Note: Descriptions are shown in the official language in which they were submitted.
CA 02714880 2010-09-15
A Novel Tumor Biomarker
Field of Invention
This invention relates to the diagnosis and therapy of tumor, and specifically
to a novel
tumor biomarker and methods and kits for the detection of cancer occurrence
and metastasis.
The invention also relates to methods and medicaments for the treatment of
cancer and/or its
metastasis.
Background of Invention
Currently, around 11 million people are diagnosed as tumor patients every year
worldwide, and it is speculated that this number will increase to more than 16
million by the
year of 2020. In 2005, among the 58 million deaths, 7.6 million are caused by
cancer
(accounting for about 13%). This number is increasing, and it is expected that
9 and 11.4
million people will die of cancer by the year of 2015 and 2030 respectively.
(World Health
Organization, 2006)
Tumor markers are the substances produced by tumor cells during the
progression
caused by gene mutation, including antigens and other bio-active substances,
which can be
used for the early detection of cancers as well as the monitoring of disease
progression and
response to a treatment (ASCO, 1996). It brings huge benefit for the clinical
treatment of
cancers, especially when it can be detected before any obvious clinical
phenomenon or when
it can be used to monitor the patients' response to certain treatment. At
present, in order to
better meet the clinical need, greater efforts on the research and development
of tumor
biomarkers are required.
The applications of most tumor biomarkers currently used in clinic are more or
less
restricted due to the not-so-good sensitivity and specificity. For example,
the AFP level and
ultra-sonic examinations are largely used for liver cancer detection. Although
their
sensitivities are not very high, they indeed prolong the survival rate of the
patients by
diagnosis of the high-risk people. The tumor antigen CA-125 has a higher
sensitivity but lacks
specificity. Similarly, the blood tumor biomarker CA15-3 which is used for the
detection of
breast cancer could hardly be used for early detection due to low sensitivity.
Therefore,
22031285.2
1
CA 02714880 2010-09-15
methods for the early detection of cancer as well as to distinguish benign and
malignant
tumors are currently not available in clinic. New technologies as well as new
methods are
required to be developed to resolve these problems.
The development of tumor proteonomics brings hope for the identification of
novel
tumor biomarkers. The malignant transformation of tumors always results in the
change of
protein expressions, which could be quantified at the protein level. Thus, a
lot of information
and data could be derived, by which potential biomarkers could be identified
and evaluated
for further development and clinical application.
Hsp90a (Heat shock protein 90a, Hsp90a) is a molecular chaperone, which
functions to
stabilize its client proteins in their active states. Hsp90a is one of the
most abundant proteins
in the eukaryotic cells accounting for about 1-2% of whole cell proteins. The
intracellular
Hsp90a mainly functions to stabilize its clients (i.e. estrogen receptor) and
assistant their
maturation (i.e. some kinases and signal proteins). However, in other
physiological
conditions, Hsp90a is also involved in mediating events such as the evolution
of mutated
proteins, rearrangement of cytoskeleton, translocation of nuclear proteins,
cell proliferation
and apoptosis, protein degradation, antigen processing and LPS recognition
etc. Hsp90a is
also related with many diseases such as cancer, autoimmune disorder and
cardiovascular
diseases. For example, the monoclonal antibody against the antigen of LKVIRK
sequence
derived from Hsp90a can be used to treat fungal-related infection, and this
clinic trial is
currently ongoing by the Neutec company (Trade name: Mycogrip).
It is also reported that Hsp90a could be secreted under some stimulus (Liao et
al. (2000)
J. Biol. Chem. 275, 189-96). As a classical intracellular protein, there is
little report regarding
the function of extracelluar Hsp90a. In previous reports, Hsp90a was
identified to help the
antigen processing in APCs and was one of the four proteins related to the
lipid raft. They can
interact with LPC thus trigger the intracellular response of cells.
(Triantafilou et al. (2002)
Trends in Immunology 23, 301-4).
Hsp90a was also found to be highly expressed in the surface of some tumor
cells,
including the small-cell-lung cancer cell, melanoma and liver cancer cells
(Ferrarini et al.
(1992) Int. J. Cancer 51, 613-19). The high expression of cell surface Hsp90a
in these cells
were speculated to be related with the antigen processing while direct
evidence is not
available.
