Abstract
The invention is based on the introduction of mutations into the amino acid sequence of human Arylsulfatase A (ASA or ARSA) in order to increase protein stability. The invention introduces amino acid mutations, such as deletions, substitutions or additions, into the C-terminal part of the human ARSA enzyme, in particular at a position around or at amino acid 424, which result in a sequence that does not comprise E424. Provided are further nucleic acids and vectors for the expression of the mutated ARSA of the invention, recombinant cells and pharmaceutical composition comprising the mutated ARSA, as well as its use in the treatment of diseases that are characterized by a reduced activity of endogenous ARSA.
Claims
1-16. (canceled)
17. A mutated arylsulfatase A (ARSA) enzyme, comprising an amino acid sequence with at least 90% sequence identity to SEQ ID NO: 1 (human ARSA enzyme), wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1.
18. The mutated ARSA enzyme according to claim 17, wherein at least one mutation is a modification, deletion or substitution of E424.
19. The mutated ARSA enzyme according to claim 17, wherein the at least one mutation is a substitution with a W, Y, A, F, R, G, L and preferably is E424R, E424G or E424L.
20. The mutated ARSA enzyme according to claim 17, wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one further mutation at amino acid positions 202, 286 and/or 291 of SEQ ID NO: 1.
21. The mutated ARSA enzyme according to any claim 17, wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one further mutation selected from M202V, T286R and/or R291N compared to SEQ ID NO: 1, preferably of at least M202V.
22. The mutated ARSA enzyme according to claim 17, wherein the mutated ARSA has an increased protein half-life compared to a wild-type human ARSA of SEQ ID NO: 1; and/or has a decreases mannose 6-phosphorylation of lysosomal proteins compared to a wild-type human ARSA of SEQ ID NO: 1.
23. The mutated ARSA enzyme according to claim 17, further comprising a C-terminally attached apoE-II protein.
24. The mutated ARSA enzyme according to claim 17, comprising compared to SEQ ID NO: 1 the mutations at positions M202, T286, R291 and E424, and a C-terminal covalently attached apoE-II protein.
25. An isolated nucleic acid, comprising a nucleotide sequence encoding a mutated ARSA enzyme, wherein the mutated ARSA enzyme is at least 90% identical to SEQ ID NO: 1 (human ARSA enzyme), and wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1.
26. A vector, comprising a nucleotide sequence encoding a mutated ARSA enzyme, wherein the mutated ARSA enzyme is at least 90% identical to SEQ ID NO: 1 (human ARSA enzyme), and wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1.
27. The vector according to claim 26, which is an expression vector, comprising promoter sequence operably linked to the nucleic acid sequence encoding the mutated ARSA enzyme.
28. A recombinant cell, comprising a mutated ARSA enzyme or a nucleic acid encoding the mutated ARSA enzyme, wherein the mutated ARSA enzyme is at least 90% identical to SEQ ID NO: 1 (human ARSA enzyme), and wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1.
29. A pharmaceutical composition, comprising a mutated ARSA enzyme or a nucleic acid encoding the mutated ARSA enzyme, or a recombinant cell comprising the mutated ARSA enzyme or the nucleic acid encoding the mutated ARSA enzyme, wherein the mutated ARSA enzyme is at least 90% identical to SEQ ID NO: 1 (human ARSA enzyme), and wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1, together with a pharmaceutically acceptable carrier, stabilizer and/or excipient.
30. A method for treating a subject suffering from a disease, the method comprising administering to the subject a mutated ARSA enzyme or a nucleic acid encoding the mutated ARSA enzyme, or administering to the subject a recombinant cell comprising the mutated ARSA enzyme or the nucleic acid encoding the mutated ARSA enzyme, wherein the mutated ARSA enzyme is at least 90% identical to SEQ ID NO: 1 (human ARSA enzyme), and wherein the mutated ARSA enzyme amino acid sequence when aligned to the sequence of SEQ ID NO: 1, comprises at least one mutation which is a mutation or modification of E424 of SEQ ID NO: 1.
