Patent Grant
7462354
The Sequence Listing written in file Sequence Listing 2060 0320003.txt, 149,504 bytes, created on May 10, 2004 on two identical copies of compact discs for application Ser. No. 09/894,018, Sette et al., Method and System for Optimizing Minigenes and Peptides Encoded Thereby, is herein incorporated-by-reference.
This present invention relates to the field of biology. In particular, it relates to multi-epitope (multi-epitope) vaccines and methods of designing such vaccines to provide increased immunogenicity. In certain embodiments, the multi-epitope vaccine is encoded by a minigene that provides optimized immunogenicity of the construct.
The technology relevant to multi-epitope (“minigene”) vaccines is developing. Several independent studies have established that induction of simultaneous immune responses against multiple epitopes can be achieved. For example, responses against a large number of T cell specificities can be induced and detected. In natural situations, Doolan et al (Immunity, Vol. 7(1):97-112 (1997)) simultaneously detected recall T cell responses, against as many as 17 different P. falciparum epitopes using PBMC from a single donor. Similarly, Bertoni and colleagues (J Clin Invest, Vol. 100(3):503-13 (1997)) detected simultaneous CTL responses against 12 different HBV-derived epitopes in a single donor. In terms of immunization with multi-epitope DNA minigene vaccines, several examples have been reported where multiple T cell responses were induced. For example, minigene vaccines composed of approximately ten MHC Class I epitopes in which all epitopes were immunogenic and/or antigenic have been reported. Specifically, minigene vaccines composed of 9 EBV (Thomson et al., Proc Natl Acad Sci USA, Vol. 92(13):5845-9 (1995)), 7HIV (Woodberry et al., J Virol, Vol. 73(7):5320-5 (1999)), 10 murine (Thomson et al., J Immunol, Vol. 160(4):1717-23 (1998)) and 10 tumor-derived (Mateo et al., J Immunol, Vol. 163(7):4058-63 (1999)) epitopes have been shown to be active. It has also been shown that a multi-epitope DNA plasmid encoding nine different HLA-A2.1- and A11-restricted epitopes derived from HBV and HIV induced CTL against all epitopes (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)).
Thus, minigene vaccines containing multiple MHC Class I and Class II (i.e., CTL) epitopes can be designed, and presentation and recognition can be obtained for all epitopes. However, the immunogenicity of multi-epitope constructs appears to be strongly influenced by a number of variables, a number of which have heretofore been unknown. For example, the immunogenicity (or antigenicity) of the same epitope expressed in the context of different vaccine constructs can vary over several orders of magnitude. Thus, there exists a need to identify strategies to optimize multi-epitope vaccine constructs. Such optimization is important in terms of induction of potent immune responses and ultimately, for clinical efficacy. Accordingly, the present invention provides strategies to optimize antigenicity and immunogenicity of multi-epitope vaccines encompassing a large number of epitopes, and optimized multi-epitope vaccines, particularly minigene vaccines, generated in accordance with these strategies.
The following paragraphs provide a brief review of some of the main variables potentially influencing minigene immunogenicity, epitope processing, and presentation on antigen presenting cells (APCs) in association with Class I and Class II MHC molecules.
Immunodominance
Of the many thousand possible peptides that are encoded by a complex foreign pathogen, only a small fraction ends up in a peptide form capable of binding to MHC Class I antigens and thus of being recognized by T cells. This phenomenon, of obvious potential impact on the development of a multi-epitope vaccine, is known as immunodominance (Yewdell et al., Annu Rev Immunol, 17:51-88 (1999)). Several major variables contribute to immunodominance. Herein, we describe variables affecting the generation of the appropriate peptides, both in qualitative and quantitative terms, as a result of intracellular processing.
Junctional Epitopes
A junctional epitope is defined as an epitope created due to the juxtaposition of two other epitopes. The new epitope is composed of a C-terminal section derived from a first epitope, and an N-terminal section derived from a second epitope. Creation of junctional epitopes is a potential problem in the design of multi-epitope minigene vaccines, for both Class I and Class II restricted epitopes for the following reasons. Firstly, when developing a minigene composed of, or containing, human epitopes, which are typically tested for immunogenicity in HLA transgenic laboratory animals, the creation of murine epitopes could create undesired immunodominance effects. Secondly, the creation of new, unintended epitopes for human HLA Class I or Class II molecules could elicit in vaccine recipients, new T cell specificities that are not expressed by infected cells or tumors that are the targets-induced T cell responses. These responses are by definition irrelevant and ineffective and could even be counterproductive, by creating undesired immunodominance effects.
The existence of junctional epitopes has been documented in a variety of different experimental situations. Gefter and collaborators first demonstrated the effect in a system in which two different Class II restricted epitopes were juxtaposed and colinearly synthesized (Perkins et al., J Immunol, Vol. 146(7):2137-44 (1991)). The effect was so marked that the immune system recognition of the epitopes could be completely “silenced” by these new junctional epitopes (Wang et al., Cell Immunol, Vol. 143(2):284-97 (1992)). Helper T cells directed against junctional epitopes were also observed in humans as a result of immunization with a synthetic lipopeptide, which was composed of an HLA-A2-restricted HBV-derived immunodominant CTL epitope, and a universal Tetanus Toxoid-derived HTL epitope (Livingston et al, J Immunol, Vol. 159(3):1383-92 (1997)). Thus, the creation of junctional epitopes are a major consideration in the design of multi-epitope constructs.
The present invention provides methods of addressing this problem and avoiding or minimizing the occurrence of junctional epitopes.
Flanking Regions
Class I restricted epitopes are generated by a complex process (Yewdell et al., Annu Rev Immunol, 17:51-88 (1999)). Limited proteolysis involving endoproteases and potential trimming by exoproteases is followed by translocation across the endoplasmic reticulum (ER) membrane by transporters associated with antigen processing (TAP) molecules. The major cytosolic protease complex involved in generation of antigenic peptides, and their precursors, is the proteosome (Niedermann et al., Immunity, Vol. 2(3):289-99 (1995)), although ER trimming of CTL precursors has also been demonstrated (Paz et al., Immunity Vol. 11(2):241-51 (1999)). It has long been debated whether or not the residues immediately flanking the C and N terminus of the epitope, have an influence on the efficiency of epitope generation.
The yield and availability of processed epitope has been implicated as a major variable in determining immunogenicity and could thus clearly have a major impact on overall minigene potency in that the magnitude of immune response can be directly proportional to the amount of epitope bound by MHC and displayed for T cell recognition. Several studies have provided evidence that this is indeed the case. For example, induction of virus-specific CTL that is essentially proportional to epitope density (Wherry et al., J Immunol, Vol. 163(7):3735-45 (1999)) has been observed. Further, recombinant minigenes, which encode a preprocessed optimal epitope, have been used to induce higher levels of epitope expression than naturally observed with full-length protein (Anton et al., J Immunol, Vol. 158(6):2535-42 (1997)). In general, minigene priming has been shown to be more effective than priming with the whole antigen (Restifo et al., J Immunol, Vol. 154(9):4414-22 (1995); Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)), even though some exceptions have been noted (Iwasaki et al., Vaccine, Vol. 17(15-16):2081-8 (1999)).
Early studies concluded that residues within the epitope (Hahn et al., J Exp Med, Vol. 176(5):1335-41 (1992)) primarily regulate immunogenicity. Similar conclusions were reached by other studies, mostly based on grafting an epitope in an unrelated gene, or in the same gene, but in a different location (Chimini et al., J Exp Med, Vol. 169(1):297-302 (1989); Hahn et al., J Exp Med, Vol. 174(3):733-6 (1991)). Other experiments however (Del Val et al., Cell, Vol. 66(6):1145-53 (1991); Hahn et al., J Exp Med, Vol. 176(5):1335-41 (1992)), suggested that residues localized directly adjacent to the CTL epitope can directly influence recognition (Couillin et al., J Exp Med, Vol. 180(3):1129-34 (1994); Bergmann et al., J Virol. Vol. 68(8):5306-10 (1994)). In the context of minigene vaccines, the controversy has been renewed. Shastri and coworkers (Shastri et al., J Immunol, Vol. 155(9):4339-46 (1995)) found that T cell responses were not significantly affected by varying the N-terminal flanking residue but were inhibited by the addition of a single C-terminal flanking residue. The most dramatic inhibition was observed with isoleucine, leucine, cysteine, and proline as the C-terminal flanking residues. In contrast, Gileadi (Gileadi et al., Eur J Immunol, Vol. 29(7):2213-22 (1999)) reported profound effects as a function of the residues located at the N terminus of mouse influenza virus epitopes. Bergmann and coworkers found that aromatic, basic and alanine residues supported efficient epitope recognition, while G and P residues were strongly inhibitory (Bergmann et al., J Immunol, Vol. 157(8):3242-9 (1996)). In contrast, Lippolis (Lippolis et al., J Virol, Vol. 69(5):3134-46 (1995)) concluded that substituting flanking residues did not effect recognition. However, only rather conservative substitutions which are unlikely to affect proteosome specificity, were tested.
It appears that the specificity of these effects, and in general of natural epitopes, roughly correlates with proteosome specificity. For example, proteosome specificity is partly trypsin-like (Niedermann et al., Immunity, Vol. 2(3):289-99 (1995)), with cleavage following basic amino acids. Nevertheless, efficient cleavage of the carboxyl side of hydrophobic and acidic residues is also possible. Consistent with these specificities are the studies of Sherman and collaborators, which found that an R to H mutation at the position following the C-terminus of a p53 epitope affects proteosome-mediated processing of the protein (Theobald et al., J Exp Med, Vol. 188(6):1017-28 (1998)). Several other studies (Hanke et al., J Gen Virol, Vol. 79 (Pt 1):83-90 (1998); Thomson et al., Proc Natl Acad Sci USA, Vol. 92(13):5845-9 (1995)) indicated that minigenes can be constructed utilizing minimal epitopes, and that these flanking sequences appear not be required, although the potential for further optimization by the use of flanking regions was also acknowledged.
In sum, for HLA Class I epitopes, the effects of flanking regions on processing and presentation of CTL epitopes is as yet undefined. A systematic analysis of the effect of modulation of flanking regions has not been performed for minigene vaccines. Thus, analysis utilizing minigene vaccines encoding epitopes restricted by human Class I in general is needed. The present invention provides such an analysis and accordingly, provides multi-epitope vaccine constructs optimized for immunogenicity and antigenicity, and methods of designing such constructs.
HLA Class II peptide complexes are also generated as a result of a complex series of events that is distinct from HLA Class I processing. The processing pathway involves association with Invariant chain (Ii), its transport to specialized compartments, the degradation of Ii to CLIP, and HLA-DM catalyzed removal of CLIP (see (Blum et al., Crit Rev Immunol, Vol. 17(5-6):411-7 (1997); Arndt et al., Immunol Res, Vol. 16(3):261-72 (1997)) for review. Moreover, there is a potentially crucial role of various cathepsins in general, and cathepsin S and L in particular, in Ii degradation (Nakagawa et al., Immunity, Vol. 10(2):207-17 (1999)). In terms of generation of functional epitopes however, the process appears to be somewhat less selective (Chapman H. A., Curr Opin Immunol, Vol. 10(1):93-102 (1998)), and peptides of many sizes can bind to MHC Class IMHC Class II (Hunt et al., Science, Vol. 256(5065):1817-20 (1992)). Most or all of the possible peptides appear to be generated (Moudgil et al., J Immunol, Vol. 159(6):2574-9 (1997); and Thomson et al., J Virol, Vol. 72(3):2246-52 (1998)). Thus, as compared to the issue of flanking regions, the creation of junctional epitopes can be a more serious concern in particular embodiments.
The invention provides a method and system for optimizing the efficacy of multi-epitope vaccines so as to minimize the number of junctional epitopes and maximize, or at least increase, the immunogenicity and/or antigenicity of multi-epitope vaccines. Also disclosed are multi-epitope constructs and nucleicic acids encoding such constructs (minigenes).
In one embodiment of the invention, a computerized method for designing a multi-epitope construct having multiple epitopes includes the steps of: storing a plurality of input parameters in a memory of a computer system, the input parameters including a plurality of epitopes, at least one motif for identifying junctional epitopes, a plurality of amino acid insertions and at least one enhancement weight value for each insertion; generating a list of epitope pairs from the plurality of epitopes; determining for each epitope pair at least one optimum combination of amino acid insertions based on the at least one motif, the plurality of insertions and the at least one enhancement weight value for each insertion; and identifying at least one optimum arrangement of the plurality of epitopes, wherein a respective one of the at least one optimum combination of amino acid insertions is inserted at a respective junction of two epitopes, so as to provide an optimized multi-epitope construct. In a preferred embodiment, the step of identifying at least one optimum arrangement of epitopes may be accomplished by performing either an exhaustive search wherein all permutations of arrangements of the plurality of epitopes are evaluated or a stochastic search wherein only a subset of all permutations of arrangements of the plurality of epitopes are evaluated.
In a further embodiment, the method determines for each epitope pair at least one optimum combination of amino acid insertions by calculating a function value (F) for each possible combination of insertions for each epitope pair, wherein the number of insertions in a combination may range from 0 to a maximum number of insertions (MaxInsertions) value input by a user, and the function value is calculated in accordance with the equation F=(C+N)/J, when J>0, and F=2(C+N), when J=0, wherein C equals the enhancement weight value of a C+1 flanking amino acid, N equals the enhancement weight value of an N−1 flanking amino acid, and J equals the number of junctional epitopes detected for each respective combination of insertions in an epitope pair based on said at least one motif.
In another embodiment of the invention, a computer system for designing a multi-epitope construct having multiple epitopes, includes: a memory for storing a plurality of input parameters such as a plurality of epitopes, at least one motif for identifying junctional epitopes, a plurality of amino acid insertions and at least one enhancement weight value for each insertion; a processor for retrieving the input parameters from memory and generating a list of epitope pairs from the plurality of epitopes; wherein the processor further determines for each epitope pair at least one optimum combination of amino acid insertions, based on the at least one motif, the plurality of insertions and the at least one enhancement weight value for each insertion. The processor further identifies at least one optimum arrangement of the plurality of epitopes, wherein a respective one of the optimum combinations of amino acid insertions are inserted at a respective junction of two epitopes, to provide an optimized multi-epitope construct; and a display monitor, coupled to the processor, for displaying at least one optimum arrangement of the plurality of epitopes to a user.
In a further embodiment, the invention provides a data storage device storing a computer program for designing a multi-epitope construct having multiple epitopes, the computer program, when executed by a computer system, performing a process that includes the steps of: retrieving a plurality of input parameters from a memory of a computer system, the input parameters including, for example, a plurality of epitopes, at least one motif for identifying junctional epitopes, a plurality of amino acid insertions and at least one enhancement weight value for each insertion; generating a list of epitope pairs from the plurality of epitopes; determining for each epitope pair at least one optimum combination of amino acid insertions based on the at least one motif, the plurality of insertions and the at least one enhancement weight value for each insertion; and identifying at least one optimum arrangement of the plurality of epitopes, wherein a respective one of the at least one optimum combination of amino acid insertions is inserted at a respective junction of two epitopes, so as to provide an optimized multi-epitope construct.
In another embodiment, the invention provides a method and system for designing a multi-epitope construct that comprises multiple epitopes. The method comprising steps of: (i) sorting the multiple epitopes to minimize the number of junctional epitopes; (ii) introducing a flanking amino acid residue at a C+1 position of an epitope to be included within the multi-epitope construct; (iii) introducing one or more amino acid spacer residues between two epitopes of the multi-epitope construct, wherein the spacer prevents the occurrence of a junctional epitope; and, (iv) selecting one or more multi-epitope constructs that have a minimal number of junctional epitopes, a minimal number of amino acid spacer residues, and a maximum number of flanking amino acid residues at a C+1 position relative to each epitope. In some embodiments, the spacer residues are independently selected from residues that are not known HLA Class II primary anchor residues. In particular embodiments, introducing the spacer residues prevents the occurrence of an HTL epitope. Such a spacer often comprises at least 5 amino acid residues independently selected from the group consisting of G, P, and N. In some embodiments the spacer is GPGPG (SEQ ID NO:369).
In some embodiments, introducing the spacer residues prevents the occurrence of a CTL epitope and further, wherein the spacer is 1, 2, 3, 4, 5, 6, 7, or 8 amino acid residues independently selected from the group consisting of A and G. Often, the flanking residue is introduced at the C+1 position of a CTL epitope and is selected from the group consisting of K, R, N, G, and A. In some embodiments, the flanking residue is adjacent to the spacer sequence. The method of the invention can also include substituting an N-terminal residue of an epitope that is adjacent to a C-terminus of an adjacent epitope within the multi-epitope construct with a residue selected from the group consisting of K, R, N, G, and A.
In some embodiments, the method of the invention can also further comprise a step of predicting a structure of the multi-epitope construct, and further, selecting one or more constructs that have a maximal structure, i.e., that are processed by an HLA processing pathway to produce all of the epitopes comprised by the construct. Often, the epitopes comprised by the multi-epitope construct are encoded by a minigene. In some embodiments, the multi-epitope construct encoded by the minigene is EP-HIV-1090 as set out in
In another embodiment of the invention, a system for optimizing multi-epitope minigenes includes a computer system having a processor (e.g., central processing unit) and at least one memory coupled to the processor for storing instructions executed by the processor and data to be manipulated (i.e., processed) by the processor. The computer system further includes an input device (e.g., keyboard) coupled to the processor and the at least one memory for allowing a user to input desired parameters and information to be accessed by the processor. The processor may be a single CPU or a plurality of different processing devices/circuits integrated onto a single integrated circuit chip. Alternatively, the processor may be a collection of discrete processing devices/circuits selectively coupled to one another via either direct wire/conductor connections or via a data bus. Similarly, the at least one memory may be one large memory device (e.g., EPROM), or a collection of a plurality of discrete memory devices (e.g., EEPROM, EPROM, RAM, DRAM, SDRAM, Flash, etc.) selectively coupled to one another for selectively storing data and/or program information (i.e., instructions executed by the processor). Those of ordinary skill in the art would easily be able to implement a desired computer system architecture to perform the operations and functions disclosed herein.
In one embodiment, the computer system includes a display monitor for displaying information, instructions, images, graphics, etc. The computer system receives user inputs via a keyboard. These user input parameters may include, for example, the number of insertions (i.e., flanking residues and spacer residues), the peptides to be processed, the C+1 and N−1 weighting values for each amino acid, and the motifs to use for searching for junctional epitopes. Based on these input values/parameters, the computer system executes a “Junctional Analyzer” software program which automatically determines the number of junctional epitope for each peptide pair and also calculates an “enhancement” value for each combination of flanking residues and spacers that may be inserted at the junction of each peptide pair. The results of the junctional analyzer program are then used in either an exhaustive or stochastic search program which determines the “optimal” combination or linkage of the entire set of peptides to create a multi-epitope polypeptide, or minigene, having a minimal number of junctional epitopes and a maximum functional (e.g., immunogenicity) value.
