Note: Descriptions are shown in the official language in which they were submitted.
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
ELUCIDATION OF GENE FUNCTION
BACKGROUND OF THE INVENTION
Genes are the blueprints of all living organisms and are physically composed
of DNA.
The collection of all genes of an organism is called a genome. When
"expressed," each gene is
translated into a distinct protein, and proteins are the physical building
blocks of all living
organisms. Each cell in an organism is composed of tens of thousands of
proteins, each of
which has a function that, collectively, defines what that cell does and how
it behaves.
Gene expression identifies which genes are used in any given cell type, and
how often
each of those genes is used. When genes are active, that is, "expressed," they
make copies of
themselves, called messenger RNAs (mRNAs), which in turn direct the production
of their
protein products. Gene expression technology identifies and quantifies all of
the mRNAs in a
cell. Different cell types use different subsets of genes. It is the subset of
genes and how often
each gene within the subset is used that defines cell function, that
constitutes its "biological
program."
Each of the 30,000-60,000 genes that each human carries (the human "genome")
encodes for a distinct biological function. These functions are carried out
not by the genes
themselves, but by their protein products (the human "proteome"). A gene
encodes for the
production of a protein, and that protein performs that gene's biological
function. Several
factors make it necessary to study proteins, rather than just genes
themselves, in order to arrive
at a complete understanding of normal biology and disease mechanisms. For
example, drugs
act on proteins, not genes, so an understanding of protein structure and
function is cxucial to
rational dxug design and optimization. In addition, the correlation between
mRNA levels and
the abundance of the encoded protein is very poor. Thus, while genomic data
can provide clues
to functional differences between two biological states, the measurement of
differences at the
pxotein level reveals true discoveries.
Proteomics is a term that describes the study of proteins-their expression,
interactions,
and structure/function relationships-within the context of the framework
provided by
genornics. Whereas genomics is devoted to identifying all human genes,
proteomics will be
crucial to the development of the higher order information necessary to
understand how these
genes function.
1
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
Until recently, researchers focused on the identification and sequencing of
genes
involved in a specific disease. However, the mere identification of a
diseasa~related gene is not
sufficient to understand its role in the disease process. Information on the
role of an individual
gene or a set of genes in the complex biology of a particular pathological
process is essential.
This is where functional genomics will play a key role.
Functional genomics aims at assigning the function to genes that are
responsible for
speciftc biological processes and diseases. It meets the challenge to identify
new and clinically
relevant drug targets and therapeutic genes. Currently, the industry is
limited in their drug
development efforts by the lack of new validated drug targets. Of the genes
identified to date,
the function of only a small number has been determined. Various techniques
have been
employed thus far to assign function to a gene. However, ascertaining the
direct link between
genes and their biological function remains a major technical hurdle. The
pharmaceutical and
biotechnology industry is looking fom a fast, efficient and automated
technology platform to
identify those genes that have diagnostic, therapeutic research, and other
relevance.
The challenge pharmaceutical companies face today is to develop drugs that act
on
novel, specific protein targets that are produced by genes. Despite
revolutionary advances
made in molecular biology and genomics, until recently only approximately 400
out of about
4,000 possible targets have been identified. Moreover, there has been no fast
and efficient way
to identify additional targets for drug development. Efforts to sequence the
complete set of
human genes have generated huge amounts of fundamentally important genetic
information,
including useful information about a handful of genes that are associated with
particular
disease conditions. However, there has been limited progress using this
information to identify
drug targets quickly and systematically. The result is a shortage of validated
drug targets and a
dearth of tools to determine which new targets have clinical promise.
One problem is that a human cell is vastly more complex than a linear
arrangement of
genes that systematically "pump out" their proteins. Only a small subset of
possible proteins is
made in a given cell at a given time and this subset changes over time and
with environmental
conditions. With 30,000 genes, the number of possible combinations of
expressed proteins is
staggering, and often the answer may lie in their interaction or regulation,
not just in their
expression. Many questions related to finding drugs have remained. How do the
different
proteins produced by these genes interact in various parts of the cell to
result in a particular
biological outcome, like releasing histamine in allergic individuals, or
multiplying without
retaining the characteristics of the cell's organ, such as a lung cell
multiplying into tumor cells?
How does the folded, three-dimensional shape of a protein (beyond its linear,
two-dimensional
2
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
sequence) effect the biology of the cell? Most importantly, what is the
function of all these
genes in relevant disease processes? An understanding of the key biological
"relationships" in
the disease process is still missing but is much needed: when, to what degree,
and under what
conditions (i.e., in what disease states) are various combinations of genes
expressed and what
are the key relationships among these genes? Many other questions of this kind
also are
extant. Tools and methods for addressing such questions are greatly desired.
Genomics in species other than man is developing as well and the need thus
exists for
ways to ascertain gene function in such systems. For example, knowledge of
gene function
and relationship in insect species will provide improved, selective,
pesticides and the like.
Understanding animal gene function will permit development of industrial,
commercial,
consumer and other products and methods having a decreased environmental
burden,than at
present, while obtaining improved efficacy and efficiency. Veterinary products
and other
materials will be improved thereby. Nor are the benefits of knowledge of gene
function
limited to animal systems. The genes of plants, fungi, viruses, bacteria and
even prion-like
constructs rnay be elucidated in this way. Great economic, therapeutic and
environmental
benefits result.
Although these approaches can tell us what genes or gene products are
"involved" in a
disease state (i.e. they were expressed in some pattern statistically related
to that phenotype),
they could not tell us which, if any, caused the condition - or - whether the
converse was
instead true. Also, because of the complex nature of the interactions of
molecules within the
cell, even if a gene that was present in a disease state could be identified,
redundancies in the
biology or slight mutations in a gene provided for almost unlimited
permutations and
combinations of outcomes. Moreover, researchers still do not know what will
reverse the
disease condition, the real goal of drug therapy.
Once any of these approaches has produced some information about the genes
that are
"involved" in a disease state, they all share a time- and resource - consuming
next step. Since
involvement is not causality, researchers do not know which gene or gene
product causes the
disease, much less which can cause its reversal. Within the context of drug
discovery, this lick
is termed "target validation."
Realizing that the answer to finding new drug targets that can reverse the
effects of
disease may lie in the interaction between proteins, not just the over- or
unde~expression of
them, researchers have begun to study the function of different genes and
proteins and how
those proteins function within the context of the entire signaling pathway to
which they belong.
As part of this effort, scientists are also beginning to elucidate the
signaling pathways (intricate
3
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
biochemical circuits of proteins that relay massages to the cell) to determine
how that interplay
affects biological outcome of the cell. This attempt to ascertain the
biological function of genes
and their protein products is known as "functional genomics." Many functional
genomics
approaches involve conducting assays (laboratory tests) to determine the
function each protein
in a pathway of interest, then moving onto the next pathway and analyzing its
members, and
searching through the complicated myriad of pathways, by process of
elimination, for a protein
target that regulates function.
Given the seemingly endless number of proteins that could be involved with a
particular disease, this approach is incredibly time intensive and
inefficient. In addition, it
frequently leads to a dead end for two primary reasons. First, often it is not
an individual
protein that controls the biological fate of a cell, but its interaction with
another protein that is
the key event. Further complicating matters is that in each cell, there exist
many possible
signaling pathways that can lead to a variety of physiologic outcomes. In
treating disease, it
may be possible to modify a signaling pathway other than the defective one and
still improve
the health of the cell. Also, the proteins and pathways selected for these
studies are based on an
assumption that they are "involved" in a disease and not any true biological
scientific evidence
that they are causally related. Given the cost (over $500 million per drug) of
the subsequent
steps from small molecule screening through animal testing to human trials and
the time used
(6-12 years), this can be an expensive and time-consuming gamble. Since these
failures are
usually because of toxicity or lack of efficacy (functional reflections of the
target's activities),
functional information at the very beginning of the discovery process could
avoid much of this
wasted time and money.
There are two general approaches to associating protein function with gene
structure.
The first approach involves deciphering a particular gene sequence from a vast
amount of
genetic data, cloning the gene, modifying the cloned gene so that it actively
expresses the
protein it encodes and then screening this protein for biological function.
The second approach
involves the initial identification of a tissue or cell type that exhibits a
biological characteristic
of interest, isolation and identification of the proteins involved, and then
identification of the
genes that encode such proteins. Major limitations of both methodologies are
that they are
typically resource-intensive, involve multiple time-consuming steps, and
generally require the
identification and cloning of the gene or knowledge of a gene's sequence in
order to produce
protein. Because the protein is the functional unit of life, the production of
protein for
functional analysis is one of the most significant bottlenecks in the
development of new gena-
based therapeutic arid diagnostic products.
4
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
Thus, there exists a great need for ways to ascertain the function and
relationships of
genes in a rapid and economical way. Such methods and the associated tools,
protocols and
materials are greatly sought in order to provide commercially valuable
information for drug
discovery, diagnostics, veterinary products, pesticides, fungicides,
industrial materials,
commercial materials, consumer products and otherwise.
SUMMARY OF THE INVENTION
The present invention is directed to achieving some or all of the foregoing
objectives as
well as other objects and benefits as will be apparent to persons skilled in
the art. In an
exemplary embodiment, arrays of viable cells are provided on a substrate. The
array comprises
elements, each of the elements comprising a subset of the cells. The cell are
transfected or
transduced with nucleic acid, preferably a preselected nucleic acid. The
identity of the nucleic
acid is known, at least in relation to the location of the array element with
which it is
associated on the substrate. The nucleic acids may be DNA or RNA, may be
encapsulated
within a gene delivery vector such as a virus, and may - and in accordance
with certain
preferred embodiments do - comprise all or part of a library.
