PHENYLALANINE-FREE PROTEIN AND DNA CODING THEREFOR

Abstract

A DNA molecule coding for a food protein, such as ovalbumin or casein, modified so that the codons for phenylalanine have been omitted or replaced by codons for one or more other metabolisable amino acids. Also a modified edible protein coded for by such a DNA molecule. Such modified proteins are useful in the nutrition of patients suffering from phenylketonuria.

Description

[0001] This invention relates to an edible protein which has been modified so that it is phenylalanine free, to DNA coding for it, and to a method of producing it. Such a protein is a useful nutrient in the treatment of diseases which are associated with difficulty in metabolising phenylalanine. A particular example of such a disease is phenylketonuria (PKU).

[0002] PKU is a genetically acquired disease that occurs in a relatively fixed proportion of new births in a human population. A defect in the enzyme carrying out the pterin-dependent hydroxylation of phenylalanine to tyrosine prevents the body from metabolizing the amino acid phenylalanine. This amino acid occurs in varying proportions in all proteins in foodstuffs and is, in the correct amount, essential for human protein synthesis, and therefore for the growth and maintenance of the body. Patients with PKU cannot remove excess phenylalanine from the blood and tissues and the failure to achieve this control over phenylalanine levels leads to grave neurological damage, especially in the growing child.

[0003] PKU patients are at present fed with a synthetic diet which contains a metabolically-correct amount of phenylalanine along with a mixture of the other amino acids needed for growth. Such a diet is unpalatable and is presented in liquid form only and therefore has difficulty in achieving patient compliance.

[0004] An object of this invention is to provide an edible protein which when pure contains no phenylalanine and which can form the basis for a diet containing the optimal nutritional phenylalanine content for PKU patients. This object may be achieved by taking the gene from a known nutritional protein and modifying it so that the codons coding for phenylalanine are deleted or are replaced by codons coding for another amino acid.

[0005] An alternative approach is to synthesise by chemical means DNA coding for a phenylalanine-free polypeptide, starting either from fragments of genes coding for existing proteins, or from the nucleotides themselves.

[0006] According to one aspect of the invention we provide a DNA molecule coding for a food protein, modified in that the codons coding for phenylalanine have been deleted or replaced by codons coding for one or more other amino acids.

[0007] According to another aspect of the invention we provide an edible polypeptide which comprises a food protein modified in that the phenylalanine residues have been omitted or have been replaced by one or more other amino acids also occurring in protein.

[0008] We further provide a nutrient material comprising an edible polypeptide as defined above and other edible substances.

[0009] The food protein is preferably a common food protein such as ovalbumin or caesin.

[0010] We also provide a nutrient material comprising an edible protein or modified food protein as hereinbefore defined, and other edible substances.

[0011] The protein according to the invention is phenylalanine free when pure, but the diet of the patient must contain some phenylalanine, i.e. the amount required for metabolism, but with substantially no excess.

[0012] An obvious approach would be to add an appropriate proportion of normal food proteins, which contain phenylalanine, to a pure phenylalanine-free protein according to the invention.

[0013] On the other hand, proteins are notoriously difficult to purify to a high level. If only partially purified, the phenylalanine-free protein will be accompanied by other protein products of the host organism containing their normal amounts of phenylalanine. Thus, if the modified protein is only partly purified (which is much easier than complete purification), a protein mixture containing overall a reduced proportion of phenylalanine will be obtained. By controlling the degree of purification, a protein mixture containing a metabolically-appropriate proportion of phenylalanine can be produced. This invention also provides such a mixture.

[0014] Although codons for phenylalanine may simply be deleted from the gene for a food protein, in order to preserve as far as possible the tertiary structure of the protein the codons coding for phenylalanine are preferably replaced by codons coding for another amino acid, preferably those having the most similar properties, e.g. tyrosine.

[0015] We also provide an expression vector into which has been incorporated DNA for an edible protein or modified food protein as described herein. The expression vector is preferably a Saccharomyces cerevisiae expression vector because this yeast has a long history as a human foodstuff and is amenable to genetic manipulation. Other yeasts, e.g. Pischia pastoris, may also be used.