In one embodiment, if the number of peptides to be processed by the computer system is less than fourteen, an exhaustive search program is executed by the computer system which examines all permutations of the peptides making up the polypeptide to find the permutation with the “best” or “optimal” function value. In one embodiment, the function value is calculated using the equation (Ce+Ne)/J when J is greater than zero and 2*(Ce+Ne) when J is equal to zero, where Ce is the enhancement “weight” value of an amino acid at the C+1 position of a peptide, Ne is the enhancement “weight” value of an amino acid at the N−1 position of a peptide, and J is the number of junctional epitopes contained in the polypeptide minigene. Thus, maximizing this function value will identify the peptide pairs having the least number of junctional epitopes and the maximum enhancement weight value for flanking residues. If the number of peptides to be processed is fourteen or more, the computer system executes a stochastic search program that uses a “Monte Carlo” technique to examine many regions of the permutation space to find the best estimate of the optimum arrangement of peptides (e.g., having the maximum function value).
In a further embodiment, the computer system allows a user to input parameter values which format or limit the output results of the exhaustive or stochastic search program. For example, a user may input the maximum number of results having the same function value (“MaxDuplicateFunctionValue=X”) to limit the number of permutations that are generated as a result of the search. Since it is possible for the search programs to find many arrangements that give the same function value, it may be desirable to prevent the output file from being filled by a large number of equivalent solutions. Once this limit is reached no more results are reported until a larger or “better” function value is found. As another example, the user may input the maximum number of “hits” per probe during a stochastic search process. This parameter prevents the stochastic search program from generating too much output on a single probe. In a preferred embodiment, the number of permutations examined in a single probe is limited by several factors: the amount of time set for each probe in the input text file; the speed of the computer, and the values of the parameters “MaxHitsPerProbe” and “MaxDuplicateFunctionValues.” The algorithms used to generate and select permutations for analysis may be in accordance with well-known recursive algorithms found in many computer science text books. For example, six permutations of three things taken three at a time would be generated in the following sequence: ABC; ACB; BAC; BCA; CBA; CAB. As a further example of an input parameter, a user may input how the stochastic search is performed, e.g., randomly, statistically or other methodology; the maximum time allowed for each probe (e.g., 5 minutes); and the number of probes to perform.
Also disclosed herein are multi-epitope constructs designed by the methods described above and hereafter. The multi-epitope constructs include spacer nucleic acids between a subset of the epitope nucleic acids or all of the epitope nucleic acids. One or more of the spacer nucleic acids may encode amino acid sequences different from amino acid sequences encoded by other spacer nucleic acids to optimize epitope processing and to minimize the presence of junctional epitopes.
The following definitions are provided to enable one of ordinary skill in the art to understand some of the preferred embodiments of invention disclosed herein. It is understood, however, that these definitions are exemplary only and should not be used to limit the scope of the invention as set forth in the claims. Those of ordinary skill in the art will be able to construct slight modifications to the definitions below and utilize such modified definitions to understand and practice the invention disclosed herein. Such modifications, which would be obvious to one of ordinary skill in the art, as they may be applicable to the claims set forth below, are considered to be within the scope of the present invention.
Throughout this disclosure, “binding data” results are often expressed in terms of “IC50's.” IC50 is the concentration of peptide in a binding assay at which 50% inhibition of binding of a reference peptide is observed. Given the conditions in which the assays are run (i.e., limiting HLA proteins and labeled peptide concentrations), these values approximate KD values. Assays for determining binding are described in detail, e.g., in PCT publications WO 94/20127 and WO 94/03205. It should be noted that IC50 values can change, often dramatically, if the assay conditions are varied, and depending on the particular reagents used (e.g., HLA preparation, etc.). For example, excessive concentrations of HLA molecules will increase the apparent measured IC50 of a given ligand. Alternatively, binding is expressed relative to a reference peptide. Although as a particular assay becomes more, or less, sensitive, the IC50's of the peptides tested may change somewhat, the binding relative to the reference peptide will not significantly change. For example, in an assay run under conditions such that the IC50 of the reference peptide increases 10-fold, the IC50 values of the test peptides will also shift approximately 10-fold. Therefore, to avoid ambiguities, the assessment of whether a peptide is a good, intermediate, weak, or negative binder is generally based on its IC50, relative to the IC50 of a standard peptide. Binding may also be determined using other assay systems including those using: live cells (e.g., Ceppellini et al., Nature 339:392, 1989; Christnick et al., Nature 352:67, 1991; Busch et al., Int. Immunol. 2:443, 19990;Hill et al., J. Immunol. 147:189,1991;del Guercioet al., J. Immunol. 154:685, 1995), cell free systems using detergent lysates (e.g., Cerundolo et al., J. Immunol. 21:2069, 1991), immobilized purified MHC (e.g., Hill et al., J. Immunol. 152, 2890, 1994; Marshall et al., J. Immunol. 152:4946, 1994), ELISA systems (e.g., Reay et al., EMBO J. 11:2829, 1992), surface plasmon resonance (e.g., Khilko et al., J. Biol. Chem. 268:15425, 1993); high flux soluble phase assays (Hammer et al., J. Exp. Med. 180:2353, 1994), and measurement of class I MHC stabilization or assembly (e.g., Ljunggren et al., Nature 346:476, 1990; Schumacher et al., Cell 62:563, 1990; Townsend et al., Cell 62:285, 1990; Parker et al., J. Immunol. 149:1896, 1992).
The designation of a residue position in an epitope as the “carboxyl terminus” or the “carboxyl terminal position” refers to the residue position at the end of the epitope that is nearest to the carboxyl terminus of a peptide, which is designated using conventional nomenclature as defined below. “C+1” refers to the residue or position immediately following the C-terminal residue of the epitope, ie., refers to the residue flanking the C-terminus of the epitope. The “carboxyl terminal position” of the epitope occurring at the carboxyl end of the multi-epitope construct may or may not actually correspond to the carboxyl terminal end of polypeptide. In preferred embodiments, the epitopes employed in the optimized multi-epitope constructs are motif-bearing epitopes and the carboxyl terminus of the epitope is defined with respect to primary anchor residues corresponding to a particular motif.
The designation of a residue position in an epitope as “amino terminus” or “amino-terminal position” refers to the residue position at the end of the epitope which is nearest to the amino terminus of a peptide, which is designated using conventional nomenclature as defined below. “N−1” refers to the residue or position immediately adjacent to the epitope at the amino terminal end (position number 1) of an eptiope. The “amino terminal position” of the epitope occurring at the amino terminal end of the multi-epitope construct may or may not actually corresponds to the amino terminal end of the polypeptide. In preferred embodiments, the epitopes employed in the optimized multi-epitope constructs are motif-bearing epitopes and the amino terminus of the epitope is defined with respect to primary anchor residues corresponding to a particular motif.
A “computer” or “computer system” generally includes: a processor; at least one information storage/retrieval apparatus such as, for example, a hard drive, a disk drive or a tape drive; at least one input apparatus such as, for example, a keyboard, a mouse, a touch screen, or a microphone; and display structure. Additionally, the computer may include a communication channel in communication with a network such that remote users may communicate with the computer via the network to perform minigene optimization functions disclosed herein. Such a computer may include more or less than what is listed above. The network may be a local area network (LAN), wide area network (WAN) or a global network such as the world wide web (e.g., the internet).
A “construct” as used herein generally denotes a composition that does not occur in nature. A construct can be produced by synthetic technologies, e.g., recombinant DNA preparation and expression or chemical synthetic techniques for nucleic or amino acids. A construct can also be produced by the addition or affiliation of one material with another such that the result is not found in nature in that form. A “multi-epitope construct” comprises multiple epitope nucleic acids that encode peptide epitopes of any length that can bind to a molecule functioning in the immune system, preferably a class I HLA and a T-cell receptor or a class II HLA and a T-cell receptor. All of the epitope nucleic acids in a multi-epitope construct can encode class I HLA epitopes or class II HLA epitopes. Class I HLA-encoding epitope nucleic acids are referred to as CTL epitope nucleic acids, and class II HLA-encoding epitope nucleic acids are referred to as HTL epitope nucleic acids. Some multi-epitope constructs can have a subset of the multi-epitope nucleic acids encoding class I HLA epitopes and another subset of the multi-epitope nucleic acids encoding class II HLA epitopes. The CTL epitope nucleic acids preferably encode an epitope peptide of about eight to about thirteen amino acids in length, more preferably about eight to about eleven amino acids in length, and most preferably about nine amino acids in length. The HTL epitope nucleic acids can encode an epitope peptide of about seven to about twenty three, preferably about seven to about seventeen, more preferably about eleven to about fifteen, and most preferably about thirteen amino acids in length. The multi-epitope constructs described herein preferably include five or more, ten or more, fifteen or more, twenty or more, or twenty-five or more epitope nucleic acids. All of the epitope nucleic acids in a multi-epitope construct may be from one organism (e.g., the nucleotide sequence of every epitope nucleic acid may be present in HIV strains), or the multi-epitope construct may include epitope nucleic acids present in two or more different organisms (e.g., some epitopes from HIV and some from HCV). As described hereafter, one or more epitope nucleic acids in the multi-epitope construct may be flanked by a spacer nucleic acid.
A “multi-epitope vaccine,” which is synonyous with a “polyepitopic vaccine,” is a vaccine comprising multiple epitopes.
“Cross-reactive binding” indicates that a peptide is bound by more than one HLA molecule; a synonym is “degenerate binding.”
A “cryptic epitope” elicits a response by immunization with an isolated peptide, but the response is not cross-reactive in vitro when intact whole protein that comprises the epitope is used as an antigen.
A “dominant epitope” is an epitope that induces an immune response upon immunization with a whole native antigen (see, e.g., Sercarz, et al., Annu. Rev. Immunol. 11:729-766, 1993). Such a response is cross-reactive in vitro with an isolated peptide epitope.
With regard to a particular amino acid sequence, an “epitope” is a set of amino acid residues which is involved in recognition by a particular immunoglobulin, or in the context of T cells, those residues necessary for recognition by T cell receptor proteins and/or Major Histocompatibility Complex (MHC) receptors. In an immune system setting, in vitro or in vivo, an epitope is the collective features of a molecule, such as primary, secondary and tertiary peptide structure, and charge, that together form a site recognized by an immunoglobulin, T cell receptor or HLA molecule. Throughout this disclosure epitope and peptide are often used interchangeably. It is to be appreciated, however, that isolated or purified protein or peptide molecules larger than and comprising an epitope of the invention are still within the bounds of the invention.
A “flanking residue” is a residue that is positioned next to an epitope. A flanking residue can be introduced or inserted at a position adjacent to the N-terminus or the C-terminus of an epitope.
An “immunogenic peptide” or “peptide epitope” is a peptide that comprises an allele-specific motif or supermotif such that the peptide will bind an HLA molecule and induce a CTL and/or HTL response. Thus, immunogenic peptides of the invention are capable of binding to an appropriate HLA molecule and thereafter inducing a cytotoxic T cell response, or a helper T cell response, to the antigen from which the immunogenic peptide is derived.
“Heteroclitic analogs” are defined herein as a peptide with increased potency for a specific T cell, as measured by increased responses to a given dose, or by a requirement of lesser amounts to achieve the same response. Advantages of heteroclitic analogs include that the epitopes can be more potent, or more economical (since a lower amount is required to achieve the same effect). In addition, modified epitopes might overcome antigen-specific T cell unresponsiveness (T cell tolerance).
“Human Leukocyte Antigen” or “HLA” is a human class I or class II Major Histocompatibility Complex (MHC) protein (see, e.g., Stites, et al., I
An “HLA supertype or HLA family,” as used herein, describes sets of HLA molecules grouped based on shared peptide-binding specificities. HLA class I molecules that share similar binding affinity for peptides bearing certain amino acid motifs are grouped into such HLA supertypes. The terms HLA superfamily, HLA supertype family, HLA family, and HLA xx-like molecules (where xx denotes a particular HLA type), are synonyms.
As used herein, “high affinity” with respect to HLA class I molecules is defined as binding with an IC50, or KD value, of 50 nM or less; “intermediate affinity” with respect to HLA class I molecules is defined as binding with an IC50 or KD value of between about 50 and about 500 nM. “High affinity” with respect to binding to HLA class II molecules is defined as binding with an IC50 or KD value of 100 nM or less; “intermediate affinity” with respect to binding to HLA class II molecules is defined as binding with an IC50 or KD value of between about 100 and about 1000 nM.
An “IC50” is the concentration of peptide in a binding assay at which 50% inhibition of binding of a reference peptide is observed. Depending on the conditions in which the assays are run (i.e., limiting HLA proteins and labeled peptide concentrations), these values may approximate KD values.
The terms “identical” or percent “identity,” in the context of two or more peptide sequences, refer to two or more sequences or subsequences that are the same or have a specified percentage of amino acid residues that are the same, when compared and aligned for maximum correspondence over a comparison window, as measured using a sequence comparison algorithm or by manual alignment and visual inspection.
“Introducing” an amino acid residue at a particular position in a multi-epitope construct, e.g., adjacent, at the C-terminal side, to the C-terminus of the epitope, encompasses configuring multiple epitopes such that a desired residue is at a particular position, e.g., adjacent to the epitope, or such that a deleterious residue is not adjacent to the C-terminus of the epitope. The term also includes inserting an amino acid residue, preferably a preferred or intermediate amino acid residue, at a particular position. An amino acid residue can also be introduced into a sequence by substituting one amino acid residue for another. Preferably, such a substitution is made in accordance with analoging principles set forth, e.g., in U.S. Ser. No. 09/260,714 filed Mar. 1, 1999, now abandoned and PCT application number PCT/US00/19774.
The phrases “isolated” or “biologically pure” refer to material that is substantially or essentially free from components which normally accompany the material as it is found in its native state. Thus, isolated peptides in accordance with the invention preferably do not contain materials normally associated with the peptides in their in situ environment.
“Link” or “join” refers to any method known in the art for functionally connecting peptides, including, without limitation, recombinant fusion, covalent bonding, disulfide bonding, ionic bonding, hydrogen bonding, and electrostatic bonding.
“Major Histocompatibility Complex” or “MHC” is a cluster of genes that plays a role in control of the cellular interactions responsible for physiologic immune responses. In humans, the MHC complex is also known as the HLA complex. For a detailed description of the MHC and HLA complexes, see, Paul, F
As used herein, “middle of the peptide” is a position in a peptide that is neither an amino or a carboxyl terminus.
A “minimal number of junctional epitopes” as used herein refers to a number of junctional epitopes that is lower than what would be created using a random selection criteria.
The term “motif” refers to the pattern of residues in a peptide of defined length, usually a peptide of from about 8 to about 13 amino acids for a class I HLA motif and from about 6 to about 25 amino acids for a class II HLA motif, which is recognized by a particular HLA molecule. Peptide motifs are typically different for each protein encoded by each human HLA allele and differ in the pattern of the primary and secondary anchor residues.
A “negative binding residue” or “deleterious residue” is an amino acid which, if present at certain positions (typically not primary anchor positions) in a peptide epitope, results in decreased binding affinity of the peptide for the peptide's corresponding HLA molecule.
“Optimizing” refers to increasing the immunogenicity or antigenicity of a multi-epitope construct having at least one epitope pair by sorting epitopes to minimize the occurrence of junctional epitopes, inserting flanking residues that flank the C-terminus or N-terminus of an epitope, and inserting spacer residue to further prevent the occurrence of junctional epitopes or to provide a flanking residue. An increase in immunogenicity or antigenicity of an optimized multi-epitope construct is measured relative to a multi-epitope construct that has not been constructed based on the optimization parameters and is using assays known to those of skill in the art, e.g., assessment of immunogenicity in HLA transgenic mice, ELISPOT, interferon-gamma release assays, tetramer staining, chromium release assays, and presentation on dendritic cells.
The term “peptide” is used interchangeably with “oligopeptide” in the present specification to designate a series of residues, typically L-amino acids, connected one to the other, typically by peptide bonds between the α-amino and carboxyl groups of adjacent amino acids. The preferred CTL-inducing peptides of the invention are 13 residues or less in length and usually consist of between about 8 and about 11 residues, preferably 9 or 10 residues. The preferred HTL-inducing oligopeptides are less than about 50 residues in length and usually consist of between about 6 and about 30 residues, more usually between about 12 and 25, and often between about 15 and 20 residues.
A “PanDR binding peptide or PADRE® peptide” is a member of a family of molecules that binds more than one HLA class II DR molecule. The pattern that defines the PADRE® family of molecules can be thought of as an HLA Class II supermotif. PADRE® binds to most HLA-DR molecules and stimulates in vitro and in vivo human helper T lymphocyte (HTL) responses.
“Pharmaceutically acceptable” refers to a generally non-toxic, inert, and/or physiologically compatible composition.
“Presented to an HLA Class I processing pathway” means that the multi-epitope constructs are introduced into a cell such that they are largely processed by an HLA Class I processing pathway. Typically, multi-epitope constructs are introduced into the cells using expression vectors that encode the multi-epitope constructs. HLA Class II epitopes that are encoded by such a minigene are also presented on Class II molecules, although the mechanism of entry of the epitopes into the Class II processing pathway is not defined.
A “primary anchor residue” or a “primary MHC anchor” is an amino acid at a specific position along a peptide sequence that is understood to provide a contact point between the immunogenic peptide and the HLA molecule. One to three, usually two, primary anchor residues within a peptide of defined length generally defines a “motif” for an immunogenic peptide. These residues are understood to fit in close contact with peptide binding grooves of an HLA molecule, with their side chains buried in specific pockets of the binding grooves themselves. In one embodiment, for example, the primary anchor residues of an HLA class I epitope are located at position 2 (from the amino terminal position) and at the carboxyl terminal position of a 9-residue peptide epitope in accordance with the invention. The primary anchor positions for each motif and supermotif are described, for example, in Tables I and III of PCT/US00/27766, or PCT/US00/19774. Preferred amino acids that can serve as in the anchors for most Class II epitopes consist of M and F in position one and V, M, S, T, A and C in position six. Tolerated amino acids that can occupy these positions for most Class II epitopes consist of L, I, V, W, and Y in position one and P, L and I in position six. The presence of these amino acids in positions one and six in Class II epitopes defines the HLA-DR1, 4, 7 supermotif. The HLA-DR3 binding motif is defined by preferred amino acids from the group of L, I, V, M, F, Y and A in position one and D, E, N, Q, S and T in position four and K, R and H in position six. Other amino acids may be tolerated in these positions but they are not preferred.
Furthermore, analog peptides can be created by altering the presence or absence of particular residues in these primary anchor positions. Such analogs are used to modulate the binding affinity of a peptide comprising a particular motif or supermotif.
“Promiscuous recognition” occurs where a distinct peptide is recognized by the same T cell clone in the context of various HLA molecules. Promiscuous recognition or binding is synonymous with cross-reactive binding.
A “protective immune response” or “therapeutic immune response” refers to a CTL and/or an HTL response to an antigen derived from an infectious agent or a tumor antigen, which in some way prevents or at least partially arrests disease symptoms, side effects or progression. The immune response may also include an antibody response that has been facilitated by the stimulation of helper T cells.
The term “residue” refers to an amino acid or amino acid mimetic incorporated into a peptide or protein by an amide bond or amide bond mimetic.