Such libraries may comprise cDNA libraries, RNA libraries, oligonucleotide
libraries,
antisense libraries, viral libraries, and other libraries and library-like
collections. In
accordance with other embodiments, the molecular identity of at least some of
the nucleic
acids is known. In accordance with other embodiments, the nucleic acids are
selected for their
likelihood of inhibiting, stimulating or otherwise affecting a disease state,
phenotype or
condition. Such may be selected for association with known or suspected
biological functions
as well. The nucleic acids may be frozen prior to their having been
transfected into the viable
cells and, indeed, the long term stability of arrays of such nucleic acids on
substrates permit
efficient and convenient elaboration of viable, transduced cell arrays upon
demand.
The nucleic acids may be mammalian, especially human, and can represent a wide
range of species including rabbit, rodent, primate and others. Norrmammalian
animals such as
insects, fish, birds, and lower creatures may also give rise to the nucleic
acids. Such may also
derive from plants, fungi, bacteria, viruses and even constructs such as
prions and the Iike.
They may be wholly-artificial as well and may include any of a wide variety of
homologies,
substitutions, variants, and chemically modified species- all known, peg se,
to persons skilled
in the art.
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
It is preferred that a gene transduction vehicle or promoter be provided
attendant to the
nucleic acids to facilitate transduction into and expression within viable
cells with high
efficiency.
This invention is also directed to arrays of nucleic acids on a substrate, the
identity of
each of the elements of the array being known at least in relation to the
location of each
element on the substrate. Such arrays include gene transduction vehicles or
promoter attendant
to the nucleic acids. It is also preferred for some embodiments that binding
of nucleic acids to
the substrate is enhanced through provision of a binding enhancement material
on the substrate
or with the nucleic acids themselves,
These arrays of nucleic acids are prepared by elaborating upon a suitable
substrate
nucleic acids in a preselected pattern or spatial arrangement. The nucleic
acids are preferably
associated with a transduction promotion vehicle or agent. These arrays can be
stored for long
periods of time, especially when frozen. The arrays of nucleic acids on a
substrate may be
further manipulated - even after passage of time - by binding viable cells to
the substrate at the
locations where elements of the array of nucleic acids are present. The cells
are caused to
become transduced by the nucleic acids such that an array of viable cells
having exogenous
nucleic acid therein arises.
It will be understood that the terms "transfection" and "transduction" are
closely related
in the art and that the former term is generally "contained within" the
latter. They will be used
generally as synonyms, but otherwise, as convention in the scientific context
suggests.
These arrays of cells can be used in a very wide number of assays, screens and
tests,
especially including protocols which elucidate the fixnction of the genes
represented by the
genetic material thus provided. Thus, in a preferred embodiment, arrays of
cells on a substrate
can be incubated under conditions selected to promote their growth in order to
give rise to the
biological products coded for by the nucleic acid incorporated into the cells.
Determination of
these products can be correlated with the identity of the nucleic acid by
virtue of the spatial
position of the respective cells on the substrate, such that the functions of
the nucleic acids can
be elucidated and attributed to the particular nucleic acid involved.
A very wide range of screens, tests and assays may be used with the arrays of
this
invention. It is preferred that such activities be conducted under the
operative control of a
computer.
The invention includes an array of viable cells on a substrate, each element
of the array
comprising a subset of the cells transduced with a preselected nucleic acid,
and the identity of
each of the transduced nucleic acids being known in relation to the location
of the element on
6
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
the substrate. In one aspect, the preselected nucleic acids comprise a cDNA
library, a viral
vector library, an RNA library, an oligonucleotide library, a library from a
virus, an
agrobacterium library, or an antisense library. In another aspect, the
preselected nucleic acids
comprise the identical gene mutated at defined genetic locations at each
element of the array.
In another aspect, the identities of at least some of the preselected nucleic
acids are known. In
another aspect, the preselected nucleic acids have been frozen on the
substrate. In another
aspect, the preselected nucleic acids are selected for their likelihood of
inhibiting an identified
disease state, phenotype or condition. In another aspect, the preselected
nucleic acids are
selected fox their being associated with a known or suspected biological
function. In another
aspect, the preselected nucleic acids encode membrane proteins. In a further
aspect, the
preselected nucleic acids encode G-protein couple receptors, ion channels, or
viral Envelope
proteins. In another aspect, the cells are mammalian, avian, insect, plant,
plant protoplast,
yeast, fungus, bacterium, or human. In another aspect, the cells stably
express a gene of knov~n
identity prior to their application to the substrate.
The invention also includes an array of nucleic acids on a substrate, the
identity of the
elements of the array being known in relation to the location of the elements
on the substrate,
and each element of the array further comprising a gene transduction vehicle.
In one aspect, the
nucleic acids are DNA. In another aspect, the nucleic acids comprise a cDNA
library, a viral
vector library, an RNA library, an oligonucleotide library, a library from a
virus, an
agrobacterium library, or an antisense library. In another aspect, the nucleic
acids comprise the
identical gene mutated at defined genetic locations at each element of the
array. In another
aspect, the identities of at least some of the nucleic acids are known. Tn
another aspect, the
array is substantially stable to freezing conditions. In another aspect, the
substrate is
compatible with use for mass spectrometry. In a further aspect, the substrate
further comprises
a gold layer.
The invention also includes a solid body having a surface, the surface being
adapted to
bind a gene transduction vehicle in a reversible manner, to permit cells to
adhere to the surface,
and to allow cells to be transduced by the transduction vehicle. In one
aspect, the adaltation
comprises an antibody directed to the transduction vehicle. In another aspect,
the adaptation
comprises an antibody directed to the exterior proteins of a viral vector. Tn
a further aspect, the
adaptation comprises an antibody directed to the Hexon or Fiber proteins of
Adenovirus.
The invention also includes a solid body having a surface treated to allow
spotting of
volumes of liquid less than about 1 microliter in an array format without
allowing the liquid to
7
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
completely desiccate. In an exemplary aspect, the treatment comprises
application of trehalose
or game-amino-propylsilane, or freezing of the array during or after spotting
of liquids.
The invention also includes a method for spotting volumes of liquid less than
about 1
microliter in an array format onto a solid surface without allowing the liquid
to completely
desiccate, comprising including in the spotting medium a sugar, trehalose or
glycerol.
The invention also includes a method of constructing an array of viable cells.
The
method comprises providing a substrate, elaborating upon the substrate an
array of nucleic
acids, binding viable cells to the substrate at the locations where elements
of the array of
nucleic acids are present, and transducing at least some of the cells present
at said locations
with the nucleic acid present at those locations. In one aspect, a gene
transduction enhancing
composition is included with the nucleic acids elaborated upon the substrate.
In another aspect,
a surface of the substrate is coated with a binding promoting composition to
enhance the
binding of the array of nucleic acids to the substrate. In another aspect, the
array is incubated
subsequent to transduction under conditions selected to promote growth of the
cells. In another
aspect, the cells are mammalian, xodent, rabbit, primate, avian, insect,
plant, plant protoplast,
yeast, fungus, bacterium, or human. In another aspect, the substrate is
inorganic or glass
material.
The invention also includes a method for determining the biological products
produced
by members of a library of nucleic acids. The method comprises constructing an
array of viable
cells by elaborating upon a substrate an array comprising at least a portion
of said library of
nucleic acids, binding viable cells to the substrate at the locatnns where
elements of the array
of nucleic acids are present, transducing at least some of the cells present
at said locations with
the nucleic acid present at those locations, incubating the array of cells
under conditions
selected to promote growth of the cells, determining the biological products
produced at
elements of the array, and relating the production of such elements with the
nucleic acid
present at said elements. In one aspect, the identities of the members of the
library of nucleic
acids are known in relation to the location of the nucleic acids on the
substrate. In another
aspect, the library of nucleic acids is selected to be related to a disease
state. In another aspect,
the library of nucleic acids is selected to be associated with a known or
suspected biological
function. In another aspect, the library of nucleic acids is used to identify
a protein mutation
with a defined phenotype or function.
In another aspect, the library of nucleic acids is selected to encode surface-
bound
monoclonal antibodies. According to one preferred embodiment, the array is
used to identify
drug candidates that bind to proteins correlated with adverse absorption,
digestion, metabolism,
8
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
excretion, toxicity, bioavailability, or cell death. In another aspect, the
library of nucleic acids
comprises the identical gene mutated at defined genetic locations at each
element of the array.
In another aspect, elaboration includes placing upon a surface of the
substrate a binding
promoting composition to enhance the binding of the nucleic acids to the
substrate. In another
aspect, elaboration includes co-depositing a gene transduction enhancing
composition with the
nucleic acids on the substrate. In another aspect, relating comprises
detecting the biological
products produced by the cells. The array can be challenged with a
predetermined chemical or
biological species during at least part of the incubation step. In another
aspect, the array is used
to identify a protein mutation with a select phenotype. In a further
embodiment, the phenotype
is improved antibody reactivity for use in designing improved vaccine
candidates. In another
aspect, the array is used to identify the target for a drug candidate where
said target is not yet
linked to a specific disease. Arrays in accordance with the inventio can be
used to identify the
target for a drug candidate of unknown specificity. In a further aspect, the
drug candidate is a
protein, monoclonal antibody, or low molecular weight organic compound. In a
further aspect,
the drug candidate has been tested for toxicity and bioavailability prior to
the identification of
its target. In another aspect, the array is used to define antibody
reactivities from an animal's
sera. In another aspect, the nucleic acid libxary is derived from one species
and the cells are
derived from a different species. In another aspect, the cells stably express
a gene of known
identity prior to their application to the substrate.
