ m/z (M+Na)+ A4G4 2393.94 2393.92 2393.99 2393.95 0.03 2393.78 2393.98 2393.81 2393.87 0.10 FA4G1L1 2418.96 2418.91 2419.04 2418.97 0.06 2418.98 2419.02 2418.93 2418.98 0.05 FA4G4 2539.94 2540.00 2539.98 2539.97 0.03 2539.82 2539.97 2539.89 2539.91 0.06 FA4G2L1 2580.97 2581.05 2580.97 2581.00 0.04 2580.78 2580.97 2580.94 2580.92 0.08 FA4G3L1 2742.95 2742.88 2742.94 2742.92 0.03 2742.77 2743.08 2742.90 2742.95 0.13 FA4G4L1 2904.83 2904.79 2904.87 2904.83 0.04 2904.79 2904.94 2904.89 2904.89 0.06 FA4G4L2 3270.47 3270.40 3270.54 3270.47 0.06 3270.63 3270.01 3269.87 3270.08 0.32 2AB labeled glycans Peak area % M3 0.05 0.18 0.05 0.09 0.08 0.17 0.11 0.02 0.10 0.08 A1 1.52 2.32 1.55 1.80 0.45 2.79 2.34 1.90 2.43 0.37 FA1 9.16 13.55 7.98 10.23 2.94 11.67 11.46 11.58 11.57 0.11 A2 0.29 0.55 0.49 0.44 0.14 0.60 0.53 0.26 0.50 0.14 FA2 25.81 22.26 20.98 23.02 2.50 24.60 25.11 25.55 24.99 0.40 M5/A3 5.3 6.61 6.1 6.00 0.66 6.44 6.30 5.18 6.13 0.54 A2G1 1.23 1.67 1.14 1.35 0.28 1.49 1.43 1.25 1.42 0.10 FA3 4.1 4.2 6.49 4.93 1.35 3.90 4.31 2.67 3.82 0.67 FA2G1 3.8 3.2 2.68 3.23 0.56 3.37 3.54 4.58 3.68 0.51 FA2G1 3.57 3.43 2.32 3.11 0.68 4.11 3.86 4.59 4.11 0.30 FA4 4.35 4.31 6.31 4.99 1.14 3.63 4.15 2.68 3.65 0.60 A2G2 2.28 2.62 2.21 2.37 0.22 2.24 2.24 2.46 2.28 0.10 FA3G1 2.99 1.61 3.69 2.76 1.06 3.05 3.19 2.25 2.95 0.40 FA2G2 7.39 6.71 5.08 6.39 1.19 5.73 6.80 7.03 6.42 0.64 FA4G1 2.33 2.46 4.51 3.10 1.22 3.73 2.13 3.42 3.03 0.83 FA3G3 2.28 2.15 3.01 2.48 0.46 2.29 2.36 1.71 2.20 0.28

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD 2AB labeled glycans FA3G3 3.68 3.45 3.78 3.64 0.17 3.19 3.41 3.01 3.24 0.17 FA4G3 2.65 2.41 2.94 2.67 0.27 2.41 2.36 2.13 2.33 0.12 FA4G3 0.99 1.15 1.51 1.22 0.27 0.84 0.93 0.89 0.89 0.05 FA4G4 5.74 5.39 5.75 5.63 0.21 4.76 4.78 5.35 4.89 0.26 FA4G4L1 3.43 3.05 3.69 3.39 0.32 2.73 2.64 3.54 2.86 0.39 FA4G4L2 1.49 1.34 1.68 1.50 0.17 1.22 1.13 1.87 1.31 0.31 UNKNOWN 0.38 0.41 0.48 0.42 0.05 0.36 0.29 0.61 0.38 0.13

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Glycopeptide Profile and Site Occupation (UPLC-QToF-MS) m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ Glycopeptide masses seen in reduced/alkylated trypsin Approximate retention time (min) Tryptic peptides Glycan Glyco-peptide m/z (M+H) ⁺ 39.7-40.7 T15-16 FA1 3078.34 3078.31 3078.33 3078.31 3078.32 0.01 3078.31 3078.33 3078.32 3078.32 0.01 FA2 3281.42 3281.40 3281.40 3281.39 3281.40 0.01 3281.38 3281.41 3281.40 3281.40 0.01 FA2G1 3443.47 3443.46 3443.46 3443.45 3443.45 0.01 3443.44 3443.47 3443.45 3443.45 0.01 FA3 3484.50 3484.48 3484.49 3484.47 3484.48 0.01 3484.46 3484.49 3484.48 3484.48 0.01 FA2G2 3605.53 3605.51 3605.51 3605.51 3605.51 0.00 3605.49 3605.52 3605.51 3605.51 0.01 FA3G1 3646.55 3646.54 3646.54 3646.53 3646.54 0.01 3646.52 3646.54 3646.54 3646.53 0.01 FA4 3687.58 3687.57 3687.56 3687.55 3687.56 0.01 3687.55 3687.57 3687.57 3687.56 0.01 FA3G2 3808.61 3808.59 3808.60 3808.59 3808.59 0.01 3808.57 3808.60 3808.60 3808.59 0.02 FA4G1 3849.63 3849.62 3849.62 3849.61 3849.62 0.01 3849.60 3849.63 3849.62 3849.62 0.01 FA3G3 3970.66 3970.65 3970.65 3970.64 3970.65 0.01 3970.62 3970.65 3970.65 3970.64 0.02 17.4-17.9 T26-27 A2 2755.12 2755.11 2755.11 2755.10 2755.11 0.01 2755.10 2755.12 2755.11 2755.11 0.01 FA2 2901.18 2901.16 2901.16 2901.16 2901.16 0.00 2901.15 2901.17 2901.16 2901.16 0.01 FA2G1 3063.23 3063.21 3063.21 3063.21 3063.21 0.00 3063.21 3063.22 3063.21 3063.22 0.01 FA3 3104.26 3104.25 3104.25 3104.24 3104.24 0.00 3104.23 3104.26 3104.25 3104.24 0.01 FA2G2 3225.28 3225.28 3225.28 3225.27 3225.28 0.00 ND 3225.28 3225.27 3225.28 0.00 FA3G1 3266.31 3266.29 3266.30 3266.30 3266.29 0.00 3266.28 3266.30 3266.29 3266.29 0.01 FA4 3307.34 3307.32 3307.33 3307.32 3307.32 0.00 3307.30 3307.33 3307.32 3307.32 0.01 FA4G1 3469.39 3469.38 3469.38 3469.37 3469.37 0.01 3469.36 3469.39 3469.38 3469.37 0.02 FA4G2 3631.43 3631.43 3631.44 3631.42 3631.43 0.01 3631.41 3631.44 3631.43 3631.43 0.01 FA4G3 3793.49 3793.48 3793.49 3793.49 3793.49 0.00 3793.46 3793.50 3793.47 3793.48 0.02 28.6-29.9 T33 A1 2191.92 2191.93 2191.98 2191.92 2191.94 0.03 2191.91 2191.93 2191.93 2191.92 0.01 A2 2394.99 2395.00 2395.06 2395.00 2395.02 0.04 2394.99 2395.01 2395.01 2395.00 0.01 A2G1 2557.05 2557.06 2557.12 2557.05 2557.08 0.04 2557.05 2557.07 2557.06 2557.06 0.01 A3 2598.07 2598.07 2598.14 2598.09 2598.10 0.03 2598.06 2598.08 2598.07 2598.07 0.01 A2G2 2719.10 2719.11 2719.18 2719.10 2719.13 0.04 2719.10 2719.12 2719.11 2719.11 0.01 A3G1 2760.13 2760.14 2760.20 2760.13 2760.16 0.04 2760.12 2760.15 2760.14 2760.14 0.01 A4 2801.15 2801.16 2801.23 2801.15 2801.18 0.04 2801.15 2801.17 2801.16 2801.16 0.01