31. The method of claim 30, wherein the subject suffers from a pathological insufficiency of endogenous ARSA.
32. The method of claim 31, wherein the disease is a leukodystrophy.
33. The method of claim 32, wherein the leukodystrophy is metachromatic leukodystrophy.
Description
BRIEF DESCRIPTION OF THE FIGURES AND SEQUENCES
[0085] The figures show:
[0086] FIG. 1: shows the extracellular half-life of human ASA (hARSA) with permutated glutamate-424 (E424). Wildtype hARSA was cloned into the eukaryotic expression plasmid pcDNA3 (Invitrogen, Carlsbad, CA, USA) and E424 was exchanged by all 19 alternative proteinogenic amino acids using site-directed mutagenesis. Then, chinese hamster ovary (CHO) K1 cells (300,000 cells/35 mm dish) were transfected with Turbofect (Thermo Fisher Scientific, Dreieich, Germany) and 4 ?g plasmid DNA each. Two days after transfection the conditioned media were harvested and incubated at 37?C. Immediately after harvesting (day 0) and in weekly intervals for up to day 28, hARSA levels were measured with an enzyme-linked immunosorbent assay (ELISA). The half-lives of the individual hARSA-mutants were extracted from the blotted kinetics of decline (see FIG. 2-5 as examples). Data were confirmed by a functional assay measuring the time-dependent decline of ASA activity (not shown).
[0087] FIG. 2: shows the kinetics of decline of wildtype hARSA in conditioned cell culture medium. For experimental details see FIG. 1. The calculated half-life is 4.6 days.
[0088] FIG. 3: shows the kinetics of decline of the hARSA-mutant E424Q in conditioned cell culture medium. For experimental details see FIG. 1. The calculated half-life is 7.5 days
[0089] FIG. 4: shows the kinetics of decline of the hARSA-mutant E424A in conditioned cell culture medium. For experimental details see FIG. 1. The calculated half-life is 15.7 days.
[0090] FIG. 5: shows the kinetics of decline of the hARSA-mutant E424R in conditioned cell culture medium. For experimental details see FIG. 1. The half-life is above 28 days and cannot be precisely calculated from the available data set.
[0091] FIG. 6: shows the ammonium chloride-induced hypersecretion of wildtype hARSA and three hARSA-mutants. CHO-K1 cells were transfected in triplicates as described in the legend of FIG. 1 and ammonium chloride (10 mM) was added to the culture medium 24 h after transfection. This causes a pH shift in the endosomal system of the cells so that newly synthesized hARSA can no longer be targeted to the lysosomal compartment, but is delivered from the cell. The hARSA concentrations in the media were measured by ELISA 24 h after addition of ammonium chloride and are shown as means+/?SD. Mutations that prevent hARSA to adopt its proper three-dimensional structure would be retained in the endoplasmic reticulum (ER) and result in very low extracellular concentrations. This is not the case proving that the hARSA-mutants are correctly folded and overcome the conformational proofreading system of the ER
[0092] FIG. 7: shows the targeting and specific activity of wildtype hARSA and three hARSA-mutants. CHO-K1 cells were transfected as described in the legend of FIG. 1. To determine the rate of physiological hARSA-secretion, the conditioned media and the cells were harvested 2 days after transfection and the extra- and intracellular amount of hARSA was measured by ELISA. Bars represent means+/?SDs of n=3 independent transfection experiments. Compared to wildtype hARSA the three hARSA-mutants are less efficiently targeted to the lysosome, but mainly secreted. In addition, the activities of the hARSA-variants were measured in the media by a functional assay. The resulting enzyme units were related to the amount of hARSA (measured by ELISA) yielding the specific activity which is expressed as units hARSA per ?g hARSA (U/?g). Compared to wildtype hARSA, the specific activity of all three hARSA-mutants is substantially increased.
[0093] FIG. 8: shows the endocytosis of wildtype hARSA and the hARSA-mutant E424A by two types of target cells as indicated. CHO-K1 cells were transfected as described in the legend of FIG. 1. Conditioned medium was collected 48 h later and added to subconfluent cultures of human hepatoma cells and murine fibroblasts, respectively. To prevent uptake via the mannose 6-phosphate (M6P) receptor (MPR300), 7.5 mM soluble M6P was added to some dishes with human hepatoma cells as indicated. After 24 h the cells were harvested by trypsinization and the amount of internalized hARSA and hASA_E424A was determined by ELISA. Human hepatoma cells express MPR300. In the absence of competitive amounts of soluble M6P uptake of hASA_E424A is reduced compared to wildtype hARSA. In the presence of soluble M6P, uptake of the hARSA-mutant is, however, increased. Increased M6P-independent uptake was confirmed by using MPR300-deficient murine fibroblasts as target cells. Bars represent means+/?SDs of n=3 independent transfection experiments per condition.