In another aspect, the cells used contain a co-transduced gene, comprising one
or more
genes introduced into all the cells used on the array. In a further aspect,
the co-transduced gene
is a modifying enzyme. In a further aspect, the co-transduced gene is a
kinase, phosphatase,
glycosidase, protease, or chaperone protein. In a further aspect, the co-
transduced gene is
identified using the methods described above to identify the function of a
gene or protein. In
such iterative usage, the array is first used to identify a gene that confers
a particular phenotype
of function upon a cell, that gene is then introduced either stably or
transiently into all or
substantially all the cells that are then placed on another cell-array to
identify a different gene
that confers a phenotype or function to the cell that is related in some way
to the first gene,
either by homology, function, phenotype, complementarity, inhibition, or
relatedness on a
pathway. The identification of proteins that are involved in the same pathway
is one example
of such iterative usage.
9
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
DETAILED DESCRIPTION OF CERTAIN PREFERRED EMBODIMENTS
The present invention has been named by the inventor the "Protein Expression
Chip" or
the "cell-array." This exemplary cell-array is a preferably disposable axray
comprising a
capture media bonded to a solid surface and an arrayed library of gene
transduction vectors.
The transduction vectors comprise nucleic acid together with the optional but
greatly preferred
transduction promotion material. Living cells are placed on the chip
overnight, the cells are
allowed to express, and any one of hundreds of different assays can then be
performed. An
exemplary cell-array can be constructed as follows:
a. Slides are preferably coated with a substrate to which mammalian cells can
bind.
b. The slides are also arrayed with a cDNA library within plasmids or viral
vectors. The
library could also be in anti-sense form to inhibit natively expressed genes
in the cell.
c. The cDNA library is mixed with a gene transduction vehicle that will allow
the DNA
vectox to adhere to the slide but also to transduce the cell.
d. Cells in suspension are plated directly onto this array. Cells that settle
onto the slide
and adhere over a particular spot (position) in the array will contact the
genetic clone in
the transduction vector.
e. Using one of several possible molecular mechanisms, cells contacting a
specific clone
will be transduced. Convenient molecular mechanisms include use of
retroviruses,
Adenoviruses, Vaccinia virus, Adeno-associated Virus, Baculovirus, Semliki
Forest
Virus, and other viruses which can mediate gene transduction. In addition,
chemical
and lipid-based transfection methods such as calcium phosphate, DEAE Dextran,
transferrin, Lipofectamine, or GenePorter can be used. Any gene transduction
vector is
possible that can allow the gene to enter the cell and be expressed.
f. The Bells, now adhered to the slide surface and transduced with a gene, are
allowed to
express the gene (e.g. growth at 37°C overnight).
g. The living cells, now over-expressing a defined functional protein, can
then be assayed
using any number of techniques. With the correct vector, very high levels of
expression
can be achieved for biochemical and functional detection.
Once transduction has occurred and the cells have taken in the nucleic acid,
e.g. cloned
cDNA, the effect of the expression of the cDNA can be observed. The library
can be such that
the nucleic acids introduced to the cultured cells will have a mechanism to
positively effect
expression (i.e. the vectors will have a gene promoter). In this manner, the
information
residing in the cDNA sequence will be expressed Any and all detection methods
can then be
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
utilized. Mechanical, optical, and laser array reading possibilities that
currently exist are
capable of detecting the different signals of output assays from slide-based
arrays. When a spot
with a desired property (i.e. signal) is detected, its position in the array
makes identification of
the gene that caused the signal straightforward.
The surfaces that can be used for gene expression and array creation can be
composed
of any number of solid or semi-solid surfaces that can support the creation of
an array and/or
the growth of cells. For example, slides can be coated with a substrate to
which mammalian
cells can bind. Some slide materials do not need to be coated, while others
may be coated to
increase cell adherence. The surface must also support the creation of an
array which can be
temporarily bonded to the surface until cells are added and gene transduction
occurs. The
surface substrate that is used for this reversible DNA adherence may be the
same or a different
chemical composition than the substrate used to promote cell adherence. The
surface may also
be created to allow alteration of assay and detection (e.g. conductive
material to control
hybridization stringency). For example, the electro-magnetic properties of
some ceramics and
metals can be tuned to enhance gene transduction, detection, optical
reflectance or
transmission, hybridization, or other uses of the array. The array has been
enabled using glass,
tissue culture plastic, Poly Lysine coated glass and plastic, and permanox
plastic. Other
examples of suxface materials that can be used include, glass, quartz,
ceramic, plastic (e.g.
polystyrene, polypropylene), pennanox, poly-lysine coated surface materials,
silanized
surfaces, tissue culture plastic (e.g. polystyrene), agar, dextran, nylon,
paper, nitrocellulose,
silicon, gold, and optical fiber.
The genetic constructs that introduce DNA into cells for expression can be
such that the
DNA introduced to the cultured cells will have a mechanism to positively
effect expression. In
other words, the vectors will preferably have a gene promoter in order to
attain efficient
expression of protein from that gene. The promoter can be any number of widely
used
constitutively active pxomoters, such as CMV, but can also be composed of
inducible
promoters, cell-type specific promoters, or any other type of transcriptional
element. In this
manner, the information residing in the cDNA sequence will be expressed. The
genetic library
can be composed of cDNA but could also be composed of a genomic library. The
use of cDNA
focuses the screen on expressed sequences and is thus superior to random
genomic sequences
which may or may not be expressed and may or may not be expressed in any given
cell type.
The library could also be in anti-sense form to inhibit natively expressed
genes in the cell. The
library could also encode peptide sequences to screen for active or inhibitory
functions of
peptides, or to measure their ability to bind molecules (e.g. antibodies,T-
cell receptors).
11
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
For the construction of arrayed cDNA libraries, total mRNA is converted into
cDNA,
cloned into a transfer vector and subsequently transformed in E. coli. A
normalization process
(e.g. multiplex hybridization with oligonucleotides) can remove the majority
of abundantly
expressed genes, resulting in a normalized library. Individual colonies of
library are arrayed in
a microtiter format (e.g. 96-well). Automated plasmid preps can then amplify
the uniform
DNA construct within each pick. The purified DNA can then be arrayed onto a
cell-array. The
use of a library from a source of interest (e.g. a tumor, a unique cell line,
or cells of phenotypic
interest) can be used to identify the genetic and proteomic determinants of
cell behavior and
phenotype.
The construct used can code for any number of types of proteins, peptides, or
gene
products, including cDNA, mRNA, ribozymes, RNA protein fusions, organic
compounds,
cofactors, secreted proteins, membrane receptors, and others.
The transduction vehicle can be composed of a number of different forms. Any
gene
transduction vehicle is possible that can allow the gene to enter the cell and
be expressed. The
critical property of the library is that it exist in the form of a
transduction vehicle that will,
upon the addition of cells, effect the introduction of the individual clones
that make up the
library, into the cells in the immediate vicinity of the spot (position in the
array). Optimal
vectors will be determined through experimentation. Alternative vehicles may
be required for
cells of different spacies or plant organisms. Several possible molecular
mechanisms of gene
transduction are possible, including viral vectors, chemical vehicles,
bacterial vectors such as
agrobacterium, and lipid-based vehicles such as, retroviruses, adenoviruses,
vaccinia virus,
adeno-associated virus, adenovirus-AAV combination viruses, murine, leukemia
virus, HIV
and SIV-based vectors, VSV (with or without a low pH-buffer pulse),
bacculovirus, semliki
forest virus, other viruses can mediate this gene transduction, transposons,
adenoviru~DNA
conjugates, peptide-MAb conjugates, calcium phosphate, DEAE Dextran,
transfenrin,
lipofectin, lipofectamine, lipofectamine plus, lipofectamine 2000, Gene
PorterTM, PEG,
phosphatidylserine and calcium, microinjection, magnetic beads, and ballistic
particles (gene
guns).
Attachment of a gene transduction vector to the substrate surface can be
accomplished
using any number of methods that have the effect of maintaining the vector in
the approximate
location where initially applied while still allowing the vector to enter
cells once cells are
added. The surface of cells is naturally negatively charged. Purified DNA is
also naturally
negatively charged. A positively charged intermediate is often used to mediate
introduction of
purified DNA across a cell membrane and into a cell for expression (i.e. most
transfection
12
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
reagents function in this manner). The chemistry used to attach the gene
iransduction vector
may also take advantage of these properties, although it is not a necessary
feature. One
approach is to dry small spots of transfection mixture on a slide surface.
Other mechanisms for
attachment include, drying, salt precipitate, crystallization capture,
biotin/avidin, polyethylene
glycol (PEG), polyethylene oxide, biotinylated lipid-doped transfection
reagent, tetrameric
avidin for binding multiple molecules at once, poly-lysine, glycerol,
trehalose, gama-
aminopropylsilane, polyethyleneimine, DEAF-dextran, gelatin, pluronics, gum
arabic, sucrose,
antibody capture, carboxylated polyvinylidene fluoride, dextran and carboxy-
dextran, lectins
and carbohydrates, cross-linking, covalent modification for attachment (e.g.
free amines to
carboxy-dextran), electrostatic (charge-mediated) attachment, physical
barriers (e.g. etched
wells), antibody-mediated attachment, and magnetic attachment. For adherence
and stability of
viral vectors to a surface, preferred embodiments include adherence using
antibodies,
positively charged compounds, coated surfaces such as GAPS-coated glass, and
the addition of
stabilizing reagents such as trehalose, gelatin, glycerol, or sucrose. For
example, an antibody
specific for the Fiber protein that composes the exterior of Adenovirus can be
used to capture
the virus to a surface but still allow cells to be infected by the vector.
Similarly, an antibody
specific for the Hexon protein of Adenovirus can accomplish the same results.