[0016] We further provide a host, for example a yeast such as S. cerevisiae or Pichia pastoris, transformed by such an expression vector.

[0017] Ovalbumin and caesin have been selected as preferred food proteins to be modified in accordance with this invention because they are naturally-occurring proteins which are commonly used as human foodstuffs, are widely acceptable, and also because the modified proteins are likely to behave in a similar manner to the native proteins when cooked or subjected to other food processing steps. A wide variety of other food proteins may, however, also be chosen.

[0018] Preferably, apart from omitting or substituting codons coding for phenylalanine, the DNA molecule coding for the edible protein is modified as necessary to ensure that the codon for each amino acid is the codon of preference for the selected host, e.g. S. cerevisiae.

[0019] DNA sequences and polypeptides embodying the invention will now be described in more detail in non-limiting manner, with reference to the Figures and Examples. FIGS. 1 to 11 relate to modified chick ovalbumin and FIGS. 12 to 24 relate to modified bovine casein.

[0020] FIG. 1 shows the sequence of cDNA for unmodified chick ovalbumin (a copy of the Genbank entry).

[0021] FIG. 2 shows the amino acid sequence (in single letter code) of the polypeptide coded for by the coding region of FIG. 1.

[0022] FIG. 3 shows a polypeptide corresponding to that of FIG. 2, but from which all phenylalanine residues have been deleted.

[0023] FIG. 4 shows the amino acid sequence of a polypeptide corresponding to FIG. 2, in which the phenylalanine residues have been replaced by tyrosine residues.

[0024] FIG. 5 shows DNA sequence coding for the polypeptide shown in FIG. 3, where the codons have been selected using a preferred pattern of codon usage for S. cerevisiae.

[0025] FIG. 6 shows a DNA sequence coding for the polypeptide shown in FIG. 4.

[0026] FIGS. 7 to 10 show, respectively, the nucleotide sequence for the constructs pl+oval−f+3 end, pl+h6oval−f+end, pl+Yoval−f+3nd, and pl+h6Yoval−f+3end.

[0027] FIG. 11 shows the nucleotide sequence of the synthetic pl+oval−f+3end gene constructed from overlapping oligonucleotides.

[0028] FIG. 12 shows the nucleotide sequence of bovine alpha-S1-casein mRNA.

[0029] FIG. 13 shows the amino acid sequence of mature alpha-S1-casein.

[0030] FIG. 14 shows a modified protein corresponding to that of FIG. 13, but from which all phenylalanine residues have been deleted.

[0031] FIG. 15. shows a DNA sequence coding for the modified protein of FIG. 14.

[0032] FIG. 16 shows a DNA sequence corresponding to that of FIG. 15 but including the non-translated regions of alpha-S1-casein.

[0033] FIG. 17 shows the nucleotide sequences of bovine casein gene blocks A and B, form which the whole gene was subsequentially assembled.

[0034] FIGS. 18a and 18b show respectively the predicted and actual DNA and protein sequences of block A.

[0035] FIGS. 19a and 19b show respectively the predicted and actual DNA and protein sequences of block B.

[0036] FIG. 20 shows the combines DNA and protein sequences of blocks A and B.

[0037] FIG. 21 shows the complete DNA and protein sequences of the synthetic casein.

[0038] FIG. 22 shows the assembly of the casein gene sequences in plasmid pMTL22.

[0039] FIG. 23 shows the construction of the E. coli yeast expression vector pMTL8133.

[0040] FIG. 24 shows the casein gene sequence cloned into pMTL8133.