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Glycopeptide Profile and Site Occupation (UPLC-QToF-MS) m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ m/z (M+H)+ Glycopeptide masses seen in reduced/alkylated trypsin Approximate hold time (min) Tryptic peptides Glycan Glyco-peptide m/z (M+H) ⁺ 99.5-102.8 T35 A2 3496.66 3496.64 3496.65 3496.63 3496.64 0.01 3496.63 3496.65 3496.65 3496.64 0.01 A3 3699.74 3699.73 3699.73 3699.72 3699.73 0.01 3699.71 3699.73 3699.73 3699.73 0.01 FA3 3845.80 3845.79 3845.78 3845.77 3845.78 0.01 3845.77 3845.79 3845.79 3845.79 0.01 A4 3902.82 3902.81 3902.80 3902.80 3902.80 0.01 3902.78 3902.81 3902.80 3902.80 0.02 FA3G1 4007.85 4007.85 4007.84 4007.83 4007.84 0.01 4007.83 4007.85 4007.85 4007.84 0.01 FA4 4048.88 4048.88 4048.87 4048.86 4048.87 0.01 4048.86 4048.88 4048.87 4048.87 0.01 FA4G1 4210.93 4210.93 4210.93 4210.91 4210.92 0.01 4210.92 4210.93 4210.93 4210.93 0.01 46.3-47.0 T45-46 FA2 4548.04 4547.99 4548.00 4547.97 4547.99 0.01 4547.97 4548.01 4547.99 4547.99 0.02 FA2G1 4710.09 4710.04 4710.05 4710.03 4710.04 0.01 4710.03 4710.06 4710.05 4710.05 0.02 FA3 4751.11 4751.07 4751.08 4751.05 4751.06 0.01 4751.05 4751.09 4751.08 4751.07 0.02 FA2G2 4872.14 4872.09 4872.10 4872.08 4872.09 0.01 4872.08 4872.11 4872.10 4872.10 0.01 FA3G1 4913.17 4913.12 4913.13 4913.11 4913.12 0.01 4913.11 4913.14 4913.13 4913.13 0.02 FA4 4954.19 4954.15 4954.16 4954.13 4954.15 0.01 4954.18 4954.17 4954.21 4954.18 0.02 FA3G2 5075.22 5075.17 5075.18 5075.16 5075.17 0.01 5075.16 5075.20 5075.18 5075.18 0.02 FA4G1 5116.25 5116.20 5116.21 5116.19 5116.20 0.01 5116.18 5116.22 5116.21 5116.21 0.02 FA4G2 5278.29 5278.25 5278.26 5278.25 5278.25 0.01 5278.23 5278.27 5278.26 5278.26 0.02 19.6-20.0 T55 A1 2284.91 2284.92 2284.92 2284.91 2284.92 0.00 2284.90 2284.92 2284.92 2284.91 0.01 A2 2487.99 2488.00 2488.00 2487.99 2488.00 0.00 2487.98 2488.00 2488.00 2487.99 0.01 FA2 2634.05 2634.06 2634.06 2634.05 2634.06 0.00 2634.05 2634.06 2634.06 2634.05 0.01 FA2G1 2796.10 2796.12 2796.12 2796.12 2796.12 0.00 2796.10 2796.12 2796.12 2796.11 0.01

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Glycopeptide Profile and Site Occupation (UPLC-QToF-MS) Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Relative contributions of glycoforms at each site of the N-linked oligosaccharide Approximate holding time (min) Tryptic peptides Glycan 39.7-40.7 T15-16 FA1 5 8 6 6.4 1.7 12 8 10 10.0 1.6 FA2 16 18 19 17.7 1.3 18 18 18 17.6 0.1 FA2G1 eleven eleven 8 9.9 1.6 12 eleven eleven 11.5 0.7 FA3 12 12 17 13.4 2.8 10 10 8 9.3 0.8 FA2G2 10 eleven 9 10.1 1.1 12 9 eleven 10.7 1.3 FA3G1 7 7 7 7.1 0.2 9 7 7 7.7 1.0 FA4 7 6 10 7.5 1.8 5 5 3 4.4 0.9 FA3G2 5 4 3 4.2 0.9 4 5 5 4.5 0.8 FA4G1 6 5 5 5.3 0.4 4 5 4 4.2 0.4 FA3G3 4 3 3 3.4 0.9 1 5 6 3.7 2.6 17.4-17.9 T26-27 A2 10 12 9 10.4 1.7 12 eleven 10 10.7 0.7 FA2 5 6 7 6.3 1.0 8 8 6 7.5 1.0 FA2G1 4 5 3 4.0 0.8 2 6 7 5.0 2.6 FA3 14 13 19 15.3 3.1 15 12 9 12.1 2.9 FA2G2 6 7 5 6.0 1.1 0 6 8 4.7 4.2 FA3G1 8 7 7 7.0 0.7 9 7 7 7.8 1.2 FA4 10 9 13 10.9 2.1 8 7 5 6.8 1.3 FA4G1 9 7 9 8.4 1.0 7 7 7 7.0 0.3 FA4G2 6 6 5 5.8 0.5 7 6 7 6.4 0.4 FA4G3 4 3 3 3.3 0.8 5 4 5 4.6 0.7 28.6-29.9 T33 A1 eleven 12 10 10.7 1.1 18 16 17 17.1 1.1 A2 23 24 24 23.5 0.5 21 22 21 21.3 0.9 A2G1 14 13 eleven 12.5 1.6 14 13 14 13.7 0.8 A3 8 7 15 10.0 4.0 5 6 5 5.3 0.5 A2G2 12 12 10 11.3 0.9 eleven 10 13 11.5 1.3 A3G1 6 5 5 5.4 0.2 5 4 5 4.8 0.4 A4 4 3 5 3.9 1.0 2 3 2 2.2 0.5 99.5-102.8 T35 A2 7 9 7 7.6 1.1 9 9 8 9.0 0.6 A3 eleven 13 14 12.7 1.4 9 eleven 9 9.8 0.7 FA3 8 8 10 8.8 1.3 8 8 7 7.9 0.7