A “secondary anchor residue” is an amino acid at a position other than a primary anchor position in a peptide that may influence peptide binding. A secondary anchor residue occurs at a significantly higher frequency amongst bound peptides than would be expected by random distribution of amino acids at one position. The secondary anchor residues are said to occur at “secondary anchor positions.” A secondary anchor residue can be identified as a residue which is present at a higher frequency among high or intermediate affinity binding peptides, or a residue otherwise associated with high or intermediate affinity binding. For example, analog peptides can be created by altering the presence or absence of particular residues in these secondary anchor positions. Such analogs are used to finely modulate the binding affinity of a peptide comprising a particular motif or supermotif. The terminology “fixed peptide” is sometimes used to refer to an analog peptide.
“Sorting epitopes” refers to determining or designing an order of the epitopes in a multi-epitope construct.
A “spacer” refers to a sequence that is inserted between two epitopes in a multi-epitope construct to prevent the occurrence of junctional epitopes and/or to increase the efficiency of processing. A multi-epitope construct may have one or more spacer nucleic acids. A spacer nucleic acid may flank each epitope nucleic acid in a construct, or the spacer nucleic acid to epitope nucleic acid ratio may be about 2 to 10, about 5 to 10, about 6 to 10, about 7 to 10, about 8 to 10, or about 9 to 10, where a ratio of about 8 to 10 has been determined to yield favorable results for some constructs.
The spacer nucleic acid may encode one or more amino acids. A spacer nucleic acid flanking a class I HLA epitope in a multi-epitope construct is preferably between one and about eight amino acids in length. A spacer nucleic acid flanking a class II HLA epitope in a multi-epitope construct is preferably greater than five, six, seven, or more amino acids in length, and more preferably five or six amino acids in length.
The number of spacers in a construct, the number of amino acids in a spacer, and the amino acid composition of a spacer can be selected to optimize epitope processing and/or minimize junctional epitopes. It is preferred that spacers are selected by concomitantly optimizing epitope processing and junctional motifs. Suitable amino acids for optimizing epitope processing are described herein. Also, suitable amino acid spacing for minimizing the number of junctional epitopes in a construct are described herein for class I and class II HLAs. For example, spacers flanking class II HLA epitopes preferably include G, P, and/or N residues as these are not generally known to be primary anchor residues (see, e.g., PCT/US00/19774). A particularly preferred spacer for flanking a class II HLA epitope includes alternating G and P residues, for example, (GP)n, (PG)n, (GP)nG, (PG)nP, and so forth, where n is an integer between one and ten, preferably two or about two, and where a specific example of such a spacer is GPGPG (SEQ ID NO:369).
In some multi-epitope constructs, it is sufficient that each spacer nucleic acid encodes the same amino acid sequence. In multi-epitope constructs having two spacer nucleic acids encoding the same amino acid sequence, the spacer nucleic acids encoding those spacers may have the same or different nucleotide sequences, where different nucleotide sequences may be preferred to decrease the likelihood of unintended recombination events when the multi-epitope construct is inserted into cells.
In other multi-epitope constructs, one or more of the spacer nucleic acids may encode different amino acid sequences. While many of the spacer nucleic acids may encode the same amino acid sequence in a multi-epitope construct, one, two, three, four, five or more spacer nucleic acids may encode different amino acid sequences, and it is possible that all of the spacer nucleic acids in a multi-epitope construct encode different amino acid sequences. Spacer nucleic acids may be optimized with respect to the epitope nucleic acids they flank by determining whether a spacer sequence will maximize epitope processing and/or minimize junctional epitopes, as described herein.
Multi-epitope constructs may be distinguished from one another according to whether the spacers in one construct optimize epitope processing or minimize junctional epitopes over another construct, and preferably, constructs may be distinguished where one construct is concomitantly optimized for epitope processing and junctional epitopes over the other. Computer assisted methods and in vitro and in vivo laboratory methods for determining whether a construct is optimized for epitope processing and junctional motifs are described herein.
A “subdominant epitope” is an epitope which evokes little or no response upon immunization with whole antigens which comprise the epitope, but for which a response can be obtained by immunization with an isolated epitope, and this response (unlike the case of cryptic epitopes) is detected when whole protein is used to recall the response in vitro or in vivo.
A “supermotif” is an amino acid sequence for a peptide that provides binding specificity shared by HLA molecules encoded by two or more HLA alleles. Preferably, a supermotif-bearing peptide is recognized with high or intermediate affinity (as defined herein) by two or more HLA antigens.
“Synthetic peptide” refers to a peptide that is not naturally occurring, but is man-made using such methods as chemical synthesis or recombinant DNA technology.
A “TCR contact residue” or “T cell receptor contact residue” is an amino acid residue in an epitope that is understood to be bound by a T cell receptor; these are defined herein as not being any primary MHC anchor. T cell receptor contact residues are defined as the position/positions in the peptide where all analogs tested induce T-cell recognition relative to that induced with a wildtype peptide.
The term “homology,” as used herein, refers to a degree of complementarity between two nucleotide sequences. The word “identity” may substitute for the word “homology” when a nucleic acid has the same nucleotide sequence as another nucleic acid. Sequence homology and sequence identity can also be determined by hybridization studies under high stringency and/or low stringency, and disclosed herein are nucleic acids that hybridize to the multi-epitope constructs under low stringency or under high stringency. Also, sequence homology and sequence identity can be determined by analyzing sequences using algorithms and computer programs known in the art. Such methods be used to assess whether a nucleic acid is identical or homologous to the multi-epitope constructs disclosed herein. The invention pertains in part to nucleotide sequences having 80% or more, 85% or more, 90% or more, 95% or more, 97% or more, 98% or more, or 99% or more identity to the nucleotide sequence of a multi-epitope construct disclosed herein.
As used herein, the term “stringent conditions” refers to conditions which permit hybridization between nucleotide sequences and the nucleotide sequences of the disclosed multi-epitope constructs. Suitable stringent conditions can be defined by, for example, the concentrations of salt or formamide in the prehybridization and hybridization solutions, or by the hybridization temperature, and are well known in the art. In particular, stringency can be increased by reducing the concentration of salt, increasing the concentration of formamide, or raising the hybridization temperature. For example, hybridization under high stringency conditions could occur in about 50% formamide at about 37° C. to 42° C. In particular, hybridization could occur under high stringency conditions at 42° C. in 50% formamide, 5× SSPE, 0.3% SDS, and 200 μg/ml sheared and denatured salmon sperm DNA or at 42° C. in a solution comprising 50% formamide, 5×SSC (750 mM NaCl, 75 mM trisodium citrate), 50 mM sodium phosphate (pH 7.6), 5× Denhardt's solution, 10% dextran sulfate, and 20 μg/ml denatured, sheared salmon sperm DNA, followed by washing the filters in 0.1×SSC at about 65° C. Hybridization could occur under reduced stringency conditions in about 35% to 25% formamide at about 30° C. to 35° C. For example, reduced stringency conditions could occur at 35° C. in 35% formamide, 5× SSPE, 0.3% SDS, and 200 μg/ml sheared and denatured salmon sperm DNA. The temperature range corresponding to a particular level of stringency can be further narrowed by calculating the purine to pyrimidine ratio of the nucleic acid of interest and adjusting the temperature accordingly. Variations on the above ranges and conditions are well known in the art.
In addition to utilizing hybridization studies to assess sequence identity or sequence homology, known computer programs may be used to determine whether a particular nucleic acid is homologous to a multi-epitope construct disclosed herein. An example of such a program is the Bestfit program (Wisconsin Sequence Analysis Package, Version 8 for Unix, Genetics Computer Group, University Research Park, 575 Science Drive, Madison, Wis. 53711), and other sequence alignment programs are known in the art and may be utilized for determining whether two or more nucleotide sequences are homologous. Bestfit uses the local homology algorithm of Smith and Waterman, Advances in Applied Mathematics 2:482-489 (1981), to find the best segment of homology between two sequences. When using Bestfit or any other sequence alignment program to determine whether a particular sequence is, for instance, 95% identical to a reference sequence, the parameters may be set such that the percentage of identity is calculated over the full length of the reference nucleotide sequence and that gaps in homology of up to 5% of the total number of nucleotides in the reference sequence are allowed.
The nomenclature used to describe peptide compounds follows the conventional practice wherein the amino group is presented to the left (the N-terminus) and the carboxyl group to the right (the C-terminus) of each amino acid residue. When amino acid residue positions are referred to in an epitope, they are numbered in an amino to carboxyl direction with position one being the position closest to the amino terminal end of the epitope, or the peptide or protein of which it may be a part. In the formulae representing selected specific embodiments of the present invention, the amino- and carboxyl-terminal groups, although not specifically shown, are in the form they would assume at physiologic pH values, unless otherwise specified. In the amino acid structure formulae, each residue is generally represented by standard three-letter or single-letter designations. The L-form of an amino acid residue is represented by a capital single letter or a capital first letter of a three-letter symbol, and the D-form for those amino acids having D-forms is represented by a lower case single letter or a lower case three letter symbol. Glycine has no asymmetric carbon atom and is simply referred to as “Gly” or G. Symbols for the amino acids are shown below.
Amino acid “chemical characteristics” are defined as: Aromatic (F,W,Y); Aliphatic-hydrophobic (L, I, V, M); Small polar (S, T, C); Large polar (Q, N); Acidic (D, E); Basic (R, H, K); Proline; Alanine; and Glycine.
Acronyms used herein are as follows:
This application may be relevant to U.S. Ser. No. 09/189,702 filed Nov. 10, 1998, currently pending, which is a CIP of U.S. Ser. No. 08/205,713 filed Mar. 4, 1994 now abandoned, which is a CIP of Ser. No. 08/159,184 filed Nov. 29, 1993 and now abandoned, which is a CIP of Ser. No. 08/073,205 filed Jun. 4, 1993 and now abandoned, which is a CIP of Ser. No. 08/027,146 filed Mar. 5, 1993 and now abandoned. The present application is also related to U.S. Ser. No. 09/226,775 now abandoned, which is a CIP of U.S. Ser. No. 08/815,396 now abandoned, which claims the benefit of U.S. Ser. No. 60/013,113 , now abandoned. Furthermore, the present application is related to U.S. Ser. No. 09/017,735 now abandoned, which is a CIP of abandoned U.S. Ser. Nos. 08/589,108; 08/753,622 now abandoned, U.S. Ser. No. 08/822,382 now abandoned, abandoned U.S. Ser. Nos. 60/013,980, 08/454,033 now abandoned, U.S. Ser. No. 09/116,424 now abandoned, and U.S. Ser. No. 08/349,177 now abandoned. The present application is also related to U.S. Ser. No. 09/017,524 now abandoned, U.S. Ser. No. 08/821,739 now abandoned, abandoned U.S. Ser. Nos. 60/013,833, 08/758,409 now abandoned, U.S. Ser. No. 08/589,107 now abandoned, U.S. Ser. No. 08/451,913 now abandoned, U.S. Ser. No. 08/186,266 now U.S. Pat. No. 5,662,907, U.S. Ser. No. 09/116,061 now abandoned, and U.S. Ser. No. 08/347,610 now abandoned, which is a CIP of U.S. Ser. No. 08/159,339 now U.S. Pat. No. 6,037,135, which is a CIP of abandoned U.S. Ser. No. 08/103,396, which is a CIP of abandoned U.S. Ser. No. 08/027,746, which is a CIP of abandoned U.S. Ser. No. 07/926,666. The present application may also be relevant to U.S. Ser. No. 09/017,743 now abandoned, U.S. Ser. No. 08/753,615 now abandoned; U.S. Ser. No. 08/590,298 now abandoned, U.S. Ser. No. 09/115,400 now abandoned, and U.S. Ser. No. 08/452,843 now abandoned, which is a CIP of U.S. Ser. No. 08/344,824 now abandoned, which is a CIP of abandoned U.S. Ser. No. 08/278,634. The present application may also be related to provisional U.S. Ser. No. 60/087,192 now abandoned and U.S. Ser. No. 09/009,953 now abandoned, which is a CIP of abandoned U.S. Ser. No. 60/036,713 and abandoned U.S. Ser. No. 60/037,432. In addition, the present application may be relevant to U.S. Ser. No. 09/098,584 now abandoned, and U.S. Ser. No. 09/239,043 now U.S. Pat. No. 6,689,363. The present application may also be relevant to co-pending U.S. Ser. No. 09/583,200 filed May 30, 2000 currently pending, U.S. Ser. No. 09/260,714 filed Mar. 1, 1999 now abandoned, and U.S. Provisional Application 60/239,008, now abandoned. All of the above applications are incorporated herein by reference.
The invention is described in detail below with reference to the figures wherein like elements are referenced with like numerals throughout. The invention relates to a method and system of designing multi-epitope vaccines with optimized immunogenicity. In preferred embodiments, the vaccine comprises CTL and HTL epitopes. Vaccines in accordance with the invention allow for significant, non-ethnically biased population coverage, and can preferably focus on epitopes conserved amongst different viral or other antigenic isolates. Through the method and system disclosed herein, vaccines can be optimized with regard to the magnitude and breadth of responses, and can allow for the simplest epitope configuration. Finally, general methods are provided to evaluate immunogenicity of a multi-epitope vaccine in humans.
The method of the invention comprises designing a multi-epitope construct based on principles identified herein. In one aspect, the invention provides for simultaneous induction of responses against specific CTL and HTL epitopes, using single promoter minigene vaccines. Such minigene constructs can contain many different epitopes, preferably greater than 10, often greater than 20, 25, 30, 25, 40, 45, 50, 55, 60, 65, 70, or more.
In a preferred embodiment, a computer system identifies one or more optimal multi-epitope constructs for a minigene by performing the following functions and/or analyses:
(i) the epitopes to be incorporated into the multi-epitope construct are sorted to provide an order that minimizes the number of junctional epitopes formed. A more detailed discussion of this sorting procedure is provided below with reference to
(ii) flanking residues that enhance immunogenicity may be inserted at the flanking positions of epitopes. In particular embodiments, flanking residues are inserted at the C+1 position of CTL epitopes.
(iii) spacer sequences may be inserted between epitopes to prevent occurance of junctional epitopes. In particular embodiments, the spacer sequences can also include a residue that promotes immunogenicity at the N-terminus of the linker such that the residue flanks the C-terminus of a CTL epitope.
In particular embodiments to prevent HTL junctional epitopes, a spacer composed of amino acid residues that do not correspond to any known HLA Class II anchor residue, are used, e.g, alternating G and P residues (a GP spacer) is included between two HTL epitopes.
Another aspect of the invention, (consideration (ii) above) involves the introduction or substitution of particular amino acid residues at positions that flank epitopes, e.g., a position immediately adjacent to the C-terminus of the epitope, thereby generating multi-epitope constructs with enhanced antigenicity and immunogenicity compared to constructs that do not contain the particular residue introduced or substituted at that site, i.e., non-optimized minigenes. The methods of optimizing multi-epitope constructs comprise a step of introducing a flanking residue, preferably K, N, G, R, or A at the C+1 position of the epitope, i.e., the position immediately adjacent to the C-terminus of the epitope. In an alternative embodiment, residues that contribute to decreased immunogenicity, i.e., negatively charged residues, e.g., D, aliphatic residues (I, L, M, V) or aromatic non-trytophan residues, are replaced. The flanking residue can be introduced by positioning appropriate epitopes to provide the favorable flanking residue, or by inserting a specific residue.
As noted in the background section, minigenes encoding up to 10 epitopes have been used to induce responses against a number of different epitopes. The data relating to an experimental minigene, pMin .1 has been published (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)). Disclosed herein, are parameters for designing and evaluating multi-epitope constructs with optimized immunogenicity that address myriad disease indications of interest.
Design parameters were identified based on a number of studies. In a preliminary evaluation of multi-epitope constructs, data on three different multi-epitope constructs, incorporating 20 to 25 different CTL epitopes each, are presented (
Thus, eleven days after a single i.m. DNA vaccine injection, responses against 8 to 14 different representative epitopes were evaluated following a single six day in vitro restimulation, utilizing assays to measure CTL activity (either chromium release or in situ IFN production, as described herein). Priming of epitope specific CTL could be demonstrated for 6/8 (75%), 10/14 (72%) and 13/14 (93%) of the epitopes tested in the case of HIV-1, HCV1 and HCV2, respectively. Thus, multi-epitope minigenes, capable of simultaneously priming CTL responses against a large number of epitopes, can be readily designed. However, it should be emphasized that CTL priming for some epitopes was not detected and, in several of the 36 cases considered, responses were infrequent, or varied significantly in magnitude over at least three orders of magnitude (1000-fold). These results strongly suggested that a more careful analysis and optimization of the minigene constructs was required to.
The possibility that the suboptimal performance of priming for certain epitopes might be related to minigene size was also examined. In fact, most of the published reports describe minigenes of up to ten epitopes, and the few instances in which 20-epitope minigenes have been reported, activity directed against only two or three epitopes was measured. To address this possibility, two smaller minigenes (HIV-1.1 and HIV-1.2) each encompassing ten epitopes, and corresponding to one half of the HIV-1 minigene, were synthesized and tested. Responses against four representative epitopes were measured.
1)Represents the fraction of independent cultures yielding positive responses
2)Lytic Units (LU)
3)Not Applicable
It was found that the responses induced by the smaller minigenes were comparable, and if anything, lower than those induced by the twenty-epitope construct (Table 1.) Accordingly, factors relating to minigene size are unlikely explanations for the observed suboptimal priming to certain epitopes and thus other parameters, disclosed herein, are used to design efficacious multi-epitope constructs.
The Minimization of Junctional Motifs
One of the considerations in designing multi-epitope constructs is the inadvertent creation of junctional epitopes when placing epitopes adjacent to each other. The presence of such epitopes in a minigene could significantly affect minigene performance. Strategies to guard against this undesired effect are disclosed herein for application to the development of multi-epitope or minigene vaccines. Junctional epitopes can first be minimized by sorting the epitopes to identify an order in which the numbers of junctional epitopes is minimized. Such a sorting procedure can be performed using a computer or by eye, if necessary, or depending on the number of epitopes to be included in the multi-epitope construct.
For example, a computer program that finds patterns, e.g., Panorama, manufactured by ProVUE Development, Huntington Beach, Calif., U.S.A., can be used in accordance with one embodiment of the invention. A very large number of different epitope arrangements can be considered in designing a particular minigene construct. A computer program accepts as input, the particular set of epitopes considered, and the motifs to be scanned in order to evaluate whether there are any junctional epitopes bearing these motifs. For example, a program can simulate building a minigene, and in an euristic computational algorithm, examine epitope pairs to avoid or minimize the occurrance of junctional motifs. The program can for example, evaluate 6×105 (about half a million) minigene configurations/second.
A complete analysis of a 10-epitope construct using a computer program as described in the preceding paragraph requires examining 10 factorial≅3.6×106 combinations and can be completed in six seconds. A fourteen-epitope construct can be completely analyzed in a couple of days. Thus, analysis time goes up very rapidly as larger minigenes are considered. However, a complete analysis is not always required and the program can be run for any desired length of time. In either case, the computer system of the present invention identifies and provides at least one configuration having a minimum number of junctional epitopes.