[0094] FIG. 9: A: Stability of recombinant hARSA and the indicated hARSA-mutant in human blood serum. The two hARSA-variants were recombinantly expressed in CHO suspension cells and purified from the conditioned media by standard procedures. Each hARSA-variant (125 ng) was mixed with 50 ?l serum and incubated at 37?C for up to 7 days. The hARSA levels were measured at time point 0 (0 h chase) and after 24, 48, 96 and 168 h. Data are related to the initial ASA level. B: Specific activity of hARSA and the indicated hARSA-mutant. The two hARSA-variants were recombinantly expressed in CHO suspension cells, purified from the conditioned media by standard procedures and diluted in an appropriate volume of buffer. The activity and the mass of the ASA-variants was determined by a functional assay and ELISA, respectively. The resulting enzyme units were related to the masses yielding the specific activity which is expressed as units hARSA per ?g hARSA (U/?g).
[0095] FIG. 10: shows the functional parameters of recombinant hARSA and the indicated hARSA-mutant. Recombinantly expressed enzyme was added to the medium of MPR300-deficient murine fibroblasts at a concentration of 5 ?g/ml. Cells were allowed to endocytose the hARSA-variants within a 24 h feeding period. After that, cells were washed with PBS and a glycine-buffer pH 3.0 to detach surface-bound enzyme and incubated with fresh medium for up to 10 days. Within this chase period, cells were harvested at different time points and the intracellular concentrations of the hARSA-variants were determined by ELISA. Left: Decline of hARSA levels within the feeding period. Middle: Intracellular concentration of the two hARSA-variants immediately after the feeding period (day 0). Right: time-dependent decline of intracellular levels.
[0096] FIG. 11: shows an oligomeric state of different hASA-variants at pH 7.0. The indicated hASA-variants were separated by size exclusion chromatography using a Superdex 200 column (Amersham Pharmacia) linked to an ?kta FPLC system (GE Healthcare). The buffer was 150 mM NaCl, 20 mM Bis-Tris, pH 7.0. Fractions (0.5 ml) were collected and analysed on ASA activity using p-nitrocatechol sulfate as a substrate. Elution profiles of the four indicated hASA-variants are superimposed. While wildtype hASA presents as a dimer at neutral pH, E424A-mutants form octamers.
[0097] FIG. 12: shows a filter binding assay. The indicated hASA-variants (1 ?g each) were separated by SDS-PAGE and transferred to a polyvinylidene difluoride (PVDF)-membrane by Western blotting. Left: The membrane was probed with a myc-tagged fragment of the human MPR300 encompassing the M6P-binding domain (domain 9). Fragment bound to ASA was detected by sequential incubation with a murine monoclonal anti-myc antibody, a peroxidase-conjugated goat anti-mouse antibody and an ECL chemiluminescent substrate. Note that the MPR300 fragment does not bind efficiently to the two E424A-mutants. Right: Control to verify loading of similar amounts of the four ASA-variants. MPR300-fragments were washed from the membrane and ASA was detected by sequential incubation with a rabbit polyclonal anti-ASA antiserum, a peroxidase-conjugated goat anti-rabbit antiserum and ECL chemiluminescent substrate.
[0098] FIG. 13: shows left: Schematic representation of the four hASA constructs that are compared in the proof-of-concept study. The positions of the amino acid exchanges and the ApoE tag are indicated. The constructs are not drawn to scale. Right: Implementation of the proof-of-concept study in an immune tolerant mouse model of MLD.
[0099] FIG. 14: shows that no weight loss can be observed during treatment with wildtype hASA and the three ApoE-tagged hASA-constructs.
[0100] FIG. 15: shows two graphs indicating significant reduction of sulfatide storage in the central nervous system; left: in the brain and right: in the spinal cord; after administration of the respective hASA constructs listed along the abscissa (for a schematic representation of the constructs see FIG. 13). Bars represent means+SD of n=3 mice per group. Percentages of mean storage reductions are shown above the bars. Asteriscs indicate statistically significant declines of sulfatide levels compared to mock-treated controls.
