Human antibody fragments (Fab 14.6.19 and Fab 14.6.20) including
polynucleotides and amino acids (SEQ ID NO: 3, SEQ ID NO: 4, SEQ
ID NO: 5 and SEQ ID NO: 6) that identify them. Both Fabs are fully
human, are affinity matured in vivo, are highly specific for breast
cancer, and target an antigen that is immumogenic in vivo. Thus,
each Fab may be a useful clinical reagent for diagnosis or therapy
of breast cancer and may also lead to the discovery of a novel immunogenic
and tumor specific breast cancer antigen.
What is claimed is:
1. An isolated polynucleotide encoding a breast cancer-specific
antibody fragment including SEQ ID NO:1.
2. A hybridization probe comprising the polynucleotide of claim
1 and a detectable label.
3. A polynucleotide fragment that is fully complementary to a corresponding
segment of the polynucleotide of claim 1.
4. An isolated polynucleotide encoding a breast cancer-specific
antibody fragment including SEQ ID NO:2.
5. A hybridization probe comprising the polynucleotide of claim
4 and a detectable label.
6. A polynucleotide fragment that is fully complementary to a corresponding
segment of the polynucleotide of claim 4.
7. An isolated antibody or antibody fragment that binds to breast
cancer cells and contains the amino acid sequence of SEQ ID NO:3.
8. An isolated antibody or antibody fragment that binds to breast
cancer cells and contains the amino acid sequence of SEQ ID NO:4.
9. An isolated antibody or antibody fragment that binds to breast
cancer cells and contains the amino acid sequence of SEQ ID NO:
10. An isolated antibody or antibody fragment that binds to breast
cancer cells and contains the amino acid sequence of SEQ ID NO:
11. A method for screening breast cancer cells, comprising the
step of contacting said breast cancer cells with one or more antibodies
that contain one or more sequences selected from the group consisting
of SEQ ID NO: 3, SEQ ID NO: 4, SEQ ID NO: 5, or SEQ ID NO: 6.
CROSS-REFERENCE TO RELATED APPLICATIONS
 This application claims priority to U.S. Provisional Application
No. 60/423,052, filed on Oct. 31, 2002.
BACKGROUND OF THE INVENTION
 1. Field of Invention
 This invention generally relates to the field of cancer
biology and in particular to novel breast-cancer specific antibody
fragments, polypeptides, and polynucleotides encoding those fragments.
 2. Description of the Related Art
 Breast cancer is the most common life-threatening malignancy
diagnosed in women. Approximately 192,000 new cases of breast cancer
were diagnosed in the United States in 2001 and roughly half of
these cases occurred in women over the age of 65. Age is the single
greatest risk factor for developing breast cancer. Despite the common
misconception that breast cancer in the elderly is a benign disease,
breast cancer kills approximately 24,000 women over 65 each year.
The elderly account for nearly 60% of all deaths due to breast cancer.
Because of their comorbidities, socioeconomic status, and limited
resources, many older patients receive inferior therapy and consequently
have decreased survival rates.
 In addition, meta-analysis of patients with early breast
cancer suggests standard chemotherapies are not as effective in
older women as in younger women. In fact, a recent study shows there
was no statistically significant benefit seen to receiving adjuvant
therapy for women over age 70. These results may be skewed by the
under-representation of elderly women who participate in or are
offered participation in clinical trials. Still, the results suggest
that breast cancer in elderly patients may respond differently to
 Novel therapies that produce less toxicity, maintain quality
of life and physical function, and target their breast cancer specifically
are needed. Unfortunately, aromatase inhibitors are associated with
elevated risk for decreased bone mass, while tamoxifen is associated
with increased risk of thromboembolism, both issues of particular
concern in the geriatric population.
 Biologic therapies can be uniquely designed to target individual
tumors with minimal toxicity. Rituxan.RTM., an antibody therapy
directed against a B cell antigen, has been found to increase response
rates and survival in elderly patients with aggressive non-Hodgkin's
lymphoma. Herceptin, an antibody targeting the her2/neu receptor
on breast cancer cells has also been found to be well-tolerated
and effective in elderly breast cancer patients. Unfortunately,
overexpression of the her2/neu receptor is seen in less than 20%
of elderly breast cancer patients. Consequently, this well tolerated
and effective therapy can be offered to only a small subset of elderly
patients. Different antigenic targets are required for elderly patients
as their tumors often express different proteins and receptors than
tumors arising in younger patients. Determining the antigens expressed
in breast cancer of the elderly may improve understanding of the
biology and natural history of the disease in this subset of patients
and could lead to the development of new diagnostic and therapeutic
 Decreasing systemic cytotoxic immunity with age has been
well documented, impairing the potential efficacy of cell based
immunotherapy in the elderly. The proliferative potential of T cells
in response to antigens challenge decreases with age, as does the
ability to generate cytotoxic precursors. However, ADCC remains
unimpaired with aging, suggesting that antibody immunotherapy will
have equal efficacy in elderly patients. Although this area has
received little attention, murine studies indicate that young and
aged animals have equal NK and ADCC activity. As far as is known,
no studies specifically addressing this issue have been published
to date, but anecdotal evidence suggests that trastuzumab is effective
in the minority of elderly patients with Her-2/neu positive tumors.
Further, rituximab has also demonstrated efficacy in older patients,
indicating functional ADCC in the elderly in vivo. Therefore, an
antibody targeting a more universally expressed breast cancer antigen
could be useful.
 Heterogeneity predominates in reports of changes in humoral
immunity with aging. However, there are a number of reports of distinct
changes in humoral immune responses with age. Production of antibody
in response to foreign antigen decreases with age, but there is
no change in overall numbers of lymphocytes. There is also a decrease
in affinity and avidity of response that may be due to reliance
upon natural antibody or memory responses in older patients rather
than adaptive immunity. For example, one research group found decreased
rates of hypermutation in kappa light chains in patients of average
age of 83 as compared to an average age of 21. In aged mice, the
number of B cell precursors decreases, supporting the hypothesis
of a decreased ability to mount new adaptive responses. However,
overall levels of VH hypermutation in peripheral blood are equivalent
in young and elderly adults. In older humans, there is a repertoire
shift to use of VH4 Ig gene family members, while VH3 dominates
in younger patient repertoires. There is also an apparent shift
away from high-affinity IgG1 responses in older patients challenged
with influenza vaccination. There is an increase in the presence
of low affinity autoantibodies. Taken together, these data suggest
that the antibody repertoire may be more limited in older individuals,
but is not clear that this translates into impaired function.
