A human monoclonal anti-melanoma single chain Fv (scFv)antibody derived from tumor-infiltrating B lymphocytes.

Item Type text; Dissertation-Reproduction (electronic)

Authors Zhang, Hua.

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this materialis made possible by the University Libraries, University of Arizona.Further transmission, reproduction or presentation (such aspublic display or performance) of protected items is prohibitedexcept with permission of the author.

Download date 22/05/2018 07:02:39

Link to Item http://hdl.handle.net/10150/187293

INFORMATION TO USERS

This manuscript ,has been reproduced from the microfilm master. UMI

films the text directly from the original or copy submitted. Thus, some

thesis and dissertation copies are in typewriter face, while others may

be from any type of computer printer.

The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality

illustrations and photographs, print bleedthrough, substandard margins,

and improper alignment can adversely affect reproduction.

In the unlikely. event that the author did not send UMI a complete

mannscript and there are missing pages, these will be noted. Also, if

unauthorized copyright material had to be removed, a note will indicate

the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by

sectioning the original, beginning at the upper left-hand comer and

contimJing from left to right in equal sections with small overlaps. Each

original is also photographed in one exposure and is included in

reduced form. at the back of the book.

Photographs included in the original manuscript have been reproduced

xerographically in this copy. Higher quality 6" x 9" black and white

photographic prints are available for any photographs or illustrations

appearing in this copy for an additional charge. Contact UMI directly

to order.

UMI A Bell & Howell Information Company

300 North Zeeb Road. Ann Arbor. MI48106-1346 USA 313/761-4700 800:521-0600

1

A HUMAN MONOCLONAL ANTI-MELANOMA SINGLE CHAIN Fv (scFv)

ANTIBODY DERIVED FROM TUMOR-INFILTRATING B LYMPHOCYTES

By

Hua Zhang

A Dissertation Submitted to the Faculty of the

DEPARTMENT OF MICROBIOLOGY AND IMMUNOLOGY

In Partial Fulfillment of the Requirements for the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSI1Y OF ARIZONA

1995

UMI Number: 9604518

OMI Microform 9604518 Copyright 1995, by OMI Company. All rights reserved.

This microform edition is protected against unauthorized copying under Title 17, United States Code.

UMI 300 North Zeeb Road Ann Arbor, MI 48103

THE UNIVERSITY OF ARIZONA GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have

2

read the dissertation prepared bY~f~u~la~7-~~h~a~n~q~ __________________________ _

entitled A lIUl'·1AN· f10NOCIDNAT.J ANTI-t·1ELANO"'JA SINGT.JE CHAIN Fv (scFv)

ANTIBODY DERIVED FR0'1 TU1'10R-INFIJ..'I'RATING B LYt1PHOCYTES.

and recommend that it be accepted as fulfilling the dissertation

requirement for the Degree of Doctor of ~ljJQSapby

l ~,\ 11 it .( .~< Date

~1131(r-r

Date

/ ~UJ/:;T-

Date'

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requ~men t. . . I

l "-' f' {\, ! I'!

CCet'l l \, r II l Llj\·./C...., Date I Dissertation D1rector \

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at the University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his or her judgement the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

. IJ. "'1/' SIG NED: __ 7::::-",=,1 "/_" '_/"-0' '_' -=' =" .... =.:=' 'F': ==..::.... __ .~. /_ ..

... --

3

ACKNOWLEDGMENTS

I would like to thank my committee for providing guidance and direction

throughout my studies, especially Evan Hersh and William Grimes who

have provided moral as well as SCientific support.

Doug Lake for all his advice, helpful discussions and molecular biology

expertise. Ralph Bernstein for helpful discussions, technical support.

Samuel Schluter for helpful discussions. Alex Barbuto for helpful

discussions.

Jin-Ming Yang for immunofluorescence photography. Norma Seaver for

FACS analysis.

4

Finally, I would like to thank my wife for her support throughout the

past more than 5 years of my graduate studies and especially mention my

daughter, Jenny, and son, Michael, who brought a special happiness and

fun into my life during the studies.

5

TABLE OF CONTENTS

LIST OF FIGURES ........................................................................ 8

LIST OF TABLES .......................................................................... 9

ABS1"RACT ................................................................................... 1 0

CHAPTER 1. INTRODUCTION ....................................................... 12

Antibody Structure and Function ......................................... 13 Immunoglobulin Genes ........................................................ 15 Monoclonal Antibody ........................................................... 17 Antibody Engineering ........................................................... 20 Molecular Cloning and Expression of Ig Genes ....................... 22 Tumor-infiltrating B Lymphocytes ......................................... 27 Melanoma -associated Antigens ............................................. 30 Immunology and Immunotherapy of Melanoma ...................... 33 Rationale for Study .............................................................. 36

CHAPTER 2. MATERIALS AND METHODS ...................................... 39

Cell Lines ............................................................................. 39 Preparation of Tumor-infiltrating B Lymphocytes (TIL-B) ......... 40

B Cell Expansion and Immunoglobulin Secretion In Vitro ........ 40

ELISA for Human Ig and Ig Classes ....................................... .41 Frequency of Response (F(r)) Calculation .............................. .42

Detection of Anti-tumor Antibodies (Tumor Cell ELISA) ......... .42 Cloning EBV-transformed B Cell Lines by Limiting Dilution ... .43 Messenger RNA Isolation and cDNA Synthesis ........................ 44 Capture and Assembly of Ig V Region by

-., Polymerase Chain Reaction (PCR) .......................................... 44 Cloning Immunoglobulin Variable Regions into

6

Sequencing Vectors ............................................................. 45 PCR Screening of Transformed Bacterial Colonies ................ .46 Mini-preparation of Plasmid DNA from E. coli ...................... .47

DNA Sequencing of the Immunoglobulin Variable Regions .... .48 Cloning AZM 1 scFv DNA into pET21 d Expression Vectors .... .49 Expression of AZM 1 scFv in E. coli.. ..................................... 49

Preparation of Bacterial Cell Lysates for AZMl scFv Isolation ...................................................... 49 Purification of AZMl scFv on a Nickel-agarose Column ......... 50 Biotinylation of AZMl scFv ................................................. 51 Fixed Cell Immunofluorescence and Flow Cytometry ............. 51 Radioimmunoprecipitation Analysis ..................................... 52 Trypsinization Assay ............................................................ 53

ClIAP1'ER 3. RESULTS .................................................................. 54

Cloning EBV-transformed B Cells by Limiting Dilution .......... 54 Capture and Assembly of Anti-melanoma Immunoglobulin Heavy and Light Variable Regions .......................................... 55 Nucleotide and Amino Acid Sequence of Immunoglobulin Heavy and Light Chain Variable Regions ................................ 56 Expression of AZMl scFv Antibody ......................................... 57 Purification of AZMl scFv Antibody ........................................ 58 Primary Analysis of Biological Activities of AZM 1 scFv ............. 59 Identification of Melanoma-associated Antigen by AZM 1 scFv ........................................................................ 60

Primary Characterization of Melanoma-associated Antigen Recognized by AZM 1 scFv .......................... 60

ClIAP1'ER 4. DISCUSSION .............................................................. 77

In Vitro Tumor-infiltrating B Cell Expansion ............................ 77

Capture of Immunoglobulin Genes .......................................... 81 Construction of AZM 1 scFv ..................................................... 81

7

Expression of AZMl scFv Antibody in E. coli ......................... 83

Purification of AZMl scFv Antibody ...................................... 84

Tumor-reactivity of AZMl scFv Antibody ............................... 84

Identification of Melanoma-associated Antigen ..................... 86

Conclusion ....................... II ••••• II •••••••• I ••••••••••••••••••••••••••••• II .89

APPENDIX A. PCR PRIMERS FOR HUMAN Ig GENES .................... 90

APPENDIX B. BUFFERS AND MEDIA ........................................... 91

REFERENCES ............................................................................ 92

8

LIST OF FIGURES

FIGURE 1. Limiting dilution analysis (LDA) .................................. 62 FIGURE 2. PCR products of VH, Vk and scFv ................................ 63 FIGURE 3A. Nucleotide and deduced amino acid sequences of AZMl

heavy chain variable region ......................................... 64 FIGURE 3B. Nucleotide and deduced amino acid sequences of AZMl

light chain variable region ........................................... 65 FIGURE 4. A diagram ofpET21d expression vector ......................... 66 FIGURE 5. SDS-PAGE analysis of purification of the scFv (AZMl)

by a Nickel-agarose column .......................................... 67 FIGURE 6. Immunofluorescence stain on fixed tumor cells ............. 68 FIGURE 7. FACS analysis of tumor cells treated with the scFv

(AZMl) ........................................................................ 69

FIGURE 8. Radioimmunoprecipitation of a melanoma-specific antigen by AZMl scFv .................................................. 70

9

LIST OF TABLES

TABLE 1. Characteristics of two EBV -transformed TIL-B cell lines (Mel-Go) ...................................................... 71

TABLE 2. Characteristics of expanded EBV-transformed B cell clones from Line 1-3 .................................................. 72

TABLE 3. Determining the anti-melanoma specificity of AZM1 scFv .......................................................................... 73

TABLE 4. FACS analysis of melanoma cells treated with AZM1 scFv .......................................................................... 74

TABLE 5. FACS analysis of non-melanoma cells treated with AZMl scFv ................................................................. 75

TABLE 6. Comparison of SK-Mel-2 cells stained with. AZM1 scFv before or after trypsin treatment .......................... 76

10

ABSTRACT

The development of recombinant DNA technology has made it

feasible to clone, construct and express fully human immunoglobulin

molecules. Here we report a novel methodology to make human anti­

tumor scFv antibodies from tumor-infiltrating B lymphocytes (TIL-B). We

isolated and expanded TIL-B from melanomas in the presence of EBV.

The transformed B cells secreting tumor-specific antibodies were

identified and cloned by limiting dilution. From one B cell clone with

specific melanoma reactivity, we captured the immunoglobulin variable

region genes, VH and Vk, by polymerase chain reaction (peR), sequenced

the genes and linked them together by peR assembly using a (GlY4Ser)3

linker to form the scFv gene that was subsequently cloned into the

pET21d vector and expressed. The scFv protein, obtained, with a

molecular weight of 29 KD was purified and biotinylated for further

characterization. The scFv demonstrated specific tumor reactivity to 21

of 24 different melanoma cell lines and not to 14 non-melanoma tumor

cell lines, including breast, ovarian and colon cancer cell lines, normal

human melanocytes as well as normal human leukocytes. These results

were obtained using 1) a tumor cell ELISA, 2) fixed cell

immunofluorescence and 3) live cell flow cytometry. The

immunoprecipitation results indicated that a protein antigen of 45 KD

was recognized by the scFv. Since we previously reported that about 70%

of human tumors of different histologic types contain tumor-infiltrating

B lymphocytes producing specific anti-tumor antibodies, this approach

11

offers a rapid. effective method by combining in vitro B cell expansion and

peR-gene cloning to elucidate the repertoire of the human anti-tumor

immune responses and to make human monoclonal anti-tumor antibody

molecules.

12

CHAPTER 1

INTRODUCTION

The first effective immunization was performed in 1796 by Edward

Jenner, who introduced vaccination with cowpox as a means of

protecting against smallpox (1). Later Louis Pasteur and his

collaborators investigated the possibility of protecting against infection

by vaccination with attenuated strains of microorganisms.

In 1890, Behring and Kitasato demonstrated the neutralizing

activity of sera from animals immunized with diphtheria or tetanus toxin

(1). This was considered the first proof of the existence of antibody. Since

then, antibodies have become a very useful tool in both diagnostic and

therapeutic fields in the fight against various infectious diseases and

cancers.

In the host anti-tumor immune defense system, cytotoxiC T

lymphocytes (CTL) are the killer cells that directly destroy tumor cells (2),

while B lymphocytes and their terminally differentiated plasma cells

produce antibodies (3) which attach to the tumor cells and may

subsequently destroy them by activation of the complement system (4) or

by mediating antibody dependent cellular cytotoxicity (ADCC) (5). In

addition, antibodies can neutralize microorganisms and their tOxins, and

modulate the immune system through idiotype and anti-idiotype

networks (6). The diversity, specificity, and biological activity of

antibodies make them potentially ideal reagents for therapeutic and

diagnostic purposes in the clinic and the laboratory.

Antibody Structure and Function

Antibodies are specific proteins of the globulin class, which are

interchangeably termed "immunoglobulins". Formally, an antibody is a

molecule with biological activity that binds to a known antigen, while

immunoglobulin refers to a group of proteins irrespective of whether or

not their binding target is known. With the possible exception of so­

called natural antibody, antibodies arise in response to foreign or

neoplastic antigens introduced into the body (7).

13

Immunoglobulins are glycoproteins composed of polypeptide and a

small amount of carbohydrate. The polypeptide component possesses

almost all of the biologic properties associated with antibody molecules,

while the carbohydrate part is thought to be important for correct

immunoglobulin folding, solubility and transportation during synthesis,

as well as for protection against protease activity, e.g. IgA (8).

The basic structure of immunoglobulin molecules consists of two

identical heavy chains plus two identical light chains that are held

together by disulfide bonds (6). The heavy chain is made up of

apprOximately 450-575 amino acids with a molecular weight of about 51-

72 KD, while the light chain consists of about 220 amino acids and has

a molecular weight of about 25 RD. The antigenic differences of the

heavy chains determine five different classes or isotypes of

immunoglobulins which are designated as IgG, IgA, IgM, IgD and IgE.

14

There are two types of light chains which are defined as Kappa and

Lambda. Disulfide bonds within each chain direct the formation of

globular domains. Light chains comprise two such domains designated

as variable region (VL) and constant region (CLl, whereas heavy chains

are made up of four or five domains in which one is the variable region

(VH) and the other constant regions (CHI, CH2, CH3 or CH4). Each

domain is apprOximately 110 amino acids in length (9). All variable

regions are located at the amino-termini of both heavy and light chains.

The carboxyl-end of heavy chains either terminates at the end of the last

globular domain for the secreted form of the protein or has an extended

hydrophobic tail, in the cell surface form, that serves to anchor the

molecule within the cellular membrane (6).

The variable regions of the heavy and light chains form the

antigen-binding site. Comparisons of amino acid sequences of many

heavy and light chain V regions have revealed three discrete locations of

hypervariability (complementarity-determining regions, CDRs) for each

chain (10, 11, 12, 13). These CDRs confer the fine specificity that enables

the antibody to bind to the epitope of a particular antigen (14).

Of five known classes (isotype) of immunoglobulins, IgG, IgM, IgA,

IgE and IgD. IgG is the most abundant immunoglobulin in the human

body and provides the most extensive and long-lived antibody responses

to various microbial and neoplastic antigens (6).

Antibodies are bifunctional molecules that exert their functions

through their amino-terminal and carboxy-terminal ends. Their primary

function is to bind antigen through the antigen-binding sites on their

15

amino-terminal ends. They also have a variety of effector functions that

are mediated through their carboxy-terminal ends (Fc region). The

effector functions of antibodies include complement fIxation, Fc receptor

binding, antibody-dependent cell-mediated cytotoxicity (ADCC),

opsonization for phagocytosis and placental transfer of IgG (6).

Immuno~lobulin Genes

The germ-line genes encoding immunoglobulins are located at

three different loci on the human chromosomes. The heavy-chain locus

is on chromosome 14, the kappa light-chain locus is on chromosome 2,

and the lambda light chain is on chromosome 22 (15). An

immunoglobulin (Ig) polypeptide chain is coded by multiple genes

scattered along a chromosome of the germ-line genome. These widely

separated gene segments are brought together (gene rearrangement)

during B lymphocyte differentiation to form a complete Ig gene.

In the process of B cell development, heavy and light chain

germline immunoglobulin (Ig) genes undergo rearrangement. In heavy

chains, one of the twelve germline D (diversity) segments is paired with

one of four germline J Uoining) segments. Then one of over 300 variable

region gene segments pairs with the DJ, resulting in a heavy chain VDJ

rearrangement (16, 17, 18). Only one allele from each V, D and J region

rearranges (19). This phenomenon is called allelic exclusion. Following

heavy chain rearrangement, the light chain genes, kappa and lambda,

undergo a Similar rearrangement except that only VL and JL genes

rearrange (20). It is generally thought that kappa chain genes rearrange

16

first. If they do not rearrange successfully, then the lambda chain genes

rearrange. Only one type, kappa or lambda is expressed and associates

with the heavy chain (15). Allelic exclusion holds true for light chains as

well as heavy chains (21).

Once both heavy and light chain genes successfully rearrange, the

pre-B cell expresses cytoplasmic IgM. Surface IgM (monomer) and IgD are

then expressed as the pre-B cell continues its ontogeny towards

becoming a mature B cell, ready for antigen stimulation. If a B cell

expressing surface Ig encounters a specific antigen during immune

surveillance, it can be stimulated to proliferate. Proliferation of a specific

B cell clone in response to a specific antigen occurs in a series of

complex events, which induce resting B cells to enter the cell cycle. The

antigen molecules that bind to surface immunoglobulin receptors are

internalized, partially digested, and the fragments are recycled to the B­

cell surface in association with Class II molecules encoded in the major

histocompatibility gene complex. The combination of antigen-fragment

and Class II molecules on the B cell surface is recognized by T cells

bearing the appropriate T cell antigen receptor. In this way, B cells can

present antigen to helper T cells which, in turn, are induced to produce

soluble factors, such as IIA, IL5 and IL6, to promote proliferation and

differentiation of the antigen-presenting B cells. Thus, the specific clone

of antibody-producing cells is expanded (22).

During ontogeny and functional differentiation, the heavy chain

genes may undergo further gene rearrangement that results in

immunoglobulin class switching (23,24,25). As the B lymphocytes

17

differentiate into plasma cells. one heavy chain constant gene segment

can be substituted for another without altering the VDJ combination. In

other words. a given variable region gene can be expressed in association

with more than one heavy chain class or subclass so that at the cellular

level. the same antibody specificity can be associated with the synthesis

of an IgM (characteristic of the primary response) or with an IgG

immunoglobulins (characteristic of the secondary response).

It has been estimated that an individual is capable of producing up

to 109 different antibody molecules. How is this vast diversity generated

from a limited number of germ line genes? The existence of a large

number of V genes and of a smaller set of D and J segments in the

germline DNA contribute to the first mechanism of antibody diversity

(26. 27). Second. imprecise Joining of various V gene segments from

either heavy or light chains with one of any J segments and one of any D

segments (for heavy chain). creates sequence variation at the pOints of

recombination. augmenting diversity significantly (28. 29). Third. random

association of Land H chains plays an important role in increasing

diversity (27). Last. somatic mutations in the V genes. including point

mutations and frameshift mutations. add to the antibody diversity (30.

31, 32).

Monoclonal Antibody

The diversity. speCificity. and biological activity of antibodies make

them potentially important reagents for therapeutic and diagnostic

purposes in the clinic and the laboratory.