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Glycopeptide Profile and Site Occupation (UPLC-QToF-MS) Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Observed Relative % Relative contributions of glycoforms at each site of the N-linked oligosaccharide Approximate holding time (min) Tryptic peptides Glycan A4 4 4 5 4.1 0.4 3 3 3 3.1 0.3 FA3G1 5 5 5 5.0 0.3 5 5 5 5.4 0.1 FA4 8 7 10 8.2 1.7 6 7 6 6.3 0.5 FA4G1 6 5 6 5.6 0.6 5 6 6 5.4 0.5 46.3-47.0 T45-46 FA2 17 17 16 16.5 0.0 17 18 22 18.7 2.6 FA2G1 9 10 7 8.6 1.8 13 10 12 11.9 1.4 FA3 17 18 21 18.5 2.2 20 18 15 17.8 2.6 FA2G2 6 7 5 6.2 1.0 7 6 8 6.8 1.1 FA3G1 eleven 10 9 9.9 1.0 12 eleven eleven 11.5 0.8 FA4 14 13 17 14.8 2.2 2 12 1 4.8 6.0 FA3G2 5 5 4 4.8 0.8 5 5 6 5.7 0.7 FA4G1 9 8 9 8.6 0.5 8 8 7 7.8 0.5 FA4G2 5 4 4 4.2 0.4 4 5 5 4.5 0.3 19.6-20.0 T55 A1 4 4 4 4.2 0.1 3 5 3 3.7 0.9 A2 38 37 33 36.0 2.6 29 44 42 38.4 8.0 FA2 36 41 44 40.1 3.7 40 29 32 33.7 5.5 FA2G1 7 6 5 5.6 0.9 5 6 8 6.2 1.4

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD Total sialic acid per monomer (mol/mol) 2.8 2.0 1.8 2.20 0.53 1.1 1.5 2.1 1.57 0.50 Phosphorylation of LC-MSe Peak area T13 non-phosph. † 2611 2190 2137 2800 2672 2579 Peak area T13 S93 phosph. ‡ 1370 1215 967 1375 1260 1324 % Phosphorylated 34.4 35.7 31.2 33.8 2.3 32.9 32.0 33.9 32.9 1.0 ICP-MS Ionic molarity factor Zn2+ 1.63 1.81 1.41 1.62 0.20 1.88 1.88 2.27 2.01 0.23 Mg2+ 0.05 0.05 0.03 0.05 0.02 0.04 0.08 0.12 0.08 0.04 Ca2+ 0.77 0.87 0.85 0.83 0.05 1.04 1.04 0.97 1.02 0.04 Specific Activity (pNPP) (U/mg) 877.0 896.0 812.0 861.7 44.0 875.0 991.0 1079.0 981.7 102.3 HA binding % Binding 92 92 89 91 2 97 92 85 91 5

†: Peptide modes: Cys102 COMe; Molecular formula: C102H159N27O38S1; Monoisotopic mass: 2402.106

‡: Peptide modes: Cys102 COMe S93 PO4; Molecular formula: C102H160N27O41S1P1; Monoisotopic mass: 2482.0723

Table 29 (Cont.). Physico-chemical data of Asfotase Alpha

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD PPi Example Km 90.8 88.9 73.5 94.5 8.1 81.0 41.1 90.3 70.8 26.1 almost equal
reference Km 81.7 81.7 62.3 75.2 11.2 69.3 38.3 79.6 62.4 21.5 K _m % of: Standard 111% 109% 118% 129% 33% 117% 107% 113% 113% 5% K _cat sample 125.7 141.5 103.7 123.6 19.0 125.7 100.0 137.0 131.4 19.0 almost equal
reference K _cat 124.0 124.0 107.6 118.5 9.5 116.0 99.0 114.9 115.5 9.5 K _cat % of reference 101% 114% 96% 104% 9% 108% 101% 119% 114% 9%

Table 30 Tests for forced thermal degradation of asphotase alfa

Analysis category Analysis Time points for each analysis T0 T12±3 T24±3 T48±3 Purity SDS-PAGE in recovery/
non-reducing conditions X X Purity and Impurities SEC HPLC X X X X impurities AEX X X X X impurities RP-HPLC X X X X impurities Peptide mapping X X X potency Specific Activity (pNPP) X X X X

Table 31. Summary of the results of forced thermal degradation of asphotase alfa

Test 2K 20K #35 #36 #38 2K Average 2K SD #40 #42 #34 20K Average 20KSD RP-HPLC % Main Peak Area (RT ~ 21.2 minutes) T0 99.07 99.61 99.45 99.38 0.28 99.61 99.64 99.65 99.63 0.02 T14 89.21 88.98 80.86 86.35 4.76 90.88 91.65 91.35 91.29 0.39 T24 82.96 83.01 75.33 80.43 4.42 87.09 86.68 86.32 86.70 0.39 T48 73.28 72.24 67.48 71.00 3.09 76.25 78.70 76.87 77.27 1.27 AEX % Main peak area T0 93.90 94.60 94.20 94.23 0.35 96.00 96.80 96.50 96.43 0.40 T14 87.00 86.40 77.70 83.70 5.20 96.10 94.10 92.60 94.27 1.76 T24 76.40 77.20 67.30 73.63 5.50 82.40 83.00 81.70 82.37 0.65 T48 61.47 63.89 56.87 60.74 3.57 68.80 72.10 69.00 69.97 1.85 GP-HPLC % Main peak area T0 96.0 96.2 95.9 96.0 0.2 97.5 98.2 97.8 97.8 0.4 T14 84.7 85.3 74.3 81.4 6.2 87.2 88.7 88.1 88.0 0.8 T24 77.5 77.6 67.4 74.2 5.9 80.7 82.2 81.2 81.4 0.8 T48 63.8 62.9 57.4 61.4 3.5 68.0 71.9 69.5 69.8 2.0 Specific Activity (pNPP) (U/mg) T0 901 900 781 861 69 938 949 1084 990 81 T14 793 795 684 757 64 613 679 801 698 95 T24 747 793 600 713 101 577 585 732 631 87 T48 524 543 458 508 45 386 505 369 420 74 SDS-PAGE Under non-reducing conditions % Main band T0 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 T48 96.0 97.5 97.0 96.8 0.8 97.1 97.1 97.8 97.3 0.4 Under reducing conditions % Main band T0 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0 T48 100.0 100.0 100.0 100.0 0.0 100.0 100.0 100.0 100.0 0.0

Purity

Analytical ultracentrifugation (AUC)

[0250] The % monomer levels determined by analytical ultracentrifugation were comparable between the 2K and 20K lots. All values ranged from 94.5% to 97.6% with a 20K mean of 96.8% and a 2K mean of 95.6%. The difference lies in the variability of the AUC assay, as previously demonstrated for the IgG1 antibody assay (see Pekar and Sukumar 2007 Quantitation of aggregates in therapeutic proteins using sedimentation velocity analytical ultracentrifugation: Practical considerations that affect precision and accuracy. Analytical Biochemistry , 367: 225- 237), which is also a glycoprotein with a similar molecular weight. The GP-HPLC data (shown below) shows a comparable % monomer, which also indicates that the small batch-to-batch difference found by AUC was due to intra-assay variation (e.g., differences between instrument elements). A small additional peak at 2.5(c) was found for asphotase alpha obtained from exemplary process #35, which was an artifact of the c(c) analysis. The offset of peak #35 corresponding to asphotase alpha dimer (approximately 11(c)) is considered to be within the margin of error of the method. Asphotase alfa AUC data were obtained from the University of Connecticut, Analytical Ultracentrifugation Center, Biotechnology Bioservice Center. This method distinguishes asphotase alpha monomer from aggregates consisting of dimeric or larger species and their relative % depending on the coefficient of continuous distribution of sedimentation.