 Isolation of fully human affinity-matured antibodies to
tumor-specific cell surface antigens has proven problematic. Most
breast cancer antigens discovered by serum antibody responses are
intracellular, and not specific to breast tumors, and are thus of
limited utility. Antigens against which a serum antibody response
occurs include p53, c-myc, and c-myb. Cell surface proteins are
often poorly immunogenic for a number of reasons including shedding,
glycosylation and low copy number. Possible exceptions to this are
Her-2/neu and MUC-1, which are overexpressed on breast tumors, and
which are weakly immunogenic in vivo.
 Her-2/neu is overexpressed on 25-50% of breast tumors as
the result of gene amplification, and overexpression is correlated
with poor prognosis. Less than 20% of geriatric breast tumors overexpress
Her-2/neu. In the subset of patients with Her-2/neu overexpression,
20% produce serum antibodies against Her-2/neu. The immunogenicity
of Her-2/neu may result from overexpression in an ectopic tissue,
in conjunction with immunological "danger" signals including
tumor necrosis. Although Her2 is expressed on cardiac tissue, and
so is not completely tumor-specific, anti-Her-2/neu antibody immunotherapy
with trastuzumab has produced excellent response rates with minimal
toxicity both alone and in combination chemotherapy.
 MUC-1 is immunogenic in an underglycosylated form occurring
on tumors. There is evidence that a naturally occurring humoral
response to MUC-1 can be protective in breast cancer. The presence
of serum MUC1/antibody complexes correlates with recurrence-free
survival. However, clinical trials targeting MUC1 in breast cancer
have thus far produced modest results. The Theratope.RTM. vaccine
(Biomira, Inc.) consists of the mucin Sialyl-Tn conjugated to Keyhole
Limpet Hemocyanin. In phase II trails of Theratope, patients with
high antibodies titers against Theratope showed a trend towards
 Anti-actin antibodies have been cloned from breast tumor
infiltrating cells in the histologic subtype medullary carcinoma.
While actin is not a cell surface protein, proteolyzed actin peptides
are displayed on the surface of apoptotic typical medullary carcinoma
cells in vivo. However, it is not clear what the significance of
this might be to other histologic types of breast cancer, as typical
medullary carcinoma has a number of anomalous immunological features
in comparison to non-medullary breast cancers including high levels
of apoptosis, elevated HLADR expression by tumor cells, the presence
of a diagnostic plasmacytic infiltrate, and elevated expression
of MIB1, ICAM1 and LFA1. Recent data do not indicate production
of anti-actin antibodies in infiltrating ductal carcinoma, the most
common histologic type of breast cancer.
 Prior to the recent discovery of breast tumor-specific TIL-B
and intratumoral germinal centers, little was known about breast
tumor infiltrating B cells. However, one study found that anti-tumor
antibodies were produced by tumor-infiltrating B cells in approximately
70% of non-breast tumors examined. Tumors from which tumor-specific
TIL-B have been cloned include melanoma, colon carcinoma, ovarian
carcinoma, lung carcinoma, glioma, sarcoma, neuroblastoma, and Hodgkin's
 Many breast adenocarcinomas contain lymphocytic infiltrates
to varying degrees, with heavy infiltrates occurring in .about.20%,
and moderate infiltrates in .about.50%. The composition of breast
tumor-infiltrating lymphocytes (TIL) varies between patients, and
is generally heterogeneous, containing both CD4 and CD8 T cells,
with fewer numbers of B cells, macrophages, and NK cells. Approximately
24% of breast adenocarcinomas contain B cells, with B cells comprising
up to 40% of the TIL. When present, B cells occur exclusively in
follicle-like aggregates, which is consistent with in situ antigen-driven
expansion. It was recently demonstrated that these aggregates are
in fact functional ectopic germinal centers.
 CD1a-+ dendritic cells are also a component of infiltrates,
and were found in the tumor bed of all breast cancer samples examined
in one study. Intratumoral dendritic cells are closely associated
with tumor cells, consistent with a possible role in antigen presentation.
In some cases, T cells are clustered around mature dendritic cells
in peritumoral areas, which is characteristic of an ongoing immune
 Geriatric breast cancer is notable for indolent growth,
expression of the estrogen receptor, normal p53, lower proliferative
rates, but no expression of Her-2/neu or EGFR. Age above 60 rather
than menopausal status distinguishes this group from others. Expression
profiling and immunohistochemical studies suggest that these Her-2/neu
negative, ER positive tumors frequently observed in older women
may represent a distinct subgroup with different antigen expression,
clinical progression, and possibly cell of origin. Human breast
ductal tissue is composed of luminal epithelium and myoepithelium,
which have distinct immunohistochemical and gene expression profiles.
Based on gene expression profiling, two clades of breast tumors
have been identified, one of which most closely resembles luminal
breast epithelium. The luminal epithelium-like tumors are largely
ER positive and Her-2/neu negative.
 In contrast, the more aggressive breast cancers observed
in younger women represent a subset of breast cancers with distinct
clinical and prognostic features including poor differentiation,
high proliferative index, low estrogen receptor expression, frequent
gene alterations, expression of Her-2/neu, altered p53 and host
immune response. Microarray expression studies confirms this aggressive
subset, and further indicates a distinct pattern of expression consistent
with breast basal endothelial cells in contrast to the more indolent
ER-positive breast cancers of older women which more closely resemble
breast luminal epithelium.
 While lymph node or spleen germinal center reactions are
the paradigm for B cell expansion, B cells can proliferate and in
some cases undergo affinity maturation in ectopic germinal centers,
although this has previously been described only in autoimmunity.