An example of the results of this type of approach is presented in Table 2. The number of junctional motifs in ten different random assortments of the same epitopes contained in the HCV1 minigene, which incorporates 25 epitopes, and is the result of a two day computer analysis, is presented in this Table. In the non-optimized assortments, a large number of A2, A11 and Kb motifs were found, in the 25 to 38 range, with an average of 31. By comparison, only two such junctional motifs are present in the HCV1 minigene assortment. In conclusion, a computer program can be utilized to effectively minimize the number of junctional motifs present in minigene constructs.
Eliminating Class II Junctional Epitopes and Testing for Class II Restricted Responses In Vivo
As a further element in eliminating junctional epitopes, spacer sequences can be inserted between two epitopes that create a junctional epitope when juxtaposed.
In one embodiment, to correct the problem of junctional epitopes for HTL epitopes, a spacer of, for example, five amino acids in length is inserted between the two epitopes. The amino acid residues incorporated into such a spacer are preferably those amino acid residues that are not known to be primary anchor residues for any of the HLA Class II binding motifs. Such residues include G, P, and N. In a preferred embodiment, a spacer with the sequence GPGPG (SEQ ID NO: 369) is inserted between two epitopes. Previous work has demonstrated that the GP spacer is particularly effective in disrupting Class II binding interactions (Sette et al., J Immunol., 143:1268-73 (1989)). All known human Class II binding motifs and the mouse IAb (the Class II expressed by HLA transgenic mice) do not tolerate either G or P at this main anchor positions, which are spaced four residues apart. This approach virtually guarantees that no Class II restricted epitopes can be formed as junctional epitopes.
In an example validating this design consideration, we synthesized polypeptides incorporating HIV-derived HTL epitopes. These epitopes are broadly cross-reactive HLA DR binding epitopes. It was then determined that these epitopes also efficiently bind the murine IAb Class II molecule. A diagram illustrating the two different synthetic polypeptides considered is shown in
The first construct incorporates four different epitopes linearly arranged, while the second construct incorporates the GPGPG (SEQ ID NO: 369) spacer. Synthetic peptides corresponding to the three potential junctional epitopes were also synthesized.
The capacity of 2 nanomoles of these different constructs to prime for proliferative responses to the various epitopes in IAb positive mice was tested, and compared to the responses induced by equimolar amounts of a pool of the same peptides (3 micrograms of each peptide). Specifically, groups of 3 mice were injected with peptides in CFA emulsions, 11 days after injection their lymph node cells were cultured in vitro for an additional 3 days, and thymidine incorporation was measured in the last 24 hours of culture. It was found (
These results demonstrate that responses against multiple HIV-derived Class II epitopes can be simultaneously induced, and also illustrate how IAb/DR crossreactivity can be utilized to investigate the immunogenicity of various constructs incorporating HTL epitope candidates. Finally, they demonstrate that appropriate spacers can be employed to effectively disrupt Class II junctional epitopes that would otherwise interfere with effective vaccine immunogenicity.
In the case of Class I restricted responses, one case of a naturally occurring junctional epitope and the consequent inhibition of epitope specific responses has been presented by McMichael and coworkers (Tussey et al., Immunity, Vol. 3(1):65-77 (1995)). To address the problem of junctional epitopes for Class I, similar analyses can be performed. For example, a specific computer program is employed to identify potential Class I restricted junctional epitopes, by screening for selected murine motifs and for the most common human Class I HLA A and B motifs.
Spacer sequences can also similarly be employed to prevent CTL junctional epitopes. Often, very small residues such as A or G are preferred spacer residues. G also occurs relatively infrequently as a preferred primary anchor residue (see, e.g., PCT/US00/24802) of an HLA Class I binding motif. These spacers can vary in length, e.g., spacer sequences can typically be 1, 2, 3, 4, 5, 6, 7, or 8 amino acid residues in length and are sometimes longer. Smaller lengths are often preferred because of physical constraints in producing the multi-epitope construct.
The Influence of Flanking Regions on CTL Minigene Immunogenicity
Another factor to be considered in designing minigenes is to insert residues that favor immunogenicity at the position flanking the C-terminus of a CTL epitope.
Disclosed herein are studies that identify residues that increase immunogenicity and, accordingly, residues that are inserted in multi-epitope constructs to optimize immunogenicity.
The molecular context in which an epitope was expressed often dramatically influenced the frequency and/or magnitude of priming of CTL specific for that epitope in HLA transgenic mice. Two examples are shown in Table 3.
1IFNγ secretory units
The immunogenicity of the HBV Core 18 epitope expressed in the pMin5 minigene was approximately 200-fold lower in magnitude than that observed in the case of the pMin1 minigene. Similarly, the immunogenicity of the HCV Core 132 epitope expressed in the context of the HCV1 minigene was marginal, with significant T cell priming demonstrable in only 2 of 12 different independent CTL experiments/cultures performed. These two positive experiments yielded responses of approximately 100 SU of IFNγ. However, when the same epitope was expressed in the context of the HCV2 minigene, positive responses were observed in 17/18 cases, and with average magnitudes approximately five-fold higher.
Immunogenicity of HIV-FT in HLA-A*0201/Kb Transgenic Mice
An HIV multi-epitope DNA vaccine, HIV-FT (
The ability of HIV-FT to prime recall CTL responses in vivo was evaluated by intramuscular immunization of HLA-A*0201/Kb transgenic mice. Splenocytes from animals immunized with 100 μg of HIV-FT plasmid DNA were stimulated with each of the HLA-A*0201 epitopes encoded in HIV-FT and assayed for peptide-specific CTL activity after six days of culture. Representative CTL responses against three of the epitopes in HIV-FT are shown in
Correlation Between Epitope Immunogenicity and Levels of HIV-FT Epitope Presentation in Transfected Cell Lines
The differential immunogenicity of the HLA-A*0201 epitopes in HIV-FT was then assessed. Differential MHC binding affinity could be excluded as all of the epitopes bind HLA-A*0201 with high affinity (
To test this, Jurkat cells, a human T cell line, expressing the HLA-A*0201/Kb gene (Vitiello et al., J. Exp. Med. 173, 1007-1015, 1991) were transfected with the HIV-FT expressed in an episomal vector. A human cell line was selected for use to eliminate any possible artifacts that may be associated with potential differences in the processing capabilities between humans and mice. This transfected cell line matches the human MHC presentation with human antigen processing capabilities and provides support for the subsequent development of CTL epitope-based DNA vaccines for use in humans.
Peptide-specific CTL lines detected presentation in the transfected targets of four of the HLA-A*0201 epitopes encoded in the HIV-FT, Pol 498, Env 134, Pol 448 and Nef 221. To quantitate the level at which each of these epitopes was produced and presented, the CTL lines specific for the various epitopes were incubated with untransfected targets and variable amounts of each epitope or peptides. These CTL dose response curves were utilized as standard curves to determine the peptide concentration inducing levels of IFNγ secretion equivalent to those observed in response to the HIV-FT transfected target cells. This value is referred to as a “peptide equivalent dose” and taken as a relative measure of the amount of epitope presented on the transfected cell.
Table 4 summarizes the findings of this analysis for eight of the HLA-A*0201 epitopes encoded in the HIV-FT. Peptide equivalent doses varied from a high of 33.3 ng/ml for Nef 221 to less than 0.4 ng/ml peptide equivalents for epitopes Gag 271, Gag 386 and Pol 774. Cumulatively these results indicate that in human cells lines transfected with HIV-FT there is at least a 100-fold variation exists in the levels of presentation of the different HLA-A*0201 restricted epitopes. All of the epitopes that were presented at detectable levels in antigenicity assays were also immunogenic in vivo. The only epitope that was immunogenic and not antigenic was Gag 271. In this case, immunization of HLA-A*0201/Kb transgenic mice with HIV-FT induced a weak CTL response in less than a third of the cultures tested. The other two epitopes, which were presented below the limit of sensitivity for the antigenicity analysis, Gag 386 and Pol 774, were non-immunogenic. In conclusion these results suggest that the heterogeneity in CTL responses induced by HIV-FT immunization can at least in part be attributed to suboptimal epitope presentation.
1magnitude expressed as LU (ref); the correlation coefficient relative to peptide equivalents R + 0.44
2frequency of positive cultures (number cultures > 2LU/total tested); the correlation coefficient relative to peptide equivalents R + 0.8.
3magnitude expressed in ng/ml
4number of independent experiments
Flanking Amino Acids Influence CTL Epitope Immunogenicity in Vivo Following Vaccination
As described herein, the particular amino acids flanking individual CTL epitopes is one factor that influences or enhances the efficiency with which an epitope is processed by altering the susceptibility of the antigen to proteolytic cleavage. To examine the influence of flanking amino acids on epitope immunogenicity, immunogenicity data was obtained from HLA-A*0201,-A*1101 and -B*0701 transgenic mice immunized with a number of unrelated experimental multi-epitope DNA constructs encoding minimal CTL epitopes without intervening sequences. A database representing 94 different epitope/flanking residue combinations was compiled to determine the possible influence the immediately flanking amino acids on epitope immunogenicity. A given epitope and flanking amino acid combination was included only once to prevent artificial skewing of the analysis because of redundancies. Epitope immunogenicity in HLA transgenic was considered optimal if greater than 100 SU or 20 LU in at least 30% of the cultures measured. CTL responses were typically scored in one of four categories: (+++), outstanding-more than 200 LU or 1000 SU; (++), good-20-200 LU or 100-1000 SU; (+), intermediate-2 to 20 LU or 10 to 100 SU; and (+/−), weak or negative-less than 2 LU or 10 SU. The numbers of optimal versus sub-optimal responses were categorized based on the chemical type of amino acid in the flanking positions and the significance of differences were determined using a chi-square test.
This analysis did not find any associations between the type of amino acids present at the amino-terminus of an epitope and immunogenicity. However, significant effects of the carboxyl-terminus flanking residue, the C+1 residue, were identified. Positively charged amino acids, K or R were most frequently associated with optimal CTL responses, a frequency of 68% (
Direct Evaluation of the Effect of the C1 Residue on Epitope Immunogenicity
To directly evaluate the effect of preferred versus deleterious types of amino acids in the C+1 flanking position, two multi-epitope constructs, referred to as HBV.1 and HBV.2 (
For HBV.1, the HIV-1 epitope directly following the highly immunogenic HBV Core 18 epitope was replaced with the HBV Pol 562 epitope. This altered the C+1 residue from a K to an F. The second construct, HBV.2, was produced by the insertion of an additional epitope, HBV Pol 629, between the HBV Core 18 and Pol 562 epitopes; a change that replaced the C+1 amino acid with a K residue. When the immunogenicity of the Core 18 epitope presented in these different contexts was evaluated in HLA-A*0201/Kb transgenic mice, it was determined that the Core 18 epitope was virtually non-immunogenic in HBV.1 but strongly immunogenic in HBV.2 (
To further test the effects of the C+1 flanking amino acid on CTL epitope immunogenicity, a set of constructs that differ from HBV.1 by the insertion of single amino acids at the C+1 position relative to the Core 18 epitope (
Variations in CTL Epitope Immunogenicity Are Correlated with the Amount Presented
If the variation of the immunogenicity of Core 18 associated with different C+1 residues was the result of differential sensitivity to proteolytic cleavage then large differences in the levels of epitope presentation should be detectable in different constructs. To test this, Jurkat cells, expressing the same HLA-A*0201/Kb gene expressed in the transgenic mice, were transfected with an episomal vector expressing either HBV.1 or HBV.1K. The Core 18 epitope was presented at >105 higher levels when a K was in the C+1 position, compared to the presence of an F in the same position (
In summary, for flanking residues, it was found that either very small residues such as A, C or G, or large residues such as Q, W, K, or R were generally associated with good or outstanding responses. The absence of a C+1 residue because of a stop codon in the minigene, or the presence of intermediate size residues such as S or T was associated with a more intermediate response pattern. Finally, in the case of a negatively charged residue, D; aliphatic (V, I, L, M) or aromatic-non tryptophan residues (Y, F), relatively poor responses were observed. These results show that the particular residue flanking the epitope's C-terminus can dramatically influence the response frequency and magnitude. Flanking residues at the C+1 position can also be introduced in combination with spacer sequences. Thus, a residue that favors immunogenicity, preferably, K, R, N, A, or G, is included as a flanking residue of a spacer.
Sorting and Optimization of Multi-epitope Constructs
To develop multi-epitope constructs using the invention, the epitopes for inclusion in the multi-epitope construct are sorted and optimized using the parameters defined herein. Sorting and optimization can be performed using a computer or, for fewer numbers of epitopes, not using a computer.
Computerized optimization can typically be performed as follows. The following provides an example of a computerized system that identifies and optimizes, e.g., provides for a minimal number of junctional epitopes and a maximal number of flanking residues, epitope combinations.
As shown in
In a preferred embodiment, the computer system 100 stores a software program, e.g., object code, in the hard drive memory of the computer system 100.This object code is executed by the CPU for performing the functions described herein. One component, or module, of the software program carries out the function of analyzing and identifying junctional epitopes at the peptide junctions of the polypeptide minigene as well as evaluating combinations of spacer and flanking residues at these junctions. This software module is referred to herein as the “Junctional Analyzer” module or program. In a preferred embodiment, the Junctional Analyzer analyzes and processes peptides entered by a user in accordance with other criteria, data and operating parameters described below.
In a preferred embodiment, a database containing different epitope/flanking residue combinations is stratified on the basis of epitope immunogenicity and the number of optimal versus suboptimal responses are sorted to rank the amino acids and assign enhancement weight values. The text file also contains a set of motifs 206 to use in detecting junctional epitopes. In a preferred embodiment, the user may also enter a maximum number of amino acids (spacers and flanking) to insert between each pair of peptides (MaxInsertions) 208 to function as spacers and/or flanking residues. Other parameters, values or commands (collectively referred to herein as “parameters”) to control the operation of the program may also be entered such as, for example: “OutputToScreen (Y/N)” 210; “OutputToFile (Y/N)” 212; the minimum function value to accept as a valid result (“MinimumAccepted”) 214; the maximum number of results having the same function value (“MaxDuplicateFunctionValue”) 216; the maximum time allowed for a search in minutes (“SearchTime”) 218; whether an Exhaustive Search is desired (“Exhaustive=Y/N”) 220; the number of Stochastic search probes (“NumStochasticProbes”) 222; the maximum number of hits allowed per single probe during a stochastic search (“MaxHitsPerProbe”) 224; and whether the start of each probe should be random or other (“RandomProbeStart(Y/N)”) 226. These parameters are provided for purposes of illustration only. Other parameters to control the operation and output format of the program may be entered as would be obvious to those of ordinary skill in the art.
The motifs 206 in the text file 200 provide a “mask” or structural model for identifying junctional epitopes. For example the first motif 206a shown in
In a preferred embodiment, the information contained in the output table 404 is used to perform either an “Exhaustive J Search” or a “Stochastic J Search” to identify a polypeptide construct linking all eleven peptides, including optimum spacer combinations. For eleven peptides, for example, there will be ten junctions. Therefore the permutation which yields the largest sum of function values taking into account all ten junctions is identified as the “optimum” permutation(s) of the multi-epitope constructs. In one embodiment, for the convenience of the user, the output text file 400 will also contain the original list of peptides/epitopes 202, the weight values used 204, the motifs used 206, and MaxInsertion value 208, and other parameter settings 402 entered into the input text file 200 of
The “Exhaustive J Search” looks at all permutations of the peptides and selects the ones that have the largest function sum. However, due to the factorial nature of permutations, as the number of peptides to be processed increases, the time required to complete an Exhaustive J Search increases almost exponentially. For example, using a standard Macintosh 333 MHz computer, the estimated running time for 13 peptides is approximately 2.9 hours and would be approximately 40 hours for 14 peptides. The “Stochastic J Search” is designed to search many areas of the permutation sequence, rather than the entire permutation space, and report the best function sum that it finds. By reporting only permutations that meet or exceed the current maximum function total, it is possible to search a much broader area of the permutation sequence. This technique has been successful with as many as 20 peptides. The time to perform an exhaustive search of 20 peptides is estimated to be on the order of 1.3×105 years.
Referring again to
If a Stochastic search has been selected, next, at step 311, the program begins the Stochastic search by initiating a probe. Next, at step 313, the program determines if the maximum search time per probe has been exceeded. If the maximum search time has not been reached, next, at step 315, the program determines whether a single probe has reached or exceeded the maximum number of “hits” per probe. In one embodiment, a probe hit is registered when a permutation's function value sum is the same as or greater than the largest function sum previously registered for one or more previously analyzed permutations. If the maximum number of hits per probe has not been reached, then, at step 317, the current stochastic probe evaluates the next permutation or set of permutations and the process returns step 313. If at step 315 it is determined that the maximum number of hits per probe has been reached or exceeded, then, the program proceeds to step 319, where the program determines whether a maximum number of probes have already been executed. Also, if at step 313, it is determined that the maximum time limit per probe has been reached or exceeded, the program proceeds to step 319 to determine if the maximum number of probes have been completed. If, at step 319, it is determined that the maximum number of probes has not been reached, the program returns to step 311 and a new search probe is initiated. If at step 319 it is determined that the maximum number of probes have been executed, the program then proceeds to step 323 where it outputs the best set of optimum permutations identified up to that point. This “best set” may consist of only those permutations having the highest function sum or, alternatively, may consist of the permutations having the top three highest function sums, for example, or any other output criteria desired by the user.
In one preferred embodiment, if a probe has received a maximum number of hits specified per probe, any unused time for that probe is divided by the remaining probes to decide how much time should be allocated to each of the remaining probes. In other words, if a probe terminates early because of finding too many hits then the remaining probes are allocated more time. Such functionality is easily implemented by those of ordinary skill in the computer programming arts.
If at step 309, an Exhaustive search has been selected, then, at step 321, an exhaustive search is initiated which analyzes every permutation, as described above. At the completion of the Exhaustive analysis, the program proceeds to step 323 where it outputs the “best set” of optimum permutations found. As mentioned above, this “best set” may include those permutations with the highest sum function values, or the top three highest sum function values, or permutations meeting any desired criteria specified by the user (e.g., top 30 permutations with the highest function values).
For each of the decision steps or determination steps discussed above (e.g., steps 313, 315 and 319), the program may be set to perform a query at periodic intervals (e.g., every five seconds) or, alternatively, the program may be set to perform a query after a specified number of permutations (e.g., five) have been analyzed or after every permutation has been analyzed. Any one of these operation and timing protocols is easily implemented and adjusted by those of ordinary skill in the art.
The Program output provides a list of the best arrangements of the epitopes. Since many permutations may have the same value of the evaluation function, several are generated so that other factors can be considered in choosing the optimum arrangement. Examples of multi-epitope constructs generated using the above-described computerized techniques are illustrated in
Macromolecular structures such as polypeptide structures can be described in terms of various levels of organization. For a general discussion of this organization, see, e.g., Alberts et al., Molecular Biology of the Cell (3rd ed., 1994) and Cantor and Schimmel, Biophysical Chemistry Part I: The Conformation of Biological Macromolecules (1980). “Primary structure” refers to the amino acid sequence of a particular peptide. “Secondary structure” refers to locally ordered, three dimensional structures within a polypeptide. These structures are commonly known as domains. Domains are portions of a polypeptide that form a compact unit of the polypeptide. Typical domains are made up of sections of lesser organization such as stretches of β-sheet and α-helices. “Tertiary structure” refers to the complete three dimensional structure of a polypeptide monomer. “Quaternary structure” refers to the three dimensional structure formed by the noncovalent association of independent tertiary units.