18

Numerous efforts have been made for many years to find a way to

generate a homogeneous, antigen-specific antibody with high affinity.

The achievement of hybridoma technology by Kohler and Milstein (33) in

1975 that reliably yields murine monoclonal antibodies directed against

a given antigen was a remarkable technological breakthrough in the

history of immunology. There are two different cell lines involved. One

line is derived from an immunized mouse and consists of

immunoglobulin-bearing splenic B cells possessing the capacity of

specific antibody secretion; the other (fusion partner) is a malignant

myeloma cell and conveys immortality in culture on the cells with which

it is hybridized. The difference between these two is the presence or

absence of the hypoxanthine guanine phosphoribosyl transferase

(HGPRT) and thymidine kinase (TK) enzymes necessary for survival in the

growth medium containing hypoxanthine, aminopterin and thymidine

(HAT) (34). The principle behind HAT selection is that cells can

synthesize DNA either using a salvage pathway or by de novo synthesis.

Aminopterin is a dihydrofolate analog that blocks the reactivation of

tetrahydrofolate which is critical for the de novo synthesis of purines,

such as AMP, GMP, TMP (34). At this point, cells are required to use

salvage pathways, which depend on the presence of DNA precursors

(hypoxanthine and thymidine) supplied in the media, and the enzymes

HGPRT and TK. The B cells are HGPRT+ and TK+, while the myeloma

cells are HGPRT- and TK-. Mter fusion. the cells are placed in the

selective HAT medium. The fused B cells provide the hybridoma with the

capacity to produce a specific antibody and the capacity to produce

19

HGPRf and TK; the fused myeloma cells provide the hybridoma with the

capacity to proliferate indefinitely. The non-fused myeloma cells, still

HGPRf deficient, are easily killed by HAT medium. The non-fused B cells

are unable to survive after a few rounds of proliferation. A screening

process is followed to select clones of interest from the large number of

hybrids produced by fusion. This process involves culturing the

hybridomas at limiting dilutions, ensuring that the antibody-producing

hybridoma is derived from a single hybrid cell. Once obtained, the

hybridoma cells can then be frozen, grown in mass culture or injected

intraperitoneally into an animal to form tumors which produce ascites

containing large amounts of the antibodies. The antibody produced by

this technique is called monoclonal antibody (mAb).

Hybridoma technology requires fusion of an antigen-specific B cell

with a malignant fusion partner cell. For humans, unfortunately, an

entirely satisfactory human fusion partner for human B cells has not

been found. The use of a mouse myeloma cell line for a human fusion

partner cell, in most cases, leads to preferential loss of human

chromosomes and instability of the hybrids (35). Although Cote et al.

(36), Fujinaga et al. (37). and Lake et al. (38) have reported that mouse

myelomas can be fused to human B cells to produce stable hybridomas,

the production of monoclonal antibodies remains ineffiCient and labor

intensive. Moreover, antigen-specific B cells suitable for fusion are not

readily obtained from humans because of the possible ethical problems

related to deliberate immunizations in humans as well as the fact that B

cells making the specific antibodies desired are relatively rare in

peripheral blood.

20

Although murine monoclonal antibodies are easier to obtain and

of value in diagnosis and therapy of human diseases, their effectiveness

is limited because they have a short survival time in humans and induce

a xenotypic human anti-mouse antibody response (HAMA) that

neutralizes their therapeutic effect (39). Furthermore, the therapeutic

effect induced by murine monoclonal antibodies is restricted because the

Fc regions of some murine antibodies are poor at mediating their effector

functions, such as complement fixation and ADCC, in humans and they

are relatively ineffective as cytotoxiC agents (39).

To circumvent these difficulties, making genetically engineered

antibody variants would be an attractive option.

Antibody ED.gineerlng

Genetic engineering has been used to creat the chimeric antibody

that combines the murine variable region with the human constant

region (40) or the humanized antibody that combines the murine

hypervariable regions (CDRs) with the human constant and variable

framework regions (41).

In construction of chimeric antibodies, the murine VH gene

segment is first combined with human heavy-chain constant region gene

segment to make the heavy-chain gene construct. Subsequently the

murine VL gene segment is connected with a human light chain constant

region. Both the heavy- and light-chain gene constructs are then

transfected into a nonsecretor myeloma cell line. The resulting

transfectoma secretes the mUrine/human chimeric antibody. In

principle, any murine variable domain can be paired with any human

constant region isotype so that the optimal combination of antigenic

specificity and effector functions can be selected.

21

However, Jaffers et aL (42) showed when the murine antibody

OKT3 is used in patients, much of the antibody response is directed

against the V region rather than the C region. In addition, Bruggemann

et al. (43) found that although the use of a murine variable region

minimizes a HAMA response to some extent, anti-idiotypic antibodies

that recognize portions of the murine variable region still arise. Hence

chimeric antibodies, in which the V region remains murine, may still be

immunogenic.

Winter and colleagues (41) developed an effective technique for

"humanizing" murine antibodies which involved the identification.

sequencing, and synthesizing of the murine complementarity­

determining region (CDR) genes. with subsequent introduction of these

genes into a completely human antibody framework. Antibodies made in

this way have been called reshaped, CDR-grafted or humanized.

The first fully humanized antibody. CAMPATH-IH (41), which

binds to antigen on lymphocytes. has been used to treat two patients

with non-Hodgkin lymphoma (44), one patient with systemic vasculitis

(45) and 8 patients with rheumatoid arthritis (46). No antibody response

against CAMPATH-IH was detected in these eleven patients and the

treatment induced disease remissions and relieved symptoms. This

suggests that humanized antibodies would be substantially less

immunogenic in human than murine and chimeric antibodies.

22

Although the method used for making "humanized antibody" may

overcome the antigenicity complications of chimeric antibodies. it is

significantly more difficult to accomplish. First. problems are

encountered because the murine and human frameworks differ

sufficiently so that the fit of mouse hypervariable loops on human

immunoglobulin may be problematic. Second. loss or alteration of

certain residues in the original murine framework that make key

contacts with CDRs to help maintain their conformation will distort the

shape of the CDRs and reduce or abolish their affinity for antigen. Third.

murine CDRs still have murine antigenic determinants. By definition.

the production of fully human monoclonal antibody proteins by genetic

engineering can obviate all of the above problems.

Molecular Cloning and Expression of Ig Genes

With development of recombinant DNA technology. it has become

feaSible to clone. construct and express fully human Ig molecules.

Rearranged immunoglobulin genes can be rescued from any given B cell

population or from a hybridoma producing a potentially useful

monoclonal antibody. Once the heavy and light immunoglobulin genes

have been cloned and sequenced. they can be genetically engineered to

possess properties not normally found in nature.

Two crucial advances in the late 1980s made this a feaSible

approach. The first was the use of the polymerase chain reaction (PCR)

23

to isolate VH and VL genes (47, 48, 49). The second was the development

of expression systems for the secretion of antibody fragments in

functional form, using E. coli as a host (50, 51).

The peR allows the specific and rapid isolation of genes or

members of gene families (52). The genes can be isolated with a high

degree of specificity by primer-directed amplification from highly

heterogeneous DNA preparations. The only requirement for isolation by

PCR is that there is some pre-existing knowledge of the gene sequences

at either, or both, of the 5' and 3' ends. For the immunoglobulin variable

domains, there are known databases, such as the Kabat claSSification

(53), that document the sequences ofVH and VL genes from different

species, including mouse, rabbit, rat and human. The homologies shared

by the VH and VL genes within a given species allow the design of

primers that can be used in PCR reaction to isolate the variable domain

genes (47, 48, 49). Thus, primers have been designed to hybridize to the

5' and 3' ends of VH and VL genes and used to isolate both clonal

antibody genes from hybridomas and diverse VH and VL gene repertoires

from antibody-producing cells. There are nucleotide bases at the 5' and 3'

ends of VH and VL genes that are not particularly well conserved, and at

these positions redundancy is incorporated into the primers during

oligonucleotide synthesis. In addition, the primer can also be designed to

have internal restriction sites to facilitate cloning of the genes into

vectors for expression (47, 48. 54).

The first advantageous host for the expression of recombinant

proteins is E. coli with such characteristics as rapid growth and the

24

availability of many different types of cloning vectors. Initial attempts to

express whole antibody molecules in E. coli were largely unsuccessful (55,

56) because prokaryotic cells lack endoplasmic reticulum (ER) and

processing Golgi apparatus. The endoplasmic reticulum is known to

perform glycosylation of Fc portion of antibody, proper folding, disulfide

bond formation and heteromeric association of heavy and light chains

that are all required to obtain a functional antibody. However, successful

expression in E. coli of antibody fragments with antigen binding

activities, such as Fv (50), single chain Fv (scFv) (57) and Fab (58). was

obtained when such fragments were secreted into the periplasmic space

or accumulated as intracellular inclusion bodies (59, 60, 61). Expression

of immunoglobulin fragments in E. coli offers a rapid, convenient route

for their large-scale production.

The single chain Fv (scFv) is composed of immunoglobulin variable

heavy (VH) and variable light (VL) chains joined together by a flexible

peptide linker. The first reports of scFv demonstrated that rearranged

heavy and light chain immunoglobulin variable regions could be

engineered for expression in bacteria as a single polypeptide which

retained the ability to bind antigen (62, 63). Different types of scFv

peptide linkers have been designed to join the heavy and light variable

chains. The most common linker used is (GlY4Ser)3 because it is

hydrophilic, flexible and relatively free of side chains that might interfere

with scFv refolding or the capacity to bind antigen (64). The advantages

of scFv are 1) the small size that enables scFv to be rapidly cleared from

circulation, thus reducing its immunogenicity (65); 2) the capability to

penetrate the micro-vasculature of solid tumors faster and more evenly

than intact IgG, F(ab')2 and Fab fragments in an experimental colon

25

carcinoma xenograft model (66). This suggests that the relatively small

size of scFv might make them more effective for solid tumor therapy.

However, the rapid clearance may be a disadvantage pharmacokinetically

because the molecule may be in the circulation for too short a time to

bind to the target tissue.

Further studies have revealed that the scFv can be fused to the

Pseudomonas exotoxin which lacks the cell-binding unit to generate an

immunotoxin using gene cloning techniques (67). This study showed

selective cytotoxicity to human tumor cells bearing the B3 antigen in cell

culture and caused complete regression of human tumors growing in

immunodeficient mice.

In addition, Chaudhary et al. (68, 69) created a single-chain Fv

immunotoxin targeted against human cells that express IL-2 receptors.

The scFv gene was attached to a portion of the Pseudomonas exotoxin

gene lacking the exon for the cellular binding domain. This gene

construct was then expressed in E. coli to form an antibody-toxin

conjugate that selectively killed human IL-2 receptor positive cells.

Single-chain Fv fragments were chosen to avoid the destruction of Fc

receptor-positive cells. When the immunotoxin was tested on cells

harvested from patients with IL-2 receptor-positive leukemia, IL-2

receptor-positive, but not -negative, cells were killed (70). Destruction of

IL-2 receptor-positive cells can be important in developing therapy for

GVHD, transplant rejection, and autoimmune diseases. Clearly, the

ability to target certain human tumor cells for destruction can have a

wide variety of clinical applications.

The other area of antibody construction technology involves the

generation of combinatorial immunoglobulin libraries displayed on the

surface of filamentous bacteriophage (71, 72, 73, 74, 75).

26

This approach for expressing the variable regions of antibodies

fused to a phage gene product is done by coexpression of libraries for

heavy and light chain sequences. It consists of two steps which are

accomplished by first producing separate heavy and light chain libraries

and then randomly combining the two libraries into a single vector

population which then assembles the polypeptides on the surface of the

phage by fusion of heavy chain with pVIII, the major capsid proteins of

the filamentous phage (76) or pIlI, the minor coat proteins on the one

end of the phage (77). Thus, the desired antibody fragments as a form of

Fab (78) or scFv (79) could be obtained by screening the libraries against

a specific target antigen bound to a solid support.

Perhaps the most exciting prospect of the approach employing

phage display libraries is the generation of a complete human

combinatorial immunoglobulin gene expression library from non­

immunized humans, that directly produces human monoclonal antibody

fragments. Mullinax et al.(80) produced a human Fab combinatorial

library. They isolated mRNA from peripheral blood lymphocytes of donors

who had been immunized 6 days before with tetanus-toxoid, and then

used peR to construct separate libraries of the genes encoding the heavy­

chain and light chain fragments. Next they randomly combined the

27

heavy- and light-chain libraries into a bacteriophage lambda vector that

was subsequently used to infect E. coli. Positive clones were selected by

making replica plaque lifts onto hybridization filters and identifying

antigen-reactive clones with radiolabeled tetanus toxoid.

A more convenient and simpler screening method was developed by

McCafferty et al. (81) when expressing the antigen-binding variable

regions as surface proteins in a bacteriophage. This was done in a

filamentous bacteriophage. in which the VI-! and VL genes from an anti­

lysozyme antibody were linked by sequences encoding a flexible peptide

linker and then connected to the N-terminal end of a gene encoding a

surface protein. The VI-! and VL domains (scFv) were expressed on the

surface of the phage. bound specifically to antigen. and were isolated

from a mixture of phages by affinity chromatography. Two unique

benefits are derived from this screening method. One is speed. The other

is increased sensitivity because it can be used to isolate a very rare phage

(l inl06).

Tumor-inflltratin~ B Lymphocytes

It is well known that the infiltration of lymphocytes. macrophages

and granulocytes within tumors is a characteristic of some tumors and

may be associated with a positive prognosis (82. 83. 84). It has been

interpreted as an indicator of the active immune defense against the

tumor (85). On the other hand. the infiltration does not curb and reduce

the growth of most malignant neoplasms (86. 87). Among tumor­

infiltrating lymphocytes. T lymphocytes have been studied intensively

28

and described as a major infiltrating population in many tumors such as

breast cancer (88), metastatic malignant melanoma (89), lung cancer (90)

and metastatic liver cancer (91). Although B lymphocytes and plasma

cells are also observed in various tumor infiltrates (92) and can account

for up to 15% of the total lymphoid infiltrate in tumors including

ovarian cancers, melanomas and breast cancers (93), the role of tumor­

infiltrating B cells (TIL-B) in the tumor-host interaction is unclear (94).

Little is known about the tumor antigens that might be recognized by

TIL-B or about the significance of their presence within human tumors.

This is due to the difficulty of analyzing the B cell repertOire in vitro. The

plaque forming assays are useful for studying the activity of Ig-secreting

cells (95), but does not allow further characterization of the cells or their

products. Epstein-Barr virus (EBV) transformation is a way of

immortalization of human B cells (96). However. B cell lines

immortalized by EBV transformation are limited and relatively unstable

(97).

A method of B cell clonal expansion has been deSCribed (98). which

utilizes IL-2. mitomycin C-treated T cell feeder layers and a monoclonal

antibody directed against the CD3 complex. B cells cultured under these

conditions are activated and secrete Ig at a high frequency (99). In the

study. a limiting dilution assay was used to demonstrate an increased

frequency of responding B cells. Purified B cells were cultured at

densities of between 1000 cells and 0.5 cell per well with fresh.

mitomycin C-treated cells or T cell clones stimulated by immobilized mAb

to CD3. After 5 days in culture. the number of wells containing Ig-

29

secreting cells was determined. and the frequency of responding B cells

was calculated. The frequency responding to anti-CD3-stimulated T cells

was very large and greatly surpassed that induced by other polyclonal

activators, such as PWM, EBV. etc. B cells cultured with anti-CD3-

stimulated T cell clones responded better than did those cultured with

mitomycin C-treated T cells. The addition of exogenous IL-2 or IL-6 to

cultures supported by activated mitomycin C-treated T cells enhanced

the frequency of responding B cells. whereas IL-4 did not increase the

generation of Ig-secreting cells and inhibited the augmentation of B cell

responses induced by IL-2. In addition. B cells activated byanti-CD3-

stimulated T cells produced all three Ig isotypes. IgG. IgM and 19A. These

results demonstrated that under optimal culture conditions, T cells

stimulated with immobilized anti-CD3 could activated the majority of

human peripheral blood B cells to produce Ig. Therefore. this method of

B cell culture could be useful in analyzing different B cell populations.

However, the use ofT cells from the B cell donor as feeder cells should be

cautious because it would introduce a factor of variability. This is due to

that cancer often causes important alterations in T cell populations

(100, 101) and malfunctioned T cells could bias the study ofB cells, by

this method, in the patients.

Recently Barbuto et al. (102) reported a novel method of in vitro B

cell expansion using MOT cells (a HTLV II-transformed human T cell line)

as feeders. MOT cells have been able to induce clonal expansion and Ig

secretion by B lymphocytes from peripheral blood, spleen, and tumor­

infiltrating populations. The use of this stable T cell line avoids one

30

factor of variability in the analysis and allows better comparisons among

different individuals. Based on this development. Punt et al. have

reported that anti-tumor antibodies were produced by human tumor­

infiltrating B lymphocytes (TIL-B) from about 70% of various tumors. The

major isotype of immunoglobulins secreted by expanded TIL-Bin vitro was

IgG (93). Therefore. we hypothesized that TIL-B could be detected in

many or perhaps all human tumors. would be naturally enriched in the

tumor. and would be a potential source for making human anti-tumor

antibodies.

However. it is unclear that what is a role played by TIL-B cells in

tumor-host interaction. It has been demonstrated in mouse model

studies that tumor antigen-specific IgM blocked H-2-restricted and non­

restricted cytotoxiC effectors and promoted tumor growth (l03). But in

many cases tumor-specific antibodies could mediate ADCC (5).

complement-dependent cell killing (104) and direct tumor cell killing

(105).

Melanoma-associated Antigens

Malignant melanoma. like other cancers. can be conSidered as a

disorder of cell differentiation and proliferation. Normal melanocytes

arising from precursor melanoblasts usually undergo a series of

differentiation events before reaching the final cell differentiation state

(106). Thus. a tumor can arise at any given stage of maturation when an

arrest in the differentiation process has occurred without loss in the

proliferative capacity.

31

Analysis of malignant melanomas has revealed that most

melanoma -associated antigens are heterogeneously expressed to varying

degrees. With the help of monoclonal antibodies using classical

hybridoma methodology, a variety of melanoma-associated antigens have

been identified in malignant melanoma. Most of these have been

identified using mouse monoclonal antibodies but a few have been

detected using human monoclonals. These antigens are classified into

two categories: one is melanocytic differentiation antigens which are

expressed during normal melanocyte differentiation: the other is

melanoma progression antigens that are expressed during malignant

transformation of melanocytes.

Melanoma differentiation antigens represent a group of molecules

expressed almost exclusively by melanocytic cells including normal naevi,

dysplastic naevi, early and advanced primary tumors and metastatic

lesions (I07). Therefore, these antigens can be used to distinguish

melanocytes from other cell types.