Extranodal B cell proliferation has been characterized in rheumatoid
arthritis, multiple sclerosis, Sjogren's syndrome, and Grave's disease.
In these cases, extranodal expansion of B cells reactive with an
autoantigen has led to a pathogenic disease state. In the case of
rheumatoid arthritis, "rheumatoid factor" B cells form
germinal centers in the synovium, undergo somatic mutation, affinity
maturation by selection, receptor revision, clonal expansion and
differentiation to plasma cells. This suggests that mechanisms to
delete autoreactive B cells may be deficient in ectopic germinal
centers. Moreover, it was recently discovered that an analogous
situation exists in breast cancer, where ectopic germinal centers
reside in the tumor bed and produce tumor-specific antibodies.
 The relevance to this study of an in situ tumor antigen-driven
B cell proliferation in ectopic germinal centers is not its role
in tumor progression but isolation of tumor-reactive and possibly
high-affinity immunoglobulins. Like those observed in autoimmune
disease, ectopic germinal centers in breast cancer may be capable
of producing antibodies against cell surface autoantigens to a degree
not allowed in lymph node follicles. The repertoire of TILB is also
very limited and enriched for tumor reactive antibodies, facilitating
the isolation of these antibodies without the need for negative
selection. Antibodies cloned from tumor infiltrating B cells that
are specific to breast cancer cell surface antigens supports the
hypothesis of immune permissiveness, and demonstrates the feasibility
of cloning antibodies specific to breast tumor cell surface antigens
from tumor infiltrating B cells. Cell surface antigens are normally
poorly immunogenic in vivo, as demonstrated by the overwhelming
bias towards cytosolic antigens in breast cancer serum antibody
responses. As a result, fully human antibodies against tumor specific
surface antigens are rare. Thus, intratumoral germinal centers provide
a unique immunological loophole for the production of antibodies
against cell surface autoantigens. Further, the produced antibodies
are affinity matured, as judged by somatic mutation patterns. Non-quantitative
assessment of cloned Fabs in flow cytometry with breast tumor cells
is also consistent with high affinity.
 Affinity maturation in ectopic germinal centers associated
with autoimmunity has been dissected at the molecular level in order
to understand the etiology of autoimmune disease. Because ectopic
germinal centers in solid tumors had not been described prior to
a recent preliminary study, much of the relevant background in this
area derives from studies of rheumatoid arthritis and other autoimmune
diseases. In the absence of any compelling evidence to suggest a
viral or other xenobiotic etiology for breast cancer, it is useful
to view native anti-tumor immunity as an autoimmune response. Because
breast tumors express few if any truly foreign antigens, any immune
response must overcome the same tolerance hurdles as in a pathogenic
autoimmune state. Germinal centers in the synovium of rheumatoid
arthritis (RA) are formed via pathways similar to those for secondary
lymphoid follicles. Lymphotoxin (LT)-(.alpha.1.beta.2 production
by both germinal center B and T cells and production of CXCL13 by
local endothelial cells and fibroblasts contribute to germinal center
formation. Like the germinal centers that have been described in
breast tumors, synovial germinal centers in RA do not have clearly
delineated light and dark zones. Instead, proliferating B cells
occur in the follicular dendritic cell zone. Despite this, antibodies
produced by germinal centers in both cases are affinity matured.
It is thought that this lack of structural organization might be
due to insufficient production of CXCL13 by follicular dendritic
cells or its receptor on B cells, CXCR5. An unusual population of
CD8+, CD40L+, IFN+, perforin-T cells are also absolutely required
for germinal center formation in RA. These cells are not observed
in tonsil germinal centers.
 While it has been observed that CD8+ T cells in the germinal
centers in breast tumors, no information regarding expression of
these markers has been forthcoming. As in RA, the humoral response
in breast tumors must overcome the normal deletion of autoimmunity.
It is possible that the presence of antiapoptotic signals such as
constitutive CD40L on local T cells may prevent apoptosis of autoreactive
B cells, perhaps through induction of cFLIP. There is also an active
component of FDC-mediated B cell apoptotic signaling via FasL that
may be lacking in ectopic germinal centers, as one does not observe
Fas (CD95) on TILB cells. In addition, unlike the germinal centers
in RA, breast tumor germinal centers do not contain CD38hi plasma
cells or CD27+ memory B cells. Blimp-1, IL-6 and XBP-1 contribute
to plasma cell differentiation, while IL-2, IL-10 and CD40L contribute
to entry into the plasma cell compartment.
 One aim of the present invention is to explore the expression
of these factors further in order to delineate differences in tumor
immunobiology between young and elderly patients and to understand
the development of intratumoral germinal center reactions.
 The success of antibody therapy directed against Her-2/neu
and the failure of other breast cancer antibody therapies directed
against a variety of non-immunogenic antigens suggests that naturally
immunogenic tumor antigens may be superior targets. Imunogenicity
entails visibility and availability to the immune system, and is
associated with varying degrees of tumor specificity. Further, immunogenic
antigens can also be targeted by other means, and their identification
gives important information about tumor immunobiology. While SEREX
has been useful in identifying immunogenic antigens, it is time-consuming,
can produce false positives, is biased towards soluble antigens,
does not identify nonprotein antigens, and does not provide cognate
antibodies, which are then often produced via murine technology.
Because murine antibodies elicit HAMA responses that prevent multiple
dosing and do not have Fc-mediated effector functions in humans,
techniques have been developed to humanize antibodies. However,
humanization is time-consuming, expensive, technically difficult,
and can result in antibodies with altered affinities. Others use
antigens discovered by SEREX or microarray to generate human antibodies
in human immunoglobulin-transgenic mice, avoiding the need for humanization.
Given the money and time demands associated with such methods, it
is believed that a simpler, more direct approach offers great promise
with considerable economy and efficiency, for example, the phage
display methods of C. Barbas (Scripps Research Institute).