[0101] FIG. 16: shows a comparison of the results of the proof-of-concept study according to the present invention with three previous studies. In the present study the construct hASA_M202V,T286L,R291N,E424A-ApoE reduced sulfatide storage in brain 7.5-fold more efficiently than wildtype hASA. For spinal cord the factor of increase was 12.2 on average. In three previous studies listed below the table, higher enzyme doses of 20 or 50 mg/kg were used for enzym replacement therapy. Still hASA_M202V,T286L,R291N,E424A-ApoE was superior to all other constructs tested before. Thus a cumulative dose of 40 mg/kg hASA_M202V,T286L,R291N,E424A-ApoE (4?10 mg/kg) almost doubled the efficacy of enzyme replacement with a 65-fold higher cumulative dose of wildtype hASA (52?50 mg/kg). Indicated references: Matzner et al. (2009) Mol Ther. 17: 600-6; Matthes et al. (2012) Hum Mol Genet. 21: 2599-609; Simonis et al. (2019) Hum Mol Genet. 28: 1810-1821.
[0102] The sequences show:
TABLE-US-00001 SEQIDNOs.1showstheaminoacidsequenceofwildtypehumanARSA protein(isoform1)includingthesignalpeptide(underlined)andmostpreferred positionsformutation(boldandunderlined): 1020304050 MGAPRSLLLALAAGLAVARPPNIVLIFADDLGYGDLGCYGHPSSTTPNLD 60708090100 QLAAGGLRFTDFYVPVSLCTPSRAALLTGRLPVRMGMYPGVLVPSSRGGL 110120130140150 PLEEVTVAEVLAARGYLTGMAGKWHLGVGPEGAFLPPHQGFHRFLGIPYS 160170180190200 HDQGPCQNLTCFPPATPCDGGCDQGLVPIPLLANLSVEAQPPWLPGLEAR 210220230240250 YMAFAHDLMADAQRQDRPFFLYYASHHTHYPQFSGQSFAERSGRGPFGDS 260270280290300 LMELDAAVGTLMTAIGDLGLLEETLVIFTADNGPETMRMSRGGCSGLLRC 310320330340350 GKGTTYEGGVREPALAFWPGHIAPGVTHELASSLDLLPTLAALAGAPLPN 360370380390400 VTLDGFDLSPLLLGTGKSPRQSLFFYPSYPDEVRGVFAVRTGKYKAHFFT 410420430440450 QGSAHSDTTADPACHASSSLTAHEPPLLYDLSKDPGENYNLLGGVAGATP 460470480490500 EVLQALKQLQLLKAQLDAAVTFGPSQVARGEDPALQICCHPGCTPRPACC HCPDPHA SEQIDNO:2showsthenucleicacidsequenceencodingwildtypehumanARSA(cDNA): ATGGGGGCACCGCGGTCCCTCCTCCTGGCCCTGGCTGCTGGCCTGGCCGTTGCCCGTCCGCCC AACATCGTGCTGATCTTTGCCGACGACCTCGGCTATGGGGACCTGGGCTGCTATGGGCACCCC AGCTCTACCACTCCCAACCTGGACCAGCTGGCGGCGGGAGGGCTGCGGTTCACAGACTTCTAC GTGCCTGTGTCTCTGTGCACACCCTCTAGGGCCGCCCTCCTGACCGGCCGGCTCCCGGTTCGG ATGGGCATGTACCCTGGCGTCCTGGTGCCCAGCTCCCGGGGGGGCCTGCCCCTGGAGGAGGT GACCGTGGCCGAAGTCCTGGCTGCCCGAGGCTACCTCACAGGAATGGCCGGCAAGTGGCACCT TGGGGTGGGGCCTGAGGGGGCCTTCCTGCCCCCCCATCAGGGCTTCCATCGATTTCTAGGCAT CCCGTACTCCCACGACCAGGGCCCCTGCCAGAACCTGACCTGCTTCCCGCCGGCCACTCCTTGC GACGGTGGCTGTGACCAGGGCCTGGTCCCCATCCCACTGTTGGCCAACCTGTCCGTGGAGGCG CAGCCCCCCTGGCTGCCCGGACTAGAGGCCCGCTACATGGCTTTCGCCCATGACCTCATGGCC GACGCCCAGCGCCAGGATCGCCCCTTCTTCCTGTACTATGCCTCTCACCACACCCACTACCCTC AGTTCAGTGGGCAGAGCTTTGCAGAGCGTTCAGGCCGCGGGCCATTTGGGGACTCCCTGATGG AGCTGGATGCAGCTGTGGGGACCCTGATGACAGCCATAGGGGACCTGGGGCTGCTTGAAGAG