Of melanoma differentiation antigens, the S 1 00 protein is an

acidic cytoplasmic calcium -binding protein with a molecular weight of 21

lID (108, 109) and used as a good diagnostic marker particularly for

those lacking typical melanoma morphology. It is expressed by a high

proportion of melanomas. HMB-45 is another cytoplasmic molecule

found in more than 97% of melanomas (110). Tyrosinase (Ill) and gp75

(112) are transmembrane glycoproteins that are located in melanosomes,

the organelles where melanin is synthesized. Other melanoma

32

differentiation antigens. such as HMW-MAA. GD2. GD3 (113) have also

been identified.

Melanoma progression antigens represent a large variety of

molecules. the expression of which varies with tumor progression as well

as with the differentiation state of the melanocytic cell. Progression

antigens can be subdivided into two separate groups. those which are

upregulated during tumor progression and those which are

downregulated during progression.

Some members of the integrin family. such as the integrin a2j31

(114). the integrin a5j33 (115) and ICAM-l (116), etc. were found to be

associated with tumor progression. since their expression increased

markedly from small primary tumors to metastatic lesions. Melanoma

cells also express a wide variety of receptors for growth factors (117),

some of which are related to tumor progression, such as epidermal

growth factor (EGF) receptor, transferrin receptor (gp97), and

transforming growth factor (TGF) receptor.

There are few molecules that are known to be downregulated

during tumor progression. One of them is HLA class I molecules which

are decreased Significantly from small primary lesions to metastatic

tumors (118). However, it is not clear that these melanoma-associated

antigens are the ones which elicit an effective immune response in

patients with melanoma.

More recently cytotoxiC tumor-infiltrating T cell lines (TIL) from

malignant melanoma have been used to identify and eventually capture

the genes of the melanoma-associated antigens through a T cell receptor

33

and HLA-restricted mechanism. These include the MAGE-l (119) and

MAGE-3 (120) antigens, the MART-I antigen (121), gpl00 (122) and

tyrosinase (123). These studies have facilitated an analysis of the

repertoire of the human T cell response to tumors and of the antigens

they recognize. The antigens, such as MAGE-I, MAGE-3, gp 100 and

MART-I, recognized by cloned cytotoxic T cell lines are presumably those

involved in natural host defense in melanoma but are of somewhat

limited potential because of MHC restriction. It is possible that the

antigens identified by TIL-B will be more generally recognized and will

elicit immune responses in the majority of patients.

Immunology and Immunotherapy of Melanoma

The ultimate purpose of the host immune system is to develop

responses that protect the individual against the growth and spread of

malignant tumors, including melanomas. The cellular immune response

involves cytotoxic T lymphocytes (CTL) which kill the target cells

specifically in association with MHC, natural killer (NK) and

lymphokine-activated killer (LAK) cells that kill tumor cells

nonspecifically by an unknown mechanism. The humoral immune

response consists of antibodies produced by B lymphocytes and plasma

cells, which bind to tumor cells and subsequently kill them via

activation of complement system, ADCC and direct killing.

Immunotherapeutic approaches against malignant melanoma are

designed to use 1) tumor antigens (vaccines) that stimulate the immune

system against the host tumor, 2) monoclonal antibodies that directly

react against the host tumor cells and work in collaboration with T

lymphocytes, 3) cytokines that show direct cytotoxicity on tumor cells,

enhance the antitumor immune response, or both, 4) stimulated and

cultured lymphocytes that specifically or nonspecifically kill the host

tumor cells, and 5) cytokine gene-transfected tumor-specific tumor­

infiltrating lymphocytes (TIL) that specifically deliver the gene products

at the site of the tumor tissue.

34

A variety of different tumor vaccines for melanoma are under

investigation for use in this specific active immunotherapy. These

include 1) irradiated allogeneic melanoma cell vaccine. prepared from

whole melanoma cells (124). 2) polyvalent melanoma antigen vaccine.

prepared from shed melanoma-associated antigen from cultured

melanoma cells (125). 3) vaccinia oncolysates. designed to increase the

immunogenicity of the melanoma antigens by the introduction of foreign

proteins (126). 4) recombinant vaccinia virus-based melanoma vaccine (v­

p97NY) where the melanoma antigen p97 is presented in context with

MHC antigens (127), and 5) immunizations with murine anti-idiotypic

mAbs (128). An immune response to the vaccinations has been observed

and the vaccinated patients had an improved outcome. The exciting

possibility for melanoma vaccines is in the prevention of cancer.

Vaccines can prevent melanomas in animals (129); if the same holds true

for humans. melanoma vaccines could be used in high-risk populations.

Antibodies against glycolipids. GD2 and GD3. appear to be

benefiCial in retarding melanoma growth, both in humans and in

animals (130, 131, 132, 133). Monoclonal antibodies conjugated to

35

toxins (ricin A chain) have also undergone testing in clinical trials with

responses of tumor regression (134. 135).

To enhance immune effectiveness of mAbs. the application of

lymphokine or cytokines is being investigated. Cytokines can augment

ADCC mediated by mAb and increase the antitumor effects of mAb in

animal models (136). IL2 is a lymphokine that induces generation of LAK

cells. proliferation of activated tumor-specific T lymphocytes, and

induction of natural killer cells secreting IFN-gamma, lymphotoxin and

TNF-alpha (137, 138). In humans, alpha-interferon (IFNa) has been

shown to alter mAb distribution and lead to increased tumor uptake of

mAb in melanoma patients (139).

Adoptive immunotherapy is defined as the transfer to the tumor­

bearing host of immunologically reactive cells which can mediate

antitumor effects either directly or indirectly. Initial investigations were

directed towards lymphokine-activated killer (IAK) cells, generated by

culturing lymphocytes in high doses of IL-2 (140, 141). Adoptive transfer

of both LAK cells and IL-2 has resulted in antitumor effects in both

murine and human systems. Later, it has been found that TILs, isolated

from a variety of solid tumors and expanded in vitro, may be significantly

more potent than LAK cells in the treatment of human solid tumors in

vivo (142). Human clinical trials have yielded similar promising results.

with TIL therapy producing a greater response rate than therapy with

LAK cells. as well as responses in patients who had previously failed LAK

cell therapy (143. 144).

36

The tumor-homing character of TIL, demonstrated by the specific

localization of III In-Iabled TIL to the tumor tissue (145), opens the

possibility of using the TIL as a vehicle to carry a cytotoxin and

specifically deliver it to the tumor tissue. The use of a TNF -alpha gene

encoded retrovirus-transduced TIL in patients with melanoma is feasible

to achieve a higher antitumor response by secreting TNF-alpha at the

site of tumor (146, 147). The trial is currently ongoing, and it is too early

to determine the benefit of this therapeutic approach. The ability to

modify TILs with gene therapy, however, allows one to introduce a variety

of genes which may be of therapeutic benefit. This includes the

introduction of other cytokines, such as IL-2, alpha-IFN, or gamma-IFN,

or the introduction of Fc receptors would allow the use of TIL in

conjunction with mAbs.

Immunotherapy and subsequently gene therapy will continue to

play an increasingly important role in the treatment of malignant

melanoma, as well as other human malignancies. Advancements have

been made and will continue to be made in these areas, including the

use of mAbs, tumor vaccines, adoptive cellular therapy, and gene

therapy.

Rationale of Study

Human anti-tumor monoclonal antibodies (mAbs) made from

tumor-reactive TIL-B could enable us to identify and characterize tumor

target antigens and the overall repertOire of anti-tumor immune

responses. The characterization of tumor antigens should in tum prOvide

37

us a knowledge base to facilitate and understand TIL-B repertoire. In

addition, the anti-tumor mAbs are ideal reagents which could be widely

applied to the clinics and the laboratories as diagnostic and therapeutic

agents.

Although there are technical difficulties in making entire

molecules of human mAbs, in vitro synthesis of immunoglobulin

fragments which retain antigen-binding activity is now possible using

recombinant DNA technology. In addition, the availability of human TIL­

B expanded in vitro gives us the opportunity of capturing the genes of Ig

variable regions via PCR.

The objective of this study was to develop a methodology for

making human anti-tumor scFv antibodies from TIL-B and use them to

1) identify and characterize tumor antigens, 2) to study the repertOire of

human anti-tumor immune responses and 3) explore the possibility of

their clinical use for diagnostic or therapeutic purposes.

To accomplish this objective, we isolated and expanded TIL-B from

various tumors, especially from melanomas, in the presence of EBV. The

transformed B cells producing tumor-specific antibody were further

cloned by limiting dilution. The tumor-reactive B cell clones, thus

obtained, were used to construct a scFv by recombinant DNA

technology.

This study demonstrates that we have found an effective way to

make human anti-tumor scFv antibody by combining in vitroTIL-B

expansion method and PCR-Ig gene cloning technique. Establishment of

this methodology will enable us to make additional human antibody

molecules against various tumors.

38

39

CHAPTER 2

MATERIALS AND METHODS

Cell Lines

The melanoma cell lines used as tumor targets were A375, 81-61,

SK-Mel-2, SK-Mel-24 and SK-Mel-28, all from American Type Culture

Collection (ATCC). The mouse melanoma cell line, BI6. was also

obtained from ATCC. The Mel-K and Mel-G lines were established from

melanoma biopsies in our laboratory. Hey (ovarian cancer), SKOV3

(ovarian cancer), SW480 (colon cancer), NC-37 (human PBL lymphoblast

cells), A549 (lung cancer), H596b (lung cancer), BT4 7 4 (breast cancer),

SK-BR-3 (breast cancer), LNCAP (prostate cancer). PC-3 (prostate

cancer), Daudi (Burkitt's B cell lymphoma). Raji (Burkitt's lymphoma),

and MCF-7 (breast cancer) were also obtained from ATCC. Human

foreskin fibroblasts and other short-term melanoma cell lines, including

JH1308, MK457, PsI273, Mel-V, Mel-D. Mel-T, Mel-A. Mel-Wr, Mel-Ws,

Mel-Th and Mel-R were obtained from the tissue culture laboratory of the

Arizona Cancer Center. These limited passage lines were established from

biopsies and stored frozen after 5-15 passages. Mel-Ke, Mel-M. Mel-C and

Mel-Vw were fresh melanoma cells made from patients' biopsies in our

laboratory. Melanoma cell lines were confirmed to be melanoma by HMB-

45 and S-100 staining. All fresh melanomas were confirmed to be

melanoma by standard histology.

40

Preparation of Tumor-infiltratinll B Lymphocytes (TIL-B)

Single cell suspensions were obtained from surgical tumor biopsies

by cutting into small pieces (0.2-0.5 cm3) and digesting with 0.015%

(w/v) DNase (Sigma Chemical Co., St Louis, MO) and 0.15% (w/v)

collagenase (Sigma Chemical Co.). These cell suspensions were incubated

overnight at apprOximately 3-5X106 cells/ml in complete medium (CM,

RPMI-1640 (Gibco BRL, Grand Island, NY) supplemented with 10% fetal

bovine serum (Tissue Culture Biologicals, Tulane, CAl. 2mM L-glutamine

(Irvine Scientific, Santa Ana, CAl. 100U/ml penicillin and 100ug/ml

streptomycin (Irvine Scientific)) in T75 tissue culture flasks (Costar,

Cambridge, MA) for removal of adherent cells. Non-adherent cells were

poured off and centrifuged over a discontinuolls Ficoll gradient to enrich

for the tumor-infiltrating lymphocytes (TILs) (148). These cells were then

rosetted once with sheep red blood cells which were pretreated with S-(2-

aminoethyl)-isothionronium bromide hydrobromide (AET) (Aldrich

Chemical Co., Milwaukee, WI) to remove T lymphocytes (149) and the

remaining cells were washed, resuspended in CM and used as TIL-B.

However, tumor-enriched fractions were kept in culture in CM for use in

the preparation of tumor cell ELISA plates, as described below.

B Cell Expansion and Immunoillobulin Secretion in vitro

B cells at concentrations ranging from 300 cells/well to 2X104

cells/well were plated in 96-well U-bottom tissue culture plates (Costar,

Cambridge, MA) in the presence of 25% crude B95-8 supernatants (a

source of EBV for B cell stimulation and transformation) (150). As

41

controls. each plate had a group of wells that received no B cells and

wells that received no EBV supernatants. These latter wells served as

controls for the possible release of Ig from the surface of contaminating

tumor cells among the TIL-B preparation. After 10 days in culture. the

supernatant of each well was harvested and assayed for human

immunoglobulin (Ig) by ELISA.

ELISA for Human Ig and Ig Classes

Falcon flexible 96 well assay plates (Becton Dickenson Labware.

Oxnard. CAl were coated overnight with 50ul/well of goat polyclonal

anti-human Ig antibody (Tago Immunologicals, Burlingame. CAl at

1 ug/ml in carbonate buffer. pH 9.6. An amount of 50ul of B cell culture

supernatant or of a human rg standard was added per well and incubated

for 1 hr. After washing. a mixture of alkaline phosphatase-labelled goat

polyclonal antibody anti-human kappa chain (Tago Immunologicals) and

alkaline phosphatase-labelled goat polyclonal anti-human lambda chain

antibody (Tago Immunologicals), at appropriate dilutions. was added to

the wells and incubated for 1 hr at room temperature. After thorough

washing. 50ul of the substrate. p-nitrophenyl-phosphate (p-NPP) (Gibco

BRL. Grand Island. NY) at a concentration of Img/ml in carbonate

buffer. pH 9.6 was added to each wcll and the plate was read in a

Microplate reader (Dynatech MR600) at 405 nm.

To determine the different Ig classes. such as IgG. IgM and IgA.

present in the supernatants, the supernatants were assayed in three

separate plates and alkaline phosphatase-labelled goat polyclonal

antibody against each class was used in each plate.

Frequency of Response (F{rll Calculation

42

Each well was scored as positive or negative for the presence of

human Ig by comparison of its absorbance with negative control wells. A

well was considered positive when its absorbance exceeded the average

absorbance of the control wells by more than 3 standard deviations. The

frequency of responding cells was calculated. according to the Poisson

distribution. based on a linear rcgrcssion curve constructed with the

frequency of negative wells for human Ig at each cell concentration (151).

Detection of Anti-tumor Antibodies by Tumor Cell ELISA

Supernatants positive for Ig were tested for direct reactivity against

autologous tumor cells and against tumor cell lines of the same and

different histologic types.

Tumor cell suspensions were plated into 96-well flat bottom tissue

culture plates (Costar). When they became confluent. the supernatant

was carefully removed and the cell monolayer was fixed by addition of

100ul/well of glutaraldehyde (0.25% in phosphate-buffered saline (PBS))

for 5 min. The glutaraldehyde was then removed. and the plate was

washed with tris-buffered saline (TI3S) followed by addition of 100ul/well

ofTBS containing 1% bovine serum albumin (BSA) and 0.1 % sodium

azide. The plates were kept at 4°C for future use.

43

The supernatants of the TIL-B culture were added to the wells and

incubated for 2-4 hr at 37°C. Then the plate was washed in TBS three

times and goat anti-human Ig-alkaline phosphatase conjugate (Tago

Immuno!ogicals). diluted at 1:5000 in TBS-BSA was added and incubated

for another 2-4 hr at 37°C. The plate was then thoroughly washed as

above. The substrate. p-NPP. was added and the plates were read at 405

nm.

Positive wells were those with absorbance readings exceeding the

average absorbance of control wells by 3 standard deviations.

Cloning EBV -Transfonned B Cell Lines by Limiting Dilution

Once a B cell population in a well was found to produce IgG. we

determined the isotype of light chain being secreted by the B cell lines.

We then tested the reactivity of IgG produced by those lines against a

panel of autologous and allogeneic tumors of the same histology. and

unrelated histology as well as against normal human tissue.

Subsequently we cloned the tumor-specific B cell lines by limiting

dilution at 0.5 cell/well in 96 V-bottomed well plates in CM. After 10

days in culture. plates were screened visually for cell proliferation. The

wells with cell growth were marked and their supernatants were collected

for further study. After another 10 days of incubation. the cells still

growing were transferred into 24-well plates for further expansion. After

7 -day growth in 24-well plates. clumps of B cell clones were visible. The

supernatants were harvested and tested for IgG secretion and tumor­

reactivity by tumor cell ELISA. Based on the broad tumor reactivity. one

44

tumor-reactive B cell clone was chosen for mRNA extraction and cDNA

synthesis.

Messenger RNA Isolation and cDNA Synthesis

The micro-FastTrack mRNA Isolation (Invitrogen Corp, San Diego,

CAl and cDNA Cycle Kits (Invitrogen) were used to isolate mRNA and

synthesize first strand cDNA respectively. according to the instructions

provided with the kits.

When one (5-3E) of the EBV -transformed tumor-specific B cell

clones was grown to a density of lXl06 cells/ml, it was harvested and

washed once in phosphate buffered saline (PBS) and resuspended in lysis

buffer in the presence of proteinase K at 50°C for 30 min. The

chromosomal DNA was fragmented with a 20-gauge needle and syringe so

that the mRNA was able to be recovered with oligodeoxythymidine (dT)

cellulose. The mRNA was eluted from the oligo dT cellulose in distilled

water on a spin-column. The first strand of cDNA was then reverse­

transcribed from the mRNA template using random primers at 42°C for 2

hours. The reverse transcriptase was subsequently inactivated for 2 min

° at 100 C.

Capture and Assembly of Ig V Region by Polymerase Chain Reaction

(PCR)

Polymerase chain reaction was performed for both immunoglobulin

heavy and light chains using cDNA as a template. Sets of sense and anti­

sense PCR primers were designed to amplify the VH region starting from

45

amino acid position 1 to 110, and also to amplify VL region starting from

amino acid position 1 to 105 (numbering according to Kabat et al. (152)).

Appendix A shows the primer sequences used for the initial amplification

and assembly. A Nco I site is embedded at the 5' end of VH primer, while

a Xho I site' was incorporated into the 3' end ofVL primer. All primers

were designed so that the resulting PCR products, once cloned into the

expression vector, were ready for "in-frame" expression of the antibody

fragments.

PCR cycling temperatures were as follows: melt at 94°C for 1 min,

primer anneal at 50°C for 2 min, extension at 72°C for 3 min. PCR was

run in a Perkin-Elmer thermo cycler (Roche Molecular Systems, Inc.,

Branchburge, NJ) for 30 cycles in a volume of 50 ul using Taq DNA

polymerase. The PCR products were visualized on a 1 % agarose gel

stained with ethidium bromide.

After 1 st round of PCR, we purified VH and VL genes respectively

on glass beads (QIAGEN, Chatsworth, CAl. The 2nd round of PCR was

conducted to link VH and VL in VH-VL order by a (GlY4Ser)3 polylinker

using splicing overlap extension (153). This technique is called "PCR

soeing". The obtained scFv gene and scFv protein thereafter were

generically named after "AZM 1 ".

Clonin~ Immunoglobulin Variable Regions into Sequencing Vectors

Once the VH, VL and assembled scFv (AZMl) were amplified by

PCR using primers with the proper restriction endonuclease sites on the

ends of the products, they were electrophoresed on a 1 % agarose gel,

46

excised from the gel and purified on glass beads. The DNA fragments were

then ligated into a sequencing vector (pT7 blue) with overhanging 5'

thymidines (154) so that the 3' adenosines on the PCR products (155)

could base-pair with the vector. The ligation reaction consisted of 5 fold

excess insert over vector concentration. Ten units of T4 DNA ligase were

added and the ligation reaction was incubated at lSoC overnight.