 Use of phage display antibody libraries allows the rapid
and inexpensive isolation of human antibodies reactive with antigen
in a variety of formats, quantities, and can selectively isolate
high-affinity immunoglobulin. Phage display allows the rapid isolation
of high affinity Fabs reactive with cell surface antigens.
 In B cell-mediated autoimmune diseases, in situ immunity
is sometimes necessary and even sufficient to produce devastating
degradation of the involved tissues. Research regarding breast cancer
suggest that in situ immunity may be capable of producing responses
that might be suppressed in a lymph node as part of the normal suppression
of autoimmunity. Indeed, it was recently discovered the presence
of ectopic intratumoral germinal centers in breast tumors. This
had not been previously described in any tumor. These germinal centers
are functional and produce phenotypically matured B cells and affinity
matured, class-switched antibodies that are specific to breast cancer-specific
cell surface antigens. In intratumoral germinal centers, naturally
occurring immunity accomplishes what has been found to be exceedingly
difficult in the laboratory: production of high affinity antibodies
against tumor-specific cell surface antigens. In combination with
phage display technology, this has been found to be a powerful tool
to study the in situ immunobiology of breast tumors and to develop
SUMMARY OF THE INVENTION
 There are few if any therapies or diagnostic tools specifically
developed for, or specific studies of, tumor immunobiology in the
elderly. Therefore, discovery of antigen/antibody pairs appropriate
for use with geriatric breast cancer patients will improve understanding
of the disease and contribute new diagnostic and therapeutic agents.
In previous research, it was discovered that lymphocyte derived
intratumoral germinal centers in geriatric breast tumors produce
affinity-matured antibodies against breast tumor-specific cell surface
antigens. Three phage displayed Fab libraries were cloned from breast
tumor infiltrating B cells from geriatric breast cancer patients.
Initial study of one library yielded three Fabs that bind highly
tumor-specific cell-surface antigens, two of which have been sequenced
at the polynucleotide and amino acid level.
 Accordingly, in one aspect of the invention, polynucleotides
encoding antibody 16.4.19 (SEQ ID NO: 1) is disclosed. In another
aspect of the invention, polynucleotides encoding antibody 16.4.20
(SEQ ID NO: 2) is disclosed. Moreover, amino acid sequences for
16.4.19 (SEQ ID NO: 3 and SEQ ID NO: 4) and 16.4.20 (SEQ ID NO:
5 and SEQ ID NO: 6) are disclosed. Both of these antibody fragments
represent significant progress over existing antibodies because
(1) they are fully human and require no molecular modification beyond
attachment to the IgGI Fc region for clinical use; (2) each Fab
is affinity matured in vivo and may have high affinity for the breast
cancer antigen it binds; (3) each Fab is highly specific for breast
cancer; and (4) each Fab targets an antigen that is immumogenic
in vivo. Thus, each Fab may be a useful clinical reagent for diagnosis
or therapy of breast cancer and may also lead to the discovery of
a novel immunogenic and tumor specific breast cancer antigen.
 Various other purposes and advantages of the invention will
become clear from its description in the specification that follows
and from the novel features particularly pointed out in the appended
claims. Therefore, to the accomplishment of the objectives described
above, this invention consists of the features hereinafter illustrated
in the drawings, fully described in the detailed description of
the preferred embodiments and particularly pointed out in the claims.
However, such drawings and description disclose only some of the
various ways in which the invention may be practiced.
BRIEF DESCRIPTION OF THE DRAWINGS
 FIG. 1 shows the enrichment of TIL-B phage displayed Fab
libraries for binding of breast cancer cell surface antigens. Libraries
were enriched by six sequential rounds of panning and regrowth of
cell binding phage. The library derived from patient 14 was panned
on MCF7 cells. The patient 16 library was panned on the breast cancer
cell lines SKBR3, MCF7, and 2087.
 FIG. 2 shows the flow cytometry analysis of phage Fab libraries
and individual Fab clones. Libraries 14.6 and 16.6 are shown after
enrichment for breast cancer cell binding; individual Fab clones
14.6.11, 14.6.19 (SEQ ID NO: 1; SEQ ID NO: 3 and SEQ ID NO: 4),
14.6.20 (SEQ ID NO: 2; SEQ ID NO: 5 and SEQ ID NO: 6), and 16.4.19)
are also shown. Patient 14 library was panned on MCF7 cells; Patient
16 library was panned on MCF7, SKBR3, and 2087 breast cancer cells.
Only data from SKBR3 panning is shown.
 FIG. 3 shows TILB Fabs are not reactive with actin or p53.
Phage displayed Fabs 14.6.11, 14.6.19 (SEQ ID NO: 1; SEQ ID NO:
3 and SEQ ID NO: 4) and 14.6.20 (SEQ ID NO: 2; SEQ ID NO: 5 and
SEQ ID NO: 6) were tested for reactivity with actin and p53 by ELISA.
Tetnus toxoid antigen and anti-tetnus toxoid phage displayed Fab
were included as controls.
 FIG. 4 shows expression of soluble Fabs. Fabs 14.6.11, 14.6.19
(SEQ ID NO: 1; SEQ ID NO: 3 and SEQ ID NO: 4), and 14.6.20 (SEQ
ID NO: 2; SEQ ID NO: 5 and SEQ ID NO: 6) were expressed as soluble
Fabs and purified by His-Tag affinity. Lanes 1, 4, 7, column flow-through;
lanes 2, 5, 8 column wash; lanes 3, 6, 9 column elute.
 FIG. 5 shows TILB clonality. Clonality was determined by
RTPCR and sequencing of IgG heavy chains. Percentages indicate %
of total IgG heavy chain sequences for which a clonal relative was
 FIG. 6 shows the pattern of clonal expansion of TILBs in
infiltrating ductal breast carcinoma. IGHV# indicates germline heavy
chain variable gene used by progenitor B cell. Empty circles indicate
deduced intermediates; numbers inside circles indicate sequenced
clones; numbers next to arrows indicate numbers of mutations in
comparison to gernline at a given branching.