DNA adheres to a number of surfaces (glass, poly-Lysine coated glass,
Permanox,
tissue culture plastic) if simply a liquid mixture (e.g. a transfection
precipitate) is placed on the
surface. These experiments were conducted by placing DNA mixtures onto the
surfaces and
staining with EtBr to visualize. Each type of DNA precipitate yields distinct
pattern formations
that may be representative of the type of precipitate formed on the surface.
Permanox plastic
has achieved the highest transduction efficiency and the greatest adherence of
all surfaces tried
so far. Tissue culture plastic has achieved nearly as high efficiency. Glass
surfaces proved
difficult because of hydrophobicity and difficulty of cell and DNA adherence.
Glass coated
with Poly-Lysine did not improve the characteristics of the plain glass
surface to a great extent.
In fact, Poly-Lysine tended to cause spots to be difficult to form on the
surface (liquids would
not form a precise spot). The gene transduction vector used (e.g. plasmid DNA,
lipid
transfection vehicles, or viral vectors) can be but need not be dried on the
surface of the array
material. For example, inclusion of glycerol or trehalose, freezing of samples
on the array, or
arraying under conditions of high humidity can be used to spot array locations
without drying
of the samples.
The amount of spill-over of gene transduction to cells outside the intended
bounds of
the spot increases over time. Cells assayed after 1 day are typically well
within the intended
13
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
boundary, which is often visible under light microscopy as a thin dark line
circling the
perimeter of the spot. Few cells expressing the marker gene outside the
perimeter are usually
visible. After two days, more cells outside the perimeter are visible, and
after 3 ~ days a
significant number of cells can be seen outside the perimeter. Improved
surface chemistry
conditions may be able to contain spill-over more accurately and over longer
periods of time.
Washing the DNA spot with buffer (e.g. PBS) can help to eliminate spill-over
of the
precipitate beyond the spot intended, but can also reduce the amount of DNA
bound to the
surface. In practice, the reduction in transfection efficiency from two 1 rnI
PBS washes was
minimal while the reduction in spill-over was significant.
Any number of cell types can be used for this technology. Cells that readily
adhere to
the surface substrate and that are efficiently transduced by the vector chosen
are the most
easily adapted to the technology. Adherent cells should first be resuspended
before adding to
the slide. An even monolayer of cells is preferred for optimal expression of
all spots in the
array. Densely-packed cells may be advantageous for covering the entire slide
and all positions
of the array.
Cells in suspension axe plated directly onto the array. Cells that settle onto
the slide
and adhere in a position over a particular spot (position) in the array will
contact the genetic
clone in the transduction vector. Cells contacting a specific clone will be
transduced. The cells,
now adhered to the slide surface and transduced with a gene, are allowed to
express the gene
(e.g. overnight incubation). 293 and 293T cells have been shown to work and
any other type of
cell type, cell line, or primary cell can also be used including, 293, 293T,
QT6, HeLa, COS,
CF2TH, CCC, CD4 cells, CD8 cells, Neurons, Astrocytes, Fibroblasts, Stem
cells,
Hematopoeitic stem cells, Progenitor cells, B-cells, and NK cells. Plant cells
and plant cell
lines may also be used either as intact cells or as protoplasts with their
cell walls removed, such
as by enzymatic digestion.
Some cell types (e.g. primary cells) may be difficult to transduce with some
vectors,
and optimal conditions and gene transduction vehicles will have to be
determined for these.
Only routine experimentation should be required, however. The assay is not
limited to existing
cell types. Cells that are developed and prepared specially for this
application (e.g. competent
mammalian cells with greater transfection efficiency) can also be used. In
addition, cell types
with specially designed markers (e.g. signal cascade markers or transcription
reporter genes)
can also be used. For example, a cell line can be prepared that has a reporter
gene (e.g. GFP)
under the control of a MAPI~-responsive promoter. When these cells axe placed
onto a celp
14
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
array, any gene on the chip that activates the MAPK pathway will activate the
reporter, which
can be easily detected.
Cells used for the cell-array can be manipulated prior to addition to the cell-
array. For
example, a vector that expresses a Tyr-kinase could be introduced into all the
cells prior to
addition of the cells to the cell-array. In this way, each cell would over-
express two (or more)
genes simultaneously - the single gene introduced into all the cells and a
specific gene at the
cell's position in the array. In this way, modification of proteins, cell
pathways, and functions
can be controlled, modified, and assayed for functional significance. For
example, cells can be
infected with a virus (e.g. Adenovirus or vaccinia virus) that expresses the
Furin protease.
Cells could also be transfected in bulls. Once these cells are prepared, they
can be placed onto
the cell-array to express the gene at each position in the array. If the
cell~array is designed, for
instance, to contain 10,000 variants of the HIV Envelope protein, the effects
of Furin on
Envelope (cleavage and activation) can be determined using functional or
chemical assays (e.g.
fusion, ability to bind radiolabeled CD4, exposure of hidden epitopes that can
be detected with
antibodies). In one preferred embodiment, functional assays can be used to
identify
immunologic characteristics of a protein of an infectious agent. For example,
an array of HIV
Envelope protein mutants can identify variants of the protein that are
recognized by broadly
cross-reactive neutralizing antibodies. The proteins encoded by such mutants
could serve as
vaccine candidates for eliciting a broad protective response.
Cells used on the cell-array need not be human in origin, Cells from other
species of
primates, mammals, insects, plants, fish, birds, fungi or bacteria may be
used. Genetic
transduction mechanisms may need to be altered based on the type of cell used.
It is possible, for some applications of the cell-array, that cell recovery
may be desired.
In this case, surfaces that allow microdissection, physical isolation of
cells, or laser-assisted
recovery may be used to allow fine recovery of cells with a specific function
or phenotype.
This may be especially useful when screening diverse pools of cells with
unique qualities (eg.
B-cells, T-cells, hybridomas). Cells with a desired phenotype can be used for
disease models or
for iterative identification of genes along a pathway.
Cell-based assays are important for functional screening of genes to identify
new drug
targets and gene therapeutics. In a cell based assay, cells are transduced
with a gene, measured
for interaction with a probe, and/or followed by determining changes in
cellular behavior or
phenotype. For example, a proliferation assay can be used to determine genes
tlat trigger
proliferation and that might be causal to a certain cancer.
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
The cells on the cell-array, living and overexpressing defined functional
proteins at
very high levels, can be assayed using any number of techniques. Once
transduction has
occurred and the cells have taken in the cloned cDNA, the effect of the
expression of the
cDNA can be observed. It is at this stage that perhaps the greatest value of
the cell-array arises.
Those skilled in the art will know many ways to screen expressed sequences for
the functions
they desire to investigate.
Hundreds of assays have already bean adapted to the standard 1x3 inch slide
format,
and a variety of parameters can be measured using automated detection systems
that have
already been developed. Some of the many functional and biochemical assays
that could be
utilized include, reporter gene expression, cell proliferation (e.g. agar
overlay), phenotypic
change, RNA transcription, cell migration, capillary formation, intracellular
localization,
differentiation, enzyme activity, cytotoxicity, infection, fusion, and
binding. All are known,
peY se, to persons skilled in the art.
Currently available fluorescent detection systems can detect a fluorescently
labeled
probe on a 1x3 inch slide in 1 minute at 5-10 um resolution. Software for
reading and
interpreting this data has also been developed by third parties for analyzing
standard gene.
based arrays. By analyzing the resultant 1 billion data points, we can rapidly
identify those few
cells that contain the probe of interest or that display the desired
phenotypic change. Other
labeling techniques, such as radioactivity, could also be employed.
Any and all detection methods can be utilized. Mechanical, optical, and laser
array
reading possibilities that currently exist and future detection technologies
that can be created
are capable of detecting the different signals of output assays. In addition,
new detection
methods that have unique applications to our technology, such as intracellular
imaging, may be
developed. The cell-array is designed to be amenable to assay and detection
using any existing
or future detection techniques that can be applied to cells, arrays, or
slides. When a spot with a
desired property (i.e. signal) is detected, its position in the array makes
identification of the
gene that caused the signal trivial. A non-exhaustive list of detection
technologies includes,
fluorescent labeling, radiolabeling, eolorimetric assays,
immunohistochemistry, optical
detection, cell staining, time-resolved fluorescence spectroscopy for real-
time binding,
fluorescence microscopy, spectroscopy, DNA/RNA hybridization (e.g. with
cellular
DNA/RNA), in situ hybridization, scanning probe potentiometry, automated
intracellular
imaging, surface plasmon resonance, confocal microscopy (e.g. automated),
atomic force
microscopy, miniaturized electronic biosensors (e.g. at each array position),
scanning electron
microscopy (e.g. automated), SELDI (surface-enhanced laser
desorption/ionization), MALDI
16
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
TOF (Matrix-assisted laser desorption/ionization time-of flight), and other
mass spectrometry-
based detection methods. Each of these are known, peg s~ to persons skilled in
the art.
Both intracellular and extracellular proteins can be assayed. Extracellular
proteins will
be directly accessible to assay and detection. Intracellular proteins can be
accessed using any
number of standard mechanisms, including the use of membrane-permeable
substrates.
Detergents are also readily available that can make all intracellular proteins
accessible.
Detergents range from strong ionic detergents (e.g. SDS) that could disrupt
all cells on the
slide to very mild, non ionic detergents (e.g. digitonin) or porin proteins
(e.g. Streptolysin-O)
that merely create small pores in the cell membrane. Cells can be fixed (e.g.
with formaldehyde
or methanol), stained (e.g. standard immunohistochemistry), probed (e.g. in
situ hybridization),
or assayed (e.g. for transcription-driven markers) as needed, either in a
living state or in a fixed
state.
The cell-array can take any physical form that can accommodate an array on
which
living cells will be placed. In the envisioned form, the cell-array will
physically be composed
of a plastic slide that measure 1x3 inches and is about 1/16 inch thick. Such
a slide can adapt to
any number of currently available readers, adapters, techniques, and devices.