The ligated DNA was subsequently used to transform competent E.

coli (Novablue cells, Novagen, Madison, WO, according to the protocols

described (156). Competent Novablue cells (50ul) were thawed from -70°C

and mixed with the ligation reaction for 30 min on ice. The cells were

heat-shocked at 42°C for 60 seconds, 100ui SOC medium was added to

the cells after which the cells were incubated at 37°C for 1 hour. Finally,

the transformed cells were plated on Luria Bertoni (LB) agar containing

50ug/ml carbenicillin, O.lmM isopropyl j3-D-thiogalacto-pyranoside

(IPfG) and 20ug/ml 5-Bromo-4-chloro-3-indolyl j3-D-galactopyranoside

(X-Gal), and incubated at 37°C overnight.

Transformants containing insert were identified as white colonies

(having inserts) versus blue colonies (no inserts) and the ability to grow

on LB plates containing 50ug/ml carbenicillin. In addition, PCR

screening was employed to identify white colonies which contain

immunoglobulin variable regions, VH and VL respectively.

PCR Screening of Transformed Bacterial Colonies

After the transformed colonies from a ligation had grown on a LB

plate containing 50ug/ml carbenicillin, a micropipet tip was used to pick

47

the colony from the plate. The colony was resuspended in SOul of water

and boiled at 100°C for 5 min. The boiled bacteria were then pelleted by

centrifugation at 13000xg. An amount of 10ul of the supernatant from

the boiled bacterial colony was used as template in a PCR with primers

specific for either heavy or light chain immunoglobulin 5' and 3' variable

regions. After a 30-cycle PCR using the conditions described above, the

PCR products were electrophoresed on a 1 % agarose gel.

If PCR screening showed the colonies had the plasmids containing

an insert with expected size, the individual colonies were picked from the

plate and grown to stationary phase in LB broth containing 50ug/ml

carbenicillin. The plasmids were isolated from those individual bacterial

suspensions, and subsequently subjected to DNA sequence analysis to

further ensure that the inserts were immunoglobulin variable regions.

Mini-preparation of Plasmid DNA from E. coli

Plasmid DNA was obtained from the bacterial cells using the

following methodology. Plasmid DNA was isolated by pelleting the

transformed Novablue cells from 3ml of stationary phase culture at

10,000xg for 5 min. The cell pellet was resuspended in lS0ul 50mM Tris­

HCl, pH8.0 containing 10mM EDTA and 100ug/ml RNase A. The cells

were then lysed with lS0ul of 1 % SDS in 0.2M NaOH after which the

chromosomal DNA and cellular proteins were precipitated by adding

170ul of 2.5SM potassium acetate, pH4.8. The precipitate was pelleted in

a microcentrifuge for 10 min at 10,OOOxg after which the supernatant

was pipetted into a separate microfuge tube. The plasmid DNA in the

48

supernatant was then precipitated by the addition of 100% ethanol and

pelleted for 15 min at 12,000xg in a microcentrifuge at 4°C. Subsequently

the DNA pellet was washed with 70% ethanol, pelleted again as above

and air-dried. The plasmid DNA was finally reconstituted with water for

future use.

DNA Sequencin~ of the Immuno~lobu1in Variable Re.aons

Double stranded sequencing (157) was performed using modified T7

DNA polymerase (United States Biochemical, Cleveland, OH) and

Sequentide (Dupont, Newton, CT). Prior to sequencing, the plasmid

template was purified on glass beads. The template was then denatured

in 0.2M NaOH, 0.2mM EDTA for 30 min, neutralized in 3M sodium

acetate, precipitated with 100% ethanol, washed in 70% ethanol, dried

and then reconstituted with 7ul distilled water. The appropriate primer

was added and the sequencing reaction carried out using Sequenase

(modified T7 DNA polymerase, United States Biochemical Corp.) and

Sequentide (dithiothreitol, 35S-deoxyATP (dATP), dideoxyATP (ddATP),

ddCTP, ddGTP and ddTTP). Finally the sequencing reactions were loaded

onto a 7.8M urea, 5% acrylamide gel and electrophoresed at 70 watts.

The gel was dried, and exposed to Kodak XAR-5 film for 24-48 hours after

which the film was developed. The DNA sequences obtained were

compared to Genbank to ensure that they were immunoglobulin variable

regions.

Clonin~ AZMl scFv DNA into pET 21d Expression Vectors

Once they were confirmed to be immunoglobulin variable region

genes, the assembled scFv (AZMI) insert was digested out of pT7 blue

vector with Nco I and Xho I, and was then separated from parent

plasmids by agarose gel electrophoresis, punfied on glass beads

(QIAGEN). The insert was finally ligated into a Nco IjXho I-digested

expression vector, pET2ld. The methods were used here as described

above.

Expression of AZMl scFv in E. coli

49

The plasmid DNA of pET 2Id vectors containing AZMI scFv insert

were then used to transform BL21 (DE3) cells (Novagen). BL21 (DE3) cells

carry a chromosomal copy of the 17 RNA polymerase which is inducible

by indolyl-pyranoside thio-galactose (IPTG) via a lambda phage lysogen

(158). BL21 (DE3) cells transformed with the expression plasmid

containing the scFv were grown at 37°C in a shaker water bath to mid

log (3-4 hrs) in shaking after which IPTG was added to a final

concentration of ImM and cells further incubated at 30°C-32°C for 3-5

hrs.

Preparation of Bacterial Cell Lysates for AZMl scFv Isolation

After 3-5 hr incubation, 50ml culture containing E. coli producing

AZM 1 scFv were subjected to centrifugation for 10min in a GSA rotor at

8000xg. The cell pellet was washed 2 times in binding buffer (O.5M NaCl,

5mM imidazole, 20mM Tris-HCI, pH 7.9). The pellet was resuspended in

50

6M urea. pHB.O in binding buffer and sonicated three times for 15

seconds each to solubilize the scFv. The lysate was then incubated at

4°C for 4 hours or overnight. Finally the urea-solubilized cell pellet was

subjected to centrifugation at 12.000xg in a SS-34 rotor for 10 min and

the supernatant was removed for purification.

Purification of AZMl scFv on a Nickel-Agarose Column

A nickel-agarose column was made to purify the AZMl scFv

containing polyhistidine tails from the bacterial cell lysates. This is

because the polyhistidine peptide has strong binding affinity to nickel

ions. this feature could be exploited for purification purpose. The nickel­

agarose column was prepared by charging the metal chelating linker on

the agarose with 50mM NiS04.6H20 (159. 160). The bacterial celllysates

were filtered through a 0.45um filter and loaded onto the nickel-agarose

column which was pre-equilibrated with the denaturing binding buffer

(6M urea. O.SM NaCI. SmM imidazole. 20mM Tris-HCl. pH 7.9). The

column was then washed with 15 bed volumes of the denaturing binding

buffer followed by 10 bed volumes of wash buffer (6M urea. 0.5M NaCI.

60mM imidazole. 20mM Tris-HCl. pH7.9). The polyhistidine-containing

scFv was finally eluted from the column in Iml fractions with eluting

buffer (6M Urea. 1M Imidazole. O.25M NaCI. 10mM Tris-HCl. pH7.9). The

purity of the fractions was analyzed by SDS-PAGE followed by Coomassie

blue staining. Fractions containing purified proteins were pooled. diluted

to a concentration of less than IS0ug/ml and dialyzed against 1 liter of

50mM sodium bicarbonate buffer (NaHC03. pH 8.0) with gradual

withdrawal of urea for 2 or 3 days.

SDS-PAGE was performed. essentially. as described by Laemmli

(161).

Biotinylation of AZMl scFv

After dialysis. the protein content of the pooled and purified

material was quantified by the BCA method (162) (Pierce. Rockford. ILl.

NHS-LC-Biotin (Pierce) was then solubilized in 50mM NaHC03. pH 8.0

51

and added to the dialYL:ed fractions such that the molar ratio between

biotin and scFv was approximately 14: 1. Biotinylation was performed at

room temperature for 2 hrs after which Tris-HC!, pH7.9 (l5mM final

concentration) was added to stop the biotinylation reaction. The

biotinylated scFv was used for characterization of the scFv and the

antigen it recognized.

Fixed Cell Immunofluorescence and Flow Cytometry

For fixed cell immunofluorescence (156). tumor cells were

harvested. washed with phosphate buffered saline (PBS) once. then

resuspended at 2X106 cells/ml in PBS. A small drop of the cell

suspensions was added onto the wells of the glass slide and air-dried.

The cells were then fixed in acetone for 10 min in a -20°C freezer. Once

the slide was dried. it can be stored at -70°C for months. An amount of

30ul/well of 1:10 diluted biotinylated AZM1 scFv (3-4ug/ml) in 1% BSA

was added to the appropriate wells. For negative wells. 1% BSA was

52

added instead. The slide was incubated for 2 hours at 37°C in a moist

chamber. After that. the slide was rinsed twice in PBS. air-dried.

overlayed with 30ul/well of 1 :50 diluted avidin-FITC (Sigma Chemical

Co.) and incubated for 2 hours at 37°C in a moist chamber. Finally the

slides were rinsed twice again in PBS. air-dried and SuI of 90% glycerol in

PBS was added. Cells were observed at 495nm and photographed.

For flow cytometry (156). tumor cells were harvested and washed in

PBS. An amount of 100ul of 1: 10 diluted biotinylated AZM 1 scFv in 1 %

BSA was added per million cells and incubated on ice for 60 min. while

the cells which served as a negative control were not treated with AZM1

scFv but with 1% BSA. After four washes with PBS. 100ul of 1:50 diluted

avidin-FITC (Sigma Chemical Co.) were added and the cells were

incubated on ice for another 60 min, washed four times in PBS and kept

on ice until analysis.

Radioimmunoprecipitation Analysis

Tumor cells were radiolabeled overnight in a 24-well plate with

[35S]-L-methionine (Du Pont). The cells were then washed and lysed in

100ul 1 % NP 40 lysis buffer. The cell lysates were added to 100 ul of

biotinylated AZM1 scFv at a final concentration of 4ug/ml and incubated

on ice for 60 min. After that. 200ul of 10% streptavidin-beads (Sigma

Chemical Co.) in PBS were added to the lysates which were further

incubated on ice for 60 min with gentle shaking. Then the beads were

spun down and washed thoroughly in PBS 4 times. SDS-PAGE and

autoradiography were performed sequentially.

53

Trypsinization Assay

Fresh SK-Mel-2 cells were mechanically removed from T75 flasks

and washed twice with HBSS. Subsequently. the cells were separated

into two groups of 2XI06 each. One was treated with Img/ml trypsin in

2ml HBSS at room temperature for 15 min. The other was just incubated

in 2ml HBSS. After that. both groups were washed twice in complete

medium (CM) and then analyzed by FACS as described above.

54

CHAPTER 3

RESULTS

Cloninfl EBV-transformed B Cells by Limitinfl Dilution

After primary enrichment, the tumor-infiltrating B lymphocyte

preparation was plated and expanded in 96-well U-bottomed plates in the

presence of EBV. Two of the EBV-transformed B cell lines from a

melanoma patient (Mel-Go) which secreted IgG for over 28 days were

identified. Because these B cell lines came from 103cells/well and the

frequency of the B cell response to EBV in terms of Ig production in this

patient was 1/1385 (Figure I), they were treated as oligoclonal. Table 1

shows the characteristics of these two TIL-B cell lines derived from Mel­

Go.

Both B cell lines were determined to produce IgG with Line 1-3

producing a kappa light chain and Line 1-4 having a lambda light chain.

Line 1-3 showed reactivity to more melanoma cell lines than Line 1-4. All

the tumor cells tested here were allogeneic to these two B cell lines. This

is due to loss of autologous melanoma cells (Mel-Go) in the begining of

study. Neither reacted to normal human foreskin fibroblasts.

TIL-B cell line 1-3 was then chosen and sub cloned by limiting

dilution at 0.5cell/well. In a total of 1200 wells, 75 wells (6.25%) showed

B cell growth by microscopic examination after 10 days of incubation.

Only 21 clones (1.75%) demonstrated IgG production and only 7 clones

(0.58%) reacted to an allogeneic malignant melanoma cell line.

55

Mter further growth for another 10 days in 96-well plates. these 7

B cell clones were transferred to a 24-well plate for further expansion.

Table 2 summarizes the characteristics of expanded B cell clones from

Mel-Go Line 1-3 in 10-day growth of the 24-well plate. One clone (5-3E)

which arose from TIL-B cell line 1-3 reacted to three melanoma cell lines

but not to the other unrelated tumor cell lines tested was selected for

further study.

Capture and Assembly of Anti-Melanoma Immunoglobulin Heavy and

Light Chain Variable Regions

Messenger RNA was isolated from clone 5-3E. using oligo-dT

cellulose followed by reverse transcription into cDNA. Polymerase chain

reaction (PCR) was employed to amplify the immunoglobulin variable

heavy and Variable light chains using sets of 5' and 3' oligonucleotide

primers (Appendix A) specific for human immunoglobulin gamma and

kappa chains. The results of the PCR amplification for both heavy and

light variable regions are shown in Figure 2. A set of heavy chain

subgroup III primers amplified IgG VH resulting in a PCR product of 350

base pairs (bp) (lane 2). while a set o[Vk primers resulted in the

amplification of an apprOximately 320 bp product (lane 3). The second

round of PCR was performed to link VH and Vk with the (GlY4Ser)3

linker as deSCribed in the methods. The scFv product was observed at

about 700 bp (lane 4).

56

Nucleotide and Amino Acid Sequence of Immunoglobulin Heavy and

Light Chain Variable Regions

Mter the heavy and light immunoglobulin variable region peR

products were purified on an agarose gel, they were directly cloned into

pT7blue plasmids (T-vector) which also served as a sequencing vector.

The T-vector system enables direct and effiCient cloning of PCR products

via an overhanging 3' adenosine (A) on the PCR product (155) which

base-paired with a 5' overhanging thymidine (T) through a modified

EcoRVrestriction endonuclease site on the T-vector (154). Transformants

containing VH and VL inserts interrupted the .J3-galactosidase gene and

appeared as white colonies on LB agar plates containing 50ug/ml

carbenicillin, O.lmM IPfG and 20ug/ml 5-Bromo-4-chloro-3-indolyl.J3-D­

galactopyranoside (X-gal).

The immunoglobulin fragments were sequenced once they were

cloned into the T-vector. Double-stranded sequencing was performed

using the Sanger dideoxy-nucleotide termination method (157). Three

clones from the heavy chain variable region and 3 clones from the light

chain variable region were sequenced in the T -vector to ensure that the

correct variable region sequence was obtained. The nucleotide and amino

acid sequences of heavy and light chain variable regions are shown in

Figures 3A and 3B. Mter data base search, both were confirmed to be

human immunoglobulin genes.

57

Expression of AZMl scFv Antibody

Based on the known human immunoglobulin sequences (53), peR

primers (Appendix A) were designed to amplifY the VH and Vk for "in

frame" protein expression in the pET21d expression vector. Initially VH

and Vk genes were separately amplified and subsequently linked together

by peR via a (GlY4Ser) 3 linker to generate AZMl scFv. The peR product

of AZMl scFv was then directly cloned into the T-vector without any

modification. Secondly, the AZMl scFv DNA, containing Nco I and Xho I

restriction sites at the 5' and 3' ends of the gene respectively, was

digested out of the T-vector by Nco I and Xho I endonucleases and directly

ligated into the pET21d expression vector, so that they were in the

correct reading frame to express functional immunoglobulin proteins.

Once AZMl scFv DNA were cloned into the pET21d expression vector, its

plasmid was used to transform Novablue E. coli cells. A diagram of the

pET21d bacterial expression vector is shown in Figure 4.

The pET21d expression vector system utilizes the bacterial host

cell, BL2l(DE3), which contains a chromosomal copy of the T7 RNA

polymerase gene controlled by lac promoter. The addition of IPTG turns

on the lac promoter that activates T7 RNA polymerase. The important

features of this expression system are 1) expression of the

immunoglobulin fragment does not occur in bacterial hosts without a

source of the T7 RNA polymerase. This results in more stable propagation

of the immunoglobulin genes in the plasmid; 2) the expressed protein

contains a polyhistidine tail at carboxyl end which can be exploited to

pUrify the expressed proteins by a metal chelating column, such as a

nickel agarose column (89).

58

The AZMl scFv-containing expression vector (pET21d) was isolated

and purified from Novablue cells and super-coiled DNA was used to

transform BL21 (DE3) cells. Because there is no signal peptide on the

expressed proteins when using pET21d expression vector, the proteins

produced in BL21 (DE3) cells are insoluble and form "inclusion bodies" in

the cytoplasm. The latter are solubilized under denaturing condition in

high molar urea. Individual colonies were picked and grown to mid log in

Superbroth and then induced with IPTG for 3-5 hours at 32°C. Finally

the cells were processed as described in Methods.

Purification of AZMl scFv Antibody

Since the scFv proteins produced from the pET21d-transformed

cells contain a carboxyl polyhistidine tail, it was possible to pUrify them

on a nickel-agarose column using immobilized metal affinity

chromatography. Bacterial cell lysates in binding buffer containing 6M

urea were loaded onto a nickel ion-charged agarose column. After

thorough washes, the proteins were eluted in 1ml fractions. Aliquotes

were collected and electrophoresed in SDS-PAGE to determine which

fractions contained the AZM 1 scFv proteins. Figure 5 shows a Coomassie

blue-stained SDS-PAGE analysis of the fractions from a nickel-agarose

column by purifying the scFv-producing BL21 (DE3) lysates. The scFv

proteins were eluted in fraction 2 (Lane 6 in Figure 5) with a molecular

weight of apprOximately 29 KD.

59

Primary Analysis of the Biological Activities of AZMl scFv

The fractions containing the AZM 1 scFv were pooled, dialY.led and

then biotinylated as described in methods. The biotinylated scFv was

used to test for specific tumor reactivity. The tumor cell ELISA showed

that the scFv bound to 6 of 7 melanoma cell lines with no reactivity to

unrelated tumor cell lines, Hey and MCF-7, or to normal human tissue

(Table 3).

To further characterize the tumor reactivity and specificity of the

scFv, immunofluorescence on fixed tumor cells and FACS analysis on

viable tumor cells were performed. Fixed cell immunofluorescence was

conducted with five different melanoma cell lines and two unrelated

ovarian cancer cell lines. As shown in Figure 6, all five melanoma cell

lines, including A375, SK-Mel-2, SK-Mel-28, Mel-K and Mel-G, were

positive in immunofluorescence staining with the biotinylated AZM 1

scFv and avidin-FITC system. The two unrelated ovarian cancer cell

lines, Hey and SKOV3 did not react. This demonstrates that our anti­

melanoma AZM 1 scFv retains the ability to bind specifically to the

melanoma cells.