 FIG. 7 shows a summary of TILB marker expression. Upper
table indicates FACS results (underneath table), lower left table
indicates results of TILB immunohistochemistry. Numbers in upper
table indicate percentage of cells that are positive for the first
antigen which are also positive for the second antigen. For example,
95% of CD19+ TIL are also IgG+. These preliminary numbers are based
on averages, and have not yet been subjected to statistical analyses.
 FIG. 8 shows somatic mutation of IgG heavy chains from TIL-B.
Patients=p1, p2, p3; tumor draining lymph node, p3node; healthy
donor PBMC, KPBMC.
DETAILED DESCRIPTION OF THE INVENTION
 Unless defined otherwise, all technical and scientific terms
used herein have the same meanings as commonly understood by those
of ordinary skill in art of the invention. For example, see the
definitions provided by U.S. Pat. No 5,955,312 by Hillman and Goli,
which is incorporated herein by reference. All publications mentioned
herein are incorporated by reference for the purpose of describing
and disclosing the cell lines, vectors, and methodologies which
might be used in connection with the invention.
 Although many different methods and materials similar or
equivalent to those described herein can be used in the practice
or testing of the present invention, the preferred methods, devices,
and material are now described.
 It will be appreciated by those skilled in the art that,
as a result of the degeneracy of the genetic code, a multitude of
Fab-encoding nucleotide sequences, some bearing minimal homology
to the nucleotide sequences of any known and naturally occurring
gene, may be produced. The invention contemplates every possible
variation of nucleotide sequence that could be made by selecting
combinations based on possible codon choices. These combinations
are made in accordance with the standard triplet genetic code as
applied to the nucleotide sequence encoding naturally occurring
Fabs, and all such variations are to be considered as being specifically
 Although nucleotide sequences which encode Fabs and their
variants are preferably capable of hybridizing to the nucleotide
sequence of the naturally occurring transcription sequences under
appropriately selected conditions of stringency, it can be advantageous
to produce nucleotide sequences encoding Fabs or their derivatives
possessing a substantially different codon usage. For example, codons
may be selected to increase the rate at which expression of the
peptide occurs in a particular prokaryotic or eukaryotic expression
host in accordance with the frequency with which particular codons
are utilized by the host. For example, the TAG sequence in Fab 14.6.20
(SEQ ID NO: 2) starting at position 685 encodes a stop codon in
mammalian expression systems. Thus, changing that codon to CAG (glutamine)
allows full expression to occur withoug affecting function. Other
reasons for substantially altering the nucleotide sequence encoding
Fabs and their derivatives without altering the encoded amino acid
sequences include the production of RNA transcripts having more
desirable properties, such as a greater stability or half-life,
than transcripts produced from the naturally occurring sequence.
Moreover, fragments of the disclosed Fabs may possess moieties that
provide breast cancer cell specific binding to take place.
 As known by one skilled in the art, a DNA sequence, or portions
thereof, encoding Fabs and their derivatives may be produced entirely
by synthetic chemistry. Subsequently, the synthetic nucleotide sequence
may be inserted into any of the many available DNA vectors and cell
systems using reagents that are commonly available. Moreover, synthetic
chemistry may be used to introduce mutations into a sequence encoding
Fabs or any portion thereof.
 Also included within the scope of the invention are polynucleotide
sequences that are capable of hybridizing to the nucleotide sequences
of SEQ ID NO:1 or SEQ ID NO:2 under various conditions of stringency.
Hybridization conditions are based on the melting temperature (Tm)
of the nucleic acid binding complex or probe, as taught in Berger
and Kimmel (1987, Guide to Molecular Cloning Techniques, Methods
in Enzymology, v.152, Academic Press, San Diego, Calif.).
 Methods well known in the art can be used to construct expression
vectors containing sequences encoding a Fab and appropriate transcriptional
and translational control elements. Methods may include in vitro
recombinant DNA techniques, synthetic techniques, and in vivo genetic
recombination in a variety of expression vector/host systems, such
as bacteria transformed with recombinant bacteriophage or plasmids
or insect cell systems infected with viral expression vectors such
as the baculovirus. These methods are described in standard laboratory
references, such as Sambrook, J. et al. Molecular Cloning, A Laboratory
Manual, Cold Spring Harbor Press, Plainview, N.Y. (1989).
 Altered nucleic acids encoding Fabs which may be used in
accordance with the invention include deletions, insertions or substitutions
of different nucleotides resulting in a polynucleotide and polypeptide
that encodes the same or a functionally equivalent Fab. The protein
may also show deletions, insertions or substitutions of amino acid
residues which produce a silent change and result in functionally
equivalent Fab. Deliberate amino acid substitutions may be made
on the basis of similarity in polarity, charge, solubility, hydrophobicity,
hydrophilicity, and/or the amphipathic nature of the residues as
long as the biological activity of Fab is retained. For example,
negatively charged amino acids aspartic acid and glutamic acid might
be substituted for one another.
 Also included within the scope of the invention are alleles
encoding Fab. As used herein, an "allele" or "allelic
sequence" is an alternative form of the nucleic acid sequence
encoding Fab. Alleles result from a mutation, i.e. a change in the
nucleic acid sequence, and generally produce altered mRNAs or polypeptides
whose structure or function may or may not be altered. Any given
gene may have none, one or many allelic forms. Common mutational
changes which give rise to natural deletions, additions or substitutions
of amino acids. Each of these types of changes may occur alone,
or in combination with the others, one or more times in a given
sequence. For example, the TAG sequence in Fab 14.6.20 (SEQ ID NO:
2) starting at position 685 may be changed to TCG (serine) to reflect
an allele found in parent B cells.
 Many ways exist in the art by which Fab may be used therapeutically.
Examples include, but are not limited to, administering Fab through
the introduction of an expression vector into a subject for in vivo
therapy or administering Fab as part of a pharmaceutical composition.
Depending on the route of administration, appropriate agents for
use in combination with Fab for therapy may include any conventional
pharmaceutical carrier such as saline or buffered saline (intravenous
dosing) and dextrose or water (oral dosing). Further details on
techniques for formulation and administration may be found in the
latest edition of Remington's Pharmaceutical Sciences (Maack Publishing
Co., Easton, Pa.).