22031285.2
2
CA 02714880 2010-09-15
It is also reported that Hsp90a could help the translocation of transmembrane
proteins
(Schlatter et al. (2002) Biochem. J. 362, 675-84), and is related with the
efflux of some
anti-leukemia, lung cancer, cervix cancer drugs (Rappa et al (2002) Oncol.
Res. 12, 113-9 and
Rappa et al (2000) Anticancer Drug Des 15,127-34).
The intracellular Hsp90a is an important target for the development of anti-
cancer drugs,
as it is involved in the regulation of many signaling pathways which are
critical for the cancer
cell transformation. Inhibition of intracellular Hsp90a could result in the
selective degradation
of proteins related with cell proliferation, cell cycle control as well as
apoptosis. Recently,
some known antibiotics such as Geldanamycin, Radicicol and Coumermycin Al are
natural
inhibitors of Hsp90a. A patent (WO 00/53169) describes this mechanism and
proposes that
preventing the interaction of chaperones with its clients could result in the
inhibition of its
chaperone activity. Among these antibiotics, Coumarin and its derivatives are
believed to have
this activity. However, these inhibitors described in patent WO 00/53169
mainly target the
intracellular Hsp90a.
The analogue of Geldanamycin 17-AAG is also an inhibitor of Hsp90a and is
currently
under clinical trials. However, some reports show that 17-AAG could have non-
specific
inhibitory effects and cell toxicity by interacting with many other cellular
components. In
addition, due to the limited knowledge on the physiological functions of
Hsp90a and its
clients, direct inhibition of intracellular Hsp90a is risky.
The patent (EP 1457499A 1) describes the function of extracellular Hsp90a in
promoting
tumor cell invasion via activating the MMP-2. Based on these mechanisms, the
patent
proposes that inhibition of extracellular Hsp90a could prevent the tumor
invasion and
metastasis, and by detecting the response of tumor cells to the treatment of
Hsp90a inhibitor
they can deduce the invasive ability of the cells and their relationship with
Hsp90a.
The inventors of patent WO/2008/070472 propose that they can monitor the anti-
tumor
efficacy of Hsp90a targeted therapy by detecting the plasma Hsp90a and other
related factors.
In this patent, they provide the relationship between the plasma Hsp90a level
and the efficacy
of the inhibitors including 17-AAG and 17-DMAG as well as the relationship
between the
level of plasma Hsp90a and tumor volume in mouse models. However, they do not
provide
any evidence about the exact form of plasma Hsp90a and do not demonstrate the
relationship
between the plasma Hsp90a level and tumor malignancy especially tumor
metastasis. They do
22031285.2
3
CA 02714880 2010-09-15
not propose the application of plasma Hsp90a as an independent tumor biomarker
in tumor
diagnosis and prognosis, either.
One group reported that serum Hsp90a level is related with the stages of non-
small-cell
lung cancer (Xu et al. (2007) J. Cancer Mol. 3,107-112). The serum level of
Hsp90a in these
lung cancer patients was significantly higher than that of normal people or
benign tumor
patients. However, again this paper did not identify the exact form of serum
Hsp90a as well
as its relationship with tumor metastasis. Besides, it only investigated non-
small-cell lung
cancer, while the relationship between serum Hsp90a level with breast cancer,
liver cancer,
and pancreatic cancer is unknown. Moreover, the serum level of Hsp90a was not
quantitatively measured, thus could hardly be translated to clinic development
and further
application.
Summary of Invention
This invention is based on the discovery that the plasma Hsp90a level is
correlated with
the development, malignancy and metastasis of many types of cancer.
Accordingly, plasma
Hsp90a can be used as a new tumor biomarker. The inventors found that the
plasma Hsp90a
is different from the intracellular Hsp90a, because the plasma Hsp90a lacks
four amino acid
residues at its C-terminus compared with the intracellular Hsp90a.
Therefore, in one aspect, this invention provides an isolated polypeptide
comprising or
consisting of the amino acid sequence of SEQ ID No.1.
The polypeptide of this invention may be phosphorylated. In particular, in the
polypeptide of this invention, one or more amino acid residues in the amino
acid sequence of
SEQ ID No.1 selected from the group consisting of Thr90, Ser231, Ser263,
Tyr309 and a
combination thereof are phosphorylated. Preferably, the Thr90 in the
polypeptide of this
invention is phosphorylated.