However, other
modalities, such as 96-well sized plates, can also be used. Kits including
such arrays may be
produced.
The cell-array is particularly suitable for robotics and automation under the
control of a
computer. High-throughput, robotic, biomaterial-dispensing systems are
available to allow
precise and accurate addressability of substrates during array "printing" of
nucleic acids. Most
of the requisite engineering has already been performed in the course of
building standard gene
arrays used by the genomics industry. For example, gene arrayers (spotters)
are commercially
available and can array 10,000-40,000 spots on a standard 1x3 inch slide. A
40,000 feature
array can be composed of a 200x200 matrix. Currently available machines are
very capable of
producing tens of thousands of spots per array and the technology is improving
at a very rapid
pace. Current array technologies include pin-based arrayers, ink jet based
arrayers,
photolithography, and piezo-electric arrayers, any of which could be used to
produce an array
on a cell-array. The use of such arrayers for production of a cell-array
capable of gene
transduction is a further aspect of the invention.
At a feature size of 50-200 microns (spot diameter), such an array readily
fits on a 1x3
inch glass slide. Since many cells range in diameter from 1 ~m to 10 Vim, each
spot in an array
can be designed to transduce anywhere between 25 and 40,000 cells.
17
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
The cell-array is differentiated, inter alia, from other forms of "bio-chips"
by the
functional expression of proteins, the physical architecture, the structural
integrity of proteins
immobilized on the surface, and the ability to measure a variety of in vitro,
in situ, and
functional assays. Uses for the cell array include protein discovery, protein
profiling, structure
determination, activity measurements, as well as the assessment of
protein~protein and protein-
small molecule interactions.
Cell-arrays allow for the rapid identification and characterization of
proteins, including
the small bioactive peptides and rare proteins missed with 2-D gel
technologies. Identification
and functional characterization of proteins that are expressed in disease
state can now be
achieved. Cell-array libraries can also be used to screen for cells that
express specific
therapeutic proteins of interest.
The present invention can be used to produce and characterize proteins of alI
types.
Even complex proteins such as G-protein Coupled Receptors ion channels, and
HIV Envelope
can be produced with ease. In one example, a human gene library can be used to
express
random proteins that will still accurately express these complex (and the
other simple) proteins.
In another example, a mutation library of a single type of protein (e.g. a
random mutagenesis
of HIV Envelope or a GPCR) can be arrayed and assayed for function or other
phenotypic
characteristics.
Expression libraries may also be used to create and isolate cell lines that
express
validated protein targets of interest. Once the appropriate cell line
expressing the protein target
of interest has been isolated, the cell-array platform can be used to apply a
variety of drug
discovery techniques to identify lead candidates for drugs that may interact
with this target.
Using automated detection technologies (e.g. in situ hybridization, DNAIRNA
hybridization), the cell-array is capable of detecting changes in the
expression or localization
of any protein. The protein of interest is not necessarily limited to the
protein encoded by the
transduced gene. In other words, the characteristics of one protein or gene
can be monitored in
the presence (or absence) of every other protein introduced using the array.
An alternative application of the cell-array methodology is to produce
proteins in situ
on a chip. mRNA can be captured, synthesized, or produced with or without a
cell at a specific
location on the chip. With the mRNA at a specific location, in vitro
translation can be initiated
using standard protocols to produce proteins or peptides directly at the site
of mRNA location.
The protein synthesized can be bonded to the same site of synthesis using
cellarray surface
chemistry, new chemistries, or affinity tags embedded in the proteins
themselves (e.g. epitope
tags and antibody-coated slides).
18
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
Invasion of cells (human or non-human) by infectious diseases requires a
cellular
receptor. These receptors are often ideal candidates for drug intervention,
and the infectious
disease protein that interacts with these receptors is often an ideal
candidate for vaccine
development. Identification of these receptors can be accomplished using the
cell-array by
allowing live infectious agents to invade living cells. If the cells axe not
normally permissive
fox entry, then expression of the correct pxotein (receptor) will allow entry
of the agent. If the
cells are naturally permissive for entry, then elimination of expression of
critical genes (e.g.
using antisense cell-arrays) will disable the agent from entering or
replicating in the cell. The
genes identified may be involved in entry or in post entry events such as
assembly, replication,
or release from the cell.
Through the sequential identification of multiple genes involved in the entry
and
replication of an infectious agent, an entire pathway can be mapped. In this
iterative fashion,
completely non-permissive Bells (e.g. murine cells) can be made permissive for
steps in
infectious agent (e.g. HIV) invasion (e.g. entry, nuclear transport,
transcription, assembly,
budding, etc.). In addition, alternative pathways of replication or blockage
of replication can be
functionally mapped.
Cell-arrays have application to numerous infectious agents, such as, HIV,
hepatitis
strains, ebola, other viral strains, tuberculosis, N. meningitis, and other
bacteria strains. As
well as viruses, retroviruses, prior-caused disease, metabolic disorders and
other conditions.
The present invention permits the discovery of a gene, the discovery of the
function of
that gene, and measurement of the functional consequences of alterations in
the gene.
Massively parallel screening of function, as is now provided, automates the
measurement of
thousands of physical and chemical characteristics of a selected organism's
genes at different
times of the organism's life cycle by profiling pxotein expression and
cellular phenotype.
Versatile, distinct assays can be used for functional screening of morphology,
cell shape,
capillary formation, invasion, motility, localization of expressed reporter
genes, NO
production, growth factors, enzyme substrates, and other factors.
Cell-arrays can be used to identify and characterize proteins that confer
resistance to
chemotherapeutic agents in tumor cells, control the growth or formation of
specific cell or
tissue types (such as nerve cells, immune cells, or other cell types), control
immune cell
function (e.g. antibody production), and affect tumor cell formation.
Other applications include rapid analysis of genetically modified plants,
glycosylation
assay development, peptide binding assays, antigen capture from natural killer
cells, beta
amyloid peptide assay, identification of substrates fox proteases, capture of
cytokines by
19
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
orphan receptors, retentate mapping of Mycoplasm to establish phylogeny, assay
for drug
effect on a HeLa cell marker protein, DNA-protein interactions, quantitation
of bioactive
peptides, actin-binding venom peptides, identification of a drug target
protein, prostate cancer
mufti-antigen immunoassay, assay for nicotinic acetylcholine receptor
activity, cell-cell
interactions (e.g. sperm-egg fusion), localization of genes to intracellular
compartments (e.g.
GFP-tagged genes, follow post-translational processing for all proteins), over-
or under-
production of protein on a pathway (e.g. vitamins, amino acids) to find genes
that regulate
metabolic pathways, and other relationships.
Small-molecule drugs act on proteins. Knowledge of a protein's structure and
how
structure encodes function is crucial to the rational design and optimization
of candidate drugs.
Structure/function studies help to identify the site on the protein that
should be targeted by a
drug. This information can be gleaned only from the direct study of proteins.
Structure/function studies, as with protein expression work, currently aye
performed principally
with decades old technology. Higher-order structural information is critical
to drug discovery
and it can only be determined by investigating proteins directly. The
celharray technology can
control and identify the precise form-splice variant and/or post translational
modification-of
a protein that confers a specific function. Examples of protein
characterization programs
include mapping of protein phosphorylation sites, B-lactoglobulin peptide
mapping and protein
ID, protein purification and protease mapping, peptide mapping of proteases
and secretases,
mapping of phosphorylation sites on proteins, protein glycosylation assays (N-
and O-linked
carbohydrates), mapping of protease cleavage sites (e.g. Furin sites), mapping
of protein
sulfation sites and their effect on function, identification of DNA, RNA, or
protein modifiers
by using detection substrates (e.g. restriction enzymes, phosphorylation), and
other things.
Mutations in complex proteins can be screened at a high rate of speed for
phenotype or
function using the cell-array. For example, random mutants of complex proteins
such as HIV
Envelope and G-protein Coupled Receptors can be generated and screened on a
customized
cell-array. Function, structure, and reactivity (e.g. MAbs) can be analyzed
and only mutants
with desired characteristics need be isolated or sequenced.
Proteins act through concerted pathways, or networks, rather than in
isolation. Many
biological pathways are of a cascade nature, where the initiating action kicks
off multiple
second-order actions, each of which, in turn, initiates multiple third-order
actions. These
pathways typically contain key regulatory junctions, where entire pathways may
be turned on
or off. It is critical to map pathways in order to identify the optimal point
of intervention, such
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
as at the initiating signal or a key regulatory juncture, e.g. of a pro-
inflammatory pathway for
an anti-inflammatory drug candidate.
Pathway and network mapping studies allow one to establish the relationships
between
the fundamental biological commands used in the cell. Gene expression studies
can identify all
of the commands used in a cell's biological program and how often each one is
used, but little
about how those instructions code for function. There hay been few fundamental
advances in
network mapping technology over the past 15 or so years. Up to now, mapping a
pathway has
taken years and even decades. New technologies such as yeast-2-hybrid are
fundamentally
speeding this mapping, but are limited in fundamental ways: mammalian pathways
can not be
mapped, extracellular interactions such as ligand-receptor binding can not be
mapped, and
modified forms of proteins (e.g. phosphorylated and glycosylated) can not be
assayed. Cell-
arrays uniquely enable proteins to be processed from a variety of cells and
species in order to
determine the pathways and networks within which they operate. A potential
ligand can be
assayed for interaction with every other protein expressed in the human
genome, both
intracellular and extracellular. With very minor modification of the cell-
array, we can control
nearly all forms of protein modification (both known and unknown) to determine
if post-
translational modification of a protein is required for interaction with its
ligands.