 Breast tumor-infiltrating B cells (TIL-B) in tumors derived
from geriatric patients have been investigated with the goal of
developing new diagnostic or therapeutic antibodies. In discovering
the present invention, the histology, phenotype, IgG repertoire
and immunoglobulin specificity of tumor infiltrating B cells (TIL-B)
from infiltrating ductal carcinomas of the breast were examined.
Patient data and work done are summarized below in table I:
1TABLE 1 Summary of patient data and work done in preliminary study.
Patients were designated numerically in order of collection; Phage
display indicates Fab library cloned, GC IHC indicates germinal
center immunohistochemistry performed; LN IHC indicated lymph node
immunohistochemistry performed; TIL FACS indicates flow cytometry
analysis of TIL performed; grade indicates tumor grade; ER/PR indicates
% cells positive for estrogen or progesterone receptor; over 2+
indicates HER2+; MIB1 proliferation index; IDC , infiltrating ductal
carcinoma; ILC , infiltrating lobular carcinoma. patient phage display
GC IHC LN IHC TIL FACS Age Grade ER/PR Her2 MIB-1 Diagnosis 3 x
5 x 61 1 95/<2 N 2% IDC 8 x x 83 2 99/65 1+ 9% IDC 10 x 12 x
39 3 N/N N 75% IDC 13 x 77 2 N/90% N <5% mixed 14 x x 78 R2 50/N
RN <2% IDC 16 x x 73 3 N/N 2+ 60% IDC 25 x x 61 2 100/86 1+ 31%
IDC/ILC 26 x x 48 3 N/N N >95% IDC 30 x 54 3 N/N N 64% IDC 32
x 70 2, 3 N/N 2+ 26% IDC 33 x 79 2 95/90 N 6% IDC 35 x 58 3 N/N
3+ 100% 12% IDC 36 x 58 3 >95/N 3+ 39% IDC 39 x x 65 3 N/<2%
3+ 52% IDC 40 x x 45 3 100/N 1+ 30% IDC
 To determine if TIL-B immunoglobulins were reactive with
tumor, 2 phage-displayed immunoglobulin Fab libraries were generated
from geriatric patient TILs by the methods of Barbas et al (Barbas,
C. F. 2000. Phage display: a laboratory manual. Cold Spring Harbor
Labortory Press, Cold Spring Harbor, N.Y.). Fab (heavy chain variable
region plus CH1 and light chain) libraries of .about.1.times.10.sup.7
Fab clones were cloned in the pCOMBX phage display vector (gift
of C. Barbas, Scripps Research Institute, San Diego, Calif.). The
inventor Julia Coronella received training in phage display from
Dr. Carlos Barbas of Scripps Research Institute at the Cold Spring
Harbor course on phage display 2000. The libraries were panned on
cultured breast cancer cells in order to enrich for Fabs that bind
breast cancer cell surface antigens. Panning consisted of 6 sequential
rounds of incubation of phage displayed Fabs with breast tumor cells
under increasingly stringent conditions to isolate high affinity
Fabs, and regrowth of the binding fraction. Enrichment of patient
14 Fab library for tumor cell binding was observed after 4 sequential
rounds of panning on MCF7 cells (FIG. 1). Enrichment of patient
16 Fab library for tumor cell binding was observed after only 1
panning on SKBR3, MCF7 and 2087 breast cancer cells, indicating
a high fraction of tumor binding Fabs in the library (FIG. 1).
 Individual Fab clones were selected from the enriched libraries
and assessed for tumor cell surface binding by flow cytometry. Fabs
were selected from all panning passages showing enrichment. Clones
with identical sequences were not multiply analyzed. Clones with
low or absent tumor cell binding were discarded. Because of the
inherent variability in phage Fab binding assays (phage Fab are
unstable and must be grown and prepared prior to each new experiment),
all flow cytometry and ELISA assays were replicated multiple times.
Flow cytometry data are summarized in table 2 below:
2TABLE 2 Flow cytometry analysis of phage display Fab libraries
and individual Fabs. Patient 14 library data is shown for MCF7 panning;
patient 16 library data is shown for SKBR3 panning. MCF7 SKBR3 3133
3199 2087 HMEC A549 HL60 FF B16NEO B16MUC1 C57MUC1 antibody 14.6
hi med lo hi med 0 med 0 0 lo lo 0 14.6.11 hi hi lo hi med 0 0 0
0 0 0 0 14.6.19 hi 0 lo lo 0 0 0 0 0 0 0 0 14.6.20 hi hi lo med
med 0 0 0 0 0 0 0 16.60 hi hi hi 0 16.4.19 hi hi 0 0 antigen HER2
0 hi med 0 med 0 MUC1 lo 0 med 0 lo 0 0 hi hi CEA 0 0 lo 0 0 med
EpCAM 0 hi hi 0
 Thus far, 3 Fabs with apparent specificity for breast tumor
cells have been isolated from the patient 14 Fab library (FIG. 2).
Fab 14.6.11 binds all breast cancer cell lines tested but not nonmalignant
healthy breast epithelium, primary fibroblasts or the leukemia cell
line HL60. Fabs 14.6.19 (SEQ ID NO: 1; SEQ ID NO: 3 and SEQ ID NO:
4) and 14.6.20 (SEQ ID NO: 2, SEQ ID NO: 5 and SEQ ID NO: 6) were
also highly specific for breast cancer, but less universal binding
of breast cancer cell lines was observed. A number of promising
Fabs were isolated from the patient 16 library and are currently
under analysis. Monovalent phage Fabs have low avidity, and the
FACS protocol utilized is a lengthy multi-step staining. Therefore,
Fab clones that exhibit binding are likely of high affinity, with
low off-rates. The 3 Fabs contain somatic mutations in the antigen-binding
CDR regions, suggesting affinity maturation.