The polypeptide of this invention may serve as a tumor biomarker. Using an
agent
specifically binding to the polypeptide of this invention, it is possible to
detect the plasma
level of this polypeptide, and thereby to determine the presence of cancers
and the stage and
metastasis of cancers.
Accordingly, in another aspect, this invention relates to a method of
determining the
presence, stage and/or metastasis of a cancer in a subject, the method
comprises the step of
22031285.2
4
CA 02714880 2010-09-15
detecting the level of a polypeptide comprising or consisting of the amino
acid sequence of
SEQ ID No.1 in a plasma sample of the subject by using an agent capable of
specifically
binding to said polypeptide.
In yet another aspect, this invention relates to a method of screening the
presence of a
cancer in a high-risk population comprising the step of detecting the level of
a polypeptide
comprising or consisting of the amino acid sequence of SEQ ID No.l in a plasma
sample by
using an agent capable of specifically binding to said polypeptide.
In still another aspect, this invention relates to a method of determining the
prognosis of
a patient having a cancer comprising the step of detecting the level of a
polypeptide
comprising or consisting of the amino acid sequence of SEQ ID No.l in a plasma
sample of
the patient by using an agent capable of specifically binding to said
polypeptide.
In a further aspect, this invention relates to method of determining the
efficiency of a
surgery, radiotherapy or chemotherapy treatment to a patient having a tumor
comprising the
step of detecting the level of a polypeptide comprising or consisting of the
amino acid
sequence of SEQ ID No.1 in a plasma sample of the patient by using an agent
capable of
specifically binding to said polypeptide.
Preferably, the agent capable of specifically binding to the polypeptide of
this invention
can be a specific antibody against the polypeptide. Preferably, the antibody
is a monoclonal
antibody or an antigen binding fragment thereof, such as scFv. Fab. Fab' and
F(ab')2. In one
specific embodiment, the antibody is MAb E9 or D10 produced by the cell line
deposited
under CGMCC No. 2903 or 2904, respectively.
According to this invention, the antibody specifically binds to the
polypeptide present in
plasma. Preferably, the antibody specifically binds to a phosphorylated form
of the
polypeptide of the invention, said phosphorylated form of the polypeptide
contains one or
more phosphorylated amino acid residues in the amino acid sequence of SEQ ID
No.1
selected from the group consisting of Thr90, Ser231, Ser263, Tyr309 and a
combination
thereof.. Preferably, the antibody specifically binds to the polypeptide which
is
phosphorylated at Thr90.
In another aspect, this invention relates to a method of preventing or
treating cancer
metastasis in a subject comprising the step of administering an inhibitor of
the polypeptide of
the invention to the subject. In one embodiment of this aspect, the inhibitor
is a specific
22031285.2
5
CA 02714880 2010-09-15
antibody against a polypeptide comprising or consisting of the amino acid
sequence of SEQ
ID No.l. Preferably, this antibody is humanized antibody or an antigen binding
fragment
thereof. In one embodiment, the antibody specifically binds to a
phosphorylated form of the
polypeptide, said phosphorylated form of the polypeptide contains one or more
phosphorylated amino acid residues in the amino acid sequence of SEQ ID No.1
selected
from the group consisting of Thr90, Ser231, Ser263, Tyr309 and a combination
thereof.. In a
preferred embodiment, the antibody specifically binds to the polypeptide which
is
phosphorylated at Thr90. In one specific embodiment, the antibody is MAb E9 or
D10
produced by the cell line deposited under CGMCC No. 2903 or 2904,
respectively.
Among the different aspects of this invention, the cancer is selected from the
group
consisting of lung cancer, liver cancer, gastric cancer, gastrointestinal
cancer, esophagus
cancer, osteosarcoma, pancreatic cancer, lymphoma, colorectal cancer, breast
cancer, prostate
cancer, oral cancer, nasopharyngeal cancer, cervical cancer, ovarian cancer,
leukemia,
malignant melanoma, sarcoma, renal cell carcinoma and cholangiocarcinoma.
This invention also involves antibodies that can specifically bind to the
polypeptide of
the invention which is present in plasma. In one specific embodiment, the
antibody is MAb
E9 or D10 produced by the cell line deposited under CGMCC No. 2903 or 2904,
respectively.