Similar results were also generated by FACS analysis in which 33

tumor cell lines were stained with the scFv (Table 4, Table 5 and Figure

7). Sixteen of Eighteen melanoma cell lines tested were positive by flow

cytometry. All 14 non-melanoma cell lines, including prostate, breast,

colon, lung and ovarian cancer cell lines as well as human melanocytes,

were negative in the assay. The mouse melanoma cell line B 16, also

60

tested negative. Table 4 and Table 5 summarize the FACS analysis and

show the percentage of positive cells stained by the AZM 1 scFv. Figure 7

shows examples of tumor cells tested in FACS analysis. Overall

melanomas ranged from 15.0% to 96.2% positive cells. while non­

melanoma cell lines were all below 6% which was defined as the non­

specific background level.

Identification of Melanoma-associated Antigen(s) by AZMl scFv

A radioimmunoprecipitation assay (RIPA) was performed to define

the melanoma-associated antigen(s) recognized by the AZM1 scFv .

Figure 8 shows a protein band of apprOximately 45 KD in the lanes of

SK-MEL-2 and SK-MEL-28 (Due to loss of autologous melanoma cells

from Mel-Go in the begining of the study. here we were not able to show

the immunoprecitpitation result from the autologous tumor cells). No

band was seen in the lanes of MCF-7. Hey and SKOV3. Thus. we have

identified a melanoma-associated antigen of apprOximately 45 KD

present in most of the melanoma cells examined.

Primary Characterization of the Melanoma-associated Antigen

Recognized by AZMl scFv

The SK-Mel-2 cell line was treated with trypsin to determine

whether or not this melanoma-associated antigen (45 lID) was trypsin­

sensitive. which would further prove that the antigen was membrane­

bound. Table 6 shows that the cells without trypsin treatment were

88.5% positive with AZM1 in FACS analysis. while the cells treated with

61

trypsin treatment were 7.9% positive in the same assay. Therefore. the

melanoma-associated antigen recognized by AZMl scFv is trypsin­

sensitive and membrane-bound. Further characterization of this tumor

antigen is underway.

62

FREQUENCY OF RESPONSE 1~----------------------------------,

n1+-------r-----~------~------~----~ o 1000 1GOO B CB..l.&WEU.

j{r)= 1/1385: R2= 0.99

Figure 1. Limiting dilution analysis (LDA) of TIL-B cells from Mel-Go in the presence of EBV. j{r) is the frequency of B cell response. R2 is

correlation coefficient.

600 bps ..

1 2 3 4

Figure 2. peR products ofVH. Vk and scFv. Lane 1: 100 bp DNA ladder: Lane 2: VH product: Lane 3: Vk product: Lane 4: scFv products.

63

64

-+FRl

GAGGTGCAACTGGTG GAGTCTGGGGGAGGC TTGGTACAGCCTGGG GGGTCCCTGAGACTC 60

E V Q L V E S G G G L V Q P G G S L R L --.CDRl ~ FR2

TCCTGTACAGCCTCT GAATTCTCCTTTAGT TTCTATGCCATGAGC TGGGTCCGCCAGCCT 120

S C T A S E F S F S F YAM S W V R Q P --..CDR2

CCAGGGAAGGGGCTG GAGTGGGTCTCAACT ATTACTGGTAGTGCT GTCGAAATATACTAC 180

P G K G LEW V S TIT GSA V ElY Y --.FR3

GCAGACTCCGTGAAG GGCCGGTTCACCGTC TCCAGAGACAATTCC AAGAACACTCTATAT 240

ADS V K G R F T V S R D N S K N T L Y -.. CDR3

CTCCAAATGAACAGC CTGCGAGTCGAGGAC ACGGCCGTATACTAT TGTGCGAAAAGTTCA 300

L Q M N S L R V EDT A V Y yeA K S S ~FR4

GCTCCCTCCCAGTAT TCGGGCCAGGACTAC TGGGGCCAAGGAACC CTGGTCACC 354

APSQY SGQDY WGQGT LVT

Figure 3A. Nucleotide and deduced amino acid sequences of AZMl heavy chain variable region. Underlined nucleotides represent the complementary determining regions CDRl, CDR2 and CDR3.

~FRl

GAAATTCAGTTGACG CAGTCTCCATCCTCC CTGTCTGCATCTGTA GGAGACAGAGTCACC 60

E I Q L T Q S P S S L S A S V G D R V T

~CDRl ~FR2

ATCACTTGTCGGGCA AGTCAGAGCATTAGC AGCTATTTAAATTGG TATCGGCAGCAACCA 120

I T C R A S Q SIS S Y L N W Y R Q Q P ~CDR2 ~FR3

GGGAAAGCCCCTAAA CTCCTGATCTATGGT ACATCCAGTTTGCAG AGTGGGGTCCCATCA 180

G K A P K L L I Y G T S S L Q S G V P S

AGGTTCAGTGGCAGT CGATCTGGGACAGAT TTCACCCTCACCATC AGCAGTCTGCAACCT 240

R F S G S R S G T D F T L TIS S L Q P

~CDR3 ~FR4

GAAGATTTTGCAACT TACTACTGTCAACAG AGTTACAGTATGCCT CTCACTTTCGGCGGT 300

E D FAT Y Y C Q Q S Y S M P L T r G G

GGGACCAAGGTGGAG

G T K V E

315

65

Figure 3B. Nucleotide and deduced amino acid sequences of AZMl kappa chain variable region. Underlined nucleotides represent the complementary determining regions CDRl, CDR2 and CDR3.

Ora IItc)lOlJ

Osall .. III pET-21a(+) (5443bp)

B$IlG iC2<",

T7 promolor primer '69340'\

Notes The maps for pET-21b(+). pET-21C(+) and pET-21d(+) are the same as pET-21a(+) (shown) with the following exceptions:

pET-21b(+) is a 5442bp plasmid; subtract 1bp from each site beyond BamH f at 198.

pET-21c(+) is a 5441bp plasmid; subtract 2bp from each site beyond BamH 1 at 198.

pET-21d(+) is a 5440bp pla;;mid; the BamH 1 site is i~ the same reading frame as 10 pET-21c(+). An Nco 1 site is substituted for the Ndel site with a net 1 bp deletion at p<)silion 238 01 pET-21c(+). As a result. Nco 1 cuts pET21d(+) at 234. and Nhe 1 cuts ~t229. For the !~st of the sites. subtract 3bp from each site beyond position 239 in pET-21a(+). Nde I does not cut pET-21d(+). Note also that Sly I is not unique in pET-21d(+).

The 11 ork,lin in pET-21-a-d(+) vectors is oriented so that infection with helper phage will produce virions containing single stranded DNA thetis the same strand as that shown below. Therelore. single stranded sequencing should be performed using the T7terminator primer.

~ ~ T7 promoter ~ tee operalor.2!!!!L ...!2!... ACA IC ICCA ICCCCCCAAA I IAATACGACTCACTATACCCGAA TTGIGAGCGGATAACU HCCCCTC1 AGAAA 1 AI. H 1 IGI I IAAC I 11 AAGAAGGAGA

Ndsl Nhel T7.TeO"· pET-210 BamHl ECIORI Sac I Sal I Hindlll iJ:J:l ::r Hla'Tao·

II. T ACA 110 TGCC I ioCCA TCAC TCCICCACACCAAA TCCGICGCCCAruCMmCAGcTcCCTCCAcAACcTTGCcCCCCcAC 1CGAGCACCACCACCACCACCAC IGA Moll. I oSorllo I 1hrG' yG' yG I nC I nM.IC I yArgC I ySorG I uPh.C I uL.uArgArgC I nA I oCysG I yArglhrArgA loProProProProProLeu WI .. ' ;r~" ~-.'. ~';;. pEI-2Ib ••• CCICCCCATCCCAATTCGACC1CCG1CGACAACCTTCCGGCCGCACICGAGCACCACCACCACCACCAC1GA

I - NoO"f;»:1;,;,~:?;;.(;.: ... G I yArgA,pProA,nS.rS.rS.rYo I A'pLy,LouA I 010 loA I oL.uG I uH I,H I,H I,HI,H I,H IsEnd

.... ~irc'~4i:: pE 1-21 c.d ••• GCICCCA 1CCCAA TTCCACCICCC1CCACAACCTTCCCCCCCCAC ICCACCACCACCACCACCACCAC ICA &i'-faS."w .. -,,, ••• C lyArgll.Argt ItArgAloProStrThrSorLtuArgProHI,SorSor1hrThrThr1hrThrThrC lu

Bpull021 T7 lermlnelor Col Teeccc T CC T AACAAACCCCCAAACCAACC TCACT I CCC ICC ICCCAC~CTCACCAA T AAC1 ioCCA I AACCCC T TCCCCCC I CTAAACCCCICT ICACCCC1 T T T T TC

• Ava I Slle nOI unique in pEJ.24ao(i(.) l7lonnlnalor primer '69337.\

Figure 4. A diagram ofpET21d expression vector

66

110

84

47

33

24

16

67

1 2 345678910111213

Figure 5. SDS-PAGE analysis of purification of the scFv (AZMl) by a Nickel-agarose column. Lane 1, Protein markers; Lane 2, Bacterial lysate; Lane 3, Flow-through of the lysate; Lane 4, Washes; Lane 5, Fraction 1 of eluate; Lane 6, Fraction 2; Lane 7, Fraction 3; Lane 8, Fraction 4; Lane 9, Fraction 5; Lane 10, Fraction 6; Lane 11, Fraction 7; Lane 12, Fraction 8; Lane 13, Fraction 9.

A375

SKOV3

Figure 6. Immunofluorescence stain on fixed tumor cells. The various tumor cells. specified above. were fixed on a glass slide. then treated with the biotinylated scFv (AZMl) and finally with avidin-FITC.

68

,

I I I ,

, I ,

69

Mel-W Melanocytes

Mel-R PC-3

" , ... , I '

-'

, I,

I , , , I ,

I ,

Mel-V RAJI

, , I I

I I

Figure 7. FACS analysis of tumor cells treated with the scFv (AZMl). x-axis is the fluorescence intensity: y-axis is the number of cells. Dot lines indicate the tumor cells treated only with avidin-FITC. while solid lines mean the tumor cells treated with the biotinylated scFv (AZMl) and avidin-FITC.

45KD

1 234 5

Figure 8. Radioimmunoprecipitation of a melanoma-specific antlgen by AZMl scFv. Lane 1: SK-MEL-2; Lane 2: SK-MEL-28; Lane 3: MCF-7; Lane 4: Hey; Lane 5: SKOV3.

70

71

Table 1 Characteristics of two EBV-transfonned TIL-B cell lines (Mel-Go)

Characteristic TIL-B Cell Lines

1-3 1-4

IgG + +

IgM

19A

Lambda Chain +

Kappa Chain +

Reactive to Mel-G + +

Reactive to Mel-K +

Reactive to MCF-7 +

Reactive to Human Skin Fibroblast

72

Table 2 Characteristics of expanded EBV-transformed B cell clones from line 1-3

Clones IgG Kappa Lambda McI·G McI·K i\.175M MCF·7 lIey Iluman Skin Fibroblast

1. 4·6G + +

2.8-3C + +

3.11·50 + + + +

4.5-11G + + +

5.8-10F + + +

6.2·6F + + +

7.5-3E + + + + +

Table 3 Determining the anti-melanoma speeijicity oj AZl\11 seFv

Type of Cells

A375M

Mel-K

Mel-G

JH1308

SK-Mel-24

MK457

Ps1273

Hey (Ovary)

MCF -7 (Breast)

Human Skin Fibroblast

The Positivity

+ + + +

+

+

73

74

Table 4. FACS Analysis oj Melanoma Cells Treated With AZM 1 scFv

Cell Type % of Positive cell

A375 66.0

Mel-K 75.9

Mel-G 56.5

SK-Mel-2 88.5

SK-Mel-28 64.7

Mel-V 96.2

Mel-R 92.2

Mel-D 56.8

Mel-Ke 17.2

Mel-T 15.0

Mel-A 22.9

Mel-Wr 35.2

Mel-Ws 24.1

Mel-Th 17.1

Mel- Vw 23.5

Mel-M 16.0

81-61 6.5

MeI-C 1.9

75

Table 5. FACSAnalysis oJNon-Melanoma Cells Treated withAZMl scFv

Cell Type Tumor Type % of Positive cell

NC-37 Human PBL Blast 2.2 A549 Lung Cancer 0

H596b Lung Cancer 0 SK-BR-3 Breast Cancer 0

BT474 Breast Cancer 1.3 MCF-7 Breast Cancer 0 Daudi Burkitt's Lymphoma 0 Raji Burkitt's Lymphoma 1.3

SKOV3 Ovarian Cancer 3.5 Hey Ovarian Cancer 5.9

LNCAP Prostate Cancer 0 PC-3 Prostate Cancer 0.2

SW480 Colon Cancer 4.1 Human Melanocytes 0.5

B16 Mouse Melanoma 0

Table 6. Comparison ofSK-Mel-2 Cells Stained withAZMl scFv Before or After Trypsin Treatment

With Trypsin

Without Trypsin

% of Positive Cells

7.9

88.5

76

Note: SK-Mel-2 cells were separated into two groups. One was treated with trypSin. The other served as a control without trypsin treatment. Both were then treated with biotinylated AZMl, avidin-FITC, and finally analyzed by FACS.

77

CHAPTER 4

DISCUSSION

Genetically engineered human mAbs may have several advantages

over mAbs derived by fusion of immunized mouse spleen cells with mouse

myeloma cells because they can easily be produced as human mAbs with

little immunogenicity when used in humans. they can be produced in

very large quantities. and. furthermore. they can be modified by site­

directed mutagenesis to alter their binding affinities and specificities

(163). In view of these potential applications and advantages. we herein

deSCribe a novel methodology that was used to construct. produce and

analyze a human monoclonal scFv antibody specific for human

malignant melanoma. In this method. tumor-infiltrating B lymphocytes

were isolated. enriched. cloned and expanded. Subsequently the VH and

Vk genes were isolated from an anti-tumor TIL-B cell clone and a single

chain Fv was constructed and expressed in E. coli. We have further

shown that the AZM1 scFv antibody constructed by this method can

identifY an antigen on human melanoma cells since this antigen is

recognized by the immune system of an affected patient.

In Vitro Expansion of Human Tumor-infiltrating B Cells

In recent years. tumor-infiltrating T lymphocytes (TIL) have been

successfully cultured from several human tumor types, particularly

malignant melanoma (164. 165. 166). These cytotoxiC TIL lines show

78

specificity in that they lyse MHC matched but not mismatched tumor

cells. TIL lines together with IL-2 have been used for the therapy of

metastatic malignant melanoma (167, 168). More recently TIL lines from

malignant melanoma have been used to identify and eventually capture

the gene of the tumor associated antigens recognized by the TIL. These

include the MAGE-l (119) and MAGE-3 antigens (120), the MARJ'-1

antigen (121), gpl00 (122) and tyrosinase (123). These studies have

facilitated an analysis of the repertOire of the human T cell response to

tumors and of the antigens they recognize.

However, the studies on tumor-infiltrating B cells lag behind those

on tumor-infiltrating T cells (92, 93, 94). The main reason for this is the

limitation of assays that are available to study B cell function in vitro.

These include the plaque forming cell assay (95) which does not allow

expansion and further study of B lymphocytes. Epstein Barr virus (EBV)

transformation which may be selective for IgM and has low efficacy (97,

169, 170). Recently a method of in vitro B cell expansion has been

reported (98,99), which demonstrated that peripheral blood B

lymphocytes could be activated by culture in the presence of anti -CD3-

stimulated human T cells, and that this response could be further

enhanced by the addition of IL-2 or IL-6. This method induced Ig

secretion at a high frequency as well as production of various

immunoglobulin isotypes by individual peripheral blood B lymphocytes

(171). Since T cells are essential to the activation of B cells in this

method, the use of T cells from the B cell donor might introduce a source

of variability when comparing different individuals, especially under some

79

pathological conditions, such as cancer (100, 101) and HIV infection

(172) which cause important alterations in T cell populations. If we were

able to use a stable, transformed human T cell line which provides the

necessary help for the B cell activation in culture, we would avoid this

extra source of variability in the analysis and allow better comparisons

among different individuals with different diseases or conditions. It

would also simplify the culture system. Consequently Barbuto et aZ. (102)

reported a novel method of in vitro B cell expansion using MOT cells, a

I-ITLV II-transformed human T cell line, as feeders. MOT cells did not

require any exogenous stimuli to induce Ig secretion by B lymphocytes,

and a higher level of B cell response could be achieved with MOT feeder

cells when compared to the use of a system with T cells, anti-CD3 mAb

and IL-2. This proves that the MOT system allows the expansion of

antigen-specific B cells, even in the absence of antigen-specific help, and

allows studies of the B cell repertoire under pathological conditions,

such as cancers and AIDS, where the T cell functions are expected to be

impaired.

Using this method of B cell expansion, Punt et al. (93)

demonstrated that B cells from cell suspension of various tumors and

peripheral blood of cancer patients could be expanded to produce Ig in

which IgG subfraction showed considerable binding to autologous and

allogeneic tumor targets of the same histology. Barbuto et al. (102) also

showed that some tumors had infiltrating B cells that were making anti­

TNF-alpha antibody (Cancer Immuno!. and Immunother. In press, 1995).

We speculated that the tumor stimulated the production of these

antibodies which then blocked host immune responses and therefore

constituted a mechanism of evasion from host control. However. more

importantly. we also inferred that tumor-infiltrating B cells naturally

enriched in tumors might provide a unique source of tumor-specific B

cells for the production of human anti-tumor monoclonal antibodies.

80

In this study. we initially stimulated tumor-infiltrating B

lymphocytes using MOT cells for in vitro expansion. Then we used EBV to

transform and immortalize tumor-specific-IgG-secreting B cells. After

several preliminary experiments. we observed that most EBV-transformed

B cells came from B cell cultures not treated with MOT cells (data not

shown). Further experiments showed that most of B cells treated by MOT

cells were resistant to subsequent EBV transformation. although the in

vitro MOT-mediated B cell expansion method was very effective in

expanding TIL-B for two to three weeks. After that. these cells died.

Contrary to previous reports (97. 169. 170). EBV proved to be a potent B

cell stimulator and a transforming agent which directly activates tumor­

infiltrating B cells and subsequently transforms these cells while

maintaining secretion of high titers of tumor-specific IgG for more than 6

months (data not shown). We believe that MOT treatment facilitates B

cell differentiation to plasma cells which die within 3-7 days. while EBV

treatment activates and subsequently transforms B cells for long-term

growth while producing Ig. Therefore. in subsequent experiments. we

decided to directly use EBV to stimulate B cells.

81

Capture of Immunoglobulin Genes

The polymerase chain reaction. with primers matching the 5' and

3' ends of rearranged VH and VL genes. provides the means to amplify.

clone. and express V genes from lymphocytes and hybridoma cells (46.