EXAMPLE 1

[0041] The gene and downstream non-translated DNA sequence for chick ovalbumin were based on the nucleotide sequence of the complementary cDNA for chick ovalbumin deposited by O'Hare et al in the GenBank database with the accession number V00383. The vector pEMBLyex4 (see Cesareni, G and Murray, J. A. H. (1987) In ‘Genetic Engineering’ (Ed. Setlow, J. K.) Volume 9 Plenum Publishing Corporation, New York, pp 135-154) was chosen for expression, as it can be used to direct the expression of genes which lack their own promoter. The vector harbours a hybrid promoter consisting of the upstream activator sequence of the GAL1 promoter and the 5′ non-translated leader of the CYC1 gene, up to position −4. The plasmid contains a translation initiation codon ATG downstream from the GAL1-CYC1 promoter. The codon ATG is followed by a unique HindIII site and is preceded by unique cloning sites for BamHI, PstI, Smal and XbaI. In addition to yeast selectable markers and origin of replication it carries ampicillin resistance and a functional E. coli origin. The complete nucleotide sequence of the vector is known.

[0042] The sequence of the cDNA for chick ovalbumin is shown in FIG. 1 and a translation of the ovalbumin coding region is shown in FIG. 2. The amino acid sequence of a polypeptide (oval−f) derived from chick ovalbumin, but lacking any phenylalanine residues, is shown in FIG. 3. To optimize expression of this gene when expressed in S. cerevisiae the polypeptide sequence was ‘backtranslated’ using the most preferred pattern of codon usage for S. cerevisiae (FIG. 5). A derivative of chick ovalbumin was also designed in which the phenylalanine residues are replaced by tyrosine residues in order to attempt to produce a protein which has as near as possible the tertiary structure of chick ovalbumin. The amino acid sequence of the polypeptide (Yoval−f) and its corresponding gene produced as for the oval−f gene are shown in FIGS. 4 and 6 respectively. To further facilitate expression, cloning procedures and protein purification the following modifications were made to the basic gene.

[0043] 1. Addition of a sequence corresponding to the 3′ end of the mRNA from the end of the coding region to the poly A site, in order to enhance expression.

[0044] 2. Addition of an extra TAA stop codon at the end of the gene, in order to ensure that no translation would take place beyond the normal coding region.

[0045] 3. In order to assist in vitro manipulation, addition at either end of the synthetic gene of polylinkers which contained restriction sites for PstI, BamHI, SmaI, EcoRI and HindIII. The synthetic genes do not contain sites for these restriction enzymes. The polylinkers have the following sequence: 1 5′CTGCAGGATCCCGGGAATTCAAGCTT 3′   [ PstI ]    [ SmaI ]   [HindIII]          [BamHI]      [EcoRI]

[0046] 4. In some versions of the synthetic gene a sequence corresponding to 6 histidine residues was added immediately downstream of the initiating methionine, in order to facilitate purification of the protein by a form of affinity chromatography.

[0047] Thus 4 basic variations on the original synthetic gene were obtained, with the following structures:

[0048] The synthetic gene is constructed via the synthesis of oligonucleotides each approximately 100 nucleotides long and designed in such a way that they overlap each other and will self-assemble by complementary base pairing into a contiguous structure which can be ligated via the appropriate sticky ends, generated by restriction endonuclease digestion into pEMBLyex4 or an appropriate E. coli vector such as pBR322 or pUC19. The sequences of the oligonucleotides and their arrangement is shown in FIG. 11. The end points of the individual oligonucleotides are marked by the character.

EXAMPLE 2

[0049] This example utilises bovine alpha-s1-casein. In this illustration only one synthetic gene was designed, but the general approach used in Example 1 can be applied to produce the other three genes analogous to those of Example 1, (i.e. those genes containing tyrosine replacements for phenylalanine and/or a run of six histidine residues immediately downstream of the N-terminal methionine).

[0050] The sequence of the mRNA for bovine alpha-s1-casein is shown in FIG. 12 and a translation of the region coding for the mature polypeptide is shown in FIG. 13. The modified form of the protein lacking phenylalanine residues and with an added N-terminal methionine (to permit translation) is shown in FIG. 14. A DNA sequence corresponding to this modified polypeptide produced using the most preferred pattern of codon usage for S. cerevisiae is shown in FIG. 15. Finally, the nucleotide sequence of the complete synthetic gene with the 3′ untranslated region from the bovine alpha-s1-casein mRNA added on as well as the polylinkers (described in section A) is shown in FIG. 16. It should be noted that this particular synthetic gene has an internal EcoR1 site, as well as those present in the polylinkers and therefore EcoR1 should not be used in any in vitro manipulations of this gene during insertion into a vector.