ACGCTGGTCATCTTCACTGCAGACAATGGACCTGAGACCATGCGTATGTCCCGAGGCGGCTGC TCCGGTCTCTTGCGGTGTGGAAAGGGAACGACCTACGAGGGCGGTGTCCGAGAGCCTGCCTTG GCCTTCTGGCCAGGTCATATCGCTCCCGGCGTGACCCACGAGCTGGCCAGCTCCCTGGACCTG CTGCCTACCCTGGCAGCCCTGGCTGGGGCCCCACTGCCCAATGTCACCTTGGATGGCTTTGACC TCAGCCCCCTGCTGCTGGGCACAGGCAAGAGCCCTCGGCAGTCTCTCTTCTTCTACCCGTCCTA CCCAGACGAGGTCCGTGGGGTTTTTGCTGTGCGGACTGGAAAGTACAAGGCTCACTTCTTCAC CCAGGGCTCTGCCCACAGTGATACCACTGCAGACCCTGCCTGCCACGCCTCCAGCTCTCTGACT GCTCATGAGCCCCCGCTGCTCTATGACCTGTCCAAGGACCCTGGTGAGAACTACAACCTGCTG GGGGGTGTGGCCGGGGCCACCCCAGAGGTGCTGCAAGCCCTGAAACAGCTTCAGCTGCTCAA GGCCCAGTTAGACGCAGCTGTGACCTTCGGCCCCAGCCAGGTGGCCCGGGGCGAGGACCCCGC CCTGCAGATCTGCTGTCATCCTGGCTGCACCCCCCGCCCAGCTTGCTGCCATTGCCCAGATCCC CATGCCTGA SEQIDNO:3showstheaminoacidsequenceofamutatedARSAwithincreased enzymaticactivity(boldandunderlinedaremutatedsequencescomparedtowild- typeARSA): MGAPRSLLLALAAGLAVARPPNIVLIFADDLGYGDLGCYGHPSSTTPNLDQLAAGGLRF TDFYVPVSLCTPSRAALLTGRLPVRMGMYPGVLVPSSRGGLPLEEVTVAEVLAARGYLTGMAGK WHLGVGPEGAFLPPHQGFHRFLGIPYSHDQGPCQNLTCFPPATPCDGGCDQGLVPIPLLANLS VEAQPPWLPGLEARYVAFAHDLMADAQRQDRPFFLYYASHHTHYPQFSGQSFAERSGRGPFG DSLMELDAAVGTLMTAIGDLGLLEETLVIFTADNGPELMRMSNGGCSGLLRCGKGTTYEGGVR EPALAFWPGHIAPGVTHELASSLDLLPTLAALAGAPLPNVTLDGFDLSPLLLGTGKSPRQSLFFY PSYPDEVRGVFAVRTGKYKAHFFTQGSAHSDTTADPACHASSSLTAHEPPLLYDLSKDPGENYN LLGGVAGATPEVLQALKQLQLLKAQLDAAVTFGPSQVARGEDPALQICCHPGCTPRPACCHCPD PHA SEQIDNO:4showstheaminoacidsequenceofamutatedARSAwithincreased enzymaticactivityandstabilityaccordingtotheinvention(boldandunderlined aremutatedsequencescomparedtowild-typeARSA;Xisanyaminoacidexcept glutamate-E): MGAPRSLLLALAAGLAVARPPNIVLIFADDLGYGDLGCYGHPSSTTPNLDQLAAGGLRF TDFYVPVSLCTPSRAALLTGRLPVRMGMYPGVLVPSSRGGLPLEEVTVAEVLAARGYLTGMAGK WHLGVGPEGAFLPPHQGFHRFLGIPYSHDQGPCQNLTCFPPATPCDGGCDQGLVPIPLLANLS VEAQPPWLPGLEARYVAFAHDLMADAQRQDRPFFLYYASHHTHYPQFSGQSFAERSGRGPFG DSLMELDAAVGTLMTAIGDLGLLEETLVIFTADNGPELMRMSNGGCSGLLRCGKGTTYEGGVR EPALAFWPGHIAPGVTHELASSLDLLPTLAALAGAPLPNVTLDGFDLSPLLLGTGKSPRQSLFFY PSYPDEVRGVFAVRTGKYKAHFFTQGSAHSDTTADPACHASSSLTAHXPPLLYDLSKDPGENYN LLGGVAGATPEVLQALKQLQLLKAQLDAAVTFGPSQVARGEDPALQICCHPGCTPRPACCHCPD PHA SEQIDNO:5showstheaminoacidsequenceofamutatedARSAwithincreased enzymaticactivityandstabilityaccordingtotheinvention(boldandunderlinedare mutatedsequencescomparedtowild-typeARSA;Xisanyaminoacidexceptglutamate-E), furtherincludingtheApoEIItag(underlinedanditalic-SEQIDNO:6).TheC-terminal apoE-IIsequencecomprisestwocopiesofthesequenceSAWSHPQFEK(SEQIDNO:7)as directtandemrepeatsseparatedbytheglycine-andserine-richlinkersequence GGGSGGGSGG(SEQIDNO:8)(otherlinkersequenceswillprobablyworkaswell).Dueto technicalreasonsanirrelevantserine(S)-residuewasincorporatedbetweentheC- terminalalanine(A)ofthehumanASAsequenceandtheapoE-IItag(theconstructwill