47,48). The V genes may be amplified from both cDNA and genomic DNA.

In this study. we chose cDNA to capture Ig V genes. This is because cDNA

are reverse-transcribed directly from mRNA which has undergone gene

rearrangement and is ready for translation to proteins. To maximize

complementarity, degeneracy was incorporated into the primers (47. 48,

49), or different primers were designed for different families of V genes

(173). For cloning of the amplified DNA into expression vectors.

restriction sites were incorporated within the peR primers.

In the work presented in this dissertation. messenger RNA was

obtained from an anti-melanoma IgG-producing EBV-transformed B cell

clone. reverse-transcribed into cDNA and subjected to peR-mediated Ig V

gene rescue. Degenerate sense and anti-sense human immunoglobulin­

specific primers designed for peR were successfully used for AZM 1

immunoglobulin amplification, including VH and Vk. The primer set used

to amplify AZMl VH corresponded to a human subgroup III. while the

primer set used to amplify AZMl Vk mixed with four human subgroups of

Vk chains according to Kabat's classification.

Construction of AZMl scFv gene

peR assembly was used to construct AZMl scFv. This technique

was extremely successful and was relatively simple to perform. After the

82

first round of peR, AZMl VH and Vk peR products were subjected to a

second round of peR so that a (GlY4Ser)3 linker could be constructed

between the two variable regions. The VH was amplified by peR using a

5' sense primer containing an Nco I site and a 3' anti-sense primer with

the first three glycines of first unit of GlY4Ser. The Vk was amplified by

peR using a 5' sense primer with last three amino acids, glycine-glycine­

serine, of the third unit of GlY4Ser and a 3' anti-sense primer containing

an Xho I site. The VH-GlY3, Gly-Gly-Ser-Vk and (GlY4Ser)3 were added to

a third and final peR along with the 5' sense VH primer with Nco I and 3'

anti-sense Vk primer with Xho I. The product from this reaction was the

assembled scF\r.

Among the various protein linkers, the major reason we used the

(GlY4Ser)3 linker here is that the glycine and serine residues in

(GlY4Ser)3 are relatively small and lack side chains, which provides

maximum flexibility and hydrophilicity so as not to prevent scFv from

forming a native conformation capable of binding its antigen in an

induced fit type of interaction. After peR assembly of AZMl scFv. a band

at the predicted molecular weight was observed. suggesting that the new

gene was assembled such that the (GlY4Ser)3 linker was flanked on both

sides by VH and Vk. Subsequent cloning. sequencing and protein

expression proved that AZM 1 scFv was assembled correctly.

Since digestion of the scF\r peR products with restriction

endonucleases is difficult, a two-step cloning method was necessary for

expression of immunoglobulin variable regions. This method allowed

rapid and efficient cloning of AZMl scFv peR products without

83

modification into the T-vector via overhanging 3' adenosines (A) on the

ends of the double stranded peR product. This constituted the first step

of cloning. In the second step, the AZMl gene was digested out of the T­

vector by appropriate restriction endonucleases, and ligated into the pre­

digested expression vector.

Expression of AZMl scFv Antibdoy in E. coli

The AZM 1 scFv was cloned into the pE1"21 d expression vector and

used to transform BL21(DE3) cell. We found that BL21(DE3), freshly

transformed with plasmid containing antibody genes, could be the

producer of the scFv antibodies. The transformed cells grown on the

carbenicillin-containing LB plate over 10 days reduced levels of

production or no production. Therefore, we routinely maintain a stock of

the plasmids for re-transformation.

It was important to induce the bacteria to produce antibody

fragments, scFv, when the culture was in early log phase and to harvest

the culture between two and three hours post induction. This is due to

potential plasmid instability and cell overgrowth (158). After several

initial experiments performed, we found it was extremely difficult to

reproduce the same yields from batch to batch.

One possible reason for such erratic expression might be plasmid

instability. The supplier of the expression vector has cautioned that

plasmids can easily be lost from bacterial host cells prod ucing

recombinant proteins (158). Another explanation for the loss of antibody

production from bacterial cells is that antibody fragments might be toxic

84

to the host cells and impair the cell growth. In addition. BL21 (DE3) cells

are recA positive (158) which means that the cells have recombinase A

and therefore they have the ability to remove any portions of genes.

including antibody fragment genes. and put them at some disadvantaged

sites which result in production of erratic proteins or loss of the

antibody fragments. Looking for a more stable expression system. such

as eukaryotic expression system. will be a better option.

Purification of AZMl scFv Antibody

Purification of AZMl scFv antibodies consisted of solubilizing the

bacterial cell pellet in 6M urea followed by sonication to release the

antibody fragments from the bacterial inclusion bodies. The scFv

antibody-containing lysate was run through a nickel-agarose column for

purification purposes. It was easily done and the column was reusable 8-

10 times. The scFv purified was run as a single protein band by SDS­

PAGE analysis (Figure 3).

After purification of AZMl scFv from the nickel-agarose column

under denaturing conditions. the scFv was dialY.led against 50mM

NaHC03-urea until the urea was completely exchanged for NaHC03 for

optimized protein refolding. This process usually lasted 2-3 days for

better protein refolding (158).

Tumor-reactivity of AZMl scFv Antibody

Twenty one of twenty four melanoma cell lines. including short­

term cultured melanoma cells and fresh melanoma cells. tested positive

85

using biotinylated AZM 1 scFv antibody by tumor cell ELISA, fixed cell

immunofluorescence staining and FACS analysis, demonstrating that

the AZM1 scFv antibody retained the biological activities by binding to

melanoma cells. Due to loss of autologous melanoma cells (Mel-Go) in

the begining of the study, we have no way to know whether or not the

AZM1 scFv reacts with autologous melanoma cells. In addition, it also

showed using the same assays that the AZM 1 scFv did not react to 14

non-melanoma cell lines, including breast cancer, colon cancer, prostate

cancer, ovarian cancer, normal human lymphoblast cells and human

melanocytes. That indicated the tumor-specificity of AZM 1 scFv.

It would be more significant to conduct an in vivo study using

AZM1 scFv. The in vivo study will tell us what the pharmacokinetics of

the AZM1 scFv is in the body; how well the scFv penetrates to the solid

tumors. The advantages of scFv are 1) the small size that enables scFv to

be rapidly cleared from Circulation, thus reducing its immunogenicity

(65): 2) the capability to penetrate the micro-vasculature of solid tumors

faster and more evenly than intact IgG, F(ab')2 and Fab fragments in an

experimental colon carcinoma xenograft model (66). This suggests that

the relatively small size of scFv might make them more effective for solid

tumor therapy. However, the rapid clearance may be a disadvantage

pharmacokinetically because the molecule may be in the circulation for

too short a time to bind to the target tissue.

86

Identification of Melanoma-associated Antigen

Enormous efforts to identify and exploit tumor-associated antigens

have been made in recent years through the use of monoclonal

antibodies. A variety of tumor-associated antigens have been identified in

malignant melanoma using classical hybridoma methodology. Most of

these have been identified using mouse monoclonal antibodies but a few

have been identified using human monoclonals. These antigens include

S100, HMB-45, HMW-MAA, gp97, GD2 and GD3. etc. (1 13). It is not clear

that these are the antigens which elicit an effective immune response in

patients with melanoma.

While antibodies recognize and bind cell surface molecules, T cells

recognize the cell surface complexes formed by antigenic peptides

produced from any protein synthesized within the cells and integral

membrane proteins encoded by the major histocompatibility complex

(MHC) (174, 175). The antigens, such as MAGE-l, MAGE-3, gp100 and

MARr -1, recognized by cloned cytotoxic T cell lines are presumably those

involved in natural host defense in melanoma but are of somewhat

limited potential because of MHC restriction.

Here we have identified a 45 KD membrane-bound melanoma­

associated antigen by radioimmunoprecipitation using AZM 1 scFv

derived from tumor-infiltrating B lymphocytes (Figure 8). Most melanoma

cell lines tested here possessed this antigen. as evidenced by binding of

the antibody. It is possible that the antigens recognized by TIL-B will be

more generally recognized and will elicit immune responses in the

majority of patients. Therefore. we speculate that the melanoma antigen

87

recognized by AZMl may be immunogenic, suggesting that it may serve

as tumor rejection antigen. This hypothesis could be established if we

find that antibodies directed against this 45 KD melanoma antigen are

present in the serum of melanoma patients. Currently we have set up a

serum bank collected from more than 50 melanoma patients. Once the

massive production of the antigen becomes available, we will screen this

serum bank for potential antibodies against this antgen. However, it

remains to be determined if antibody responses to melanoma are effective

anti-tumor responses and if the antigens which elicit an antibody

response can also elicit a cytotoxiC T cell response. To answer these

questions, first, we should find the cases in clinic that the spontaneous

tumor regression in melanoma patients is caused by the antibodies

against this 45 KD melanoma antigen, which certainly suggests that this

antibody response is an effective antitumor response. However, the

induction of antibody responses against tumor antigens may be an

evasive mechanism used by tumors to escape the surveillance of host

immune system. Second, to elicit CTL response by this 45 KD melanoma

antigen, it must be processed into small antigenic pep tides, which are

subsequently associated with HLA class I molecules and presented onto

the cell surface. However, it has been well documented that many

tumors, including melanomas, fail to express HLA class I antigens or

reduce the expression of HLA antigens (l13, 176, 177). Therefore, it is less

likely that the antigen will elicit an effective CTL response in melanoma

patients.

88

In addition. we have recently established a cDNA library

constructed from a melanoma cell line (SK-Mel-28) expressing the

melanoma antigen recognized by AZMl. Screening of this library is on­

going by using AZMl scFv. We will soon know the nature of the antigen.

Interestingly. it was demonstrated by FACS analysis with AZMl

that the fresh malignant melanoma cells and the short-teml melanoma

cell lines from patients' biopsies can be divided into two population. one

of which showed strong positive reactions (for example. Mel-Wr in Figure

7). It hints that the melanoma antigen recognized by AZMl may be a

melanoma differentiation antigen or a stage-specific antigen expressed at

certain stages of tumor growth. Since the antigen recognized by AZM 1 is

widely expressed in melanoma tumor cells. it may constitute a useful

target for speCific immunotherapy against melanoma. The

characterization of this tumor antigen certainly needs further studies.

Conclusion

The methodology and results demonstrated herein show that not

only can the anti-tumor Ig genes be captured. but also they can be

functionally expressed as a scFv in E. coli. The scFv antibody proteins

expressed can be used to identify the putative tumor-associated antigen.

Although our efforts to clone and evaluate this melanoma-associated

antigen are still in early stage. the significance of this study includes the

followings: 1) human monoclonal antibody molecules. such as scFv.

derived from TIL-B may be effective diagnostic and therapeutic agents

(when developed to immunotoxins) in the clinic; 2) they may be more

89

efficient reagents for gene cloning of tumor-associated antigens in

human; 3} the cDNA sequence of the antigen identified by AZM 1 may be

used to determine the location of the antigenic epitope. to evaluate its

biological function and pathogenic significance for the disease; 4} the

presence of the antigen in normal and cancer tissues can be determined

by in situ hybridization using an oligonucleotide probe derived from the

cDNA sequence; 5} the tumor-associated antigen and its peptide may be

clinically very important as effective vaccines in inducing anti-tumor

humoral and cell-mediated immune responses in cancer patients; and

finally 6} with establishment of this methodology. a systematic analysis

of the repertOire of the B cell response to melanoma and possibly other

tumors as well could be carried out.

To our best knowledge. this is one of the first reports of a

genetically engineered human monoclonal antibody molecule that shows

a specific binding activity to melanoma. We conclude that the expression

of a human anti-melanoma scFv antibody in E. coli could produce an

effiCiently folded, biologically active molecule that retains antigen­

binding activity. The availability of human antibody variable region genes

now permits us to engineer antibodies to make them better therapeutic

agents. stabilize their structure. use site-directed mutagenesis to explore

and improve specifiCity. With establishment of this methodology. we

could make more human scFv antibodies to other melanoma associated

antigens as well as to other human tumor antigens. In return, we will

discover more and more tumor-associated antigens which could be

further developed into tumor vaccines in the future.

APPENDIX A

OLIGONUCLEOTIDE PCR PRIMERS FOR AMPLIFICATION OF

HUMAN IMMUNOGLOBULIN VARIABLE REGIONS

5' Human VH Chain Primer I with Nco I site (sense)

5'-CC ATG GAG GT(GT) CAG CTG GT(AGC) (GC)(AC)G TCT GG-3'(1-8)

5' Human VH Chain Primer II with Nco I site (sense)

5'-CC ATG GAG GT(AG) CAG CTG CAG (GC)AG TC(AG) GG-3'(1-8)

5' Human VH Chain Primer III with Nco I site (sense)

5'-CC ATG GAG GTG CA(AG) CTG (GT)TG GAG TCT GGG-3'(1-8)

3' Human VH Chain Primer with GlyGlyGly (anti-sense)

90

5'-GCC TCC GCC GGT GAC CA(GT) (GT)GT (CT)CC (CT)TG GCC CCA3'(1 03-

110)

5' Human Vk Chain Primer with GlyGlySer (sense)

5'-GGA GGA TCA GA(ACT) AT(CT) (GC)(AT)G (AT)TG AC(AGCT) CAG TCT-3'

(1-7)

3' Human Vk Chain Primer with Xho 1 site (anti-sense)

5'-CTC GAG (CT)TC (CT)AC CTT GGT CCC-3'(1 01-1 05)

The (GlY4Ser)3 Polylinker (sense)

5'-GGC GGA GGC GGA TeA GGA GGA GGA GGA TCA GGC GGA GGA GGA

TCA-3'

91

APPENDIX B

BUFFERS AND MEDIA

Phosphate-buffered saline (PBS): 136mM NaCI, 10mM Na2HP04, 2.7mM KCI, 1.8mM KH2P04 in dH20.

Tris-buffered saline (TBS): 136mM NaCI, 2.7mM KCl, 24.8mM Tris-base in dH20, pH 8.0.

TBS-1\veen (TBS-T): O.OS% Tween-20 in TBS.

Carbonate butTer: 3SmM NaHC03, ISmM Na2C03, pH 9.6.

Superbroth (SB): 30 g Tryptone, 20 g Yeast extract, 10 g MOPS, pH 7.0 in 1 liter dH20; autoclave for sterility.

Luria Bertoni medium (LB): 10 g Tryptone, S g Yeast extract, S g NaCI, Iml IN NaOH in 1 liter dH20; autoclave for sterility.

Luria Bertoni Agar: 10 g Tryptone, S g Yeast extract, S g NaCI, IS g agar, Iml IN NaOH in 1 liter dH20; autoclave to dissolve agar and for

sterility. Carbenicillin may be added when temperature cools to SO-SSoC

at a final concentration of SOug/ml.

S-Bromo-4-chloro-3-indolyl 13-D-galactopyranoside (X-gal): reagent may be added to agar for blue/white colony screening at a final concentration

of20ug/ml.

Isopropyl B-D-thiogalactopyranoside (IPTG): reagent was added to mid log

phase cultures at ImM to induce cultures to produce the desired protein.

REFERENCES

l. Grabar, P. The historical background of immunology. In: Basic and Clinical Immunology (5th edition) edited by Stites, D. P., Stobo, J. D., Fudenberg, H. H. and Wells, J. V. LANGE ppl-12, 1984

92

2. Nabholz, M. and MacDonald, H.R. Cytolytic T lymphocytes. Annu. Rev. Immuno. 1 :273-296, 1983

3. ROitt, I.M., Torrigiani, G., Greaves, M.F., Brostoff. J. and Playfair, J.H. The Cellular Basis of Immunological Response. Lancet 2:367-371, 1969

4. Borsos, T. Immunoglobulin Classes and Complement Fixation, In: Progress in Immunology edited by B. Amos, p841. Academic, New York, 1971

5. Kodera, Y. and Bean, M.A. Antibody-Dependent Cell-Mediated CytotoxiCity for Human Monolayer Target Cells Bearing Blood Group and Transplantation Antigens and for Melanoma Cells. Int. J. Cancer 16:579-592, 1975

6. Hasemann, C. A. and Capra, J. D. Immunoglobulins: Structure and function. In: Fundamental Immunology (2nd edition). pp. 209-233. Raven Press, New York, 1989

7. Goodman, J. W. Immunoglobulin I: Structure and function. In: Basic and Clinical Immunogy (5th edition). LANGE pp30-42. 1984

8. Brown, P. H., and Hickman, S. Oligosaccharide processing at individual glycosylation sites on MOPC 104E immunoglobulin M. J. Biol. Chem 261:2575-2582, 1986

9. Edelman, G. M. The covalent structure of a human gammaG­immunoglobulin. XI. Functional implications. Biochem. 9:3197-3204

10. Capra, J. D. Hypervariable region of human immunoglobulin heavy chain. Nature 230:62, 1971:

93

11. Wu, T. T., and Kabat, E. A. An analysis of variable regions of Bence­Jones proteins and myeloma light chains and their implications of antibody complementarity. J. Exp. Med. 132:211-219, 1970;

12. Kabat, E. A., Wu, T. T., and Bilofsky, H. Unusual distribution of amino acids in complementarity-determining (hypervariable) segments of heavy and light chains of immunoglobulins and their possible roles in specificity of antibody-combining sites. J. Biol. Chern. 252:6609-6618. 1977;

13. Kabat, E. A. The structural basis of antibody complementarity. Advan. Protein Chem. 32: 1-13. 1978

14. Kabat, E. A.: Wu, T. T. Attempts to Locate Complementarity­determining Residuals in Variable Positions of Light and Heavy Chains. Ann. NY Acad. ScL 190:382-389, 1971

15. Korsmeyer, S. J. and Waldmann, T. A. Immunoglobulin II: Gene organization and assembly. In: Basic and Clinical Immunology (5th edition) LANGE pp.43-54, 1884

16. Early, P., Huang, H., Davic, M., Calame, K., and Hood, L. An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH. Cell 19:981-992, 1980

17. Sakano, H., Maki, R., Kurosawa, Y. W. Roeder, and Tonegawa, S. '!\vo types of somatic recombination are necessary for the generation of complete immunoglobulin heavy-chain genes. Nature 286:676-683, 1980

18. Alt, F. W. and Baltimore, D. Joining of immunoglobulin heavy chain gene segments: Implications from a chromosome with evidence of three D-JH fusions. Proc. Natl. Acad. Sci. USA 79:4118-4122, 1982

19. Early, P., and Hood, L. Allelic exclusion and nonproductive immunoglobulin gene rearrangements. Cell 24: 1-3, 1981

94

20. Alt, F. W., Enea, V., Bothwell, A. L. M., and Brltimore, D. Activity of multiple light chain genes in murine myeloma cells producing a single, functional light chain. Cell 21 : 1-12, 1980

21. Ritchie, K. A., Brinster, R. L., and Storb, U. Allelic exclusion and control of endogenous immunoglobulin gene rearrangement in k transgenic mice. Nature 312:517-520, 1984