EXAMPLE 3

[0051] This example concerns the construction of a bovine casein gene modified in that the codons for phenylalamine are replaced by codons for tyrosine.

[0052] Synthetic Gene Design

[0053] Eight C.100'mer oligonucleotides were designed, synthesised, and purified. These oligonucleotides (casein 1-8. see FIG. 16) formed the basis of two self-priming block assemblies in which the two blocks (designated A and B) overlapped by about 100 bp.

[0054] Following an initial round of PCR-mediated extension of the self-primed oligonucleotides as separate Blocks (A & B), a second round of PCR amplification using terminal flanking c. 20'mer primers (AL1 & AR1, BL1 & BR2; see FIG. 17 generated the two independent c.380 bp gene blocks A and B.

[0055] As mentioned above the design of the casein 1-8 c. 100'mer oligonucleotides was such that the encoded gene contained no phenylalanine codons, all these being substituted with tyrosine codons. A further feature was the incorporation of a number of unique restriction sites to facilitate in the final assembly of the whole gene from components of the two overlapping gene blocks. This duplication facilitates correction of erroneous PCR-mediated DNA synthesis.

[0056] Gene Block Synthesis

[0057] Using the 2-step PCR stategy described above both casein gene blocks A and B were amplified as discrete c. 380 bp products using Stratagene's native pfu DNA polymerase. Little success was achieved with the cloned enzyme. This particular enzyme was used because of its apparently superior fidelity properties.

[0058] Cloning and Sequencing of the Gene Blocks

[0059] Both blocks A & B were successfully cloned into Invitrogen's PCRII-TA cloning vector. Plasmid DNA was prepared from numerous isolates and these subjected to DNA sequence analysis using both universal and reverse sequencing primers. For the majority of clones full c. 380 bp reads were obtained. All these sequences were computer aligned against the “desired” sequence and against each other. Representative sample alignments for block A and B are shown in FIGS. 18a and b, and 19a and b, respectively.

[0060] PCRII-TA clones A100 and B69 were chosen as primary DNA sources. Two mutagenic c. 60'mer mutagenic oligo nucleotides, casein 9 and casein 10, were synthesised, purified, and used to amplify a “corrected” c.200 bp HpaI/HindIII C-terminus. This product was cloned into PCRII-TA vector and the sequence of several clones analysed using universal and reverse sequence primers. No perfect sequences were obtained but one clone (C20) which had only one base change, a G to T conversion resulting in a single amino acid change of trp to leu, was chosen.

[0061] The strategy taken was to assemble the gene sequence in pMTL22 by cloning the c. 200 bp hpa/HindIII C-terminal fragment of clone C20 next to the remainder of the gene derived from cloning of the c.265 bp BamHI/AatII of clone A100 and the c. 270 bp AatII/KpnI of clone B69 (See FIG. 22). This has been achieved and the final nucleotide sequence verified yielding the “casein” gene sequence with a TTG triplet deletion at nt. pos. 258 and a G to T base change at nt. pos.

[0062] Cloning of the casein sequence into pMTL8133

[0063] The casein gene sequence was sub-cloned from the pMTL22 construct above into the “in house” E. coli/yeast expression vector pMTL8133 (see FIG. 23). This vector is based on chloramphenicol resistance and has a hybrid PGK::REP 2 promoter element which has been shown to elicit high expression levels of other heterologous genes in both E. coli and Saccharomyces cerevisiae. The casein sequence was cloned as a PstI(flush-ended)/HindIII fragment into SspI/HindIII cleaved pMTL8133 as outlined in FIG. 24, such that it is correctly juxtaposed to the 5′-UTR sequence for elevated expression in yeast. The correct sequence at the cloning junction was verified by sequence analysis.