Structural predictions such as charge distribution, hydrophobic/hydrophilic region analysis, or folding predictions can be performed using sequence analysis programs known to those of skill in the art, for example, hydrophobic and hydrophilic domains can be identified (see, e.g., Kyte & Doolittle, J. Mol. Biol. 157:105-132 (1982) and Stryer, Biochemistry (3rd ed. 1988); see also any of a number of Internet based sequence analysis programs, such as those found at dot.imgen.bcm.tmc.edu.
A three-dimensional structural model of a multi-epitope construct can also be generated. This is generally performed by entering amino acid sequence to be analyzed into the computer system. The amino acid sequence represents the primary sequence or subsequence of the protein, which encodes the structural information of the protein. The three-dimensional structural model of the protein is then generated by the interaction of the computer system, using software known to those of skill in the art.
The amino acid sequence represents a primary structure that encodes the information necessary to form the secondary, tertiary and quaternary structure of the protein of interest. The software looks at certain parameters encoded by the primary sequence to generate the structural model. These parameters are referred to as “energy tenns,” and primarily include electrostatic potentials, hydrophobic potentials, solvent accessible surfaces, and hydrogen bonding. Secondary energy terms include van der Waals potentials. Biological molecules form the structures that minimize the energy terms in a cumulative fashion. The computer program is therefore using these terms encoded by the primary structure or amino acid sequence to create the secondary structural model. The tertiary structure of the protein encoded by the secondary structure is then formed on the basis of the energy terms of the secondary structure. The user can enter additional variables such as whether the protein is membrane bound or soluble, its location in the body, and its cellular location, e.g., cytoplasmic, surface, or nuclear. These variables along with the energy terms of the secondary structure are used to form the model of the tertiary structure. In modeling the tertiary structure, the computer program matches hydrophobic faces of secondary structure with like, and hydrophilic faces of secondary structure with like. Those multi-epitope constructs that are most readily accessible to the HLA processing apparatus are then selected.
Assessment of Immunogenicity of Multi-epitope Vaccines
The development of multi-epitope minigenes represents a unique challenge, because the species-specificity of the peptide binding to MHC. Different MHC types from different species tend to bind different sets of peptides (Rammensee et al., Immunogenetics, Vol. 41(4):178-228 (1995)). As a result, it is not possible to test in regular laboratory animals a construct composed of human epitopes. Alternatives to overcome this limitation are generally available. They include: 1) testing analogous constructs incorporating epitopes restricted by non-human MHC; 2) reliance on control epitopes restricted by non human MHC; 3) reliance on crossreactivity between human and non-human MHC; 4) the use of HLA transgenic animals; and 5) antigenicity assays utilizing human cells in vivo. The following is a brief overview of the development of the technology for analyzing antigenicity and immunogenicity.
Class I HLA Transgenics
The utility of HLA transgenic mice for the purpose of epitope identification (Sette et al., J Immunol, Vol. 153(12):5586-92 (1994); Wentworth et al., Int Immunol, Vol. 8(5):651-9 (1996); Engelhard et al., J Immunol, Vol. 146(4):1226-32 (1991); Man et al., Int Immunol, Vol. 7(4):597-605 (1995); Shirai et al., J Immunol, Vol. 154(6):2733-42 (1995)), and vaccine development (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)) has been established. Most of the published reports have investigated the use of HLA A2.1Kb mice but it should be noted that B*27, and B*3501 mice are also available. Furthermore, HLA A*11/Kb mice (Alexander et al., J Immunol, Vol. 159(10):4753-61 (1997)), and HLA B7/Kb and HLA A1/Kb mice have also been generated.
Data from 38 different potential epitopes was analyzed to determine the level of overlap between the A2.1-restricted CTL repertoire of A2.1/Kb-transgenic mice and A2.1+ humans (Wentworth et al., Eur J Immunol, Vol. 26(1):97-101 (1996)). In both humans and mice, an MHC peptide binding affinity threshold of approximately 500 nM correlates with the capacity of a peptide to elicit a CTL response in vivo. A high level of concordance between the human data in vivo and mouse data in vivo was observed for 85% of the high-binding peptides, 58% of the intermediate binders, and 83% of the low/negative binders. Similar results were also obtained with HLA A11 and HLA B7 transgenic mice (Alexander et al., J Immunol, Vol. 159(10):4753-61 (1997)). Thus, because of the extensive overlap that exists between T cell receptor repertoires of HLA transgenic mouse and human CTLs, transgenic mice are valuable for assessing immunogenicity of the multi-epitope constructs described herein.
The different specificities of TAP transport as it relates to HLA A11 mice does not prevent the use of HLA-A11 transgenic mice of evaluation of immunogenicity. While both murine and human TAP efficiently transport peptides with an hydrophobic end, only human TAP has been reported to efficiently transport peptides with positively charged C terminal ends, such as the ones bound by A3, A11 and other members of the A3 supertype. This concern does not apply to A2, A1 or B7 because both murine and human TAP should be equally capable of transporting peptides bound by A2, B7 or A1. Consistent with this understanding, Vitiello (Vitiello et al., J Exp Med, Vol. 173(4):1007-15 (1991)) and Rotzschke (Rotzschke O, Falk K., Curr Opin Immunol, Vol. 6(1):45-51 (1994)) suggested that processing is similar in mouse and human cells, while Cerundolo (Rotzschke O, Falk K., Curr Opin Immunol, Vol. 6(1):45-51 (1994)) suggested differences in murine versus human cells, both expressing HLA A3 molecules. However, using HLA A11 transgenics, expression of HLA molecules on T and B cells in vivo has been observed, suggesting that the reported unfavorable specificity of murine TAP did not prevent stabilization and transport of A11/Kb molecules in vivo (Alexander et al., J Immunol, Vol. 159(10):4753-61 (1997)). These data are in agreement with the previous observation that peptides with a charged C termini could be eluted from murine cells transfected with A11 molecules (Maier et al., Immunogenetics; Vol. 40(4):306-8 (1994)). Responses in HLA A11 mice to complex antigens, such as influenza, and most importantly to A11 restricted epitopes encoded by multi-epitope minigenes (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)) has also been detected. Thus, the TAP issue appears to be of minor concern with transgenic mice.
Another issue of potential relevance in the use of HLA transgenic mice is the possible influence of β2 microglobulin on HLA expression and binding specificity. It is well known that human β2 binds both human and mouse MHC with higher affinity and stability than mouse β2 microglobulin (Shields et al., Mol Immunol Vol. 35(14-15):919-28 (1998)). It is also well known that more stable complexes of MHC heavy chain and β2 facilitate exogenous loading of MHC Class I (Vitiello et al., Science, Vol. 250(4986):1423-6 (1990)). We have examined the potential effect of this variable by generating mice that are double transgenics for HLA/Kb and human β2. Expression of human β2 was beneficial in the case HLA B7/Kb mice, and was absolutely essential to achieve good expression levels in the case of HLA A1 transgenic mice. Accordingly, HLA/Kb and β2 double transgenic mice are currently and routinely bred and utilized by the present inventors. Thus, HLA transgenic mice can be used to model HLA-restricted recognition of four major HLA specificities (namely A2, A11, B7 and A1) and transgenic mice for other HLA specificities can be developed as suitable models for evaluation of immunogenicity.
Antigenicity Testing for Class I Epitopes
Several independent lines of experimentation indicate that the density of Class I/peptide complexes on the cell surface may correlate with the level of T cell priming. Thus, measuring the levels at which an epitope is generated and presented on an APC's surface provides an avenue to indirectly evaluate the potency of minigene vaccines in human cells in vitro. As a complement to the use of HLA Class I transgenic mice, this approach has the advantage of examining processing in human cells. (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999))
Several possible approaches to experimentally quantitate processed peptides are available. The amount of peptide on the cell surface can be quantitated by measuring the amount of peptide eluted from the APC surface (Sijts et al., J Immunol, Vol. 156(2):683-92 (1996); Demotz et al., Nature, Vol. 342(6250):682-4 (1989)). Alternatively, the number of peptide-MHC complexes can be estimated by measuring the amount of lysis or lymphokine release induced by infected or transfected target cells, and then determining the concentration of peptide necessary to obtain equivalent levels of lysis or lymphokine release (Kageyama et al., J Immunol, Vol. 154(2):567-76 (1995)).
A similar approach has also been used to measure epitope presentation in minigene-transfected cell lines. Specifically, minigene constructs that are immunogenic in HLA transgenic mice are also processed into optimal epitopes by human cells transfected with the same minigene, and the magnitude of the response observed in transgenic mice correlates with the antigenicity observed with the transfected human target cells (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)).
Using antigenicity assays, a number of related minigenes differing in epitope order or flanking residues can be transfected into APCs, and the impact of the aforementioned variables on epitope presentation can be evaluated. This can be a preferred system for testing where a relatively large number of different constructs need to be evaluated. Although it requires large numbers of epitope-specific CTLs protocols that allow for the generation of highly sensitive CTL lines (Alexander-Miller et al., Proc Natl Acad Sci U S A, Vol. 93(9):4102-7 (1996)) and also for their expansion to large numbers (Greenberg P. D., Riddell S. R., Science, Vol. 285(5427):546-51 (1999)) have been developed to address this potential problem.
It should also be kept in mind that, if the cell selected for the transfection is not reflective of the cell performing APC function in vivo, misleading results could be obtained. Cells of the B cell lineage, which are known “professional” APCs, are typically employed as transfection recipients. The use of transfected B cells of this type is an accepted practice in the field. Furthermore, a good correlation has already been noted between in vitro data utilizing transfected human B cells and in vivo results utilizing HLA transgenic mice, as described in more detail herein.
Measuring HTL Responses
In preferred embodiments, vaccine constructs are optimized to induce Class II restricted immune responses. One method of evaluating multi-epitope constructs including Class II epitopes, is to use HLA-DR transgenic mice. Several groups have produced and characterized HLA-DR transgenic mice (Taneja V., David C. S., Immunol Rev, Vol. 169:67-79 (1999)).
An alternative also exists which relies on crossreactivity between certain human MHC molecules and particular MHC molecules expressed by laboratory animals. Bertoni and colleagues (Bertoni et al., J Immunol, Vol. 161(8):4447-55 (1998)) have noted that appreciable crossreactivity can be demonstrated between certain HLA Class I supertypes and certain PATR molecules expressed by chimpanzees. Crossreactivity between human and macaques at the level of Class II (Geluk et al., J Exp Med, Vol. 177(4):979-87 (1993)) and Class I molecules (Dzuris, et al., J Immunol., July 1999) has also been noted. Finally, it can also be noted that the motif recognized by human HLA B7 supertype is essentially the same as the one recognized by the murine Class I Ld (Rammensee et al., Immunogenetics, Vol. 41(4):178-228 (1995)). Of relevance to testing HLA DR restricted epitopes in mice, it has been shown by Wall, et al (Wall et al., J Immunol., 152:4526-36 (1994)) that similarities exist in the motif of DR1 and IAb. We routinely breed our transgenic mice to take advantage of this fortuitous similarity. Furthermore, we have also shown that most of our peptides bind to IAb, so that we use these mice for the study of CTL and HTL immunogenicity.
Measuring and Quantitating Immune Responses from Clinical Samples
A crucial element to assess vaccine performance is to evaluate its capacity to induce immune responses in vivo. Analyses of CTL and HTL responses against the immunogen, as well as against common recall antigens are commonly used and are known in the art. Assays employed included chromium release, lymphokine secretion and lymphoproliferation assays.
More sensitive techniques such as the ELISPOT assay, intracellular cytokine staining, and tetramer staining have become available in the art. It is estimated that these newer methods are 10- to 100-fold more sensitive than the common CTL and HTL assays (Murali-Krishna et al., Immunity, Vol. 8(2):177-87 (1998)), because the traditional methods measure only the subset of T cells that can proliferate in vitro, and may, in fact, be representative of only a fraction of the memory T cell compartment (Ogg G.S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998)). Specifically in the case of HIV, these techniques have been used to measure antigen-specific CTL responses from patients that would have been undetectable with previous techniques (Ogg et al., Science, Vol. 279(5359):2103-6 (1998); Gray et al., J Immunol, Vol. 162(3):1780-8 (1999); Ogg et al., J Virol, Vol. 73(11):9153-60 (1999); Kalams et al., J Virol, Vol. 73(8):6721-8 (1999); Larsson et al., AIDS, Vol. 13(7):767-77 (1999); Come et al., J Acquir Immune Defic Syndr Hum Retrovirol, Vol. 20(5):442-7 (1999)).
With relatively few exceptions, direct activity of freshly isolated cells has been difficult to demonstrate by the means of traditional assays (Ogg G. S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998)). However, the increased sensitivity of the newer techniques has allowed investigators to detect responses from cells freshly isolated from infected humans or experimental animals (Murali-Krishna et al., Immunity, Vol. 8(2):177-87 (1998); Ogg G. S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998)). The availability of these sensitive assays, that do not depend on an in vitro restimulation step, has greatly facilitated the study of CTL function in natural infection and cancer. In contrast, assays utilized as an endpoint to judge effectiveness of experimental vaccines are usually performed in conjunction with one or more in vitro restimulation steps (Ogg G.S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998)). In fact, with few exceptions (Hanke et al., Vaccine, Vol. 16(4):426-35 (1998)), freshly isolated Class I-restricted CD8+T cells have been difficult to demonstrate in response to immunization with experimental vaccines designed to elicit CTL responses. The use of sensitive assays, such as ELISPOT or in situ IFNγ ELISA, have been combined with a restimulation step to achieve maximum sensitivity; MHC tetramers are also used for this purpose.
MHC tetramers were first described in 1996 by Altman and colleagues. They produced soluble HLA-A2 Class I molecules which were folded with HIV-specific peptides containing a CTL epitope complexed together into tetramers tagged with fluorescent markers. These are used to label populations of T cells from HIV-infected individuals that recognize the epitope (Ogg G. S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998)). These cells were then quantified by flow cytometry, providing a frequency measurement for the T cells that are specific for the epitope. This technique has become very popular in HIV research as well as in other infectious diseases (Ogg G.S., McMichael A. J., Curr Opin Immunol, Vol. 10(4):393-6 (1998); Ogg et al., Science, Vol. 279(5359):2103-6 (1998); Gray et al., J Immunol, Vol. 162(3):1780-8 (1999); Ogg et al., J Virol, Vol. 73(11):9153-60 (1999); Kalams et al., J Virol, Vol. 73(8):6721-8 (1999)). However, HLA polymorphism can limit the general applicability of this technique, in that the tetramer technology relies on defined HLA/peptide combinations. However, it has been shown that a variety of peptides, including HIV-derived peptides, are recognized by peptide-specific CTL lines in the context of different members of the A2, A3 and B7 supertypes (Threlkeld et al., J Immunol, Vol. 159(4):1648-57 (1997); Bertoni et al., J Clin Invest, Vol. 100(3):503-13 (1997)). Taken together these observations demonstrate that a T cell receptor (TCR) for a given MHC/peptide combination can have detectable affinity for the same peptide presented by a different MHC molecule from the same supertype.
In circumstances in which efficacy of a prophylactic vaccine is primarily correlated with the induction of a long-lasting memory response, restimulation assays can be the most appropriate and sensitive measures to monitor vaccine-induced immunological responses. Conversely, in the case of therapeutic vaccines, the main immunological correlate of activity can be the induction of effector T cell function, most aptly measured by primary assays. Thus, the use of sensitive assays allows for the most appropriate testing strategy for immunological monitoring of vaccine efficacy.
Antigenicity of Multi-epitope Minigenes in Transfected Human APC's
Antigenicity assays are performed to evaluate epitope processing and presentation in human cells. An episomal vector to efficiently transfect human target cells with epitope-based minigene vaccines is used to perform such an analysis.
For example, 221 A2Kb target cells were transfected with an HIV-1 minigene vaccine. The 221 A2Kb target cell expresses the A2Kb gene that is expressed in HLA transgenic mice, but expresses no endogenous Class I (Shimizu Y, DeMars R., J Immunol, Vol. 142(9):3320-8 (1989)). These transfected cells were assayed for their capacity to present antigen to CTL lines derived from HLA transgenic mice and specific for various HIV-derived CTL epitopes. To correct for differences in antigen sensitivity of different CTL lines, peptide dose titrations, using untransfected cells as APC, were run in parallel. Representative data is presented in
1)ng/ml;
2)number of independent experiments;
3)% of CTL cultures yielding positive results;
4)Lytic Units
By comparison, less than 25 pg/ml IFNY was detected utilizing the CTL specific for the Gag 271 epitope. The control peptide titration with untransfected target cells revealed that this negative result could not be ascribed to poor sensitivity of the particular CTL line utilized, because as little as 0.2 pg/ml of “peptide equivalents” (PE) could be detected. Thus, it appears that the Gag 271 epitope is not efficiently processed and presented in the HIV-1 transfected target cells. Utilizing the “peptide equivalents” figure as an approximate quantitation of the efficiency of processing, it can be estimated that at least 200-fold less Gag 271 is presented by the transfected targets, compared to the Pol 498 epitope.
The results of various independent determinations for four different epitopes contained within HIV-FT are compiled in Table 5. The amount of each epitope produced from the HIV-FT transfected cells ranged from 30.5 PE for Pol 498, to a low of less than 0.2 PE for Gag 271. The two epitopes Env 134 and Nef 221 were associated with intermediate values, of 6.1 and 2.1 PE, respectively.
These results were next correlated with the in vivo immunogenicity values observed for each epitope after immunization with the HIV-FT construct. The Pol 498 epitope was also the most immunogenic, as would be predicted. The Env 134 and Nef 221 epitopes, for which intermediate immunogenicity was observed in vivo, were also processed in vitro with intermediate efficiency by the transfected human cells. Finally, the Gag 271, for which no detectable in vitro processing was observed was also associated with in vivo immunogenicity suboptimal in both frequency and magnitude.
These data have several important implications. First, they suggest that different epitopes contained within a given minigene may be processed and presented with differential efficiency. Second, they suggest that immunogenicity is proportional to the amount of processed epitope generated. Finally, these results provide an important validation of the use of transgenic mice for the purpose of optimization of multi-epitope vaccines destined for human use.
III. Preparation of Multi-Epitope Constructs
Epitopes for inclusion in the multi-epitope constructs typically bear HLA Class I or Class II binding motifs as described, for example, in PCT applications PCT/US00/27766, or PCT/US00/19774.
Multiple HLA class II or class I epitopes present in a multi-epitope construct can be derived from the same antigen, or from different antigens. For example, a multi-epitope construct can contain one or more HLA epitopes that can be derived from two different antigens of the same virus or from two different antigens of different viruses. Epitopes for inclusion in a multi-epitope construct can be selected by one of skill in the art, e.g., by using a computer to select epitopes that contain HLA allele-specific motifs or supermotifs. The multi-epitope constructs of the invention can also encode one or more broadly cross-reactive binding, or universal, HLA class II epitopes, e.g., PADRE® (Epimmune, San Diego, Calif.), (described, for example, in U.S. Pat. No. 5,736,142) or a PADRE® family molecule.