Preferably, the antibody of the invention is a humanized antibody or an
antigen binding
fragment thereof. In one embodiment, the antibody specifically binds to a
phosphorylated
form of the polypeptide, said phosphorylated form of the polypeptide contains
one or more
phosphorylated amino acid residues in the amino acid sequence of SEQ ID No.1
selected
from the group consisting of Thr90, Ser231, Ser263, Tyr309 and a combination
thereof. In a
preferred embodiment, the antibody specifically binds to the polypeptide which
is
phosphorylated at Thr90.
In another aspect, this invention relates to a method of inhibiting cancer
invasiveness and
metastasis, comprising the step of inhibiting the phosphorylation of
intracellular Hsp90a in
cancer cells. In one embodiment, this invention relates to a method of
inhibiting cancer
invasiveness and metastasis, comprising the step of inhibiting the
phosphorylation of Hsp90a
at Thr90 in cancer cells. In one specific embodiment of this aspect, the
method comprises a
step of overexpressing a nucleic acid molecule encoding protein phosphatase 5
in cancer cells.
22031285.2
6
CA 02714880 2010-09-15
Brief Description of the Drawings
Figure 1: The plasma Hsp90a level in tumor-bearing mice is significantly
increased than
in that in normal mice.
Figure 2: The plasma Hsp90a level in cancer patients is significantly
increased compared
with that in normal people.
Figure 3: The titer assay of the mouse monoclonal antibody of E9 and D 10.
Figure 4: The standard curve of plasma Hsp90a using mouse monoclonal antibody
E9
and rabbit polyclonal antibody S2 (sandwich ELISA).
Figure 5: Quantitative measurement of plasma Hsp90a level in lung, liver,
pancreatic
cancer patients as well as benign galactoma and myoma of uterus patients using
sandwich
ELISA:
A: The plasma Hsp90a level of liver cancer patients is measured by sandwich
ELISA.
The plasma Hsp90a level in benign tumor patients ranges from 2-10 ng/ml,
mostly 2-5 ng/ml,
while the plasma Hsp90a level in 69% (20/29) liver cancer patients is above 50
ng/ml, the
average of which has more than 10-fold increment compared with that of benign
tumor
patients, which is consistent with the result of Western blotting. This
indicates that the plasma
Hsp90a level is positively correlated with tumor malignancy.
B: The plasma Hsp90a level of lung cancer patients is measured by sandwich
ELISA.
The plasma Hsp90a level in 64% (9/14) lung cancer patients is above 50 ng/ml,
the average
of which has more than 10-fold increment compared with that of benign tumor
patients. This
indicates the plasma Hsp90a level is positively correlated with tumor
malignancy.
C: The plasma Hsp90a level of breast cancer patients is measured by sandwich
ELISA.
Compared with benign tumor patients, the highest plasma Hsp90a level is
increased by more
than 5-fold. Overall, the average plasma Hsp90a level in breast cancer
patients shows
significantly increment when compared with that in benign tumor patients.
D: The plasma Hsp90a level of pancreatic cancer patients is measured by
sandwich
ELISA. The plasma Hsp90a level in 100% (10/10) pancreatic cancer patients is
above 50
ng/ml, the average of which has more than 10-fold increment compared with that
of benign
tumor patients. This indicates the plasma Hsp90a level is positively
correlated with tumor
malignancy.
22031285.2
7
CA 02714880 2010-09-15
Figure 6: Quantitative measurement of plasma Hsp90a level in the cancer
patients with
or without metastasis using sandwich ELISA:
A: The liver cancer patients are divided into two groups, one with metastasis
and the
other one without metastasis, then the plasma Hsp90a levels in these two
groups are
compared. The plasma Hsp90a levels in cancer patients with metastasis are all
above 200
ng/ml while those in patients without metastasis range from 50-200 ng/ml.
B: The lung cancer patients are divided into two groups, one with metastasis
and the
other one without metastasis, then the plasma Hsp90a levels in these two
groups are
compared. The plasma Hsp90a levels in cancer patients with metastasis are all
above 200
ng/ml while those in patients without metastasis range from 50-200 ng/ml.
C: The breast cancer patients are divided into two groups, one with metastasis
and the
other one without metastasis, then the plasma Hsp90a levels in these two
groups are
compared. The plasma Hsp90a levels in cancer patients with metastasis are
significantly
elevated compared with that in patients without metastasis.