 As the first stage of antigen identification, Her-2/neu,
MUC1, CEA, EpCAM, B-actin and p53 were eliminated as possible antigens.
While it is not possible to test all known breast cancer antigens
prior to proceeding to antigen identification, it is desirable to
eliminate the most obvious. Binding of the Fab clones 14.6.11, 14.6.19
(SEQ ID NO: 1; SEQ ID NO: 3 and SEQ ID NO: 4) and 14.6.20 (SEQ ID
NO: 2; SEQ ID NO: 5 and SEQ ID NO: 6) to cell lines was compared
to expression levels of antigens on these cell lines, summarized
in table 2 (above).
 These data suggests that Her-2/neu, MUC1, CEA and EpCAM
are not the antigens with which any of these Fabs react. The lack
of MUC1 binding was confirmed by FACS with MUC1-transfected cell
lines (gift of S. Gendler, Mayo Clinic, Scottsdale, Ariz.). Fabs
were also assayed for binding of B-actin and p53 by ELISA. No binding
was observed (FIG. 3). Fab clones 14.6.11, 14.6.19 (SEQ ID NO: 1;
SEQ ID NO: 3 and SEQ ID NO: 4) and 14.6.20 (SEQ ID NO: 2; SEQ ID
NO: 5 and SEQ ID NO: 6) utilize unique germline immunoglobulin genes
(table 3, below), and have non-identical patterns of cell line binding,
consistent with binding of unique antigens.
3TABLE 3 Patient 14 Fab germline genes. Heavy Chain Light Chain
Fab V D J V J 14.6.11 IGHV4-59*01 IGHD3-10*01 IGHJ5*02 IGKV1-5*03
IGKJ1*01 14.6.19 IGHV3-74*01 IGHD3-16*01 IGHJ4*02 IGKV2D-28*01 IGKJ5*01
14.6.20 IGHV3-23*01 IGHD3-3*01 IGHJ4*02 IGKV1-5*03 IGKJ4*01
 One strategy to identify the breast cancer cell surface
antigens that are reactive with the cloned Fabs is immunoprecipitation
followed by mass spectrometry. To this end, Fab clones 14.6.11,
14.6.19 and 14.6.20 were transformed into the non-supressor cell
line Top10'F (Invitrogen) to allow expression of soluble Fabs. The
pCombX phage display vector contains an amber codon between the
Fab and gene III capsid protein to allow expression of Fabs without
gene m fusion and display on the capsid. Fab clones 14.6.11, 14.6.19
(SEQ ID NO: 3 and SEQ ID NO: 4) and 14.6.20 (SEQ ID NO: 5 and SEQ
ID NO: 6) were expressed and purified by His-tag affinity (FIG.
4). Following purification, Fabs were linked to protein A-agarose.
Cell membrane proteins were solubilized from the breast cancer cell
lines MCF7 and 3199 for immunoprecipitation with the Fabs. Immunoprecipitations
are in progress, and will be analyzed by SDS-PAGE, followed by mass
spectrometry (a service of the AZCC core proteomics facility).
 Unlike T cells, the anti-tumor activity of B cells normally
occurs in a lymph node (activation/affinity maturation/differentiation/antibody
production) or bone marrow (antibody production) rather than at
the site of antigen in the periphery (with the exception of IgA-producing
plasma cells at the mucosa, which produce low-affinity antimicrobial
"natural antibodies"). With the exception of autoimmune
diseases, no precedent exists for local B cell-mediated immune reactions
in peripheral tissues. TIL-B were studied by immunohistochemistry,
flow cytometry, and immunoglobulin sequencing in order to understand
the interaction between tumor and B lymphocytes in the tumor microenvironment.
 To determine if the aggregates of B cells observed in tumors
were the result of random recruitment from the periphery or the
proliferation of tumor-infiltrating lymphocytes, IgG heavy chain
libraries were generated by RTPCR and random clones sequenced from
a total of six breast tumors, a tumor-draining lymph node, and the
peripheral blood of a healthy donor. Between 12 and 58 IgG heavy
chains were sequenced for each sample.
 Peripheral blood lymphocytes were included in order to control
for poor PCR methodology that could result in repetitive cloning
of single PCR products rather than as a measure of the peripheral
repertoire, which has been extensively characterized elsewhere.
The inventor's peripheral repertoire results are equivalent to those
published previously. B cell clones occur very infrequently (<1/20,000)
in the peripheral blood of both young and elderly humans, and are
not detectable by sampling and sequencing methods such as those
employed in this study. Oligoclonal expansion of TIL-B was established
by the presence of intratumoral clonal groups derived from common
progenitor B cells in all tumors examined (FIG. 5). Between 18 and
68% of IgG heavy chain sequences from TIL-B belonged to clonal groups,
while only 7% of tumor-draining lymph node sequences and 0% of peripheral
blood sequences were clonal, consistent with the expected large
repertoires of those populations. As occurs in lymph node germinal
centers, TIL-B lineages accumulated unique somatic mutations during
proliferation, allowing the derivation of genealogical trees and
calculation of cell doubling numbers (FIG. 6). For example, clone
36 in FIG. 6 contains 14 unique mutations in the 296 base pair VH
region in comparison with clone 30, a linear predecessor in this
TIL-B lineage. Based on a somatic hypermutation level of 1 base
pair per 10.sup.2 to 10.sup.3 bases per generation for immunoglobulins,
between 4.7 and 47 cell divisions would be required.
 Based on immunohistochemistry, it was determined that approximately
20% of breast tumors contain significant numbers of B cells, in
agreement with previously published percentages. Seven tumors containing
TIL-B were selected for further immunohistochemistry. Although tumor-infiltrating
lymphocytes were found scattered throughout the stroma and interspersed
between tumor cells in all tumors, CD20+ B lymphocytes occurred
exclusively in dense aggregates. In most cases, TILB aggregates
occurred in stromal areas immediately adjoining tumor nests, and
were not observed outside the tumor margins. All TILB aggregates
contained interdigitating CD21+ follicular dendritic cells. CD21+
cells were not observed outside of B cell germinal centers. Germinal
centers were surrounded by CD3+ T lymphocytes, the majority of which
were CD4+, although a component of CD8+ cells were also present.