Universal HLA Class II epitopes can be advantageously combined with other HLA Class I and Class II epitopes to increase the number of cells that are activated in response to a given antigen and provide broader population coverage of HLA-reactive alleles. Thus, the multi-epitope constructs of the invention can include HLA epitopes specific for an antigen, universal HLA class II epitopes, or a combination of specific HLA epitopes and at least one universal HLA class II epitope.
HLA Class I epitopes are generally about 8 to about 13 amino acids in length, in particular 8, 9, 10, or 11 amino acids in length. HLA Class II epitopes are generally about 6 to 25 amino acids in length, in particular about 13 to 21 amino acids in length. An HLA Class I or II epitope can be derived from any desired antigen of interest. The antigen of interest can be a viral antigen, surface receptor, tumor antigen, oncogene, enzyme, or any pathogen, cell or molecule for which an immune response is desired. Epitopes can be selected based on their ability to bind one or multiple HLA alleles. Epitopes that are analogs of naturally occuring sequences can also be included in the multi-epitope constructs described herein. Such analog peptides are described, for example, in PCT applications PCT/US97/03778, PCT/US00/19774, and co-pending U.S. Ser. No. 09/260,714 filed Mar. 1, 1999, now abandoned.
Multi-epitope constructs can be generated using methodology well known in the art. For example, polypeptides comprising the multi-epitope constructs can be synthesized and linked. Typically, multi-epitope minigenes are constructed using recombinant DNA technology.
IV. Expression Vectors and Construction of a Minigene
The multi-epitope constructs of the invention are typically provided as an expression vector comprising a minigene encoding the multi-epitope construct. Construction of such expression vectors is described, for example in PCT/US99/10646. The expression vectors contain at least one promoter element that is capable of expressing a transcription unit encoding the minigene in the appropriate cells of an organism so that the antigen is expressed and targeted to the appropriate HLA molecule. For example, for administration to a human, a promoter element that functions in a human cell is incorporated into the expression vector.
In preferred embodiments, the invention utilizes routine techniques in the field of recombinant genetics. Basic texts disclosing the general methods of use in this invention include Sambrook et al., Molecular Cloning, A Laboratory Manual (2nd ed. 1989); Kriegler, Gene Transfer and Expression: A Laboratory Manual (1990); and Current Protocols in Molecular Biology (Ausubel et al., eds., 1994); Oligonucleotide Synthesis: A Practical Approach (Gait, ed., 1984); Kuijpers, Nucleic Acids Research 18(17):5197 (1994); Duehohn, J Org. Chem. 59:5767-5773 (1994); Methods in Molecular Biology, volume 20 (Agrawal, ed.); and Tijssen, Laboratory Techniques in Biochemistry and Molecular Biology—Hybridization with Nucleic Acid Probes, e.g., Part I, chapter 2 “Overview of principles of hybridization and the strategy of nucleic acid probe assays” (1993)).
The nucleic acids encoding the epitopes are assembled in a minigene according to standard techniques. In general, the nucleic acid sequences encoding minigene epitopes are isolated using amplification techniques with oligonucleotide primers, or are chemically synthesized. Recombinant cloning techniques can also be used when appropriate. Oligonucleotide sequences are selected which either amplify (when using PCR to assemble the minigene) or encode (when using synthetic oligonucleotides to assemble the minigene) the desired epitopes.
Amplification techniques using primers are typically used to amplify and isolate sequences encoding the epitopes of choice from DNA or RNA (see U.S. Pat. Nos. 4,683,195 and 4,683,202; PCR Protocols: A Guide to Methods and Applications (Innis et al., eds, 1990)). Methods such as polymerase chain reaction (PCR) and ligase chain reaction (LCR) can be used to amplify epitope nucleic acid sequences directly from mRNA, from cDNA, from genomic libraries or cDNA libraries. Restriction endonuclease sites can be incorporated into the primers. Minigenes amplified by the PCR reaction can be purified from agarose gels and cloned into an appropriate vector.
Synthetic oligonucleotides can also be used to construct minigenes. This method is performed using a series of overlapping oligonucleotides, representing both the sense and non-sense strands of the gene. These DNA fragments are then annealed, ligated and cloned. Oligonucleotides that are not commercially available can be chemically synthesized according to the solid phase phosphoramidite triester method first described by Beaucage & Caruthers, Tetrahedron Letts. 22:1859-1862 (1981), using an automated synthesizer, as described in Van Devanter et. al., Nucleic Acids Res. 12:6159-6168 (1984). Purification of oligonucleotides is by either native acrylamide gel electrophoresis or by anion-exchange HPLC as described in Pearson & Reanier, J. Chrom. 255:137-149 (1983).
The epitopes of the minigene are typically subcloned into an expression vector that contains a strong promoter to direct transcription, as well as other regulatory sequences such as enhancers and polyadenylation sites. Suitable promoters are well known in the art and described, e.g., in Sambrook et al. and Ausubel et al. Eukaryotic expression systems for mammalian cells are well known in the art and are commercially available. Such promoter elements include, for example, cytomegalovirus (CMV), Rous sarcoma virus LTR and SV40.
The expression vector typically contains a transcription unit or expression cassette that contains all the additional elements required for the expression of the minigene in host cells. A typical expression cassette thus contains a promoter operably linked to the minigene and signals required for efficient polyadenylation of the transcript. Additional elements of the cassette may include enhancers and introns with functional splice donor and acceptor sites.
In addition to a promoter sequence, the expression cassette can also contain a transcription termination region downstream of the structural gene to provide for efficient termination. The termination region may be obtained from the same gene as the promoter sequence or may be obtained from different genes.
The particular expression vector used to transport the genetic information into the cell is not particularly critical. Any of the conventional vectors used for expression in eukaryotic cells may be used. Expression vectors containing regulatory elements from eukaryotic viruses are typically used in eukaryotic expression vectors, e.g., SV40 vectors, CMV vectors, papilloma virus vectors, and vectors derived from Epstein Bar virus.
The multi-epitope constructs of the invention can be expressed from a variety of vectors including plasmid vectors as well as viral or bacterial vectors. Examples of viral expression vectors include attenuated viral hosts, such as vaccinia or fowlpox. As an example of this approach, vaccinia virus is used as a vector to express nucleotide sequences that encode the peptides of the invention. Upon introduction into a host bearing a tumor, the recombinant vaccinia virus expresses the immunogenic peptide, and thereby elicits a host CTL and/or HTL response. Vaccinia vectors and methods useful in immunization protocols are described in, e.g., U.S. Pat. No. 4,722,848.
A wide variety of other vectors useful for therapeutic administration or immunization, e.g. adeno and adeno-associated virus vectors, retroviral vectors, non-viral vectors such as BCG (Bacille Calmette Guerin), Salmonella typhi vectors, detoxified anthrax toxin vectors, and the like, will be apparent to those skilled in the art.
Immunogenicity and antigenicity of the multi-epitope constructs are evaluated as described herein.
Targeting Sequences
The expression vectors of the invention may encode one or more MHC epitopes operably linked to a MHC targeting sequence, and are referred to herein as “targeting nucleic acids” or “targeting sequences.” The use of a MHC targeting sequence enhances the immune response to an antigen, relative to delivery of antigen alone, by directing the peptide epitope to the site of MHC molecule assembly and transport to the cell surface, thereby providing an increased number of MHC molecule-peptide epitope complexes available for binding to and activation of T cells.
MHC Class I targeting sequences can be used in the present invention, e.g., those sequences that target an MHC Class I epitope peptide to a cytosolic pathway or to the endoplasmic reticulum (see, e.g., Rammensee et al., Immunogenetics 41:178-228 (1995)). For example, the cytosolic pathway processes endogenous antigens that are expressed inside the cell. Although not wishing to be bound by any particular theory, cytosolic proteins are thought to be at least partially degraded by an endopeptidase activity of a proteasome and then transported to the endoplasmic reticulum by the TAP molecule (transporter associated with processing). In the endoplasmic reticulum, the antigen binds to MHC Class I molecules. Endoplasmic reticulum signal sequences bypass the cytosolic processing pathway and directly target endogenous antigens to the endoplasmic reticulum, where proteolytic degradation into peptide fragments occurs. Such MHC Class I targeting sequences are well known in the art, and include, e.g., signal sequences such as those from Ig kappa, tissue plasminogen activator or insulin. A preferred signal peptide is the human Ig kappa chain sequence. Endoplasmic reticulum signal sequences can also be used to target MHC Class II epitopes to the endoplasmic reticulum, the site of MHC Class I molecule assembly. MHC Class II targeting sequences can also be used in the invention, e.g., those that target a peptide to the endocytic pathway. These targeting sequences typically direct extracellular antigens to enter the endocytic pathway, which results in the antigen being transferred to the lysosomal compartment where the antigen is proteolytically cleaved into antigen peptides for binding to MHC Class II molecules. As with the normal processing of exogenous antigen, a sequence that directs a MHC Class II epitope to the endosomes of the endocytic pathway and/or subsequently to lysosomes, where the MHC Class II epitope can bind to a MHC Class II molecule, is a MHC Class II targeting sequence. For example, group of MHC Class II targeting sequences useful in the invention are lysosomal targeting sequences, which localize polypeptides to lysosomes. Since MHC Class II molecules typically bind to antigen peptides derived from proteolytic processing of endocytosed antigens in lysosomes, a lysosomal targeting sequence can function as a MHC Class II targeting sequence. Lysosomal targeting sequences are well known in the art and include sequences found in the lysosomal proteins LAMP-1 and LAMP-2 as described by August et al. (U.S. Pat. No. 5,633,234, issued May 27, 1997), which is incorporated herein by reference.
Other lysosomal proteins that contain lysosomal targeting sequences include HLA-DM. HLA-DM is an endosomal/lysosomal protein that functions in facilitating binding of antigen peptides to MHC Class II molecules. Since it is located in the lysosome, HLA-DM has a lysosomal targeting sequence that can function as a MHC Class II molecule targeting sequence (Copier et al., J Immunol. 157:1017-1027 (1996), which is incorporated herein by reference).
The resident lysosomal protein HLA-DO can also function as a lysosomal targeting sequence. In contrast to the above described resident lysosomal proteins LAMP-I and HLA-DM, which encode specific Tyr-containing motifs that target proteins to lysosomes, HLA-DO is targeted to lysosomes by association with HLA-DM (Liljedahl et al., EMBO J. 15:4817-4824 (1996)), which is incorporated herein by reference. Therefore, the sequences of HLA-DO that cause association with HLA-DM and, consequently, translocation of HLA-DO to lysosomes can be used as MHC Class II targeting sequences. Similarly, the murine homolog of HLA-DO, H2-DO, can be used to derive a MHC Class II targeting sequence. A MHC Class II epitope can be fused to HLA-DO or H2-DO and targeted to lysosomes.
In another example, the cytoplasmic domains of B cell receptor subunits Ig-α and Ig-β mediate antigen internalization and increase the efficiency of antigen presentation as described in, for example, Bonnerot et al., Immunity 3:335-347 (1995). Therefore, the cytoplasmic domains of the Ig-α and Ig-β proteins can function as MHC Class II targeting sequences that target a MHC Class II epitope to the endocytic pathway for processing and binding to MHC Class II molecules.
Another example of a MHC Class II targeting sequence that directs MHC Class II epitopes to the endocytic pathway is a sequence that directs polypeptides to be secreted, where the polypeptide can enter the endosomal pathway. These MHC Class II targeting sequences that direct polypeptides to be secreted mimic the normal pathway by which exogenous, extracellular antigens are processed into peptides that bind to MHC Class II molecules. Any signal sequence that functions to direct a polypeptide through the endoplasmic reticulum and ultimately to be secreted can function as a MHC Class II targeting sequence so long as the secreted polypeptide can enter the endosomal/lysosomal pathway and be cleaved into peptides that can bind to MHC Class II molecules.
In another example, the Ii protein binds to MHC Class IMHC Class II molecules in the endoplasmic reticulum, where it functions to prevent peptides present in the endoplasmic reticulum from binding to the MHC Class II molecules. Therefore, fusion of a MHC Class II epitope to the Ii protein targets the MHC Class II epitope to the endoplasmic reticulum and a MHC Class II molecule. For example, the CLIP sequence of the Ii protein can be removed and replaced with a MHC Class IMHC Class II epitope sequence so that the MHC Class II epitope is directed to the endoplasmic reticulum, where the epitope binds to a MHC Class II molecule.
In some cases, antigens themselves can serve as MHC Class II or I targeting sequences and can be fused to a universal MHC Class II epitope to stimulate an immune response. Although cytoplasmic viral antigens are generally processed and presented as complexes with MHC Class I molecules, long-lived cytoplasmic proteins such as the influenza matrix protein can enter the MHC Class IMHC Class II molecule processing pathway as described in, for example, Gueguen & Long, Proc. Natl. Acad. Sci. USA 93:14692-14697 (1996). Therefore, long-lived cytoplasmic proteins can function as a MHC Class IMHC Class II targeting sequence. For example, an expression vector encoding influenza matrix protein fused to a universal MHC Class IMHC Class II epitope can be advantageously used to target influenza antigen and the universal MHC Class IMHC Class II epitope to the MHC Class IMHC Class II pathway for stimulating an immune response to influenza.
Other examples of antigens functioning as MHC Class IMHC Class II targeting sequences include polypeptides that spontaneously form particles. The polypeptides are secreted from the cell that produces them and spontaneously form particles, which are taken up into an antigen-presenting cell by endocytosis such as receptor-mediated endocytosis or are engulfed by phagocytosis. The particles are proteolytically cleaved into antigen peptides after entering the endosomal/lysosomal pathway.
One such polypeptide that spontaneously forms particles is HBV surface antigen (HBV-S) as described in, for example, Diminsky et al., Vaccine 15:637-647 (1997) or Le Borgne et al., Virology 240:304-315 (1998). Another polypeptide that spontaneously forms particles is HBV core antigen as described in, for example, Kuhröber et al., International immunol. 9:1203-1212 (1997). Still another polypeptide that spontaneously forms particles is the yeast Ty protein as described in, for example, Weber et al., Vaccine 13:831-834 (1995). For example, an expression vector containing HBV-S antigen fused to a universal MHC Class IMHC Class II epitope can be advantageously used to target HBV-S antigen and the universal MHC Class IMHC Class II epitope to the MHC Class IMHC Class II pathway for stimulating an immune response to HBV.
Administration In Vivo
The invention also provides methods for stimulating an immune response by administering an expression vector of the invention to an individual. Administration of an expression vector of the invention for stimulating an immune response is advantageous because the expression vectors of the invention target MHC epitopes to MHC molecules, thus increasing the number of CTL and HTL activated by the antigens encoded by the expression vector.
Initially, the expression vectors of the invention are screened in mouse to determine the expression vectors having optimal activity in stimulating a desired immune response. Initial studies are therefore carried out, where possible, with mouse genes of the MHC targeting sequences. Methods of determining the activity of the expression vectors of the invention are well known in the art and include, for example, the uptake of 3H-thymidine to measure T cell activation and the release of 51Cr to measure CTL activity as described below in Examples II and III. Experiments similar to those described in Example IV are performed to determine the expression vectors having activity at stimulating an immune response. The expression vectors having activity are further tested in human. To circumvent potential adverse immunological responses to encoded mouse sequences, the expression vectors having activity are modified so that the MHC Class IMHC Class II targeting sequences are derived from human genes. For example, substitution of the analogous regions of the human homologs of genes containing various MHC Class IMHC Class II targeting sequences are substituted into the expression vectors of the invention. Expression vectors containing human MHC Class IMHC Class II targeting sequences, such as those described in Example I below, are tested for activity at stimulating an immune response in human.
The invention also relates to pharmaceutical compositions comprising a pharmaceutically acceptable carrier and an expression vector of the invention. Pharmaceutically acceptable carriers are well known in the art and include aqueous or non-aqueous solutions, suspensions and emulsions, including physiologically buffered saline, alcohol/aqueous solutions or other solvents or vehicles such as glycols, glycerol, oils such as olive oil or injectable organic esters.
A pharmaceutically acceptable carrier can contain physiologically acceptable compounds that act, for example, to stabilize the expression vector or increase the absorption of the expression vector. Such physiologically acceptable compounds include, for example, carbohydrates, such as glucose, sucrose or dextrans, antioxidants such as ascorbic acid or glutathione, chelating agents, low molecular weight polypeptides, antimicrobial agents, inert gases or other stabilizers or excipients. Expression vectors can additionally be complexed with other components such as peptides, polypeptides and carbohydrates. Expression vectors can also be complexed to particles or beads that can be administered to an individual, for example, using a vaccine gun. One skilled in the art would know that the choice of a pharmaceutically acceptable carrier, including a physiologically acceptable compound, depends, for example, on the route of administration of the expression vector.
The invention further relates to methods of administering a pharmaceutical composition comprising an expression vector of the invention to stimulate an immune response. The expression vectors are administered by methods well known in the art as described in, for example, Donnelly et al. (Ann. Rev. Immunol. 15:617-648 (1997)); Felgner et al. (U.S. Pat. No. 5,580,859, issued Dec. 3, 1996); Felgner (U.S. Pat. No. 5,703,055, issued Dec. 30, 1997); and Carson et al. (U.S. Pat. No. 5,679,647, issued Oct. 21, 1997). In one embodiment, the minigene is administered as naked nucleic acid.
A pharmaceutical composition comprising an expression vector of the invention can be administered to stimulate an immune response in a subject by various routes including, for example, orally, intravaginally, rectally, or parenterally, such as intravenously, intramuscularly, subcutaneously, intraorbitally, intracapsularly, intraperitoneally, intracistemally or by passive or facilitated absorption through the skin using, for example, a skin patch or transdermal iontophoresis, respectively. Furthermore, the composition can be administered by injection, intubation or topically, the latter of which can be passive, for example, by direct application of an ointment or powder, or active, for example, using a nasal spray or inhalant. An expression vector also can be administered as a topical spray, in which case one component of the composition is an appropriate propellant. The pharmaceutical composition also can be incorporated, if desired, into liposomes, microspheres or other polymer matrices as described in, for example, Felgner et al., U.S. Pat. No. 5,703,055; Gregoriadis, Liposome Technology, Vols. I to III (2nd ed. 1993). Liposomes, for example, which consist of phospholipids or other lipids, are nontoxic, physiologically acceptable and metabolizable carriers that are relatively simple to make and administer.
The expression vectors of the invention can be delivered to the interstitial spaces of tissues of an animal body as described in, for example, Felgner et al., U.S. Pat. Nos. 5,580,859 and 5,703,055. Administration of expression vectors of the invention to muscle is a particularly effective method of administration, including intradermal and subcutaneous injections and transdermal administration. Transdermal administration, such as by iontophoresis, is also an effective method to deliver expression vectors of the invention to muscle. Epidermal administration of expression vectors of the invention can also be employed. Epidermal administration involves mechanically or chemically irritating the outermost layer of epidermis to stimulate an immune response to the irritant (Carson et al., U.S. Pat. No. 5,679,647).