It is important to note that a fundamental advance inherent to the cell-array
is the ability
to map extracellular functional pathways of the human proteome. Since
approximately 50% of
drugs are targeted to extracellular, membrane-embedded receptors (e.g,. GPCRs
and ion
channels), the current efforts to map human protein interactions are lacking
an efficient
enabling technology. Cell-arrays and other aspects of this invention permits
one to map entire
protein-protein intracellular and extracellular functional pathways, find new
proteins
interacting with other new and known proteins, and eliminate potential targets
rapidly because
they interact with multiple signaling pathways, thus identifying the protein
as a less desirable
target.
The interactions of proteins can also be assessed by co-expressing proteins in
the same
cell. For example, every cell added can be transduced with a single specific
gene such as a
Tyrosine kinase (e.g. by transfection in bulk, creation of a stable cell line,
or by infection with
a designed virus). Alternatively, every location in the array can have this
gene for transduction.
When each spot in the array expresses a different gene (in addition to the
first one), the result
will be an array that has two genes expressed in every cell in the array - one
defined (e.g. the
Tyr-kinase) and the other specific to the position in the array. The
interaction of the two
proteins can be assessed using visual colocation, transcriptional reporting,
or other detection
21
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
techniques. Alternatively, the functional effect of the coexpression can be
monitored using any
number of functional assays. Comparison of identical arrays, only with or
without the constant
gene, allows controlled experiments to be run. If the modifying (constant)
gene encodes for a
protein modifying enzyme (e.g. kinase, phosphatase, glycosidase, etc.), the
post~translational
regulation and modification of proteins can be assayed.
The interaction of proteins (and small molecules) can also be applied using
the cell-
array for the purpose of identifying unwanted interactions. For example, many
therapeutic
proteins, antibodies, and chemicals interact with proteins other than the
targets they are
intended to interact with. Using the cell-array, these unwanted targets can be
identified in
advanced and correlated with clinical side-effects, toxicity, or
bioavailability. In this way, the
cell-array can enhance the probability of late-stage clinical success.
CeII-arrays can express tens of thousands of proteins simultaneously,
providing an
efficient substrate for determining what antibodies are currently active in
the human body (the
human "immunome"). A human cell-array enables the targets of auto-antigenic
antibodies to be
determined. Cell-arrays from other species allows the diagnostic ability to
detect antibodies
directed against proteins of other, potentially pathogenic, organisms.
Quantitative description
of the antibodies present in an individual may make an important diagnostic
tool to describe
what happens in an immune system over time, at stasis, when perturbed by
infections (e.g.
IiIV, rhinovirus), or when responding to cancer, a vaccine, etc. Autoirnmune
disorders (e.g.
arthritis, lupus, etc.) may be particularly amenable to detection using the
cell-array with a
human gene library, and diagnostics may be a key market for this application.
Cells can be
permeabilized to detect intracellular and extracellular proteins.
The cell-array system can be used for the selection of peptides, proteins, and
small
molecules with desired properties. Cell arrays and libraries constructed from
human cells and
tissues allow analysis of protein:protein, enzymeaubstrate, and drug:protein
interactions.
Molecules can bind to or cause a functional response and be detected using the
cell-array.
Targets may be involved in a variety of important biological processes,
including the
production of proteins that function in central nervous system function;
function in cell growth
and differentiation; regulate immune cell function; control metabolic
functions, such as glucose
metabolism; relate to viral infection; and affect other key biological
processes.
Cell-array technology in accordance with the invention facilitates the
discovery and
characterization of novel human genes, which might otherwise be difficult to
identify using
alternative approaches. The protocols can activate and isolate specific types
of protein families,
such as receptors and secreted proteins, that may have particular relevance to
the drug
22
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
discovery and development process. Of special note, cell-arrays can be used to
screen for
proteins that reside in the membrane surface of a cell, commonly referred to
as integral
membrane proteins. This class of proteins has accounted for approximately 50%
of the drug
targets that have been identified and utilized by the pharmaceutical industry
to date.
Mutant or diseased cells can also be screened on cell-arrays for alteration of
function or
phenotype that may indicate limes to disease or cures for a phenotype/disease
that may be
achieved directly through gene therapy, anti-sense therapy, peptide therapy,
or protein
therapeutics or through small molecules. The ability to screen for phenotypic
and functional
changes in cells that relate directly to disease is a fundamental approach for
identifying and
validating important proteins and genes in the context of disease.
Biologists can screen proteins on the chip for interaction with organic
molecules, non-
organic molecules, peptides, proteins, DNA, RNA, metal ions, lipids,
membranes, whole
families of receptors, entire classes of enzymes, complete categories of
antibodies, whole cells,
antibodies, and many other species.
The effects of candidate drugs intended to reverse a disease process, and the
determination of the degree to which this objective is achieved free of
adverse side effects on
cells or interaction with other proteins is another aspect of the invention.
In addition, cell
arrays are not limited to detecting interactions with cellular proteins. They
can be used to
screen any substance contained within, on the outside, or released by a cell,
including DNA,
RNA, ions, organic molecules, enzyme cofactors, organelles, membranes,
peptides, proteins,
and other species.
In another embodiment, the cell-array can be used to identify substrates for
drug
targets. For example, starting with a protease target, the protease can be
expressed in every cell
on the cell-array which then co-expresses potential substrates or modified
substrates. New
substrates that are cleaved by the protease or mutant substrates that are
resi~ant to the protease
can then be identified and used for drug development.
Cell-arrays and other embodiments of the invention can be used to identify and
define
the proteome, the array of proteins expressed in a human cell. With each cell
in the cellarray
expressing a defined gene, the effects of that gene on the rest of cell's
proteome can be defined.
For this purpose, special chip surfaces may need to be utilized that allow
gene transduction and
cell growth but that also allow capture of proteins via mass spectrometry.
Techniques such as
SELDI (surface-enhanced laser desorption/ionization) that can ionize specific
spots within an
array could be suited for analysis of the proteome. Applying such an analysis
across an entire
23
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
cell-array expressing the human genome would allow a researcher to define how
each gene in
the human genome effects every other protein in a cell.
The cell-array can be designed in a manner that allows microfluidic channels
to be
incorporated into the chip. In this manner, infusions of molecules directly to
cells of interest
can be discretely controlled. Alternatively, proteins released from discrete
subsets of cells
could be harvested and analyzed. In one embodiment, cells overlaying a
microfluidic channel
could receive a continuous stream of reagents, such as chemicals, antibodies,
or potential
ligands, that could then be used to detect a cellular response.
Cell-arrays enable scientists to conduct differential diagnosis of the
immunome, the
complete set of immunologic targets in a human. Protein expression of the
human genome will
enable the diagnosis of immune and inflammatory diseases that are directed to
self antigens.
Rapid identification of multiple protein disease markers simultaneously
represents a
tremendous improvement over existing assays. Single protein disease markers,
such as PSA for
prostate cancer or CA125 for ovarian cancer, have limited reliability for
early detection, and
their use remains controversial. Some examples of cell-array uses for
diagnostics include
biomaxker discovery, schizophrenia diagnostic markers, kidney stone disease
marker, protein
profiling of cell lysates, validation of protein markers, prostate cancer
markers, bladder cancer
markers from urine, toxicology correlation of drug use with immunological
response,
expression profiling, for target identification and validation, toxicology
profiling, for drug lead
selection, diagnostic evaluation, for patient management, disease management,
for therapy
selection, and others.
Because the cell-array and other aspects of the invention can express an
entire genome
simultaneously, small molecules can be screened against the proteins from the
genome to
identify reactions. In this manner, purified monoclonal antibodies and small-
molecules (e.g.
organic drugs) can be identified that target proteins of specific structures
or phenotypes of
defined function, even without knowing the precise target of interest.
For example, a random, purified monoclonal antibody from a defined hybridoma
clone
can be used to screen a cell-array. The same could be done for a chemically
pure small-
molecule. The protein that the antibody reacts with can then be defined.
Monoclonal antibodies
should react specifically with only one protein. This may be useful if
antibodies (or small
molecules) have effects of interest, but their targets are not known.
Moreover, if a random
antibody or chemical binds to a small number of targets on the cell-array
(ideally a single
target), then the specificity of that antibody or chemical is defined.
Screening a large number
of purified monoclonal antibodies or chemically pure small-molecules will
enable the
24
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
development of a library of antibodies/chemicals that have already been pre-
screened for the
specificity desired. Moreover, if these antibodies or chemicals are also
prerscreened for their
toxicity, bioavailability, etc., a library of compounds will arise that has
lcnown specificity and
that are pre-screened to be better compounds for drug development - i.e. a
library of lead
compounds to defined targets. Targets of complex nature (e.g. membrane
receptors,
glycosylated proteins) axe particularly amenable to the cell-array.
A large library can be built even before specific targets are linked to
specific diseases
or phenotypes. For example, a biotechnology company may discover that a new
gene (X) is
involved in cancer. Rather than begin screening for small-molecule lead
compounds or
antibodies to that new gene, a compound and/or antibody specific to that gene
that has already
been screened for desirable characteristics will be identified for use. In
this manner, the early
stages of drug development can be hastened and better molecules for human
application (e.g.
toxicity, bioavailability) can enter drug discovery.
In another application to achieve a similar result, purified panels of
monoclonal
antibodies (to unknown epitopes or target proteins) can be spotted on a
customized cell-array.
The chip can then be screened against proteins of interest in order to
identify which, if any, of
the antibodies on the chip bind the protein of interest. MAb supernatants can
be spotted on the
cell-array for this purpose. Alternatively, genes encoding for MAbs (e.g.
random and
mutagenized) can be used for screening purposes. Completely human MAbs can be
generated
and isolated in this manner. Similar results can potentially be obtained for
small-molecule
compounds if they are spotted directly on the chip.