SDS-PAGE/ LoC (Lab on Chip)

[0251] % main band as determined by SDS-PAGE: ^LABCHIP® (PerkinElmer) GXII Protein analysis for both non-reducing and reducing conditions was comparable between 2K lots and 20K lots. As summarized in Table 29, the mean % of the main band was 100.0% (non-reducing conditions) and 99.8% (reducing conditions) for the three batches of 2K and 100.0% (non-reducing conditions) and 99.8% (reducing conditions) for batches of 20K. In addition, comparable band patterns were observed for batches 2K and 20K as demonstrated by electropherograms of both non-reducing and reducing assays. This test separates a protein based on its molecular weight and provides an analysis of the purity of the intact protein, expressed as a percentage of the main band. The assay was performed according to the LabChip GXII Protein Assay Quick Guide, HT Protein Express LabChip Kit, version 2 (revised March 2010).

SDS-PAGE

[0252] Three batches of 2K and three batches of 20K were analyzed in parallel using SDS-PAGE. Densitometry analysis indicated that the baseband was 100% for all batches for both SDS-PAGE under non-reducing and reducing conditions. In addition to the main band, a faint band below the limit of quantitation was observed with an apparent molecular weight higher than that of asphotase alpha dimer for all batches. Only one band was observed for all batches using SDS-PAGE under reducing conditions. This test separates a protein based on its molecular weight and provides an analysis of the purity of the intact protein, expressed as a percentage of the main band. The protein sample was separated on the gel, stained with Coomassie blue for visualization, washed out the color and analyzed by densitometry. The molecular weights of asphotase alpha, degradation products and process-related impurities were evaluated against known molecular weight standards. Quantitative measurements of separated proteins are performed by scanning and analysis using densitometric analysis.

GP-HPLC

[0253] Asphotase alpha purity was analyzed in parallel using GP-HPLC. The mean 2K (96.5%) and mean 20K (98.6%) are shown in Table 29. The % asphotase alpha from the 20K lots is slightly higher than the 2K lots. However, lot data showed a similar % dimer, suggesting that the slightly lower % dimer from 2K lots is mainly due to differences in sample history such as material age. The GP-HPLC chromatogram also showed comparable products from lots 2K and 20K. The same peaks at the same retention times indicate the same species. This method distinguishes asphotase alpha from larger species and smaller fragments.

Anion exchange chromatography (AEX)

[0254] The % main peak AEX for all 2K lots and 20K lots were in a narrow range from 93.62% to 96.84% as shown in Table 29. The mean 2K was 94.17% and the mean 20K was 96 .38% of the area of the main peak. A slightly higher % of the main peak was observed for the 20K batches due to a decrease in the late elution peak. The slightly higher level of the main peak in the 20K batches demonstrates the higher purity of the product. The AEX chromatogram showed superimposed peaks for lots 2K and 20K with the same retention time, indicating that the products from all lots have the same appearance. This method separates the protein in order of increasing net surface anionic charge. Separated proteins are detected by ultraviolet absorbance at 280 nm and then quantified as a percentage of the peak area.

Charge profile

Capillary isoelectric focusing (cIEF)

[0255] The peak profiles and relative % of each peak of the 2K and 20K lots were comparable when asphotase alpha was analyzed by cIEF. All samples show six prominent peaks in the isoelectric point (pI) range of ~6.57 to ~6.85, and these peaks are referred to as peaks 1, 2, 3, 4, 5, and 6 from left to right. The peak areas found in lots 2K and 20K for each charged species variant are comparable as shown in Table 29. The comparability between % peak cIEF from lots 2K and 20K was confirmed using T-test analysis of all lots as shown in Table 32 ( all p-values were >0.05, indicating that any observed differences between batches were not significant). The cIEF electrophoregrams from batches 2K and 20K showed the same peak profiles in the same pI ranges. cIEF provides a semi-quantitative separation of charge-variant species based on the pI of protein isoforms. Separated proteins were detected at 280 nm absorption.

Table 32. T-test % peak cIEF from Lots 2K and 20K

Peak cIEF Peak 1 Peak 2 Peak 3 Peak 4 Peak 5 Peak 6 T-test % peak from 2K and 20K batches 2K parties #35 33.55 10.79 12.36 17.21 10.29 15.81 #36 29.01 9.54 13.49 18.71 11.78 17.47 #37 27.97 11.38 12.25 18.96 11.97 17.47 #38 33.09 10.24 13.62 17.09 11.21 14.74 #39 27.87 6.02 11.96 12.93 15.33 25.90 20K batch #40 24.46 8.53 12.97 18.79 14.62 20.62 #41 35.52 6.87 15.94 11.42 12.69 17.56 #42 24.40 8.96 10.60 20.25 13.27 22.53 #43 32.50 9.65 11.60 18.00 11.64 16.60 #34 36.64 8.66 11.78 16.71 11.24 15.27 #44 24.90 10.48 15.39 18.20 9.74 21.30 #45 27.40 10.10 11.65 20.94 9.50 20.40 T-test P -value 0.74 0.57 0.91 0.65 0.79 0.67

Molecular size

Size Exclusion Chromatography - Multiple Light Scattering (SEC-MALS)

[0256] Asphotase alpha molecular weights determined by SEC-MALS of lots 2K and 20K were comparable. The average molecular weight of 2K lots is 189.2 kDa and the average molecular weight of 20K lots is 189.2 kDa, as shown in Table 29. Similar peak profiles were observed for all lots on UV chromatograms (sample from lot #36 was run later, due to availability of samples, resulting in inconsistencies between its peak profiles and other samples). The SEC-MALS method combines a SEC column to separate protein species by size using a MALS instrument to determine the molecular weight of the peaks. The high variation in molecular weights of high molecular weight species (see Table 29) as measured by scatter signal and refractive index (RI) is due to the higher method variance for low abundance species. Based on the UV chromatograms, the relative retention times of the high molecular weight asphotase alpha species are the same for all six batches, indicating that the molecular weights of the high molecular weight species should also be the same.

Intact molecular weight analysis (MALDI-ToF-MS)

[0257] The average molecular weight of the 2K lots was 180,902 kDa, and the average molecular weight of the 20K lots was 180,019 kDa, as shown in Table 29. The molecular weights of the 20K lots and 2K lots are within 1% mass accuracy of an externally calibrated MALDI-ToF and therefore , is comparable. The slightly higher 2K molecular weight trend is likely due to slightly more large oligosaccharides such as sialic acid oligosaccharides. Intact molecular weight (IMW) MALDI-ToF spectra for asphotase alpha 2K and 20K batches also showed the same peak profiles. Lot #36 was analyzed on a separate occasion using a different instrument calibration but was included for completeness of comparability. Additional masses present in the spectra are due to analysis-induced fragmentation and or to a few charged ions. The intensity differences observed for the identified species are the result of matrix crystallization and laser effects. This method identifies a molecule based on molecular weight. Test samples were extracted with solid phase and mixed with sDHB matrix solution and co-deposited on MALDI targets. This dried sample was laser ionized in a ToF mass spectrometer. The m/z spectrum was taken from each sample, and the intact m/z was converted to molecular weight.