22. Kishimoto, T. and Hirano, T. B lymphocyte activation, proliferation, and immunoglobulin secretion. In: Fundamental immunology (2nd edition). pp385-411, Raven Press. New York, 1989

23. Coleclough, C., Cooper, D., and Perry, R. P. Rearrangement of immunoglobulin heavy chain genes during B lymphocyte development as revealed by studies of mouse plasmacytoma cells. Proc. Natl. Acad. Sci. 77:1422-1426, 1980

24. Hurwitz, J. L., Coleclough, C., and Cebra, J. J. CH gene rearrangements in IgM-bearing B cells and in the normal splenic DNA component of hybridomas making different isotypes of antibody. Cell 22:349-359, 1980

25. Marcu, K. B., Lang, R. B., Stanton, L. W., and Harris, L. J. A model for the molecular requirements of immunoglobulin heavy chain class swicthing. Nature 298:87-89, 1982

26. Weigert, M., Gatmaitan, L., Loh, E., Schilling, J. and Hood, L. Rearrangement of genetic information may produce immunoglobulin diversity. Nature 276:785-790, 1978:

27. Tonegawa, S. Somatic generation of antibody diversity. Nature 302:575-581, 1983

28. Weigert, M., Perry, R, Kelly, D., Hunkapiller, T .. Schilling, J. and Hood, L. The joining of V and J gene segments creates antibody diversity. Nature 283:497-499, 1980

95

29. Early, P., Huang, H., Davic, M., Calame, K., and Hood, L. An immunoglobulin heavy chain variable region gene is generated from three segments of DNA: VH, D and JH. Cell 19:981-992, 1980

30. Bernard, 0., Hozumi, N., and Tonegawa, S. Sequences of mouse immunoglobulin light chain genes before and after somatic changes. Cell 15:113-1144,1978

31. Kim, S., Davis, M., Sinn, E., Pattern, P. and Hood, L. Antibody diversity: Somatic hypermutation of rearranged VH genes. Cell 27:573-581, 1981

32. Selsing, E. and Storb, U. Somatic mutation of immunoglobulin light­chain variable-region genes. Cell 25:47-58, 1981

33. Kohler, G. and Milstein, C. Continuous Cultures of Fused Cells Secreting Antibody of Predefined SpeCificity. Nature 256:495-497, 1975

34. Littlefield, J. W. Selection of hybrids from matings of fibroblasts in vitro and their presumed recombinants. Science 145:709-710, 1964

35. Winter, G. and Milstein, C. Man-made Antibodies. Nature 349:293-299, 1991

36. Cote, R.J., Morrisey, D.M., Houghtan. A.N., Thomason, T.M .. Daly, M.E., Oettgen, H.F., Old, L.J. Specificity Analysis of Human Monoclonal Antibodies Reactive with Cell Surface and Intracellular An tigens. Proc. Natl. Acad. ScL USA 83:2959-2963, 1986

37. Fujinaga, S., Sugano, T., Matsumoto, Y-I., Masuho, Y. and MOri, R Antiviral Activities of Human Monoclonal Antibodies to Herpes Simplex Virus. J. Inject. Dis. 155:45-53, 1987

96

38. Lake. D.F .. Tomlyama. T., Robinson Jr., W.E., Matsumoto, Y., Masuho, Y. and Hersh, E.M. Generation and Characterization of a Human Monoclonal Antibody That Neutralizes Diverse HIV-1 Isolates In Vitro. AIDS 6: 17-24, 1992

39. Waldman. T.A. Monoclonal Antibodies in Diagnosis and Therapy. Science 252:1657-1662. 1991

40. Morrison. S.L .. Johnson. M.J .. Herzenberg, L.A. and Oi. V.T. Chimeric Human Antibody Molecules: Mouse Antigen-binding Domains with Human Constant Region Domains. Proc. Natl. Acad. Sci. USA 81:6851-6855, 1984

41. Reichmann. L .. Clark. M .. Walsdmann. H. and Winter. G. Reshaping Human Antibodies for Therapy. Nature 332:323-327, 1988

42. Jaffers. G.J .. Fuller. J.C .. Cosimi, B .. Russell. P. S .. Winn. H.J. and Colvin. R.B. Monoclonal Antibody Therapy. Transplantation 41 :572-578. 1986

43. Bruggemann. M .. Winter. G .. Waldmann. H .. and Neuberger. M.S. The immunogenicity of chimeric antibodies. J. Exp. Med. 170:2153-2157. 1989

44. Hale. G .. Clark, M. R .. Marcus. R .. Winter. G .. Dyer. M. J. S .. Phillips. J. M .. Riechmann. L. and waldmann. H. Remission induction in non-Hodgkin lymphoma with reshaped human monoclonal antibody COMPATH-1H. Lancet 17:1394-1399,1988

45. Mathieson. P. W .. Cobbold, S. P .. Hale, G .. Clark. M. R .. Oliveria. D. B. G .. Lockwood. C. M. and Waldmann, H. Monoclonal-antibody therapy in systemic vasculitis. New EngL J. Med. 323:250-254, 1990

46. Isaacs. J. D .. Watts. R. A .. Hazlemen. B. L .. Hale. G .. Keogan. M. T .. Cobbold. S. P .• Waldmann. J. Humanised monoclonal antibody therapy for rheumatoid arthritis. Lancet 340:748-752. 1992

47. Orlandi, R., Gussow, D. H., Jones, P. T., and Winter, G. Cloning immunoglobulin variable domains for expression by the polymerase chain reaction. Proc. Natl. Acad. Sci. USA 86:3833-3837, 1989

48. Huse, W.D., Sastry, L., Iverson, S.A., Kang, A.S., Alting-Mees, M., Burton, D.R., Benkovic, S.J., Lerner, R.A.: Generation of a Large Combinatorial Library of the Immunoglobulin Repertoire in Phage Lambda. Science 246:1275-1281, 1989

49. Larrick, J.W., Danielsson, L., Brenner, C.A., Abrahamson, M., Fry, K.E. and Borreback C.A.K.: Rapid Cloning of Rearranged Immunoglobulin Genes from Human Hybridoma Cells Using Mixed Primers and the Polymerase Chain Reaction. Biochem. and Biophy. Res. Commu. 160: 1250-1256, 1989

97

50. Skerra, A. and Pluckthun, A.: Assembly of A Functional Immunoglobulin Fv Fragment in Escherichia coli. Science 240: 1038-1041, 1988

51. Beetter, M., Chang, C. P., Robinson, R. R., and Horwitz, A. H. Escherichia coli secretion of an active chimeric antibody fragment. Science 240:1041-1043

52. Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf. S. J., Higuchi, R., Hom, G. T., Mullis, K. B., and Erlich, H. A. Primer-directed enzymatic amplification of DNA with a thermostable DNA polymerase. Science 239:487-491, 1988

53. Kabat. E. A., Wu, T. T .. Reid-Miller, M., Perry, H. M., and Gottesmann, K. S. "Sequences of Proteins of Immunological Interest." U. S. Department of Health and Human Services, U. S. government Printing Office, Washington, DC., 1991

54. Persson, M. A. A., Caothien, R. H., and Burton, D. R. Generation of diverse high-affinity human monoclonal antibodies by repertoire cloning. Proc. Natl. Acad. ScL USA 88:2432-2436, 1991

55. Boss, A., Kenten, J.H., Wood, C.R. and Entage, J.S.: Assembly of Functional Antibodies from Immunoglobulin Heavy Chain and Light Chains Synthesized in E. coli. Nucl. Acids Res. 12:3791-3800, 1984

98

56. Cabilly, S., Rigga, A.D., Parde, H., Shively, J.E., Holmes, W.E., Rey, M., Perry, L.J., Wetzel, R. and Heyneker, H.L.: Generation of Antibody Activity from Immunoglobulin polypeptide Chains Produced in E. coli. Proc. Natl. Acad. ScL USA 81 :3273-3277, 1984

57. Bird, R.E., Hardman, K.D., Jacobson, J.W., Johnson, S., Kaufman, B.M., Lee, S-M., Lee, T., Pope, S.H., Riordan, G.S., Whitlow, M.: Single­chain Antigen-Binding proteins. Science 242:423-426, 1988

58. Better, M., Chang, C.P., Robinson, R.R. and Horwitz, A. H.: Escherichia coli Secretion of An Active Chimeric Antibody Fragment. Science 240: 1041-1042, 1988

59. Liu, F-T., Albrandt, K. A., Bry, C. G., and Ishizaka, T. Expression of a biologically active fragment of human IgE e chain in Escherichia coli. Proc. Natl. Acad. Sci. USA 77:2138-2142, 1984;

60. Cabilly, S. Growth at sub-optimal temperatures allows the production of functional, antigen-binding Fab fragments in Escherichia coIL Gene 85:553-557, 1989;

61. Buchner, J., and Rudolph, R. Renaturation, purification and characterization of recombinant Fab-fragments produced in Escherichia coli. Biotech. 9: 157-162, 1991

62. Bird R. E., Hardman K. D., Jacobson J.W., Johnson S., Kaufman B. M., Lee S.-M., Lee T., Pope S. H., Riordan G. S., Whitlow M.: Single­chain Antigen-binding Proteins. Science 242:423-426, 1988

63. Huston J.S., Levinson D., Mudgett-Humter M., Tai M.-S., Novotny J., Margolies M.N., Ridge R.J., Bruccoleri R.E., Haber E., Crea R., Oppermann H.: Protein Engineering of Antibody Binding Sites: Recovery of SpeCific Activity in An Anti-digoxin Single-chain Fv Analogue Produced in E. coli. Proc. Natl. Acad. Sci. USA 85:5879-5883, 1988

64. Huston J.S., Mudgett-Hunter M., Tai M.-S., McCartney J., Warren F., Haber E., Oppermann H.: Protein Engineering of Single-chain Fv Analogs and Fusion Proteins. Meth. EnzymoL 203:47-88. 1991

99

65. Schlom, J. Antibodies in cancer therapy: BasiC principles of monoclonal antibodies. In: Biologic Therapy of Cancer. ed. DeVita. J. V. T., Hellman, S., Rosenberg. S. A. Philadelphia, J. B. Lippincott Compant, 1:464-481,1991

66. Yokota T., Milenic D. E., Whitlow M., Schlom J.: Rapid Tumor Penetration of A Single-chain Fv and Comparison with Other Immunoglobulin Forms. Cancer Res. 52:3402-3408, 1992

67. Brinkmann, U., Pai, L. H., Fitzgerald, D. J., Pastan I. B3 (Fv)­PE38KDEL, a single-chain immunotoxin that causes complete regression of a human carcinoma in mice. Proc. Natl. Acad. Sci. USA 88:8616-8620, 1991

68. Chaudhary, V. R., Queen, C., Junghans, R P., Walsmann, T. A., FitzGerald, D. J., and Pastan, I. A recombinant immunotoxin consisting of two antibody variable domains fused to Pseudomonas exotoxin. Nature 339:394-397, 1989:

69. Chaudhary, V. R., Batra, J. K., gHo, M. G., Willingham, M. C., FitzGerald, D. J., and Pastan, I. A rapid method of cloning functional variable-region antibody genes in Escherichia coli as single-chain immunotoxins. Proc. Natl. Acad. Sci. USA 87: 1066-1070, 1990

70. Kreitman, K.J, Chaudhary, V. R, Waldmann, I., Willingham, M. U., FitzGerald, D. J., and Pastan, I. The recombinant immunotoxin anti-Tac {Fv}-Pseudomonas exotoxin 40 is cytotoxiC toward peripheral blood malignant cells from patients with adult T ceHleukemia. Pro. Natl. Acad. SeL USA 87:8291-8295, 1990

71. Hoogenboom, H. R, Griffiths, A. D., Johnson, K. S., Chiswell, D. J., Hudson, P., Winter, G. Multi-subunit proteins on the surface of filamentous phage: Methodologies for displaying antibody (Fab) heavy and light chains. Nucl. Acids Res. 19:4133-4137, 1991

100

72. Gram, H., Marconi, L., Barbas III, C. F., Collet, T. A., Lerner, R. A., Kang, A. S. In vitro selection and affinity maturation of antibodies from a naive combinatorial immunoglobulin library. Proc. Natl. Acad. Sci. USA 89:3576-3580, 1992

73. Burton, D. R., Barbas III, C. F., Persson, M. A. A., Koening S., Chanock R. M., Lerner, R. A. A large array of human monoclonal antibodies to type 1 human immunodeficiency virus from combinatorial libraries of asymptomatic seropositive individuals. Proc. Natl. Acad. Sci. USA. 88:10134-10137, 1991;

74. Marks, J. D., Hoogenboom, H. R., Griffiths, A. D., \\Tinter, G. Molecular evolution of proteins on filamentous phage. Mimicking the strategy of the immune system. J. BioI. Chern. 267:16007-16010, 1992

75. McCafferty, J., Griffiths, A. D., Winter, G., Chiswell, D. J. Phage antibodies: filamentous phage displaying antibody variable domains. Nature 348:552-554, 1990

76. Kang, A.S., Barbas, C.F., Janda, K., Benkovic, S.J. and Lerner, R.A.: Linkage of Recognition and Replication Functions by Assembling CombinatOrial Antibody Fab Libraries Along Phage Surfaces. Proc. Natl. Acad. Sci. USA 88:4363-4366, 1991

77. Barbas, C.F. III, Kang, A. S., Lerner, R. and Benkovic, S.J.: Assembly of combinatOrial Antibody Libraries on Phage Surfaces: the Gene III sites. Proc. Natl. Acad. ScL USA 88:7978-7982, 1991

78. Huse, W. D., Sastry, L., Iverson, S., Kang, A. S., Alting-Mees, M., Burton, D. R., Benkovic, S. J., Lerner, R. A. Generation of a large combinatorial library of the immunoglobulin repertoire in phage lambda. Science 246:1275-1281, 1989

79. Marks, J. D., Hoogenboom, H. R., Bonnert, T. P., McCafferty, J., Griffiths, A. D., Winter, G. By-passing immunization. Human antibodies for V-gene libraries displayed on phage. J. Mol. BioI. 222: 581-597, 1991

101

80. Mullinax, D. E., Gross, E. A., Amberg, J. R., Hay, B. N., Hogrefe, H. H., Kubitz, M. M., Greener. A., Alting-Mees, M., Ardourel, D., and Short, J. M. Identification of human antibody fragment clones specific for tetanus toxoid in a bacteriophage lambda immunoexpression library. Proc. Natl. Acad. Sci. USA 87:8095-8099, 1990

81. McCafferty, J., Griffiths, A. D., Winter. G .. and Chiswell. D. J. Phage antibodies: filamentous phage displaying antibody variable domains. Nature 348:552-554. 1990

82. Kreider. J.W .. Bartlett, G.L. and Butkiewicz. B.L.: Relationship of Tumor Leucocytic Infiltration to Host Defense Mechanisms and Prognosis. Cancer Met. Rev. 3:53-74. 1984

83. Pretlow. T.P .. Keith. E.F., Cryar. A.K.. Bartolucci. A.A .. Pitts. A. M .. Pretlow II. T.G .. Kimball. P.M. and Boohaker. E.A.: Eosinophil Infiltration of Human Colonic CarCinomas as A Prognostic Indicator. Cancer Res 43:2997-3003, 1983

84. Svennevig, J.L. and Svaar. H.: Content and Distribution of Macrophages and Lymphocytes in Solid Malignant Human Tumors. Int. J. Cancer 24:754-758, 1979

85. Barbuto, J.A.M., Verastegui, E. and Hersh, E.M.: The Use of Transformed T Cell Lines for Clonal Expansion of Human B Cells for Peripheral Blood, Spleen. and Tumor-infiltrating lymphocytes. Hybrid. 12: 115-125. 1993

86. Fisher, E.R.. Paik, S.M., Rockette, H .. Jones, J .. Caplan. R., Fisher, B. and other NSABP collaborators: Prognostic Significance of Eosiophills and Mast Cells in Rectal Cancer. Hum. Pathol20:159-163. 1989

87. RiIke, F., Colnaghi, M.L, Cashinelli. N., Anderola. S., Baldini, M.T., Butalino., R. Della Porta, G., Menard, S .. Pierotli, M.A. and Testori, A,.: Prognostic Significance of HER-2/NEU ExpreSSion in Breast Cancer and Its Relationship to Other Prognostic Factors. Int. J. Cancer 49:44-49, 1991

102

88. Whiteside. T.L .. Miescher. S .. Hurlimann. J .. Moretta, L. and Von Fliedner. V.: Clonal Analysis and in situ Characterization of Lymphocytes Infiltrating Human Breast Carcinomas. Cancer Immunol. Immunother. 23: 169-178. 1986

89. Itoh. K .. Tilden. A.B. and Bulch. C.M.: IL-2Activation of Cytotoxic T Lymphocytes Infiltrating into Human Metastatic Melanoma. Cancer Res. 46:3011-3017. 1987

90. Rabinowich. H .. Cohen. R .. Bruderman. I.. Steiner. Z. and Klajman. A.: Functional Analysis of Mononuclear Cells Infiltrating into Tumors: Lysis of Autologous Tumor Cells by Cultured Infiltrating Lymphocytes. Cancer Res. 47: 173-177. 1987

91. Shimizu. Y .. Iwatsuki. S .. Herberman. R.B. and Whiteside. T.L .. : Clonal Analysis of Tumor Infiltrating Lymphoctes from Human Primary and Metastatic Liver Tumors. Int. J. Cancer 46:878-883, 1990

92. Vose. B.M. and Moore, M.: Human Tumor-infiltrating Lymphocytes: A Marker for Host Response. Seminars in Hematology 22:27-40. 1985

93. Punt. C.J.A .. Barbuto. J. A. M., Zhang, H .. Grimes. W.J .. Hersh. E.M.: Anti-tumor Antibodies Produced by Human Tumor-infiltrating and Peripheral Blood B Lymphocytes. Cancer Immunol. Immunother. 38:225-232. 1994

94. Shimokawara. I.. Imamura, M .. Yamanaka. N., Ishii. Y. and Kikuchi, K.: Identification of Lymphocyte-subpopulations in Human Breat Cancer Tissue and Its Significance: An Immunoperoxidase Study with Anti­human T and B cell Sera. Cancer 49: 1456-1460. 1982

95. Jerne. N. K.. Henry. c .. Nordin. A. A., Fuji, H .. koros, A. M. C .. and Lefkovits. I. Plaque forming cells: Methodology and theory. Transplant Rev. 18:130-191. 1974

96. Steinitz. M .. Klein, G., Koskimies, S., and Makela, O. E-B virus induced B lymphocyte cell lines producing specific antibody, Nature 269:420-422. 1977

103

97. Roder. J. C .. Cole. S. P. C .. Atlaw. T .. Campling B. G .. McGarry. R. C .• and Kozbor. D. The Epstein-Barr virus-hybridoma technique. in: Human hybridomas and monoclonal antibodies. E. G. Engleman. S. K. H. Foung. J. larrick and A. Raubitschek (edl. Plenum Press: New York. pp55-70. 1985