Antibody and T-cell receptor responses can also be generated in a similar way
if genes
encoding for proteins or epitopes are arrayed on the cell array and then
hybridomas or T-cells
are used as the cells on the array. The cells will be transduced with the
protein and will respond
appropriately by producing the protein. If the cell also produces a T-cell
receptor or antibody
that reacts with the expressed protein, it can be detected using a number of
techniques. The
cells may be recovered by laser ablation, dissection, or otherwise.
Cell-array libraries can be used to search for cells that exhibit specific
biological
properties. When a cell with a desired feature is detected, we can rapidly and
directly associate
this specific characteristic with the expressed gene by its location in the
array. One strategy
avoids the less efficient extrapolation of gene function from gene sequence
that has, up to now,
been the industry paradigm. The cell-array can be used with cells from a wide
variety of
species that are of commercial interest. Cell lines, specially prepared cell
lines (e.g. with gene
markers or transcriptional signals), and primary cells can be used.
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
The cell-array is an easy to use platform for target discovery and validation.
The cell-
array can be shipped to scientists ready to use and can be stored for months
or years in a
standard laboratory freezer. Any number of cell types, including primary
cells, can be used on
the chip, and achieving expression of every gene on the chip can be
accomplished overnight.
The cell-array technology offers a number of superior characteristics, these
include
simultaneous expression of thousands or tens of thousands of genes, expression
libraries can
include an entire organism's genome or tens of thousands of mutants of a
single gene that
could then be selected for function, assays can be performed within days (most
assays will
typically take 2-4 days, but an assay can be performed in as little as one
day), multiple chips
can be processed simultaneously, allowing comparative treatments and
conditions, and others.
Protein expression libraries express tissue-specific and rarely expressed
genes as well
as abundantly expressed genes at comparable frequencies. As a result,
significant biases toward
genes that are ordinarily expressed at high levels or in many tissues can be
minimized or
avoided. Libraries used can ensure significant coverage of the entire genome,
including rarely
expressed genes encoding key biological regulators, which axe believed to be
valuable drug
targets or therapeutic candidates.
The present invention is compatible with a variety of different biological
model
systems, or assays, including biochemical, cellular or even animal assays. In
addition, libraries
may be created from a variety of cell types, including human, animal, plant,
or prokaryotic
cells. Cell-arrays can be used to generate cell lines that express activated
genes at high levels.
These cell lines can be used to produce large quantities of proteins for
biochemical studies, in
cell-based assays for screening therapeutic compounds, or for functional
genomics studies. In
addition, genes may be permanently or temporarily expressed, depending on the
goals of the
research project. Cell-arrays can also be used to activate genes in a manner
that does not
require the isolation and cloning of individual genes or the use of gene
sequence information.
Post-translational modifications of proteins can be monitored and controlled
using the
cell-array. For example, the functional form of a protein can be recovered
from the celparray
to ascertain any post-translational modification. Even further, cells that are
co-expressing
genes that affect post-translational modification can be used to control and
measure the
function of proteins when they are modified. For example, every cell used in a
cell-array can
be made to express a Tyr-kinase just before the cells are added to the cell-
array. In another
example, the modifying gene does not need to be known - a single, even random,
gene can be
over-expressed in every cell just before the cells are added to the cel~array.
Defined functional
effects of the co-expression can then be measured.
26
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
Many proteins are modified after they are made. These modifications-the
addition of
a phosphate or complex carbohydrate, for example-often critically affect
protein function.
Many proteins are only slightly active to completely inactive until they are
appropriately
modified. Gene expression studies indicate nothing about post translational
modifications.
The breakdown products of proteins often have unique functions in their ovv~1
right. For
instance, the well-known anti-angiogenesis drug candidates, angiostatin and
endostatin are
each fragments of other proteins-plasminogen and collagen Type XVIII,
respectively. Gene
expression studies cannot identify potentially bioactive fragments of
proteins; only protein
expression studies can make this distinction. Gene expresaon studies do not
provide a
complete picture of normal or disease biology, but merely the outline of the
cell's biological
program. Protein expression studies complement gene expression information for
multiple
reasons. Thus, mRNA and protein levels arenot always correlated and splice
variants of genes
can produce multiple forms of proteins. Protein expression studies can
identify which splice
variants are being made, and whether or not the splice variant produced by a
given gene
changes in disease. Importantly, these variants can be identified after the
phenotype of interest
is uncovered, saving time by reducing the human proteome an order of magnitude
to the size of
the genome. '
The cellular architecture of the cell-array offers a number of advantageous
attributes
including stability of expressed proteins, ease of manufacture, ease of
detection using standard
assays, ability to control binding and assay conditions, high packing density
for massively
parallel protein expression, and structurally intact conformation and
orientation of proteins.
The use of cell-arrays, and other aspects of the invention for the
identiftcation of
protein:protein interactions is an attractive alternative to traditional yeast
two-hybrid systems
because they can utilize proteins derived from any type of organism- human,
microbial, plant,
etc. - and the technology can express the entire, structurally intact version
of membrane-bound
proteins and receptors. Cell-arrays have numerous advantages over the commonly
employed protein separation/purification technology (giant 2-D gel
electrophoresis), which is a
decades-old technology. In particular, cell-arrays are rapid, reproducible,
and can be used to
probe the function of even very rarely expressed genes. They enable follow-up
investigations,
such as structure/function studies, to be performed directly on chip-bound
proteins. Giant 2-D
gels, on the other hand, are slow, not terribly reproducible, require large
sample sizes, and
require significant further purification work (liquid chromatography or some
equivalent means)
before proteins can either be identified or investigated further. Moreover,
the array format of
cell-arrays, coupled with different libraries representing different types of
genes, allows
27
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
unparalleled flexibility in determining gene function. A single chip can
identify and determine
the function of more proteins than can be separated on a single giant 2-D gel.
Cel)-array
libraries can also be created from numerous sources, and can identify genes
that are lost on
giant 2- D gels due to size limitations and problems with handling membrane-
spanning regions
of proteins.
Cell-arrays and other embodiments of the invention have a diverse range of
applications for understanding the basic functions of the human genome.
Several diseases will
be immediately amenable to drug development using the cell-array. We can
isolate hundreds of
proteins directly associated with the cellular phenotype that causes major
chronic and acute
diseases. Examples of functional pathways that can be studied with the cell-
array and the
associated disease applications include pathways of cellular proliferation
(Cancer),
personalized vaccines (Non-Hodgkin's lymphoma), cell lysis (antimicrobial
peptides),
stimulation of hematopoietic growth factors (bone marrow transplants),
differentiation of cells,
cell proliferation and oncogene identification, and fat deposition increase or
decrease (obesity).
Other important diseases include asthma/allergy, autoimmunity, cardiovascular
disease,
diabetes, osteoporosis, osteoarthritis, obesity, rheumatoid arthritis,
transplant rejection, tumor
growth programs, viral infectious agents, bacterial infectious agents, fungal
infectious agents,
and metabolic profiling.
When applied to crop production, functional genomics can enhance the
nutritional
content of foods, select for enhanced phenotypes, reduce the effects of
farming on the
environment, and develop foods that can enhance food production. Arabidopsis
thalania, rice,
corn, and soy will be prime agricultural targets for functional genomics.
Arabidopsis is a useful
model organism because it is related to soybeans, cotton, vegetables and oil
seed crops. Rice is
an important target and model organism because it is one of the world's most
important grains
and commodity crops, and it is closely related to corn, wheat, barley,
sugarcane, oats and rye.
One of the unique attributes of Protein Chip Expression technology is the
ability to
rapidly identify antibodies that are differentially expressed in immune-
related diseases. We
will exploit this capability in proof of principle studies to identify novel
auto antigens that are
targeted in human immune disorders such as arthritis, asthma, and allergy. The
applications
that this capability enables is two-fold: 1) to identify and patent novel
disease markers as
diagnostics, differential diagnostics and patient management tools; and 2) to
establish the cell-
array as the platform technology to perform diagnostic testing with novel
protein markers,
which would translate into chip sales.
28
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
One example of the application of a diagnostic use for the cell-array involves
lymphoma. The cell-array can be used first to identify what B-cells have
mutated based on the
over-production of a specific antibody and the reactivity of antibodies
produced by that B-cell
to a protein on the cell-array. Next, a peptide antigen directed to that
antibody can be designed
and used as a radiological marker or drug (e.g. radiolabeled or linked with a
toxic gene).
Viral vectors allow the expression of known and unknown genes in a large range
of
host organisms and cell types in order to determine gene function (functional
genomics), and
can enable the expression of genes in cells used for the production of
therapeutics
(biomanufacturing).
Retrovirus-based technologies can introduce a large library of genes or gene
fragments
into cells of all types. Each retrovirus can be designed to encode a specific
gene and each virus
with a unique gene can be placed on the cell array. A defined virus is then
used to infect a cell
of interest in order to ovex-express a specific gene. The methods described
herein enable the
creation of a library of such retroviruses and placing tens of thousands of
them on a 1x3 inch
glass slide. Alternatively, some libraries are conunercially available, either
in arrayed or non-
arrayed format.
One advantage of using retroviruses is that once a gene or function is
identified, the
retroviral probe that caused the desired phenotypic change can be transferred
to other cells,
including animal models, and used in further development. Retroviruses are
also capable of
infecting many cell types, including cell lines, primary cells, and non-
dividing cells.
Adenoviral vectors (Ad) are a commonly used gene delivery system for gene
transduction into human cells and tissues because of their high transduction
efficiency. The
adenoviral vector carnes the transgene into the target cell, but does not
integrate it into the
target cell genome. Arrayed adenoviral libraries in a cell array format, as
described herein,
could enable high levels of expression in cells and provide a high gene
transduction rate at
each spot on an array. The greatest benefit of an arrayed library format is
the ability to perform
versatile, functional assays with a wide variety of human cell types,
including primary cells.