Identity

Deglycosylated/Reduced and Deglycosylated Molecular Weight Analysis (MALDI-ToF-MS)

[0258] The molecular weight of deglycosylated asphotase alpha was calculated from the primary amino acid sequence of two identical 726 amino acid polypeptide amino acid sequences using NIST molecular weights and isotope percentages with 20 cysteines forming 10 disulfide bonds. The deglycosylated molecular weight of asphotase alpha was calculated to be 161135.20 Da.

[0259] The molecular weights of the deglycosylated asphotase alpha from the 20K lots and from the 2K lots were within 1% of the calculated deglycosylated molecular weight (mass accuracy of externally calibrated MALDI-TOF) as shown in Table 29. The average molecular weight of the 20K lots was 160881 kDa, and the average molecular weight of the 2K lots was 161,050 kDa as shown in Table 29. The MALDI-TOF spectra for the 2K and 20K lots showed similar peak profiles. Additional masses present in the spectra were due to assay-induced fragmentation, incomplete deglycosylation, and/or multiply charged ions. The intensity differences observed for the identified species are the result of matrix crystallization and laser effects.

[0260] The molecular weight of reduced and deglycosylated asphotase alpha was calculated from the primary amino acid sequence of the 726 amino acid polypeptide amino acid sequence using NIST molecular weights and isotope percentages. The calculated molecular weight of reduced and deglycosylated asphotase alpha is 80577.68 Da.

[0261] Molecular weight values for reconstituted and deglycosylated lots of asphotase alpha 20K are compared to lots of 2K, and all values were within 1% of the calculated reconstituted and deglycosylated molecular weight (mass accuracy of externally calibrated MALDI-ToF) as shown in Table 29. The mean number of 2K batches was 80530 Da and the mean of 20K batches was 80539 Da. The MALDI-ToF spectra for batches 2K and 20K showed comparable peak profiles. Additional masses present in the spectra are due to analysis-induced fragmentation and/or multiple charged ions. The intensity differences observed for the identified species are the result of matrix crystallization and laser effects. This method identifies a molecule based on molecular weight. Test samples were extracted with solid phase and mixed with sDHB matrix solution and co-deposited on MALDI targets. This dried sample was laser ionized in a ToF mass spectrometer. The m/z spectrum was taken from each sample, and the intact m/z was converted to molecular weight.

[0262] The molecular weights of the deglycosylated asphotase alpha batches 2K and 20K were comparable and all values were within 100 ppm. (ppm) of the calculated deglycosylated molecular weight (which is the mass precision of an externally calibrated ESI-ToF-MS) as shown in Table 29. The average molecular weight of the 2K batches was 161137.39 Da and the average molecular weight of the 20K batches was 161137.28 Yes, and the values were consistent with the calculated deglycosylated molecular weight of 161135.20 Da. The reduced and deglycosylated molecular weight values for the main peak of the 2K lots were comparable to those of the 20K lots, and all values were within 100 ppm of the calculated reduced and deglycosylated molecular weight (which is the mass accuracy of an externally calibrated ESI-ToF-MS) as shown in Table 29. The average molecular weight of the 2K lots was 80575.52 Da and the average molecular weight of the 20K lots was 80574.76 Da. The values are consistent with the calculated reduced and deglycosylated molecular weight value for asphotase alpha of 80577.68 Da. This method identifies a molecule based on its intact molecular weight. Test samples were injected into a C4 RP-HPLC column and eluted with an aqueous:organic solvent gradient. The eluate was then electrosprayed into a ToF mass spectrometer and the spectrum from the upper half of the chromatographic peak was subjected to a deconvolution operation to ensure intact molecular weight.

[0263] The molar ratios of zinc and magnesium ions determined by inductively coupled plasma mass spectrometry (ICP-MS) for three lots of 20K and three lots of 2K were comparable, as shown in Table 29. Plasma Atomic Emission Spectroscopy (ICP-AES) for batches 20K and 2K were comparable, as shown in Table 29. The molecular ratios of ions for zinc and calcium are also consistent with the literature, where two zinc and one calcium were associated with each alkaline enzyme monomer. phosphatases (see E. Mornet et al., 2001 ʺStructural evidence for a functional role of human tissue non-specific alkaline phosphatase in bone mineralization.ʺ JBC , 276 : 31171-31178; and B. Stec, KM. Holtz and ER. Kantrowitz 2000 "A revised mechanism for the alkaline phosphatase reaction involving three metal ions " JMB 299 : 1303-1311). The molar ratio of magnesium ions for all parties was unexpectedly below the expected value indicated in the literature. ICP-AES was chosen for calcium analysis because ICP-AES does not suffer from the argon polyatomic interference seen with calcium isotopes in ICP-MS. Asphotase alfa ICP data were purchased commercially. These methods determined the amount of magnesium, zinc and calcium (µg/ml), from which the ratios of molar ions for asphotase alpha were calculated.

Phosphorylation site identification and quantification using UPLC-MSe

[0264] The site and relative percentage of phosphorylation was determined by UPLC-MS tryptic peptide mapping. It is known that alkaline phosphatase contains a serine residue in the active site, which can form a serine phosphate intermediate (see pp. 28-29 Millán, José Luis. 2006 "Mammalian Alkaline Phosphatases." Wiley-VCH). The same 93 serine residue was found to be phosphorylated in all batches in percentages ranging from 35.7% to 31.2% as shown in Table 29. The percentage of phosphorylation was found not to correlate with the specific activity of asphotase alpha (pNPP ), binding to HA, K _m and K _cat (data not shown). Based on the lack of correlation, the phosphorylated intermediate is not a rate-limiting step in asphotase alpha activity. Confirmation and identification of the phosphorylation site at S93 in asfotase alpha was confirmed by UPLC-MSe peptide mapping. The theoretical asphotase alpha sequence and predicted tryptic cleavages were used for analysis. The resulting LC-MS ^e data were collected and processed using Waters Biopharmalynx v.1.2. Biopharmalynx Software (Waters) identified a 13th tryptic peptide from the primary sequence (T13) that contains S93, an active site serine from the enzymatic moiety having Cys102 terminated with ac, and S93 existing both phosphorylated and unphosphorylated. Once identified as part of a peptide map, extracted ion chromatograms (EICs) were pulled for each of these peptides and smoothed and integrated. From the resulting EIC peak areas, the ratio of the phosphorylated peptide peak area to the total peak area gives the relative percentage of phosphorylation on S93. Confirmation of +80 Da associated with the phosphorylated serine side chain was obtained by examining the MS ^e fragmentation pattern generated for the proposed T13 phosphorylated peptide as shown in FIG. 36. This fragmentation pattern confirmed that S93 is shifted by +80 Da (phosphorylation mass), which is observed with y12 and y13 ions. Each batch of asphotase alpha was reduced and alkylated, then digested with trypsin. Detection was performed by ESI-Q-ToF-MS using a Waters Acquity UPLC and Synapt Q-ToF mass spectrometer operating in positive ion electrospray mode.