Other effective methods of administering an expression vector of the invention to stimulate an immune response include mucosal administration as described in, for example, Carson et al., U.S. Pat. No. 5,679,647. For mucosal administration, the most effective method of administration includes intranasal administration of an appropriate aerosol containing the expression vector and a pharmaceutical composition. Suppositories and topical preparations are also effective for delivery of expression vectors to mucosal tissues of genital, vaginal and ocular sites. Additionally, expression vectors can be complexed to particles and administered by a vaccine gun.
The dosage to be administered is dependent on the method of administration and will generally be between about 0.1 μg up to about 200 μg. For example, the dosage can be from about 0.05 μg/kg to about 50 mg/kg, in particular about 0.005-5 mg/kg. An effective dose can be determined, for example, by measuring the immune response after administration of an expression vector. For example, the production of antibodies specific for the MHC Class II epitopes or MHC Class I epitopes encoded by the expression vector can be measured by methods well known in the art, including ELISA or other immunological assays. In addition, the activation of T helper cells or a CTL response can be measured by methods well known in the art including, for example, the uptake of 3H-thymidine to measure T cell activation and the release of 51Cr to measure CTL activity (see Examples II and III below).
The pharmaceutical compositions comprising an expression vector of the invention can be administered to mammals, particularly humans, for prophylactic or therapeutic purposes. Examples of diseases that can be treated or prevented using the expression vectors of the invention include infection with HBV, HCV, HIV and CMV as well as prostate cancer, renal carcinoma, cervical carcinoma, lymphoma, condyloma acuminatum and acquired immunodeficiency syndrome (AIDS).
In therapeutic applications, the expression vectors of the invention are administered to an individual already suffering from cancer, autoimmune disease or infected with a virus. Those in the incubation phase or acute phase of the disease can be treated with expression vectors of the invention, including those expressing all universal MHC Class II epitopes, separately or in conjunction with other treatments, as appropriate.
In therapeutic and prophylactic applications, pharmaceutical compositions comprising expression vectors of the invention are administered to a patient in an amount sufficient to elicit an effective immune response to an antigen and to ameliorate the signs or symptoms of a disease. The amount of expression vector to administer that is sufficient to ameliorate the signs or symptoms of a disease is termed a therapeutically effective dose. The amount of expression vector sufficient to achieve a therapeutically effective dose will depend on the pharmaceutical composition comprising an expression vector of the invention, the manner of administration, the state and severity of the disease being treated, the weight and general state of health of the patient and the judgment of the prescribing physician.
The following examples are offered to illustrate, but not to limit the claimed invention. It is understood that the examples and embodiments described herein are for illustrative purposes only and that various modifications or changes in light thereof are suggested to persons skilled in the art and are to be included within the spirit and purview of this application and scope of the appended claims.
Examples 1-9 provide examples of assays for evaluating the immunogenicity and antigenicity of multi-epitope constructs.
Antigenicity Assays
High-affinity peptide-specific CTL lines can be generated from splenocytes of transgenic mice that have been primed with DNA, peptide/IFA, or lipopeptide. Briefly, splenocytes from transgenic mice are stimulated 0.1 μg/ml peptide and LPS blasts. Ten days after the initial stimulation, and weekly thereafter, cells are restimulated with LPS blasts pulsed for 1 hour with 0.1 μg/ml peptide. CTL lines are assayed 5 days following restimulation in an in situ IFNγ ELISA as described above. Alternatively, CTL lines that are derived from, e.g., patients infected with the targeted pathogen or who have the targeted disease, e.g., cancer, can be used. Specific CTL lines that are not available either from transgenic mice or from patients are generated from PBMC of normal donors, drawing on the expertise in the art.
Target cells to be used in these assays are Jurkat or 0.221 cells transfected with A2.1/Kb, A11/Kb, A1/Kb, or B7/Kb for CTL lines derived from transgenic mice. All these cell lines are currently available to us (Epimmune Inc., San Diego, Calif.). In the case of human CTL lines, 0.221 cells transfected with the appropriate human HLA allele are utilized. We currently have 0.221 cells transfected with A2 and A1, and are generating A11, A24 and B7 transfectants. In an alternative embodiment, if unforeseen problems arise in respect to target cells, LPS blasts and EBV-transformed lines are utilized for murine and human CTL lines, respectively.
To assay for antigenicity, serially diluted CTLs are incubated with 105 target cells and multiple peptide concentrations ranging from 1 to 10−6 μg/ml. In addition, CTLs are also incubated with target cells transfected with an episomal vector containing a minigene of interest. Episomal vectors are known in the art.
The relative amount of peptide generated by natural processing within the minigene-transfected APCs is quantitated as follows. The amount of IFNγ generated by the CTL lines upon recognition of the transfected target cells are recorded. The amount of synthetic peptide necessary to yield the same amount of IFNγ are interpolated from a standard curve generated when the same CTL line is incubated in parallel with known concentrations of peptide.
Mice, Immunizations and Cell Cultures
The derivation of the HLA-A2.1/Kb (Vitiello et al., J Exp Med, Vol. 173(4):1007-15 (1991)) and A11/Kb (Alexander et al., J Immunol, Vol. 159(10):4753-61 (1997)) transgenic mice used in this study has been described. HLA B7 Kb transgenic mice are available in house (Epimmune Inc., San Diego, Calif.). HLA DR2, DR3 and DR4 transgenic mice are obtained from C. David (Mayo Clinic). Non-transgenic H-2b mice are purchased from Charles River Laboratories or other commercial vendors. Immunizations are performed as described in (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)). All cells are grown in culture medium consisting of RPMI 1640 medium with HEPES (Gibco Life Technologies) supplemented with 10% FBS, 4 mM L-glutamine, 50 μM 2-ME, 0.5 mM sodium pyruvate, 100 μg/ml streptomycin and 100 U/ml penicillin.
HLA transgenic mice and antigenicity assays are used for the purpose of testing and optimization CTL responses. The natural crossreactivity between HLA-DR and IAb can also be exploited to test HTL responses. This evaluation provides an assessment of the antigenicity and immunogenicity of multi-epitope constructs.
Proliferation Assays
To assess the ability of HTL epitopes to induce an immune response, assays such as proliferation assays are often performed. For example, mouse CD4 T lymphocytes are immunomagnetically isolated from splenic single cell suspensions using DynaBeads Mouse CD4 (L3T4) (Dynal). Briefly, 2×107 spleen cells are incubated with 5.6×107 magnetic beads for 40 minutes at 4° C., and then washed 3 times. Magnetic beads are detached using DetachaBead Mouse CD4 (Dynal). Isolated CD4 T lymphocytes (2×105 cells/well) are cultured with 5×105 irradiated (3500 rad) syngeneic spleen cells in triplicate in flat-bottom 96-well microtiter plates. Purified peptides are added to wells at a final concentration of 20, 1, 0.05 and 0 μg/ml and cells are cultured for a total of 4 days. Approximately 14 hour before harvesting, 1 ,μCi of 3H-thymidine (ICN) is added to each well. The wells are harvested onto Unifilter GF/B plates (Packard) using the Filtermate Harvester (Packard). 3H-Thymidine incorporation is determined by liquid scintillation counting using the TopCount™ microplate scintillation counter (Packard).
51Chromium Release Assay
This assay to measure CTL activity is well known in the art. The assay quantifies the lytic activity of the T cell population by measuring the percent 51Cr released from a 51Cr-labeled target population (Brunner et al., Immunology, Vol. 14(2):181-96 (1968)). Data derived from the chromium release assay is usually expressed either as a CTL frequency/106 cell (limiting dilution analysis, LDA; (Current Protocols in Immunology, Vol 1, John Wiley & Sons, Inc., USA 1991 Chapter 3; Manual of Clinical Laboratory Immunology, Fifth edition, ASM Press, 1997 Section R), or by a less cumbersome quantitative assessment of bulk CTL activity (lytic Units; LU assay). In a LU assay, the standard E:T ratio versus percent cytotoxicity data curves generated in a 51Cr-release assay are converted into lytic units (LU) per 106 effector cells, with 1 LU defined as the lytic activity required to achieve 30% lysis of target cells (Wunderlick, J., Shearer, G., and Livingston, A. In: J. Coligan, A. Kruisbeek, D. Margulies, E. Shevach, and W. Strober (Eds.), Current Protocols in Immunology, Vol 1, “Assays for T cell function: induction and measurement of cytotoxic T lymphocyte activity.” John Wiley & Sons, Inc., USA, p. 3.11.18). The LU calculation allows quantifying responses and thus readily comparing different experimental values.
In Situ IFNγ ELISA
An in situ IFNγ ELISA assay has been developed and optimized for both freshly isolated and peptide-restimulated splenocytes (see, e.g., McKinney et al., J. Immunol. Meth. 237 (1-2):105-117 (2000))IFN. This assay is based on the ELISPOT assay, but utilizes a soluble chromagen, making it readily adaptable to high-throughput analysis. In both the primary and restimulation assays, this technique is more sensitive than either a traditional supernatant ELISA or the 51Cr-release assay, in that responses are observed in the in situ ELISA that are not detectable in these other assays. On a per cell basis, the sensitivity of the in situ ELISA is approximately one IFNγ secreting cell/104 plated cells.
96-well ELISA plates are coated with anti-IFNα (rat anti-mouse IFNα MAb, Clone R4-6A2, Pharmingen) overnight at 4° C., and then blocked for 2 hours at room temperature with 10% FBS in PBS. Serially diluted primary splenocytes or CTLs are cultured for 20 hours with peptide and 105 Jurkat A2.1/Kb cells/well at 37° C. with 5% CO2. The following day, the cells are washed out and the amount of IFNγ that had been secreted into the wells is detected in a sandwich ELISA, using biotinylated α-IFNγ (rat anti-mouse IFNγ mAb, Clone XMG1.2, Pharningen) to detect the secreted IFNγ. HRP-coupled strepavidin (Zymed) and TMB (Immunopure® TMB Substrate Kit, Pierce) are used according to the manufacturer's directions for color development. The absorbance is read at 450 nm on a Labsystems Multiskan RC ELISA plate reader. In situ IFNγ ELISA data is evaluated in secretory units (SU), based on the number of cells that secrete 100 pg of IFNγ in response to a particular peptide, corrected for the background amount of IFN in the absence of peptide.
ELISPOT Assay
The ELISPOT assay quantifies the frequency of T cells specific for a given peptide by measuring the capacity of individual cells to be induced to produce and release specific lymphokines, usually IFNγ. The increased sensitivity of the ELISPOT assay has allowed investigators to detect responses from cells freshly isolated from infected humans or experimental animals (Murali-Krishna et al., Immunity, Vol. 8(2):177-87 (1998); Ogg et al., Science, Vol. 279(5359):2103-6 (1998)). The ELISPOT assays are conducted as described above for the IFNγ ELISA until the final steps, where ExtrAvidin-AP (Sigma, 1:500 dilution) is used in place HRP-strepavidin. Color is developed using the substrate 5-BCIP (BioRad) according to the manufacturer's directions. Spots are counted using a phase contrast microscope. Alternatively, spots are counted utilizing the Zeiss KS ELISPOT reader. In this case the BCIP/NBT substrate is used.
The ELISPOT assay is routinely utilized to quantitate immune responses. The spots can be manually counted, however, in a preferred mode, a KS ELISPOT reader from Zeiss, a microscope-based system with software specifically designed to recognize and count spots is used.
Tetramer Staining
Tetramer staining is a flow cytometric technique that detects epitope-specific human CD8+ T-lymphocytes based on the interaction between the peptide epitope, class I antigen and the T-cell receptor specific for the epitope. This assay allows for the rapid quantitation of epitope specific human CD8+ T-lymphocytes in freshly isolated blood samples. MHC tetramers for various HIV peptide/HLA combinations, obtained, e.g., from the NIH repository (Tetramer Core Facility: http://www.miaid.nih.gov/reposit/tetramer/index.html). To label epitope-specific cells, 1×106 PBMC in a 100 μl volume are incubated in the dark for 40 minutes with 5 μg/ml of the appropriate tetramer plus monoclonal antibodies that recognize human CD3 and CD8 (available in different fluorochrome-conjugated forms from commercial sources including PharMingen, San Diego, Calif. or BioSource, Camarillo, Calif.). The cells are washed and paraformaldehyde fixed prior to analysis using a FACsan or FACSCalibur flow cytometer (Becton Dickinson Immunocytometry Systems, San Jose, Calif.). Sample data are analyzed using CellQuest software.
Assays from Clinical Samples
Various assays to evaluate the specific CD8+ CTL activity in frozen PBMC samples from patients or volunteers can be used. ELISPOT, chromium release, in situ IFNγ release, proliferation and tetramer assays are all useful to evaluate responses from various experimental models, e.g., those of murine and/or primate origin.
Experimental methods for the murine version of these assays are described above, and these are adapted to human systems as described (Livingston et al, J Immunol, Vol. 159(3):1383-92 (1997); Heathcote et al., Hepatology, Vol. 30(2):531-6 (1999); Livingston et al., J Immunol, Vol. 162(5):3088-95 (1999)) and can be further adapted a recognized by one of ordinary skill in the art. Calculations on the amounts of frozen PBMC samples necessary to complete the assays are also described greater detail in Example 14.
Transgenic Animals
Transgenic mice (HLA-A2.1/Kb H2b; HLA-A11/Kb; HLA-B7/Kb) are immunized intramuscularly in the anterior tibialis muscle or subcutaneously in the base of the tail with doses up to 100 μg of DNA or peptide in 10-100 μl volumes. DNA is formulated in saline, and peptides in IFA. 11-21 days later, the animals are sacrificed using CO2 asphyxiation, their spleens removed and used as the source of cells for in vitro determination of CTL function. Typically, 3-6 mice per experimental group are used. In addition, spleens from non-immunized mice are used as a source of APC for restimulation of CTL cultures. Both males and females of 8-12 weeks of age are used.
Demonstration of Simultaneous Induction of Responses Against Multiple CTL and HTL Epitopes
Construction and testing of CTL epitope strings:
This example provides an example of testing multiple CTL and HTL epitopes. For example, epitope strings encompassing 10-12 different CTL epitopes under the control of a single promoter are synthesized and incorporated in a standard plasmid, pcDNA 3.1 (Invitrogen, San Diego). These constructs include a standard signal sequence and a universal HTL epitope, PADRE®. Each set of epitopes is chosen to allow balanced population coverage. To facilitate testing and optimization, a balanced representation of epitopes that have been shown to be immunogenic in transgenic mice, and/or antigenic in humans are included.
The specific order of these CTL epitopes is chosen to minimize Class I junctional motifs by the use of the computer program, as described herein. If, despite best efforts regarding order optimization, potential junctional epitopes are still present in a construct in accordance with the invention, corresponding peptides are synthesized to monitor for CTL responses against such epitopes in HLA transgenic mice. Generally, minimization of junctional motifs is successful and adequate. However, if responses against any junctional epitopes are detected, these junctional epitopes are disrupted by the use of short one to two residue spacers, such as K, AK, KA, KK, or A, compatible with expected proteolytic cleavage preferences discussed in the previous sections.
Since the ultimate use of optimized constructs is a human vaccine, optimized human codons are utilized. However, to facilitate the optimization process in HLA transgenic mice, care are applied to select, whenever possible, human codons that are also optimal for mice. Human and murine codon usage is very similar. See, for example, Codon usage database at http://www.kazusa.or.jp/codon/.
Human cells transfected with the various minigene vaccine constructs can be used in antigenicity assays, conducted in parallel with in vivo testing in HLA transgenic mice. Any potential discrepancy between minigene vaccine efficacy, due to the differential codon usage, is addressed by the availability of these two different assay systems.
Typically, antigenicity and immunogenicity testing of plasmid constructs is conducted in parallel. In vivo testing in transgenic mice are performed for A2, A11, and B7 HLA transgenic mice. Following a protocol well established in our laboratory, cardiotoxin pretreated mice are injected i.m. with 100 μg of each plasmid and responses evaluated eleven days later (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)). Assays will include ELISPOT from freshly isolated cells, as well as interferon gamma release and cytotoxicity chromium release assays from restimulated cell cultures. All of the above mentioned techniques are well established in the art. The simultaneous measurement of responses against epitopes is not problematic, as large colonies of transgenic mice are already established “in house” for these HLA types. Groups of four to six mice are adequate to measure responses against six to ten different epitopes, in multiple readout assays. Testing of HLA A2-restricted, HIV-derived epitopes in HLA A2 transgenic mice is typically employed. However, should problems be encountered, antigenicity testing using human APC can be used as an alternative strategy, or, can be used to complement the transgenic mice studies.
For the purpose of extending the correlation between immunogenicity in transgenic animals and antigenicity, as noted in the studies reported herein, antigenicity testing is utilized to evaluate responses against epitopes such as Pol 498, Env 134, Nef 221, Gag 271, for which high affinity CTL lines are already available in house. For the purpose of generating additional suitable CTL lines, direct immunization of HLA transgenic mice with peptides emulsified in adjuvant, or lipopeptides are utilized, as described herein, and routinely applied in our laboratory, to generate lines for use in antigenicity assays.
Antigenicity assays are also used, as a primary readout for epitopes for which in vivo optimization experiments are not feasible. These epitopes include A24 and possibly A1 restricted epitopes, as well as any epitope which is non-immunogenic in HLA transgenic mice. In any such cases, we use human CTL lines, generated from pathogen exposed individuals. Alternatively, we generate CTL lines for in vitro CTL induction, using GMCSF/IL4-induced dendritic cells and peripheral blood lymphocytes (Celis et al., Proc Natl Acad Sci USA, Vol. 91(6):2105-9 (1994)).
Episomal vectors encoding the minigenes are generated and transfected into appropriate human cell lines to generate target cells. For example, the human T cell line Jurkat can be used, but lymphoblastoid cell lines have also been successfully utilized. For experiments utilizing CTL lines of human origin, well-characterized HLA-matched, homozygous, EBV cell lines are commonly used as a source of purified-MHC and as target cells and are used as recipients of the minigene transfections. For experiments utilizing CTL lines derived from HLA transgenic mice, a collection of Class I negative, EBV-transformed, mutant cell lines 0.221 (Shimizu Y, DeMars R., J Immunol, Vol. 142(9):3320-8 (1989)) transfected with matching HLA/Kb chimeric constructs are used as the recipient of the minigene transfections. Such cells effectively present peptide antigens to CTL lines (Celis et al., Proc Natl Acad Sci USA, Vol. 91(6):2105-9 (1994)).
Construction and Testing of HTL Epitope Strings:
Epitope strings encompassing 3-20 different HTL epitopes under the control of a single promoter are synthesized and incorporated into a standard plasmid, pcDNA 3.1 (Invitrogen, San Diego). To facilitate testing and optimization, each set of epitopes for a given minigene is chosen to provide a balanced representation of epitopes which are already known to be immunogenic in IAb mice. In addition, all the peptides corresponding to junctions are synthesized and tested for binding to IAb and, most importantly, for binding to a panel of fourteen different DR molecules, representative of the most common DR alleles worldwide (Southwood et al., J Immunol, Vol. 160(7):3363-73 (1998)). Thus, HTL epitopes that are not directed to an antigen of interest are not created within these plasmids. However, should junctional regions with good DR binding potential (and hence, potential DR restricted immunogenicity in vivo) be detected, a spacer such as GPGPG (SEQ ID NO: 369) is introduced to eliminate them. In all constructs, the number of Class I junctional motifs will also be minimized, as described herein.