mostlikelyworkalsowithoutthisserineorwithotheraminoacidsoraminoacid sequencesatthisposition). MGAPRSLLLALAAGLAVARPPNIVLIFADDLGYGDLGCYGHPSSTTPNLDQLAAGGLRF TDFYVPVSLCTPSRAALLTGRLPVRMGMYPGVLVPSSRGGLPLEEVTVAEVLAARGYLTGMAGK WHLGVGPEGAFLPPHQGFHRFLGIPYSHDQGPCQNLTCFPPATPCDGGCDQGLVPIPLLANLS VEAQPPWLPGLEARYVAFAHDLMADAQRQDRPFFLYYASHHTHYPQFSGQSFAERSGRGPFG DSLMELDAAVGTLMTAIGDLGLLEETLVIFTADNGPELMRMSNGGCSGLLRCGKGTTYEGGVR EPALAFWPGHIAPGVTHELASSLDLLPTLAALAGAPLPNVTLDGFDLSPLLLGTGKSPRQSLFFY PSYPDEVRGVFAVRTGKYKAHFFTQGSAHSDTTADPACHASSSLTAHXPPLLYDLSKDPGENYN LLGGVAGATPEVLQALKQLQLLKAQLDAAVTFGPSQVARGEDPALQICCHPGCTPRPACCHCPD PHASSAWSHPQFEKGGGSGGGSGGSAWSHPQFEK
EXAMPLES
[0103] Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the description, figures and tables set out herein. Such examples of the methods, uses and other aspects of the present invention are representative only, and should not be taken to limit the scope of the present invention to only such representative examples.
[0104] The examples show:
Example 1: Any Mutation of Human ARSA at Position E424 Increases Enzyme Half-Life
[0105] Substitution of E424 by any other proteinogenic amino acid consistently increases the half-life hARSA in cell culture medium (FIG. 1). It can, therefore, be concluded that the presence of glutamate at this specific position has a destabilizing effect that is abrogated if it is exchanged. Interestingly, the destabilizing effect of E424 is also diminished if aspartate, the second proteinogenic amino acid with negatively charged side chain, is inserted (see E424D). Substitution of E424 by proline reduces the amount of hARSA being delivered to the medium possibly due to misfolding of the enzyme and retention in the endoplasmic reticulum (not shown). Still, the half-life of the small amount of secreted mutant is increased (see E424P). According on the extent of stabilization the amino acid substitutions can be divided into three groups. Group-1 comprises substitutions that increase the half-life 1.3 to 2.6-fold, group-2 around 3.0 to 4.5-fold and group-3 more than 5-fold. The amino acids of a group do not share common biochemical properties such as size, lipophilicity or charge. Within group-3, for example, arginine (R) is a positively charged amino acid with large side chain, whereas glycine (G) is uncharged and has a minimal side chain of only one hydrogen atom. Leucine (L) on the contrary, is a non-polar aliphatic amino acid.
[0106] The half life of wild-type hARSA in cell culture medium is approximately 4.6 days (see FIG. 2). Substitution of E424 by glutamine (Q) increases the half-life to more than 7 days, whereas alanine (A) and arginine (R) increase it to more than 15 and 28 days (FIGS. 3 to 5), respectively.