Plasma cells (CD38) and NK cells (CD56) were rare, and occurred
randomly in relation to other lymphocytes (data not shown). Most
B cells were Ki-67-, indicating that the clonal groups observed
in immunoglobulin sequencing were the result of slow or previous
proliferation. Most germinal center B cells were positive for BCL2
and HLA-DR, but negative for CD10, CD27, and CD38, suggesting an
activated but not memory or plasma cell phenotype.
 In addition, flow cytometry was performed to assess the
presence of IgG, IgM, IgD, CD38, CD5 (associated with autoimmunity),
CD95, and CD40 on CD19+ B cells. Although CD38 was not detected
by immunohistochemistry, flow cytometry is more sensitive and can
detect CD381o cells. Six breast tumors, 5 tumor-draining lymph nodes,
peripheral blood from 5 breast cancer patients, and peripheral blood
from 4 healthy donors were analyzed. Data is summarized in FIG.
7. As occurs in a lymph node follicle, the populations of TILB were
heterogeneous; Most TILB were IgG+, IgM-, and approximately half
were IgD+. All TILB were CD5-. Some B cells expressed low levels
of CD38, as do centroblasts and centrocytes. However, no CD38hi
cells were observed, consistent with the lack of plasma cells determined
by immunohistochemistry. The absence of plasma cells can be explained
by apoptosis or alternatively, by a deficit in the plasma cell differentiation
pathway. Centroblasts and centrocytes are normally CD95+, and very
sensitive to apoptosis. In contrast, most TILB express CD40, but
not CD95 (Fas). Strong BCL2 expression of TILB was determined by
immunohistochemistry, and is unusual in that only plasma cells and
germinal center founder cells normally express this antiapoptotic
protein. BCL2 expression protects germinal center B cells from apoptosis
in vivo. Through expression of the antiapoptotic proteins CD40 and
BCL2, and lack of CD95 expression, TILB may be unusually resistant
to apoptosis. Resistance to apoptosis is consistent with the inventor's
demonstration of autoreactive antibodies from TILB, as autoreactive
B cells are normally deleted during the germinal center reaction.
One can speculate that cytokines from local activated T lymphocytes
or perhaps cytokine secretion from the tumor itself may induce this
 TIL-B IgG heavy chain mutation levels, patterns and germline
gene usage suggest that TIL-B undergo affinity maturation intratumorally,
presenting the possibility of production of high-affinity anti-tumor
immunoglobulin. However, this conclusion stems from indirect evidence
of affinity maturation, which can only be resolved through antigen
affinity studies. TIL-B IgG heavy chains contained somatic mutations
that clustered in the antigen-contacting CDR regions, as previously
observed in affinity-matured antibodies, and as was also seen in
tumor-draining lymph node but not peripheral blood IgG (FIG. 8).
Although the peripheral blood control was from a younger individual
(39 years of age), previous studies demonstrate that overall levels
of VH hypermutation in peripheral blood is equivalent in young and
elderly adults. As calculated by the polynomial algorithm of Lossos
et al., replacement and silent mutations occurred nonrandomly in
36-84% of TIL-derived IgG heavy chains. Further, low levels of TIL-B
IgG heavy chain nonsense mutation and a modest bias in germline
gene usage suggested clonal selection.
 In order to further illustrate the invention, the following
example is provided. While this example is contemplated to be the
preferred mode, it will be understood by those in the art that numerous
alternative methodologies may be successfully practiced in lieu
of the preferred method described herein. Therefore, this example
is not intended in any way to limit the invention.
 Experimental Procedures for Fabs 14.6.19, and 14.6.20. Cloning
history: A Fab library was cloned from breast tumor-infiltrating
B cells by RTPCR, as published in Coronella, J. A. et al., 2002.
Antigen-driven oligoclonal expansion of tumor-infiltrating B cells
in infiltrating ductal carcinoma of the breast. J. Immunology 169:1829.
 The library was subcloned into the pCOMBX phage display
vector (gift of C. Barbas, Scripps Research Institute, La Jolla,
Calif.). Fabs were isolated from the library on the basis of cell-surface
reactivity with MCF7 cells. Two Fabs so isolated were 14.6.19, and
14.6.20, the nucleotide and peptide sequences of which are hereinafter
 The Fabs were subsequently sent to IDEC Pharmaceuticals,
and subcloned into the N5 mKm vector (property of IDEC). A change
was made to the 14.6.20 Fab, mutating the TAG amber codon in the
VH region (in white text above) to CAG, encoding Gln.
 Flow cytometry analysis of Fabs: below are the binding profiles
of the two Fabs with a number of cancer and non-cancer cell lines.
4 MCF7 SKBR3 2087 3133 3199 MDA HMEC Breast cancer Breast cancer
Breast cancer Breast cancer Breast cancer Breast cancer primary
breast epithelium 14.6.19 + + - + + - 14.6.20 + + + - + + - B16NEO
B16MUC1 C57MUC1 FF HeLa PANC-1 murine murine + MUC1 murine + MUC1
primary foreskin fibroblasts cervical ca pancreatic ca 14.6.19 -
- - - - 14.6.20 - - - - + - A549 HL60 U251 SW480 JORP DU-145 OVCAR
lung cancer leukemia glioma colon ca melanoma ca prostate ca ovarian
ca 14.6.19 + - - + + + + 14.6.20 - - - - + + +/-
 No information exists regarding the 14.6.19 antigen. The
14.6.20 antigen is resistant to trypsin and glycopeptidase F, but
partially sensitive to periodate treatment, suggesting a protein
epitope on a cell surface glycoprotein. Based on preliminary Western
blot analysis, the antigen may be .about.129 kDa.
 Although the invention has been described with reference
to various applications, methods, and compositions, it will be appreciated
that various changes and modifications may be made without departing
from the invention. The foregoing examples are provided to better
illustrate the invention and are not intended to limit the scope
of the invention.