Experimental vaccine plasmids are tested for immunogenicity using HLA DR transgenic mice and/or mice of the H2b haplotype. Proliferation and/or cytokine production are measured (IL5, IFNγ). In a typical protocol, cardiotoxin treated mice are injected i.m. with 100 μg of each plasmid and responses evaluated eleven days later (Ishioka et al., J Immunol, Vol. 162(7):3915-25 (1999)).
Testing for Interactions between CTL and HTL Epitopes
The activities described above yield small, functional blocks of epitopes, which are utilized to demonstrate simultaneous responses/antigenicity against all epitopes analyzable. These blocks are the subject to further optimization, as described in the next example. Using these same minigenes, immunodominance is assessed. Specifically, all the CTL plasmids are mixed together, or all the HTL plasmids are mixed together. The results obtained with the minigene pool are then compared with the results obtained with the same minigene, injected separately.
These minigene plasmids are also used to determine the effects of HTL epitopes on responses to CTL epitopes. Specifically, HTL and CTL containing plasmids are pooled and injected in mice, and CTL and HTL responses to selected epitopes are measured as described herein. Often, it is determined whether the presence, e.g., of HTL epitopes derived from the target antigen enhances CTL responses beyond the level of response attained using a plasmid-containing a pan DR binding epitope, e.g., PADRE® or a PADRE® family molecule, in the CTL minigene. Typically, it is also determined whether PADRE® inhibits or augments responses to target antigen-derived HTL epitopes or conversely, if HTL epitopes derived from the antigen of interest inhibit or augment responses to PADRE® .
Responses to a large number of epitopes is attainable using this methodology. It is possible that the pooling of constructs may inhibit responses against some of the weaker epitopes. In this case, the pooling experiments are repeated after optimization.
Optimization of CTL and HTL Minigene Constructs
This example describes the optimization the CTL and HTL constructs described in Example 10. The potential influence of flanking residues on antigenicity and immunogenicity is also assessed in optimizing minigen constructs. These studies involve the inclusion of flanking residues, a synonym for which is “spacers,” which have been designed to facilitate effective processing.
Such an analysis can be performed as follows. First, the results of testing of the CTL multi-epitope constructs described in Example 10 are analyzed for trends and correlations between activity and the presence of specific residues at the 3 residues flanking the epitope's N— and C-termini. Epitopes for which suboptimal CTL priming is noted, that are suboptimal with respect to magnitude of response, are the targets for flanking region optimization. For each of the CTL minigene vaccines, encoding 10-12 different CTL epitopes, ‘second generation’ minigene vaccines, with optimized configuration, are produced.
In one embodiment, the first optimization design is to introduce either an Alanine (A) or Lysine (K) residue at position C+1 for all epitopes associated with suboptimal performance. A second optimization design is to introduce in the C+1 position, the residue naturally occurring in the target antigen, e.g., HIV, that are associated with antigenicity.
For selected epitopes, additional modifications are also introduced. Specifically, the effect of introducing other residue spacers at the epitope C− and N− termini are also investigated. Depending on the results of the analysis of the minigene vaccines described in Example 10, residues investigated may further include, for example, G, Q, W, S and T. If junctional epitopes are created by these modifications, alternative epitope orders eliminating the junctional epitopes, are rationally designed, as described herein. All second generation constructs are tested for antigenicity and immunogenicity, as described herein.
As a result of these modifications, variations in activity that correspond to specific modifications of the minigenes are identified. Certain modifications are found that have general, beneficial effects. To confirm this, generation and testing of additional minigene vaccines in which all epitopes (also the ones which displayed acceptable antigenicity and immunogenicity) are subject to the same modification are conducted. In some instances, increased activity is noted for some epitopes but not others, or less desirably that certain modifications increase the activity of some, but decrease the activity of other epitopes. In such cases, additional minigene vaccines are designed and tested, to retain the beneficial modifications, while excluding those alterations that proved to be detrimental or have no effect.
These minigene vaccines are designated so that: a) a minimum of predicted junctional epitopes are present; and, b) the epitopes which were not functional in the previous minigene vaccines are now in a new more efficacious context.
For HTL minigene vaccines, the data obtained from the “first generation” minigene vaccines are inspected for trends, in terms of junctional epitopes, and epitope position within the minigene, and proximity to spacers, e.g. GPGPG (SEQ ID NO: 369) spacers. If specific trends are detected, second generation minigene vaccines are designed based on these trends. Alternatively, in case of minigenes yielding suboptimal activity, the potential effectiveness of other targeting strategies, such as the ones based on Ii and LAMP are reevaluated, and compared to no targeting and simple, leader sequence targeting.
When large variations in activity of either the CTL or HTL minigene vaccines described in this section are detected, the results are consistent with influences such as conformational or “long-range” effects impacting minigene activity. These variables can be analyzed by means of current molecular and cellular biology techniques. For example, cell lines transfected with the various minigenes could be analyzed for mRNA expression levels, and stability by Northern analysis or primer extension assays (Current Protocols in Molecular Biology, Vol 3, John Wiley & Sons, Inc. USA 1999).
In all minigene vaccines, an antibody tag such as MYC/his can also be included. This tag allows for testing of protein expression levels. The inclusion of MYC/his tag (Manstein et al., Gene, Vol. 162(l):129-34 (1995)) also allows determination of the stability of the expressed products, by pulse-chase experiments. The results of these assays can then be compared with the results of the antigenicity and immunogenicity experiments. The results are inspected for the presence of trends and general rules, and correlation between the different variables examined.
Determination of the Simplest Plasmid Configuration Capable of Effective Delivery of Selected Epitopes
The experiments described in Examples 11 and 12 are designed to address variables concerning minigene vaccine design. Ideally, a vector that can be used in humans is used through the entire program, but one DNA vaccine plasmid for the vaccine epitope optimization studies can be used and then switched to a vector suitable for human use. Actual vector selection is dependent on several variables. For example, the availability of vectors, suitable for human use, through a reliable source, such as the National Gene Vector Laboratory (University of Michigan) is a factor.
In this example, the optimized minigenes are also ligated to form larger blocks of epitopes. All constructs are preferably designed to incorporate PADRE and leader sequence targeting in the case of CTL minigenes. Specifically, two pairs of the 10-12 CTL epitope minigenes are ligated to generate two 20-24 CTL epitope minigenes. In a situation where ligation of epitopes yields suboptimal (decreased) activity compared to the smaller minigenes, alternative combinations and orders of ligation are investigated. The specific pair of 20-24 CTL epitope minigenes yielding optimal activity are then ligated and the resulting minigene encompassing all CTL epitopes evaluated for activity. Once again up to two alternative orientations are investigated. Because of the relatively large size of this construct, the specific effect of targeting sequences are confirmed, since it is possible that leader sequence targeting are more effective on small size minigenes, while larger size constructs may be most effectively targeted by ubiquitin signals. Specifically, one construct without any specific targeting sequences is generated and compared to a construct that is targeted for degradation by the addition of a ubiquitin molecule.
A similar strategy is used for HTL. Two pairs of the 3-5 HTL epitope minigenes are ligated to generate two 7-9 HTL epitope minigenes. Once again, in a situation where ligation of these epitopes yields suboptimal (decreased) activity, alternative combinations and order of ligation are investigated. The specific pair of 7-9 CTL epitope minigenes yielding optimal activity are ligated and the resulting minigene, encompassing all HTL epitopes, is evaluated for activity. Once again, up to two alternative orientations are investigated.
Based on these results an optimized plasmid configuration capable of effective delivery of a panel, e.g., of HIV epitopes, are selected for clinical trial evaluation. Of course, epitopes from any antigen of interest (infectious or disease-associated) can be used alone or in combination. This configuration will entail one or more HTL epitope minigene(s) and one or more CTL epitope minigene(s). A combination of one long CTL and one long HTL minigene capable of effectively delivering all epitopes, is most preferable, as it simplifies further clinical development of the vaccine. In case undesirable interactions between the two minigenes are observed when co-injected, injection of the different plasmids in the same animals, but in different injection sites, or at different points in time are examined. Alternatively, if a single CTL minigene and HTL minigene encoding all the desired epitopes is not identified, pools of minigenes are considered for further development.
Evaluation and Characterization of CD8+ Lymphocyte Responses Induce Following Immunization with Multi-Epitope Vaccine
CD8+ lymphocyte responses were measured mostly relying on the ELISPOT technique. The ELISPOT assay is known in the art, and is regularly used in our laboratory. An automated Zeiss ELISPOT reader is also used as set forth herein. The assays utilized to measure CD8+ responses are primarily the IFNγ ELISPOT assay on freshly isolated cells as well as cells restimulated in vitro with peptide. In addition, in selected instances, chromium release assays are utilized. The results were correlated with the ones observed in the case of the ELISPOT assays. Tetramer staining on selected peptide/MHC combinations was also performed.
The clinical assay was developed and validated. The timing of this activity coincides with the period of time that follows selection of a clinical vaccine minigene, and precedes the availability of actual samples from individuals enrolled in the clinical trial. Assays for CTL evaluation can be established based on experience in the art, for example, experience in establishing assays for CTL evaluations in the Phase I and II trials of the experimental HBV vaccine, Theradigm (Livingston et al, J Immunol, Vol. 159(3):1383-92 (1997); Heathcote et al., Hepatology, Vol. 30(2):531-6 (1999); Livingston et al., J Immunol, Vol. 162(5):3088-95 (1999)). Specifically, Ficoll-purified PBMC derived from normal subjects, as well from, e.g., unvaccinated volunteers can be used. As noted previously, other antigenic target(s) can be used in accordance with the invention.
Design of Optimized Multi-Epitope DNA-Based Vaccine Constructs
Optimized constructs were designed with the aid of the computer-assisted methods described above which simultaneously minimize the formation of junctional epitopes and optimize C+1 processing efficiency. The following motifs were utilized for junctional minimization: murine Kb [XXXX(FY)X2-3(LIMV)] (SEQ ID NOS: 370 and 371 ); Db[XXXXNX2-3LIMV)] (SEQ ID NOS:372 and 373); human A2 [X(LM)X6-7V] (SEQ ID NOS: 374 and 375); human A3/A11 [X(LIMV)X6-7(KRY)] (SEQ ID NOS: 376 and 377); and human B7 [XPX6-7(LIMVF)] (SEQ ID NOS: 378 and 379). The C+1 propensity values were calculated from the data presented in
Immunogenicity Testing of Multi-epitope CTL Constructs and Influence of Flanking Amino Acids
HLA transgenic mice were used for immunogenicity testing of different multi-epitope constructs. One group of mice were pretreated by injecting 50 μl of 10 μM cardiotoxin bilaterally into the tibialis anterior muscle, and then four or five days later, 100 μg of a DNA construct diluted in PBS was administered to the same muscle. In another group, each mouse was injected with a peptide emulsified in CFA, wherein the peptide corresponds to an epitope within the DNA construct administered to mice in the DNA injection group. Eleven to fourteen days after immunization, splenocytes from DNA vaccinated animals and peptide vaccinated animals were recovered and CTL activity was measured in one of several assays, including a standard 51Cr-release assay, an ELISPOT assay that measured γ-IFN production by purified CD8+ T-lymphocytes without peptide epitope-specific restimulation, and an in situ ELISA, which included an in vitro epitope-specific restimulation step with a peptide epitope. Examples of CTL activity induced by the EP-HIV-1090 construct upon stimulation with peptide epitopes are shown in
The effect of different amino acids in the C+1 flanking position was directly evaluated by inserting different amino acids at the C+1 position relative to the Core 18 epitope in the HBV.1 construct. The immunogenicity data clearly demonstrate reduced immunogenicity of the Core 18 epitope when it was flanked at the C+1 position by W, Y, or L (
Immunogenicity Testing of Multi-epitope HTL Constructs and Influence of Spacer Sequences
A universal spacer consisting of GPGPG (SEQ ID NO: 369) was developed to separate HTL epitopes, thus disrupting junctional epitopes. The logic behind the design of this spacer is that neither G nor P are used as primary anchors, positions 1 and 6 in the core region of an HTL peptide epitope, by any known murine or human MHC Class I or MHC Class II molecule. The gap of five amino acids introduced by this spacer separates adjacent epitopes so the amino acids of two epitopes cannot physically serve as anchors in the 1 and 6 positions. The utility of the GPGPG (SEQ ID NO: 369) spacer was tested using synthetic peptides composed of four HIV-1 epitopes, one having three spacers and the other lacking spacers, known to bind mouse IAb. HIV 75mer was the construct having three GPGPG (SEQ ID NO: 369) spacers and HIV 60mer was the construct lacking the three spacers. Immunization of CB6F1 mice with the peptide in CFA induced HTL responses against 3 of 4 of the epitopes in the absence of the spacer but all epitopes were immunogenic when the spacer was present (
The ability of multi-epitope HTL DNA-based constructs to induce an HTL response in vivo was evaluated by intramuscular immunization of H2bxd mice with an HIV-1043-PADRE® construct. The HIV-1043-PADRE® construct is set forth in
Thus, as described above, the invention provides a novel method and system for automatically analyzing polypeptide junctions, eliminating or reducing the number of junctional epitopes, and identifying spacer combinations to optimize the efficacy of multi-epitope minigenes. Those skilled in the art will know, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. These equivalents are intended to be encompassed by the following claims.
This application is a continuation-in-part of International Application No. PCT/US00/35568 PCT, entitled “Optimized Minigenes and Peptides Encoded Thereby,” filed on 28 Dec. 2000, to Sette et al., which claims priority to U.S. Application Ser. No. 60/173,390 filed 28 Dec. 1999. This application also claims the benefit of the 16 Apr. 2001 filing date for U.S. Application Ser. No. 60/284,221, entitled “Optimized Minigenes and Peptides Encoded Thereby,” to Sette et al. Each of these documents is hereby incorporated by reference in its entirety.
This invention was made with Federal Funding assistance. Accordingly, the U.S. government may have certain rights to the invention.
| Number | Name | Date | Kind |
|---|---|---|---|
| 4235877 | Fullerton | Nov 1980 | A |
| 4487715 | Nitecki et al. | Dec 1984 | A |
| 4599230 | Milich et al. | Jul 1986 | A |
| 4599231 | Milich et al. | Jul 1986 | A |
| 4837028 | Allen | Jun 1989 | A |
| 5013548 | Haynes et al. | May 1991 | A |
| 5128319 | Arlinghaus | Jul 1992 | A |
| 5200320 | Sette et al. | Apr 1993 | A |
| 5503829 | Ladant et al. | Apr 1996 | A |
| 5633234 | August et al. | May 1997 | A |
| 5662907 | Kubo et al. | Sep 1997 | A |
| 5736142 | Sette et al. | Apr 1998 | A |
| 5783567 | Hedley et al. | Jul 1998 | A |
| 5846827 | Celis et al. | Dec 1998 | A |
| 6034214 | Boon et al. | Mar 2000 | A |
| 6037135 | Kubo et al. | Mar 2000 | A |
| 6413517 | Sette et al. | Jul 2002 | B1 |
| 6413935 | Sette et al. | Jul 2002 | B1 |
| 6419931 | Vitiello et al. | Jul 2002 | B1 |
| 6534482 | Fikes et al. | Mar 2003 | B1 |
| 6602510 | Fikes et al. | Aug 2003 | B1 |
| 6689363 | Sette et al. | Feb 2004 | B1 |
| 20020098197 | Sette et al. | Jul 2002 | A1 |
| 20020119127 | Sette et al. | Aug 2002 | A1 |
| 20020160019 | Sette et al. | Oct 2002 | A1 |
| 20020160960 | Sette et al. | Oct 2002 | A1 |
| 20020168374 | Kubo et al. | Nov 2002 | A1 |
| 20020177694 | Sette et al. | Nov 2002 | A1 |
| 20030099634 | Vitiello et al. | May 2003 | A1 |
| 20030143672 | Tangri et al. | Jul 2003 | A1 |
| 20030152580 | Sette et al. | Aug 2003 | A1 |
| 20030185822 | Grey et al. | Oct 2003 | A1 |
| 20030203869 | Fikes et al. | Oct 2003 | A1 |
| 20030216342 | Fikes et al. | Nov 2003 | A1 |
| 20030216343 | Fikes et al. | Nov 2003 | A1 |
| 20030220285 | Fikes et al. | Nov 2003 | A1 |
| 20040096445 | Sidney et al. | May 2004 | A1 |
| 20040157273 | Sidney et al. | Aug 2004 | A1 |
| 20040248113 | Sette et al. | Dec 2004 | A1 |
| 20050049197 | Sette et al. | Mar 2005 | A1 |
| Number | Date | Country |
|---|---|---|
| 0 044 710 | Jan 1982 | EP |
| 0 226 513 | Jun 1987 | EP |
| 0 429 816 | Jun 1991 | EP |
| 0 433 242 | Jun 1991 | EP |
| 0 378 881 | Jun 1993 | EP |
| WO 9202543 | Feb 1992 | WO |
| WO 9212996 | Aug 1992 | WO |
| WO 9221033 | Nov 1992 | WO |
| WO 9303764 | Mar 1993 | WO |
| WO 9303764 | Mar 1993 | WO |
| WO 9403205 | Feb 1994 | WO |
| WO 9411738 | May 1994 | WO |
| WO 9420127 | Sep 1994 | WO |
| WO 9420127 | Sep 1994 | WO |
| WO 9504817 | Feb 1995 | WO |
| WO 9507707 | Mar 1995 | WO |
| WO 9522317 | Aug 1995 | WO |
| WO 9603140 | Feb 1996 | WO |
| WO 962067 | Jul 1996 | WO |
| WO 9703778 | Feb 1997 | WO |
| WO 9733602 | Sep 1997 | WO |
| WO 9734617 | Sep 1997 | WO |
| WO 9734621 | Sep 1997 | WO |
| WO 9741440 | Nov 1997 | WO |
| WO 9832456 | Jul 1998 | WO |
| WO 9910646 | Mar 1999 | WO |
| WO 9945954 | Sep 1999 | WO |
| WO 9958658 | Nov 1999 | WO |
| WO 9961916 | Dec 1999 | WO |
| WO 9965522 | Dec 1999 | WO |
| WO 0019774 | Apr 2000 | WO |
| WO 0027766 | May 2000 | WO |
| WO 0100225 | Jan 2001 | WO |
| WO 0220616 | Mar 2002 | WO |
| WO 02061435 | Aug 2002 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 20020119127 A1 | Aug 2002 | US |
| Number | Date | Country | |
|---|---|---|---|
| 60173390 | Dec 1999 | US | |
| 60284221 | Apr 2001 | US |
| Number | Date | Country | |
|---|---|---|---|
| Parent | PCT/US00/35568 | Dec 2000 | US |
| Child | 09894018 | US |