GLYCOSYLATION

Mass profiling of N-linked oligosaccharides (MALDI-ToF-MS)

[0265] Reconstituted oligosaccharides from three 2K lots and three 20K lots were analyzed by MALDI-ToF, the results are shown in Table 29. The same glycan species with similar relative percentages were observed in 20K lots and 2K lots. This method identifies drug-associated glycans using molecular weights. Glycans were enzymatically cleaved from the drug using PNGase-F digestion and subsequent acidification. The glycans were then solid phase extracted and mixed with super-3, 4-dihydroxybenzoic acid matrix solution and co-precipitated onto MALDI targets. This dried sample was laser ionized in a ToF mass spectrometer. Collected spectra m/z (M+Na) +.

Profiling of 2-AB-labeled N-linked oligosaccharides (HPLC)

[0266] Glycoform types and the relative percentage of each glycoform were analyzed by 2-AB labeling and HPLC analysis. The results are shown in Table 29. A peak with the same retention time and with a similar relative percentage was observed in all batches (data not shown) suggesting the same types of oligosaccharides and similar levels. Batch number 36 was launched later. As an example, the percentage of basic FA2 glycoform was plotted against enzymatic activity. No correlation was observed between FA2 percentage and specific activity (pNPP), HA binding, K _m and K _cat (data not shown).

[0267] This assay characterizes the pattern of glycosylation by determining the oligosaccharides associated with a drug molecule based on retention time and peak area of fluorescently labeled free oligosaccharides. Free oligosaccharides from the drug were released enzymatically and labeled with 2-aminobenzamide (2-AB) by reductive amination. Samples were then injected into a HPLC system with a Ludger column and 2AB labeled oligosaccharides were separated and detected with a fluorescence detector; 330 nm excitation, 420 nm emission. The chromatograms of each batch were compared and the peak area of each resolved peak was used to determine the relative amount of oligosaccharide from each batch.

Glycopeptide profile by UPLC-QToF-MS

[0268] The oligosaccharide profile at each glycosylation site was assessed by glycopeptide analysis using UPLC-MS. Tissue non-specific alkaline phosphatase (TNAP) has five N-glycosylation sites (see pp. 54-55 Millán, 2006 Wiley-VCH). Asfotase alfa contains an additional N-glycosylation site in the Fc region, providing a total of six N-glycosylation sites per monomer. In FIG. 44 in FIG. 49 shows the glycopeptide fingerprint profiles for N123 in T15-16, N213 in T26-27, N254 in T33, N286 in T35, N413 in T45-46, and N564 in T55. UPLC-Q-ToF-MS glycopeptide fingerprints of summarized N-linked oligosaccharides provide a detailed characterization of N-glycosylation. Identification of glycopeptide species by mass analysis and relative quantification of each species was performed by transforming the multiple charge states observed in the raw glycopeptide fingerprint spectra into singly charged and deisotopic spectra for each site. The results of these transformations are shown in Table 29. It lists all the masses observed for glycoforms and their relative % for glycoforms constituting ≥5% of the total glycoform content in each observed N-linked site of the oligosaccharide. In this assay, asphotase alpha was reduced and alkylated, then digested with trypsin. Detection was performed by ESI-Q-ToF-MS using a Waters Acquity UPLC and Synapt Q-ToF mass spectrometer operating in positive ion electrospray mode. The theoretical asphotase alpha sequence and predicted tryptic cleavages were used for analysis. Molecular weights for theoretical glycopeptides were calculated using NIST molecular weights and isotope percentages. Glycopeptides are observed as collections of related glycoforms with close elution times and therefore require spectrum stacking to ensure complete addition of the glycopeptide for a particular site. For each of the most common peptide species containing each of the N-linked oligosaccharide sites in asphotase alpha, site-specific summation spectra (glycopeptide mass fingerprints) were obtained for each batch tested.

Total Sialic Acid (TSAC)

[0269] The total content of sialic acid was determined using HPAE-PAD. As shown in Table 29, the sialic acid content for each batch is within the specification of 0.9-3.5 moles of sialic acid per mole of monomer. Considering the change in sialic acid content due to its relatively lower amount, batches 2K and batches 20K were comparable. No correlation was observed between TSAC value and specific activity (pNPP), HA binding, K _m and K _cat .

ACTIVITY

Specific Activity (pNPP)

[0270] Asphotase alpha specific activity was determined for three 20K and three batches of 2K. The average specific activity of the 20K lots is 981.7 U/mg and 861.7 U/mg for the three 2K lots. The specific activity of all lots was in the range of 620-1250 U/mg and was comparable between the 20K and 2K lots. This method is used to determine the enzymatic activity of asfotase alpha using pNPP as a substrate. Asfotase alfa is a recombinant protein that has a single domain from the tissue-specific human alkaline phosphatase enzyme. This domain is catalytically active and hydrolyzes phosphate esters. The analysis is carried out at saturation of the enzyme to achieve and maintain V _max for the duration of the measurement. pNPP is hydrolysed to a yellow product (maximum absorbance at 405 nm). The reaction rate is directly proportional to the activity of the enzyme. One unit (U) corresponds to 1 μmol pNPP produced per minute under the conditions of the method used. Specific activity with pNPP (enzymatic activity per mg protein) was calculated based on the enzyme activity and protein concentration per A ₂₈₀ . The results are shown in Table 29.

Binding of hydroxyapatite

[0271] The % hydroxyapatite binding as determined by the hydroxyapatite (HA) binding assay in 2K asphotase alpha lots is comparable to 20K lots with all values in the 85-97% range. The average % HA binding is 91% for both batches of 20K and 2K (Table 29). The HA binding assay measures the enzymatic activity of the asphotase alpha complex of pNPP upon binding to hydroxyapatite. The synthetic substrate, p-nitrophenyl phosphate (pNPP), when hydrolyzed by asphotase alpha, produces a yellow product that can be quantified by absorption at 405 nm. This analysis was performed while saturating the substrate to achieve and maintain Vmax during the reaction.

PPi enzyme kinetic analysis using a physiologically relevant substance

[0272] the Kinetic parameters K _m and K _cat were determined using the kinetic analysis of PPi. The samples were run in some cases in parallel with the reference standard. For comparison, K _m and K _cat are plotted as a percentage of the reference standard value during sample testing (Table 29). All K _m values were within 30% of the parallel run reference standard and are considered comparable. All K _cat values were within 20% of the reference standard parallel run and are considered comparable. K _m % reference and K _cat % reference for 2K lots were comparable to 20K lots. The PPi kinetic assay measures the activity of purified asphotase alpha (enzymatic activity) using PPi, an alkaline phosphatase substrate, and determines the kinetic constants _Km and _Kcat of PPi hydrolysis under physiological conditions (37°C, pH 7.4).

FORCED TEMPERATURE DEGRADATION

[0273] To further ensure comparability of batches 2K and 20K, a parallel forced degradation study was conducted to compare degradation pathways and kinetics. Three lots of 2K (#35, #36, and #38) and three lots of 20K (#40, #42, and #34) shown in Table 33 were used for the study. Some lots required the use of drug product material (DP) due to limited availability of sample BDS.