98. Hirohata. S .. Jelinek. D. F .. and Lipsky. P. E. T cell dependent activation of B cell proliferation and differentiation by immobilized monoclonal antibodies to CD3. J.lmmunol. 140:3736-3744. 1988

99. Armoroso. K. and Lipsky, P. E. Frequency of human B cells that differentiate in reponse to anti-CD3 activated T cells. J. Immunol. 145:3155-3161. 1990

100. Mizoguchi. H .. O'Shea. J.J .. Longo. D. L .. Loeffler. C. M .• McVicar. D. W .. and Ochoa. A. C. Alterations in signal transduction molecules in T lymphocytes from tumor-bearing mice. Science 258: 1795-1798. 1992;

101. Hersh. E. M .. Mavligit. G. M .. and Gutterman. J. U. Immunodeficiency in cancer and the importance of immune evaluation of the cancer patient. Med. Clin. N. Am. 60:623-639, 1976

102. Barbuto. J. A. M .. Verastegui. E. L. and Hersh. E. M. The use of transformed T cell lines for clonal Expansion of human B cells from peripheral blood. spleen. and tumor-infiltrating lymphocytes. Hybrid. 12: 115-123, 1993

103. Manson. L.A.: Does Antibody-dependent Epitope Masking Permit Progressive Tumor Growth in the Face of Cell-mediated Cytotoxicity? Imm. Today 12:352-355. 1991

104. Shetye. J .. Frodin. J-E. Christensson. B .• Grant. C .• jacobson. B .• Sundelius. S .. Sylven. M .. Biberfeld, P. and Mellstedt. H.: Immunohistochemical Monitoring of Metastatic Colorectal Carcinoma in Pationts Treated with Monoclonal Antibodies(MAb 17 -IAl. Cancer Immunol. Immunother. 27: 154-162. 1988

104

105. Trauth, B.C., Klas, C., Peters, A.M.J., Mutzku, S. Moller, P., Falk, W., Debatin, K-M and Krammer, P.H.: Monoclonal Antibody Mediated Tumor Regression by Induction of Apoptosis. Science 245:301-305, 1989

106. Houghton, A. N., Eisinger, M., Albino, A. P., Cairncross, J. G .• Old. L. J. Surface antigens of melanocytes and melanomas. Markers of melanocytes differentiation and melanoma subset. J. Exp. Med. 156: 1755-1766, 1982

107. Herlyn. M., Koprowski. H. Melanoma antigens: immunological and biological characterization and clinical significance. Ann. Rev. Immunol. 6:283-308, 1988

108. Cocharan, A. J. Melanoma markers: biological and diagnostic considerations. Monogr. Pathol. 30:35-49, 1988

109. Cochran, A. J .. Lu. H-F, Li. P-X, Saxton, R., \Ven, D-R. S-100 protein remains a practical marker for melanocytic and other tumors. Melanoma Res. 3:325-330. 1993

110. Gown, A. M .. Vogel, A. M., Hoak. D .. Gough. F., McNutt. M A. Monoclonal antibodies specific for melanocytic tumors distinguish subpopulations ofmelanocytes. Am. J. PathoL 123:195-203, 1986

111. Jimbow. K., Fitzpatrick, T. B .. Quevedo, W. C .. Casade of melanogenesis in epidermal melanin pigmentation. The melanosome as a programmed organelle in structure and function of melanin. In: Structure and function of melanin. (Edited by Jimbow, K.) New York. Oxford Press1985, pp71-82

112. Vijayasaradhi. S., Bouchard, B., Houghton, A. N. The melanoma antigen gp75 is the human bomologue of the mouse blocus gene product. J. Exp. Med. 171:1375-1380, 1990

113. Carrel. S. and Rimold, D. Melanoma-associated Antigens. Eur. J. Cancer 29A: 1903-1907. 1993

105

114. Anichini, A., Mortarini. R, Berti. E .• Parmiani. G. Multiple VLA antigens on a subset of melanoma clones. Human Immunol. 28:119-122. 1990

115. Albelda. S. M .. Mette. S. A .• Elder. D. E .. et al. Integrin distribution in malignant melanoma: association of the.133 subunit with tumor progression. Cancer Res. 50:6757-6764. 1990

116. Johnson. J. P .. Stade. B. G .. holzmann. B., Schwable. W .. Riethmuller. G. De novo expression of intercellular-adhesion molecule 1 in melanoma correlates with increased risk of metastasis. Proc. Natl. Acad. Sci. USA 86:641-744. 1989

117. Halaban. R. Growth factors regulating normal and malignant melanocytes. In: Melanoma research: Genetics. growth factors, metastases and antigens. (edited by Nathanson. L.) Boston. Kluwer Academic Publishers 30: 19-40. 1991

118. Carrel. S .. Dore. J-F. Ruiter. D, J .. et al. The EORTC melanoma group exchange program: evaluation of a multicenter monoclonal antibody study. Int. J. Cancer. 48:836-847. 1991

119. Coulie. P. G .. Weynants. P .. Lehmann. F" herman. J .. Brichard. V., Wolfel. T .. Van Pel. A .. De Plaen. E .• Brasseur. F .. Boon. T. Genes coding for tumor antigens recognized by human cytolytic T lymphocytes. J. ImmwlOther. 14:104-109. 1993

120. Gaugler. B .. Vande Eynde. B,. Van Der Bruggen. P .. Romero, p,. Gaforio. J, J .. De Plaen. E .. Lethe. B,. Brasseur. F .. Boon. T. Human gene MAGE-3 codes for an antigen recognized on a melanoma by autologous cytolytic T lymphocytes. J. Exp. Med. 179:921-930. 1994

121. Kawakami. Y., Eliyahu. S .. Delgado. C. H .. Robbins. P. F .. Rivoltini. L.l Topalian. S. L .. Miki, T .. Rosenberg. S. A. Cloning of the gene coding for a shared human melanoma antigen recognized by autologous T cells infiltrating into tumor. Proc. Natl. Acad. Sci. USA 91:3515-3519. 1994

122. Bakker. A. B .. Schreurs. M. W .. De Boer. A. J .. Kawakami. Y .. Rosenberg. S. A .. Adema. G. J .. Figdor. C. G. Melanocyte lineage-specific

antigen gp100 is recognized by melanoma-derived tumor-infiltrating lymphocytes. J. Exp. Med. 179:1005-1009, 1994

106

123. Wolfel, T., Van Pel, A., Brichard, V., Schneider, J., Seliger, B., Meyer, Z., Buschenfelde, K. H., Boon, T. 1\vo tyrosinase nonapeptides recognized on HLA-A2 melanomas by autologous cytolytic T lymphocytes. Eur. J. Immu. 24:759-764, 1994

124. Morton, D. L. Active immunotherapy against cancer: Present status. Semin. Oneal. 13: 180-187, 1986

125. Bystryn, J.C., Jacobsen, S, Harris, M., et al. Preparation and characterization of a polyvalent human melanoma vaCCine. J. Bial. Res. Mod. 5:211-218, 1986

126. Wallack, M. K., McNally, K. R., Leftheriotis, E., et al. A Southeastern Cancer Study Group phase 1/11 trial using vaccinia melanoma oncolysates. Cancer 57:649-658, 1986

127. Estin, C. D., Stevenson, U. S., Plowman, G. D., et al. Recombinant vaccinia virus vaccine against the human melanoma antigen p97 for use in immunotherapy. Proc. Natl. Aead. Sci. USA 85: 1052-1059, 1988

128. Kusama, M., Kageshita, T., Chen, Z. J. ferone, S. Characterization of syngenetic antiidiotypic MAb to murine anti-human high molecular weight melanoma-associated antigen (HMW-MAA) MAb. J. Immunal. 143:3844-3851, 1989

129. Hellstrom, K E., Hellstrom, I., Morton, D. L., et al. Melanoma vaccines. In: Cutaneous Melanoma. pp.542-559 (edited by balch, C. M., Houghton, A. N., Milton, G. W., et al.) Philadelphia, PA: J. B. Lippincott, 1992

130. Cheung H.-K. V. Lazarus, H., Miraldi, F. D., et al. Ganglioside GD2 specific monoclonal antibody 3F8: A phase I study in patients with neuroblastoma and malignant melanoma. J. CUn. Oneal. 5: 1430-1435, 1987

107

131. Houghton. A. N .. Mintzer. D .. Cordon-Cardo. C .. et al. Mouse monoclonal IgG3 antibody detecting GD3 ganglioside: A phase I trial in potients with malignant melanoma. Proc. Natl. Acad. Sci. USA 82: 1242-1246. 1985

132. Cheresh. D. A .. Honsik. C. J .. Staffileno. L. K.. et al. Disialoganglioside GD3 on human melanoma serveds as a relevent target antigen for monoclonal antibody-mediated tumor cytolysis. Proc. Natl. Acad. Sci. USA 82:5155-5159. 1985

133. Vadhan-Raj. S. Cordon-Cardo. C .. Carswell. E .. et al. Phase I trial of a mouse monoclonal antibody against GD3 ganglioside in patients with melanoma: Induction of inflammatory responses at tumor sites. J. CUn. Oncol. 6: 1636-1642. 1988

134. Spitler. L. E .. del Rio. M .. Khentigan. A .. et al. Therapy of patients with melanoma using a monoclonal antimelanoma antibody-ricin A chain immunotoXin. Cancer Res. 47: 1717 -1723. 1987

135. Spitler. L. E .. Minor. D. R. Monoclonal antimelanoma ricin A chain immunotoXin therapy of melanoma in outpatients. Pmc. Am. Soc. CUn. Oncol. 8:1117-1124.1989

136. Munn. D. H .. Cheung. N.-K. Interleukin-2 enhancement of monoclonal antibody-mediated cellular cytotoXicity agarinst human melanoma. Cancer Res. 47:6600-6606. 1987

137. Balkwill. F. R. and Burke. F. The cytokine network. Imm. Today 10:299-302. 1989

138. Mizel. S. B. The interleukins. FASEB 3:2379-2383. 1989

139. Murray. J. L .. Rosenblum. M. G .. Lamki. L .. et al. Enhancement of tumor uptake of indium-111-labled antimelanoma monoclonal antibody 96.6 in melanoma patients receiving partially purified alpha interferon. Proc. Am. Soc. CUn. Oncol. 5:A883-889. 1989

108

140. Grimm, E. A., Mazumder, A., Zhang, H. Z., Rosenberg, S. A. Lymphokine activated killer cell phenomenon: Lysis of natural killer­resistant fresh solid tumor cells by interleukine-2 activated autologous human peripheral blood lymphocytes. J. Exp. Med. 155:823-830, 1982

141. Rosenberg, S. A., Lotze, M. T., Muul, L. M., et al. A new approach to the therapy of cancer based on the systemiC administration of autologous lymphokine-activated kinller cells and recombinant IL-2. Surgery 100:262-269, 1986

142. Rosenberg, S. A., Spiess, P, Lafreniere, R. A new approach to the adoptive immunotherapy of cancer with tumor-infiltrating lymphocytes. Scienee 233:1318-1320, 1986

143. Topalian, S. L., Solomon, D., Avis, F. P., et al. Immunotherapy of patients with advanced cancer using tumor-infiltrating lymphocytes and recombinant interleukin-2: A pilot study. J. CUn. Oneal. 6:839-845, 1988

144. Rosenberg, S. A., Packard, B. S., Aebersold, P. M., et al. use of tumor-infiltrating lymphocytes and interleukin-2 in the immunotherapy of patients with metastatic melanoma. N. Engl. J. Med. 319: 1676-1681, 1988

145. Fisher, B., Backard, B. S., Read, E. J., et al. Tumor localization of adoptively transferred indium-Ill labeled tumor infiltrating lymphocytes in patients with metastatic melanoma. J. CUn. Oneo. 7:250-256, 1989

146. Rosenberg, S. A., Aebersold, P., Cornetta, K., et al. Gene transfer into humans-Immunotherapy of patients with advanced melanoma, using tumor-infiltrating lymphocytes modified by retroviral gene transduction. N. Engl. J. Med. 323:570-576, 1990

147. Rosenberg, S. A. The immunotherapy and gene therapy of cancer. J. CUn. Oneol.l0:180-188, 1992

148. Whiteside, T.L., Miescher, S., MacDonald, H.R. and Von Fliedner, V. Separation of Tumor Infiltrating Lymphocytes from Tumor Cells in Human Solid Tumors. J. Immunal. Meth. 90:221-233, 1986

109

149. Saxon, A., Feldhaus, J. and Robins, R. A. Single Step Separation of Human T and B Cells Using AET-treated Sheep Red Cells. J. Immunol. Methods. 12:285-289, 1976

150.Raubitscheck, A.A. Epstein-Barr Virus Transformation. In: Human Hybridomas and Monoclonal Antibodies, edited by Engleman, E.G., Foung, S.K.H., Larrick, J. and Raubitscheck, A. New York: Plenum Press. 1985, p. 454-455.

151. Herry, C., Marbrook, J., Vann, D.C., Kondlin, D. and Wo fsy , C. Limitting Dilution Analysis. in: Selected Methods in Cellular Immunology. B. B. Mishell and S. M. Shiigi(ed). W. H. Freeman and Company, San Francisco. pp. 138-152.

152. Kabat, E.A., Wu, T.T., Perry, H.M., Gottesham, K.S., Foeller, C. Sequences of Proteins of Immunological Interest(5th Ed). Washington, D.C., United States Dept. of Health and Human Services, 1991

153. Horton, R. M., Hunt, H. D., Ho, S. N., Pullen, J. K., and Pease, L. R. Engineering hybrid gene without the use of restriction enzymes: gene splicing by overlap extension. Gene 77:61-68, 1991

154. Marchuck D., Drumm M., Sanlino A., and Collins F.S. Construction ofT-vectors, A Rapid and General System for Direct Cloning of Unmodified PCR Products. Nucl. Acid Res. 19:1154. 1991

155. Clark, J. M. novel non-template nucleotide addition reaction catalY.led by procaryotic and eucaryotic DNA polymerase. Nucl. Acids Res. 16:9677-9686, 1988

156. Ausubel F.M., Brent R., Kingston R.E. Moore D.D., Seidman J.G., Smith J .A. and Struhl K. Current Protocols in Molecular Biology. New York, Greene Publishing Associates and Wiley Intersciences.

157. Sanger F., Nickler S. and Coulson A. R. DNA Sequence with Chain­terminating Inhibitors. Proc. Natl. Acad. Sci. USA. 74:5463-5467, 1977

110

158. Navy R. pET System Manual, Novagen Inc. 1992

159. Proath J. and Olin B. Immobilized Metal Ion Affinity Adsorption and Immobilized Metal Ion Affinity Chromatography of Biomaterials: Serum Protein Affinities for Gel-immobilized Iron and Nickel Ions. Biochern. 22:1621-1630.,1983,

160. Smith M. C., Furman T. C., Ingolia T. D. and Pidgon C. Chelating Peptide-immobilized Metal Ion Affinity Chromatography. J. Biol. Chern. 263:7211-7215, 1988

161. Laemmli, E.K. Cleavage of Structural Proteins During the Assembly of the Head of Bacteriophage T4. Nature 227:680 -685, 1970

162. Redinbaugh, M.G. and Turley, R.B. Adaptation of the Bicinchoninic Acid Protein Assay for Use with Microtiter Plates and Sucrose Gradient Fractions. Anul. Biochem 153:267-271, 1986

163. Roberts, S.J., Cheetham, C. and Rees, A.R. Generation of an antibody with enhanced affinity and specificity for its antigen by protein engineering. Nature 328:731-734, 1988

164. Salgaller, M.L., Weber, J.S., Koenig, S., Yannelli, J.R., Rosenberg, S.A. Generation of specific anti-melanoma reactivity by stimulation of human tumor-infiltrating lymphocytes with MAGE-l synthetic peptide. Cancer Immun., Immunother. 39:105-116, 1994

165. Bakker, A.B., Schreurs, M.W., De Boer, A.J., Kawakami, Y., Rosenberg, S.A., Adema, GJ., Figdor, C.G. Melanocyte lineage-specific antigen gp100 is recognized by melanoma-derived tumor-infiltrating lymphocytes. J. Exp. Med. 179:1005-1009, 1994

166. Chen, g., Smith, M., Nguyen, T., Maher, D.W., Hersey, P. T cell recognition of melanoma antigens in association with HLA-A1 on allogeneic melanoma cells. Cancer Immun., Immunother. 38:385-393, 1994

111

167. Baars, J.W., Fonk, J.C., Scheper, R.J., von Blomberg-van der Flier, B.M., Bril, H., von Valk, P., Pinedo, H.M., Wagstaff, J. Treatment with tumor infiltrating lymphocytes and interleukin-2 in patients with metastatic melanoma: a pilot study. Biother. 4:289-297, 1992

168. Schwartzentruber, D.J., Hom, S.S., Dadmarz, R., White, D.E., Yannelli, J.R., Steinberg, S.M., Roserberg, S.A., Topalian, S.L. In vitro predictors of therapeutic response in melanoma patients receiving tumor­infiltrating lymphocytes and interleukin-2. J. CUn. Oneo. 12:1475-1483, 1994

169. Henderson, E., Miller, G., Robinson, J., Heston, L. Efficiency of transformation oflymphocytes by Epstein-Barr virus. Viral. 76: 152-157, 1977

170. Arnan, P., Ehlin-Henriksson, B., Klein, G. Epstein-Barr virus susceptibility of normal human B lymphocyte populations. J Exp. Med. 159:208-212, 1984

171. Kelly, P. J., Pascual, V., Capra, J. D., Lipsky, P. E. Anti-CD3-stimulated T cells induce the production of multiple Ig H chain isotypes by individual human peripheral B lymphocytes. J. Immunol. 148: 1294-1305, 1992

172. Lane, H. C. and Fauci, A. S. Immunologic abnormalities in the acquired immunodeficiency syndrome. Ann. Rev. Immunol. 3:477-500, 1985

173. Marks, J. D., Tristrem, M., Karpas, A., Winter, G. Oligonucleotide primers for polymerase chain reaction amplication of human immunoglobulin variable genes and design of family-specific oligonucleotide probes. Eur. J. Immunol. 21 :985-991, 1991

174. Townsend, A. and bodmer, H. Antigen recognition by class 1-restricted T lymphocytes. Annu. Rev. Immunol. 7:601-624, 1989,

175. Bjorkman, P. J. and Parham, P. Structure, function, and diversity of class I major histocompatibility complex molecules. Annu. Rev. Biochem 59:253-288, 1990

112

176. Hersey, P. Cellular therapy. Curro Oppi. in Oneo. 5:1049-1054,1993

177. Restifo, N. P., Esquivel, F., Kawakami, Y., Yewdell, J. W., Mule, J. J., et aI. Identification of human cancers deficient in antigen processing. J. Exp. Med. 177:265-272, 1993