Method of humanizing immune system molecules
AssigneeSunol Molecular Corporation
InventorsHing C. Wong, Jeffrey R. Stinson, Luis A. Mosquera
Life of this patent
Estimating the term…
Abstract
Disclosed are methods for making immune system molecules, particularly humanized antibodies and fragments thereof. A typical method optimizes similarity between individual antibody framework regions to help identify human framework regions suitable for making the molecules. Comparisons between larger antibody frameworks, hypervariable domains or both is avoided. Also disclosed are antibodies that have been humanized by the method. The invention has a wide spectrum of applications including use in the production of humanized monoclonal antibodies with suitable binding affinity and minimized human immunogenicity.
Claims41
A method for producing a humanized antibody variable (V) domain or a fragment thereof, wherein the method comprises:
- a) comparing the amino acid sequence of a framework region (FR) of a non-human antibody variable (V) domain to a collection of human antibody framework amino acid sequences, or fragments thereof,
- b) selecting a human FR sequence from the collection having the greatest amino acid sequence identity to the non-human FR,
- c) mutagenizing DNA of the non-human FR to encode a humanized FR (huFR) having an amino acid sequence substantially identical to the selected human FR from step b),
- d) repeating steps a) thru c) for each of the FRs in the non-human V domain to produce a plurality of DNA sequences in which each DNA sequence encodes a humanized FR;
- and
- e) substituting into a first vector encoding at least the V domain of the non-human antibody, each of the huFR DNA sequences from step d) for the corresponding non-human FRs encoded by the vector;
- wherein the substitution operatively links each of the huFRs to its corresponding complimentarity determining region (CDR);
- and
- f) expressing the first vector in host cells and under conditions conducive to making the humanized antibody V domain or the fragment thereof.
7 The method of claim 6, wherein step d) further comprises comparing a fourth framework region (FR4) of the non-human light chain V domain to the collection and selecting a human FR having at least about 70% sequence identity.
The method of claim 7, wherein the humanized light chain V domain comprises covalently linked in sequence:
- huFR1-CDR1-huFR2-CDR2-huFR3-CDR3-huFR4;
- or a fragment thereof.
11a. The method of claim 11, wherein the humanized light chain fragment has an amino acid length of between from about 95 to about 235 amino acids.
17 The method of claim 16, wherein step d) further comprises comparing a fourth framework region (FR4) of the non-human heavy chain V domain to the collection and selecting a human FR having at least about 70% sequence identity.
The method of claim 17, wherein the humanized light chain V domain comprises covalently linked in sequence:
- huFR1-CDR1-huFR2-CDR2-huFR3-CDR3-huFR4 or a fragment thereof.
A method for making a humanized antibody or a fragment thereof, wherein the method comprises:
- a) comparing the amino acid sequence of a framework of a non-human antibody light chain variable (V) domain (1-FR) to a collection of human antibody light chain amino acid sequences, or fragments thereof,
- b) selecting a human FR sequence from the collection having the greatest amino acid sequence identity to the 1-FR,
- c) mutagenizing DNA of the 1-FR to encode a light chain humanized FR (L-huFR) having an amino acid sequence substantially identical to the selected human FR from step b),
- d) repeating steps a) thru c) for each of the FRs in the light chain V domain to produce a plurality of DNA sequences in which each DNA sequence encodes an L-huFR,
- e) substituting into a first vector encoding at least the light chain V domain of the non-human antibody, each of the L-huFR DNA sequences from step d) for the corresponding l-FRs encoded by the vector;
- wherein the substitution operatively links each of the L-huFRs to a corresponding complimentarily determining region (CDR),
- f) comparing the amino acid sequence of a framework region of a non-human antibody heavy chain variable (V) domain (h-FR) to a collection of human antibody heavy chain amino acid sequences, or fragments thereof,
- g) selecting a human FR sequence from the collection having the greatest amino acid sequence identity to the h-FR,
- h) mutagenizing DNA of the h-FR to encode a humanized heavy chain FR(H-huFR) having an amino acid sequence substantially identical to the selected human FR from step g),
- i) repeating steps f) thru h) for each of the h-FRs in the non-human heavy chain V region to produce a plurality of DNA sequences in which each DNA sequence encodes a H-huFR,
- j) substituting into a second vector encoding at least the heavy chain V domain of the non-human antibody, each of the H-huFR DNA sequences from step i) for the corresponding h-FRs encoded by the vector;
- wherein the substitution operatively links each of the H-huFRs to a corresponding heavy chain CDR;
- and
- k) expressing the first and second vectors in the same host cells and under conditions conducive to producing humanized light and heavy chains and making the humanized antibody or the fragment thereof.
Dependencies are inferred from the claim text.