Example 2: Mutated hARSAs are not Retained in the Endoplasmic Reticulum (ER)
[0107] Many missense mutations of hASA result in a misfolded enzyme that is retained by the conformational proofreading machinery of the ER. It is, therefore, important to mention that amino acid exchanges at position 424 are well tolerated and do not cause alterations of the enzyme's three-dimensional structure interfering with activity and passage of the ER. This has been investigated in more detail for the three substitutions E424A, E424Q and E424R by blocking the M6P-receptors in the trans Golgi network with ammonium chloride. As a consequence of the receptor failure newly synthesized wildtype hASA passes the ER and Golgi apparatus, but then follows the default pathway of secretory proteins and is released into the medium instead of being delivered to lysosomes. Under the same conditions, a misfolded hASA-mutant that is retained in the ER cannot does not reach the trans Golgi network and can therefore not be secreted resulting in undetectable to very low extracellular enzyme levels. This does, however, not apply to hASA-E424A, -E424Q and -E424R (FIG. 6). It can be concluded that the exchanges at position E424 do not cause retention in the ER. The reduced extracellular concentration of hASA-E424R might be due to a lower transfection efficacy rather than to partial ER-retention. Correct folding of the mutants is also indicated by a functional assay (not shown) showing considerable activities of all mutants in the medium of transfected cells (not shown).
Example 3: Mutated hARSA has Increased Extracellular Stability
[0108] Measurements of extra- and intracellular hARSA-levels suggest that substitution of E424 alters the intracellular targeting of newly synthesized hARSA. Around two third of the wildtype hARSA is intracellularly retained under conditions of overexpression (FIG. 7). On contrast, only around 20% of the hARSA-mutants hARSA-E424A, -E424Q and -E424R is intracellularly retained. Thus, a higher percentage of mutant enzyme is delivered from the cell. Interestingly, the specific activity (milliunit enzyme activity per ?g enzyme mass) of the hypersecreted mutants is substantially increased compared to wildtype hARSA. This is probably due to the stabilizing effect of the substitutions reducing the rate of enzyme degradation in the medium.
Example 4: Mutated hARSA Shows Decreased Liver Uptake and Increased M6P-Independent BBB Transcytosis
[0109] Enzyme replacement therapy using intravenous injection of hARSA has the potential to mitigate the MLD-like disease of ASA knockout mice. High enzyme doses are, however, needed. This is due to a preferential uptake of hARSA by hepatocytes. To analyse uptake of hARSA-mutants by liver cells, we incubated the human hepatoma cell line HuH7 with conditioned medium containing hARSA-E424A. Compared to wildtype hARSA uptake was reduced to approximately 20% (FIG. 8). This result suggests, that less of the hARSA-mutants might get lost by liver uptake during enzyme replacement therapy. As a consequence, higher enzyme concentrations would persist in the circulation promoting transfer across the blood-brain barrier into the brain parenchyma. A redistribution of enzyme in favour of the brain is expected to result in higher therapeutic efficacy. In the presence of competitive amounts of M6P uptake of wildtype hARSA by HuH7 cells is substantially reduced indicating that cellular uptake is predominantly mediated by the MPR300. Around 5-fold more hARSA-E424A is endocytosed under the same conditions indicating that the mutant gets access to the cell via M6P-independent pathways. This result was confirmed by feeding MPR300-deficient murine fibroblasts. Also, in this case cellular uptake of hARSA-E424A is substantially higher than uptake of wildtype hARSA.
Example 5: Hyperactive and Hyperstable hARSA Mutations and APO-EII Functionalization can be Combined with Each Other
[0110] As shown in FIG. 1, substitution of E424 by any other proteinogenic amino acid increases the stability of hARSA in cell culture medium. Following intravenous enzyme replacement therapy hARSA is present in blood serum. To investigate the stabilizing effect in the presence of serum, a hyperactive hARSA-variant (hARSA-M2020V,T286L,R291N) with additional E424A substitution was incubated for 7 days in human serum at 37? C. The half-life of the stabilized mutant was 108 h compared to only 15 h of wildtype hARSA used as a control (FIG. 9A). Thus, the half-life in serum was increased by a factor of 7.2. In murine serum the factor of increase was greater than 7.6 (not shown). The data also show that the three amino acid exchanges M2020V, T286L, R291N, mutations previously shown in hyperactivated hARSA-variant, do not neutralize the stabilizing effect of E424A.
[0111] To determine if the stabilizing mutation E424A interferes with the hyperactivity of the previously patented hARSA triple mutant hARSA-M2020V,T286L,R291N a hARSA-variant combining all four amino acid substitutions was recombinantly expressed and analysed. As shown previously the specific activity of hARSA-M2020V,T286L, R291N is around 4.7-fold normal. The quadruple mutant with additional E424-exchange displays a specific activity of 3.3-fold normal (FIG. 9B). Consequently, introduction of E424A does not abrogate hyperactivity. Taken together, the three mutations increasing the catalytic rate constant and the single mutation increasing the half-life do not neutralize each other and can be combined in one hyperactive and superstable hARSA-variant.