Figure 7: Quantitative measurement of plasma Hsp90a level in the patients with
inflammation (pneumonia and hepatitis), normal people and tumor patients using
sandwich
ELISA:
A: To ensure that the elevated plasma Hsp90a level of cancer patients is tumor
specific,
we compared the plasma Hsp90a level in pneumonia patients, normal people and
tumor
patients and found that the plasma Hsp90a level in pneumonia patients ranges
from 2-10
ng/ml, indicating no significant changes compared with that in normal people.
B: The plasma Hsp90a level in hepatitis patients (Hepatitis A and B) ranges
from 2-10
ng/ml, indicating no significant changes compared with that in normal people.
Figure 8: The plasma Hsp90a is secreted by tumor cells.
Figure 9: The secreted Hsp90a by tumor cells is in a C-terminal truncated
form.
Figure 10: The plasma Hsp90a lacks the C-terminal four amino acid residues.
Figure 11: The plasma Hsp90a is phosphorylated.
Figure 12: The Thr90 phosphorylated Hsp90a level in the tumor patient plasma
is
elevated.
Figure 13: In the plasma of tumor patient, the increment of Thr90
phosphorylated
Hsp90a level is consistent with that of Hsp90a level.
22031285.2
8
CA 02714880 2010-09-15
Figure 14: Thr90 Phosphorylation of Hsp90a is a pre-requisite for Hsp90a
secretion.
Figure 15: PP5 dephosphorylates Thr9O phosphorylated f-Isp90a.
A: Purified PP5 and Thr90 phosphorylated Hsp90a (pT90-Hsp9Oa) are incubated
together, and the released free phosphate group is quantitatified. Peptide is
a positive control.
The result shows PP5 can directly dephosphorylate Thr90 phosphorylated Hsp90a.
B: In the human breast cancer cell line MCF-7, overexpression of human PP5
results in
the inhibition of Hsp90a Thr90 phosphorylation (0.55 of that in control group)
and siRNA of
endogenous human PP5 results in the increase of Hsp90a Thr90 phosphorylation
(1.58 of that
in control group)
Figure 16: PP5 regulates the secretion of Hsp90a.
A: In the human breast cancer cell line MCF-7, overexpression of human PP5
results in
the inhibition of Hsp90a secretion.
B: In the human breast cancer cell line MCF-7, siRNA of endogenous human PP5
results
in the increase of Hsp90a secretion.
Figure 17: The correlation of PP5 expression level and the secretion level of
Hsp90a.
Figure 18: The correlation of PP5 expression level and the invasive ability of
the tumor
cells.
Figure 19: The specific antibody of Hsp90a could inhibit the migration of
tumor cells.
Figure 20: The specific antibody of Hsp90a could inhibit tumor metastasis.
Information of Biological Material Deposition
Mouse hybridoma cell line SP2/0-Agl4 which produces monoclonal antibody E9 was
deposited at China General Microbiological Culture Collection Center (CGMCC,
Chinese
Academy of Sciences Institute of Microbiology, Datun Road, Chaoyang District,
Beijing) on
February 24, 2009 with the Deposition No. CGMCC No.2903.
Mouse hybridoma cell line SP2/0-Ag14 which produces monoclonal antibody D10
was
deposited at China General Microbiological Culture Collection Center (CGMCC,
Chinese
Academy of Sciences Institute of Microbiology, Datun Road, Chaoyang District,
Beijing) on
February 24, 2009 with the Deposition No. CGMCC No.2904.
Detailed Description of the Invention
22031285.2
9
CA 02714880 2010-09-15
Carcinogenesis is caused by changes of certain intracellular signal
transduction
pathways, accompanied by changes of protein expression, modification and
distribution.
These changes could be used to monitoring of tumor development and
progression, and these
proteins are named as tumor marker. With the development of proteomics
technology, it
becomes possible to monitor the changes of tumor proteome qualitatively or
quantitatively. So
many new tumor markers have been found to provide more accurate and credible
evidences
for the tumor clinical diagnosis and prognosis.