[0274] Lots #35, #36, #40, #42, and #34 were diluted using asphotase alpha buffer (25 mM sodium phosphate and 150 mM sodium chloride, pH 7.4) to 40 mg/mL to match the concentration protein in Lot #38. To achieve significant levels of degradation to establish degradation kinetics, forced degradation conditions were selected based on a preliminary screening of forced degradation conditions. Aliquots (200 µl) were created for each test run at each time point (0, 14, 24 and 48 hours). Aliquots of T0 were kept at 5°C until testing. The remaining aliquots of the sample were incubated at 40°C for 14, 24 and 48 hours. Samples were analyzed by SDS-PAGE, GP-HPLC, AEX, RP-HPLC, peptide mapping and specific activity as indicated in Table 30 to determine comparability. Test results for forced thermal degradation studies are shown in Table 31 for three lots of 20K and three lots of 2K.

Table 33. Lots Used for Degradation Studies

Name Concentration (mg/ml) #35 90.9 #36 96.4 #38 40.0 #40 107.6 #42 92.9 #34 101.6

SDS-PAGE (Non-reducing and reducing conditions)

[0275] T0 and T48 hours of samples were analyzed by SDS-PAGE (under non-reducing and reducing conditions) for purity. The % main band SDS-PAGE under non-reducing conditions was 100% for all lots at T=0 and decreased to a mean of 97.3% for 20K lots and 96.8% for 2K lots, as shown in Table 31. The degradation trend is comparable for batches 20K and 2K. High molecular weight species were detected in T48 hours of the sample for all six batches. This species disappeared during restoration. This result suggests that the formation of disulfide bonded aggregates is a degradation pathway for all batches. No additional bands were observed upon SDS-PAGE reduction for batches 20K and 2K at T=0 and T=48 hours, indicating no degradation of the peptide bonds. This test separates a protein based on its molecular weight and provides an analysis of the purity of the intact protein, expressed as a percentage of the main band. The protein sample is isolated on a gel, stained with Coomassie blue for visualization, stain washed out and analyzed by densitometry. The molecular weight of asphotase alpha, degradation products, and process-related impurities is estimated by comparison to known molecular weight standards. Quantitative measurements of separated proteins are performed by scanning and analysis using densitometric analysis.

GP-HPLC

[0276] Aggregate and fragment degradation caused by thermal stress was monitored by GP-HPLC and the results are shown in Table 31. The peak profiles for all batches at T=0 were similar, indicating the presence of the same species. Thermal stress resulted in a peak at about 6 minutes, corresponding to large aggregates that increased over time. The peak profiles of all batches at T=14 hours, T=24 hours and T=48 hours were similar, indicating a similar degradation path. The decline slopes of the main peak over time for all batches were similar, given that the materials used were of different ages. A similar slope indicates similar degradation kinetics.

Anion exchange chromatography (AEX)

[0277] Asphotase alpha degradation was also monitored using AEX, which separates proteins based on their charge. The same peak profiles were observed for all lots at T=0, indicating the same appearance in all lots. Thermal stress resulted in increased peaks in a retention time window of approximately 22-26 minutes. Comparable peak profiles for all lots were observed for all lots at T=14 hours, T=24 hours and at T=48 hours, indicating the formation of the same species by thermal decomposition. Similar degradation slopes, controlled by the main peak reduction, were observed for all batches, considering the small difference at T=0 (as described earlier) and the difference in age between batches.

RP-HPLC

[0278] In RP-HPLC analysis, an average peak in a retention time window of approximately 21.2-21.3 minutes was observed for all lots, indicating the same species across all lots. The main peak was the only main peak observed for all batches at T=0. Thermal stress resulted in an increase in the peak eluting immediately after the main peak. The same peak from thermal degradation observed for all batches at T=14, T=24 and T=48 hours indicates the same degradation path. The degradation kinetics, controlled by the decrease in the main peak, was comparable for all batches, as evidenced by the similar slopes of all batches.

Peptide mapping

[0279] The T0, T24, and T48 times of the samples for batches 20K and 2K were analyzed by peptide mapping to assess potential chemical degradation. No significant difference was found between 20K and 2K at all times (T=0 hour and T=48 hours). This method denatures, reduces, and alkylates the protein with guanidine hydrochloride, dithiothreitol, and iodoacetate, followed by protease trypsin digestion. The resulting peptide fragments were separated by reverse phase gradient HPLC and detected at 210 nm.

Specific Activity (pNPP)

[0280] The specific activity (pNPP) of 20K lots and 2K lots of asphotase alpha was determined by enzymatic analysis after incubation at 40°C at each time point (T0, T14, T24 and T48). The activity of the 20K parties was comparable to the 2K parties at each time point. The decrease in activity caused by thermal stress follows similar slopes for all batches over time. Protein phosphatase is an enzyme that controls the removal of the phosphate (PO ₄ ^3- ) group from protein molecules. The pNPP phosphatase assay was used to determine phosphatase activity and compare lots of 20K and 2K. This method is used to determine the enzymatic activity of asfotase alpha using pNPP as a substrate. Asfotase alfa is a recombinant protein that has a single domain from the tissue-specific human alkaline phosphatase enzyme. This domain is catalytically active and hydrolyzes phosphate esters. The analysis is carried out at saturation of the enzyme to achieve and maintain V _max for the duration of the measurement. pNPP is hydrolysed to a yellow product (maximum absorbance at 405 nm). The reaction rate is directly proportional to the activity of the enzyme. One unit (U) corresponds to 1 μmol pNPP produced per minute under the conditions of the method used. Specific activity with pNPP (enzymatic activity per mg protein) was calculated based on the enzyme activity and protein concentration per A ₂₈₀ .

[0281] The comparability of asphotase alpha, made in 20K scale, and asfotase alpha, made in 2K scale, was established. Where possible, three batches on each scale were analyzed in parallel using physicochemical methods to evaluate asphotase alpha purity, impurities, charge variants, size, structure, glycosylation, disulfide bond, free thiols, and activity. In addition, a parallel forced degradation study was conducted to evaluate degradation pathways and kinetics. The results show that the products produced at the 20K and 2K scales were comparable.

[0282] Lot purity was assessed by analytical ultracentrifugation, SDS-PAGE/LoC, SDS-PAGE, GP-HPLC and AEX. The results showed comparable purity for all lots. The same types of aggregates at very low comparable levels were found in all batches by AUC, SDS-PAGE and GP-HPLC.

[0283] The charge variants common to recombinant proteins were assessed using cIEF. The same appearance at comparable levels was observed for all batches, indicating a consistent level of post-translational modifications.

[0284] Molecular size of batches was assessed using SEC-MALS and intact molecular weight MALDI-ToF. Identity was confirmed by deglycosylated/reduced and deglycosylated MALDI-ToF molecular weight and deglycosylated/reduced and deglycosylated ESI-ToF molecular weight. Structure was also assessed by CD, disulfide binding and free thiols (LC/MS), complete free thiol analysis, metal content analysis (ICP-MS/ICP-AES), and phosphorylation site population. Comparable molecular weights and comparable hydrodynamic sizes were obtained for all lots, indicating similar modifications and similar asphotase alpha conformation from all lots.

[0285] Asphotase alpha glycosylation was assessed using MALDI-ToF free glycan, fHPLC 2AB labeled oligosaccharide, LC-MS glycopeptide analysis and total sialic acid. The same oligosaccharide species at comparable levels were obtained for all lots.

[0286] The activity of the drug substance was assessed by specific activity (pNPP), HA binding and PPi activity. The results showed that batches of 20K asphotase alfa were comparable to batches of 2K asphotase alfa.