[0112] To investigate the intracellular half-life of the hyperactive and superstable hARSA-variant MPR300-deficient murine fibroblasts were fed with hASA-M2020V,T286L,R291N,E424A and the degradation of the internalized enzyme was measured over 10 days. Whereas wildtype hARSA used as a control has an intracellular half-life of 15 h, it was 55 h for the hyperactive and superstable hARSA-variant (FIG. 10). Thus, the factor of increase is around 3.7-fold. A very similar factor of 3.6 was obtained for CHO-K1 cells (not shown).
[0113] In addition, in FIG. 11, the mutated ARSA of the invention assembles as a protein octamer having therefore 8 identical mutated ARSA monomers. Further, as shown by filter binding assays, the MPR300 cannot bind to ASA-mutants harboring E424A (FIG. 12). This can be ascribed to the lack of M6P-residues. M6P groups are added to the N-glycans of ASA as it passes through the cis-Golgi network by a reaction involving two different enzymes: UDP-N-acetylglucosamine 1-phosphotransferase and a-N-acetylglucosamine-1-phosphodiester a-N-acetylglucosaminidase. The phosphotransferase recognizes ASA and other soluble lysosomal enzymes via a patch comprising several lysine residues that are correctly spaced relative to each other. While the negative charge of E424 prevents octamerization of ASA at the near neutral pH of the endoplasmic reticulum and Golgi apparatus, E424-mutants are likely to arrive in the Golgi apparatus as octamers. It can be concluded from the known three-dimensional structure of ASA hat the lysine patch is then no longer surface exposed but buried in the octamer. As a consequence, E424-mutants are not recognized by the phosphotransferase and M6P-residues cannot be formed. This hypothesis, however which shall not be understood to restrict the invention in all embodiments, explains not only the lack of MPR300-binding to E424A-mutants in filter binding assays (FIG. 12), but also the hypersecretion of E424-mutants from cells (FIG. 7). The lack of M6P-residues allows retargeting of recombinant enzyme from the M6P-receptors mainly of liver and spleen to the blood-brain barrier and brain cells. To enhance such targeting, the apoEII tag binding to receptors of the low-density lipoprotein receptor family was linked to the E424-mutants. It was shown previously by us that the apoEII-tag further promotes transcytosis across the blood-brain barrier and increases therapeutic efficacy of enzyme replacement therapy in a mouse model of MLD (Bockenhoff et al., 2014, J Neurosci. 2014 Feb. 26; 34(9):3122-9).
[0114] The amino acid sequence of an apoEII tagged superstable and superactive ARSA is shown in SEQ ID NO: 5.
Example 6: Proof-of-Concept of Enzyme Replacement Therapy with Mutated Enzyme in a MLD Mouse Model
[0115] The aim of the proof-of-concept was to evaluate side effects and to analyse and determine the reduction of storage in the CNS. Therefore, an amount of 4?10 mg/kg was intravenously injected into MLD mice with immune tolerance to wildtype hASA. For the proof-of-concept study the wild-type hASA, ApoE-tagged hASA and two stable mutants E424A (both ApoE-tagged) were investigated. Thus in total, four constructs were compared, which are shown in FIG. 13.
[0116] In a first step, no side effects were observed over the four subsequent administrations of the enzymes. As FIG. 14 shows, no weight loss was observed in any mouse. Furthermore, none of the mice died. In addition, no signs of discomfort being indicative of anaphylactic reactions were observed such as tachypnea, fatigue, bristled fur, fur licking, or reduced cage activity. This indicates that E424A does not generate new epitopes that trigger anaphylactic reactions.
[0117] The exchange on position E424A has only a medium stabilization effect and far more efficient exchanges exist (FIG. 1). However, the exchange on position E424A increased brain sulfatide depletion by a factor of 1.5. In combination with the hyperactivating triple mutation M202V, T286L,R291N, a 60% mean depletion was achieved. Wildtype hASA only led to a mean decrease of 8% in the same experiment. (FIG. 15).
[0118] This exceeded the efficiency of ASA tested in current clinical trials by a factor of 7.5 in brain and 12.2 in spinal cord, respectively. (FIG. 16) The results of the present proof-of-concept study were compared to the results of three previous studies. A significant reduction in the brain is shown in comparison with the previous studies, although the MLD mice were treated with a significantly lower dose of 10 mg/kg of the respective mutant enzyme.