Adenoviral vectors have advantages including broad host range and low
pathogenicity.
Adenoviruses can infect a broad range of mammalian cells and therefore permit
the expression
of recombinant proteins in most mammalian cell lines and tissues and ilfection
and expression
of genes in both replicative and non-replicative cells. Additionally,
adenoviruses can infect
virtually alI cell types with the exception of some lymphoid cells. This
allows for a direct
comparison of results obtained with transformed cell lines and primary cells.
They replicate
efficiently to high titers. The Ad system allows production of 101° to
1011 VP/mL which can be
29
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
concentrated up to 1013 VP/mL. This feature makes it a very good vector system
for large-scale
applications. Helper-independent Ad can accommodate up to 7.5 kb of foreign
DNA. To
provide additional cloning space, the E1 and E3 early regions of Ad can be
deleted.
Additionally, Ad can normally encapsidate a viral DNA molecule slightly bigger
than the
normal DNA (105%). These combined features allow for the insertion of an
expression
cassette containing a gene or multiple genes of up to 7.5 kb into one
recombinant Ad. The Ad
expression system can be designed to express multiple genes in the same cell
line or tissue.
The Ad can contain two genes in a double expression cassette of the Ad
transfer vectors.
Alternatively, using different recombinant viruses each expressing a different
protein, a co-
infection of the desired cell lines can be performed. Determining the MOI
ratio of the different
recombinant viruses will provide the proper relative co-expression of the
recombinant proteins.
Moreover, there is no insertional mutagenesis; Ad remains epichromosomal in
all known cells
except eggs and therefore does not interfere with other host genes. The
integration of only one
copy of virus in zona-free eggs is a better system to produce transgenic
animals with specific
characteristics. The Ad vector system uses a human virus as a vector and human
cells as a host.
It therefore provides the ideal environment for proper folding and exact post
translational
modifications of human proteins.
E~~AMPLES
The present invention, as embodied in cell-arrays, integrates protein
biochemistry with
advanced materials science and microfabrication to create a miniaturized chip
containing high-
density arrays of functional proteins to quickly and accurately correlate
protein function with
genetic composition. The cell-array technology has been constructed from a
single-use,
disposable plastic slide expressing functional and structurally intact
proteins in cells that are
bonded to the surface. The primary components of the technology have now been
demonstrated to function in accordance with the invention.
The library is contained within a gene transduction vehicle that will allow
the vector to
adhere to the slide but to also transduce the cell. The current technology has
been enabled,
iyatey~ alia, using a cDNA construct expressing an easily measured molecular
marker (Green
Fluorescent Protein (GFP) in a pcDNA3 vector with a CMV promoter). However,
any
construct, plasmid, gene, or gene fragment could have been used as well.
A present embodiment has been prepared using lipid-based transfection vectors
Lipofectamine, Lipofectamine Plus, Lipofectamine 2000, and Gene PorterTM.
Calcium
phosphate has also been used. The precipitate formed by each of these
methodologies was
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
allowed to air-dry on a surface before placing cells on the surface. The
liquid precipitate
mixture is also placed on the surface, allowing the precipitate to form and
settle on the surface.
The rest of the liquid is washed from the transfection precipitate (i.e. no
air dry step). If the
liquid precipitate is left on the slide for sufficient time (e.g. 1 h), the
precipitatesettles, adheres
to the suxface sufftciently to withstand washing, and can then be used
directly for gene
transduction without a drying step. The precipitate can also be allowed to
adhere to the surface,
the media is replaced, and then cells added. A spotof diluted Lipofectamine
can also be placed
directly on the slide whereupon a spot of diluted DNA is placed on top. This
methodology
allows an effective precipitate to form directly at the array position of
interest rather than being
formed in a tube prior to placement on the surface.
Gene expression using this methodology increases significantly over time.
Cells
assayed 2 days following gene transduction can express double or triple the
amount of marker
gene than cells assayed the day after gene transduction. Cells assayed 3 days
following gene
transduction express incrementally more (e.g. 20-40%) marker. This increase in
gene
transduction efficiency, however, is offset in part by increased spil~over of
expxession outside
the intended bounds of gene transduction.
Optimem media was used for transfection precipitate formation although other
medias
may be employed. 10% DMEM with 1% Pen-Strep was used for growth of cells
because of its
wide-spread use for standard tissue culture growth. Other media types will
work similarly,
although some types may yield different efficiency. Antibiotics and serum may
have a
particularly strong effect on gene transduction efficiency.
CeII arrays in accordance with certain preferred embodiments can be
demonstrated. A
monolayer of cells, such as HEIR-293T cells, all of which are identical can be
deposited on a
substrate surface. The array may be such as to have a marker gene, e.g. Green
Fluorescence
Protein, that has been transduced into a defined subset of the cells. Clearly
defined
fluorescence gives a visual indication of the controlled demarcation of gene
transduction.
While optimal conditions for gene expression will be determined for each
particular
circumstance and system, as an example, a protocol follows that has been used
for prototype
development. Variations are included in the other sections discussing each
component of the
technology. The current protocol for the use of the cell-array technology is
performed, in one
exemplary embodiment, as follows:
a. 100 g,I Optimem media was combined with 1 ~.g DNA (pcDNA3-GFP) and 6 ~,l
Plus
reagent (part of the Lipofectarnine Plus commercial reagent package from Life
Technologies)
31
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
b. In a separate tube, 100 ~.1 Optimem media was combined With 4 ~1
Lipofectamine
c. Both tubes were allowed to incubate at room temperature for 15 minutes
d. The DNA mixture was then added to the Lipofectamine mixture with gentle
vortexing
e. The tube was allowed to incubate at room temperature for 15 minutes
f. 10 ~,1 of the mixture was placed as a spot on a Permanox cell-azlture slide
forming a
spot of approximately 3 mm diameter
g. The spot was allowed to air-dry in a sterile tissue culture environment
without a lid and
at room temperature for approximately 2 hours until the liquid had evaporated
and dry
residue was visible where the liquid mixture had been placed
h. The well was washed twice with 1 ml of PBS
i. 2x105 293T cells were resuspended in 0.5 ml 10% DMEM media, added to the
well
(the size of a 24-well), and allowed to settle onto the surface
j. Cells were incubated 1-2 days at 37°C and allowed to express the
gene
k. Gene expression was monitored using an inverted epi-fluorescent microscope
with a
light filter that allowed detection of GFP expression
Cells on a cell-array made in this way can be assayed. Spots representing
cells
transduced with the pcDNA3-GFP vector and distinguished from dark spaces
between the
spots, which contain cells that have not been transduced (visible under normal
white light
illumination). The vector chosen represents a convenient marker, but any
plasmid or gene could
have been chosen for any or all the spots in the array.
Spots have been formed on a surface ranging from 0.1 ~1 to 20 ~,1, thus
achieving the
lowest limit possible with manual pipetting. In each case, cells were observed
within spots
expressing the marker gene (GFP). Spots of decreasing size achieved diminished
gene
transduction frequency (1-20%), while the larger spots (10-20 ~l) could
achieve gene
transduction frequencies of over 50%. This response is likely a result of the
amount of
precipitate able to be placed within a spot (smaller drops of transfection
mixture have less
volume of precipitate).
Methods for simplifying the automated placement of transfection mixture have
been
developed. Rather than mixing lipid with DNA to obtain a precipitate prior to
addition to cells
or to a spot on a slide (a normal transfection protocol), a spot of diluted
Lipofectamine was
placed on the slide and then a spot of DNA was subsequently placed on top.
This methodology
'allows an effective precipitate to form directly at the array position of
interest, and avoids the
problem of the precipitate not staying well mixed in solution during a
prolonged arraying
32
CA 02445884 2003-10-29
WO 02/088165 PCT/US02/13432
procedure. This methodology thus represents one possible mechanism for
automated
production of a cell-array array. A modified gene arrayer might be necessary
for producing
cell-arrays using such a technique. For example, an arrayer with a duahpin
slide could first
drop Lipofectamine onto a slide, then drop the DNA onto the first drop.
Alternatively, the slide
may first be coated with a Lipofectamine layer. Alternatively, spots of DNA
can be arrayed on
a slide and then lipid-based transfectant can be placed over the DNA to form a
precipitate at
the location of the DNA.
A second method for high throughput transfection spotting has also been
enabled. A
lipid transfection mixture is prepared using only the lipid (e.g.
Lipofectamine) and media. This
was spotted onto a slide and allowed to dry. DNA-containing transfection
mixture was then
placed on top of the dried lipid and allowed to precipitate and dry at the
spot of interest. In this
way, an entire slide can be coated with a dried lipid mixture and then
individual spots of DNA
would merely have to be spotted onto the slide where they could precipitate in
place.
Exemplary variations of the foregoing procedures are as follows:
%Cell Transduction Day
1 2 3
Standard ~
200 u1 transfection 15% 45% 65%
mix
100 u1 transfection 10% 25% 50%
mix
20 u1 transfection 3% 15% 25%
mix
2x10(5) cells 15% 40% 50%
0.2x10(5) cells 15% 25% 25%
Optimem 5% 20% 20%
2% DMEM 15% 40% 40%
Washed 2x with PBS 15% 45% 45%
Lipofectamine Plus 15% 45% 55%
Lipofectamine 10% 35% 50%
u1 spot size 10% 30% 40%
1 u1 spot size 3% 10% 20%
0.25 u1 spot size 3% 25% 20%
0.1 u1 spot size 1 % 5% 5%
33