[0287] Finally, the same degradation pathway and similar degradation kinetics of batches 20K and 2K were observed in a parallel forced degradation study using thermal stress.

[0288] Thus, batches of 20K asphotase alpha and batches of 2K are comparable, as evidenced by extended characterization using multiple orthogonal methods and a parallel forced degradation study. Thus, material from 20K lots should have comparable safety and efficacy to material from 2K lots.

Example 10 Additional Comparability of Asphotase Alfa Produced in 2000 L (2K) and 20000 L (20K) Scales

[0289] Seven batches of asphotase alfa were generated at the 20,000 L scale, and five additional batches of asphotase alfa were generated using the 2000 L scale. Products for both scales were analyzed and found to be comparable (data not shown).

Claims (48)

1. A recombinant mammalian cell cultured polypeptide, wherein the recombinant polypeptide contains asphotase alpha (SEQ ID NO: 1) and has an average total sialic acid content (TSAC) of between about 0.9 to about 3.5 mol sialic acid/mol protein monomer and at least one characteristic selected from the group consisting of: a) isoelectric focus (IEF) between about 5.2 and about 6.7; b) a large glycan structure as shown in FIG. 41 or FIG. 42; c) chromatogram profile of the 2-AB-labeled oligosaccharide as shown in FIG. 38 or FIG. 39; d) MALDI-ToF fingerprint profile of the glycopeptide shown in FIG. 40 or FIG. 44-49; e) a major band on SDS-PAGE under reducing conditions having a molecular weight of about 88-108 kDa and not less than about 85% of the total recombinant polypeptide produced; f) a major band on SDS-PAGE under non-reducing conditions having a molecular weight of about 194 kDa to about 273 kDa and not less than about 85% of the total recombinant polypeptide produced; g) not less than about 95.0% for recombinant polypeptide dimers and not more than about 5.0% for aggregates, as determined by high performance liquid size exclusion chromatography (HPLC); h) a purity of at least about 95.0% as determined by reverse phase high performance liquid chromatography (RP-HPLC); i) not less than about 90.0% for the main peak, not more than about 6.0% for the acidic peaks, and not more than about 4.0% for the main peaks, as determined by anion exchange chromatography (AEX); j) percentage binding of hydroxyapatite (HA) from about 75% to about 125%; k) product specific activity (pNPP) from about 620 to about 1250 units/mg; l) K _m from about 13 to about 69 μM in an inorganic pyrophosphate (PPi) hydrolysis assay; m) K _cat from about 65 to about 165 s ^-1 in the analysis of the hydrolysis of inorganic pyrophosphate (PPi); n) pI range from about 6.45 to about 6.95 for all peaks in capillary electrophoresis; o) peaks in the MALDI-TOF mass spectrum as shown in FIG. 34A after deglycosylation; p) peaks in the MALDI-TOF mass spectrum as shown in FIG. 34B after reduction and deglycosylation; q) peaks in the MALDI-ToF mass spectrum as shown in FIG. 35; r) a phosphorylation profile as shown in FIG. 36; s) profile of the sialylated glycan on a negative MALDI-ToF mass spectrum as shown in FIG. 37A; t) profile of neutral glycans on a positive MALDI-ToF mass spectrum as shown in FIG. 37B; u) a molar ratio of magnesium per mole of recombinant polypeptide less than about 0.15; v) a molar ratio of calcium per mole of recombinant polypeptide from about 0.5 to about 1.5; And w) a molar ratio of zinc per mole of recombinant polypeptide from about 0.5 to about 3.0. 2. The recombinant polypeptide of claim 1 having from about 0.7 to about 1.19 free cysteine per half molecule. 3. A recombinant polypeptide according to claim 1 or 2 having a phosphorylation on Ser 93 in percent from about 13.5% to about 35.7%. 4. Recombinant polypeptide according to any one of paragraphs. 1-3 having not less than about 90.0% of the main peak, not more than about 6.0% for the acid peaks, and not more than about 4.0% for the main peaks on the AEX chromatogram. 5. The recombinant polypeptide of claim 4 having a major peak of at least about 93.7%, no greater than about 4.9% for acid peaks, and no greater than about 3.4% for major peaks on an AEX chromatogram. 6. The recombinant polypeptide of claim 1 having an average TSAC of about 1.2 to about 3.0 mol sialic acid/mol per monomer. 7. The recombinant polypeptide of claim 6 having an average TSAC of about 1.9 to about 2.7 mol sialic acid/mol per monomer. 8. The recombinant polypeptide of claim 7 having an average TSAC of about 1.85 to about 2.28 mol sialic acid/mol per monomer. 9. Recombinant polypeptide according to any one of paragraphs. 1-8 having a molar ratio of magnesium from about 0.03 to about 0.15. 10. The recombinant polypeptide of claim 9 having a molar ratio of magnesium from about 0.05 to about 0.10. 11. The recombinant polypeptide of claim 9 having a magnesium molar ratio of about 0.12. 12. Recombinant polypeptide according to any one of paragraphs. 1-11 containing at least about 95.0% recombinant polypeptide dimers and about 5.0% or less polypeptide aggregates, as determined by size exclusion HPLC. 13. The recombinant polypeptide of claim 12, comprising at least about 96.8% recombinant polypeptide dimers and about 3.2% or less polypeptide aggregates, specifically as determined by size exclusion HPLC. 14. The recombinant polypeptide of claim 13, comprising at least about 97.6% recombinant polypeptide dimers and about 2.4% or less aggregates, specifically as determined by size exclusion HPLC. 15. Recombinant polypeptide according to any one of paragraphs. 1-14 having at least about 95.0% purity, specifically as determined by RP-HPLC. 16. A recombinant polypeptide according to claim 15 having at least about 97.6% purity as determined by RP-HPLC. 17. Recombinant polypeptide according to any one of paragraphs. 1-16 having a hydroxyapatite (HA) binding percentage of about 75 to about 125%. 18. The recombinant polypeptide of claim 17 having an average percentage of hydroxyapatite binding from about 85% to about 97%. 19. The recombinant polypeptide of claim 18 having an average percentage of hydroxyapatite binding from about 90% to about 91%. 20. Recombinant polypeptide according to any one of paragraphs. 1-19 having a specific activity (pNPP) of about 620 to about 1250 units/mg. 21. The recombinant polypeptide of claim 20 having a mean specific activity (pNPP) of about 904.0 to about 907.7 U/mg. 22. Recombinant polypeptide according to any one of paragraphs. 1-21, wherein said recombinant polypeptide is encoded by a polynucleotide encoding a polypeptide comprising the sequence of SEQ ID NO: 1 or a sequence fully complementary to SEQ ID NO: 1. 23. Recombinant polypeptide according to any one of paragraphs. 1-22, where the recombinant polypeptide is produced in a periodically fed reactor having a volume of 200 to 20,000 liters. 24. The recombinant polypeptide of claim 23, wherein the fed batch reactor has a volume of 2,000 to 20,000 liters. 25. Pharmaceutical composition for the treatment of a subject suffering from a condition associated with a deficiency of alkaline phosphatase, containing a recombinant polypeptide, as indicated in any one of paragraphs. 1-24 and at least one pharmaceutically acceptable carrier, diluent or excipient.