[0064] The modified gene has been clone into the E. coli/yeast expression vector pMTL8133 which has previously been shown to elicit expression of heterologous genes in both Escherichia coli and Sacharomyces cerevisiae.

[0065] Casein expression studies

[0066] E. coli strain INV alpha F′ (endA1, recA1, hsdR17(r−k, m+k), suPE44, -, thi-1, gyrA, relA1, &phgr;80 lacZ&agr;&Dgr;M15&Dgr; (lacZYA-argF), deoR+, F genotype) has been transformed with the pMTL8133-casein recombinant plasmid and cultured in the presence of chloramphenicol (30 &mgr;g ml−1) to maintain selection for the plasmid. Sonic extracts have been prepared from this culture and subjected to polyacrylamide gel electrophoresis alongside native bovine alpha casein (purchased from Sigma). Blotting of this gel onto nitrocellulose membrane followed by probing of the membrane sequentially with rabbit anti-casein and peroxidase-conjugated goat anti-rabbit antibody has revealed the presence of a polypeptide equal in size to the bovine alpha casein control. This polypeptide has a predicted molecular weight of 22 kDa. This protein product is detectable by means of antibody probing. No product is visible in coomassie blue stained polyacrylamide gels.

[0067] The pMTL8133-casein recombinant plasmid is used to transform a yeast (e.g. S. cerevisiae) in order to obtain expression of the modified casein encoded thereby.

[0068] If necessary or desirable, the base change at nt. pos. 512 can be corrected using a two-step strategy as follows. Firstly, the major part of the casein gene, 510 bp PstII/KpnI (nt. pos. 15 to 530) fragment is sub-cloned into PstI/KpnI cleaved pMTL20 with concomitant loss of AatII and NcoI polylinker sites. This enables the substitution of the c. 100 bp AatII/NcoI fragment containing the TTG triplet deletion a correct sequence derived from the annealing of complementary c. 100 bp oligonucleotides (nt. pos. 178 to 279). Such a clone is used for the second step involving mutagenic PCR using oligonucleotide primers AL2 and casein 15 whereby the base change at nt. pos. 512 is corrected.

Claims

1. A DNA molecule coding for a food protein, modified in that the codons coding for phenylalanine have been omitted or replaced by codons coding for one or more other metabolisable amino acids.

2. An edible polypeptide which comprises a food protein modified in that the phenylalanine residues have been omitted or have been replaced by one or more other metabolisable amino acids.

3. A nutrient material comprising a polypeptide as defined in claim 2 and other edible substances.

4. A DNA molecule according to claim 1 wherein codons for phenylalanine have been replaced by codons for tyrosine.

5. A nutrient material according to claim 2 obtained by partially-purifying said edible polypeptide.

6. A nutrient material according to claim 5 wherein the purification has been carried out to such a degree that the material contains substantially the metabolically-required proportion of phenylalanine.

7. An expression vector into which has been incorporated a DNA molecule according to either of claims 1 or 4.

8. A DNA molecule according to claim 1 or 4, wherein the food protein is ovalbumin or casein.

9. An edible polypeptide according to claim 2, wherein the food protein is ovalbumin or casein.

10. A host transformed by an expression vector according to claim 7.

11. A host according to claim 10, which is S. cerevisiae.

12. A host according to claim 11 which is a yeast.

13. A host according to claim 12, which is S. cerevisiae or Pichia pastoris.

14. A method of producing an edible polypeptide according to claim 2 comprising transforming a host with an expression vector according to claim 8, culturing the transformed host, and harvesting the edible polypeptide.

15. A method according to claim 14, wherein the edible polypeptide is partially purified so that it still contains some phenylalanine-containing proteins from the host, the proportion of phenylalanine in the product being the metabolic amount required by a phenylketonuria patient.

16. A nutrient material comprising a partially-purified edible polypeptide obtained by the method of claim 15.