This invention is based on a discovery of a novel blood tumor marker, i.e.,
plasma
Hsp90a. Compared to intracellular Hsp90a (the amino acid sequence is SEQ ID
No.3, coding
nucleic acid sequence is SEQ ID No.4), the Hsp90a in plasma has a deletion of
4 amino acids
in the C terminus.
Hence, in one aspect, this invention provides an isolated polypeptide, which
is Hsp90a
found in serum or plasma. In this application, the term "polypeptide of the
invention" used
herein refers to Hsp90a found in serum or plasma, which comprises or consists
of the amino
acid sequence of SEQ ID No.1. Preferably, the term "polypeptide of the
invention" used
herein refers to the polypeptide consisting of the amino acid sequence of SEQ
ID No.l
sequence. In this application, the term "Hsp90a in plasma" or "Hsp90a in
serum" can be used
equally to refer to the protein Hsp90a in the blood but not intracellular or
on cell surface. In
this application, term "polypeptide" and "protein" can be used
interchangeably.
The present invention also provides a polynucleotide encoding a polypeptide
composing
or consisting of the amino acid sequence of SEQ ID No.1. In a specific
embodiment, the
polynucleotide comprises or consists of the amino acid sequence of SEQ ID
No.2.
The inventors also found that the polypeptide of the invention is in a
phosphorylated
form in plasma. Specifically, one or more amino acid residues in the amino
acid sequence of
SEQ ID No.1 selected from the group consisting of Thr90, Ser231, Ser263,
Tyr309 and a
combination thereof are phosphorylated. Preferably, the Thr90 in the
polypeptide of this
invention is phosphorylated.
The special form of plasma Hsp90a(C-terminal truncated and phosphorylated) has
not
been described. Meanwhile, plasma Hsp90a also has never been reported to be
related to
tumor development and progression. EP1457499A1 describes the extracellular
form of
Hsp90a, and suggests that the inhibitors of Hsp90a can be used to treat tumor
metastasis, to
22031285.2
CA 02714880 2010-09-15
detect tumor invasiveness and to determine the dependence of tumor
invasiveness on Hsp90a.
However, the method described in EP1457499A1 is used to detect cell surface
Hsp90a, but
not related to plasma Hsp90a. Nor did EP1457499A1 disclose that plasma Hsp90a
can be
used to determine the degree and stage of malignancy, or to monitor the
therapeutic response
and prognosis of cancer by measuring the level of Hsp90a in plasma.
WO/2008/070472
reported that the effects of the anti-cancer treatment targeting Hsp90a could
be determined
through detecting Hsp90a and its related factors in plasma, but it did not
mention Hsp90a as
an independent tumor marker in cancer diagnosis and prognosis.
The inventors examined the blood samples from nearly a hundred tumor patients
(with
breast cancer, liver cancer, pancreatic cancer, lung cancer and so on), and
found that the level
of Hsp90a in plasma is correlated to tumor malignancy, particularly
metastasis; while the
inflammatory response does not influence plasma level of Hsp90a. Therefore,
plasma Hsp90a
can serve as a tumor marker, which can be used for the diagnosis and prognosis
of tumors and
metastasis.
Additonally, the invention provides a kit examining the plasma levels of the
polypeptide
of the invention. The kit of the invention contains an agent capable of
specifically binding to
the polypeptide of the invention, which can be used to detect the level of
plasma Hsp90a.
The invention also provides a method of detecting the level of plasma Hsp90a
in a
plasma sample of the subject by using an agent capable of specifically binding
to said
polypeptide. The method of the invention can be used to establish the
diagnosis of
carcinogenesis, malignancy and metastasis; to screening cancer in high-risk
population; to
determine the prognosis of cancer patients; and to determine the efficacy of
surgery,
radiotherapy or drug therapy.
As used herein, the term "an agent capable of specifically binding to a
polypeptide of the
invention" refers to molecules which can bind to the polypeptide of this
invention with
high-affinity. These agents also include molecules which can bind
intracellular and cell
surface Hsp90a. An agent capable of specifically binding to the polypeptide of
the invention
can be proteins, especially Hsp90a specific antibodies. In the preferred
examples, the above
antibody is a monoclonal antibody or antigen binding fragment, such as scFv,
Fab, Fab 'and F
(ab') 2. In a specific embodiment, the antibodies are monoclonal antibody E9
or D10 which is
produced by the cell line with deposition number of CGMCC No. 2903 or 2904,
respectively.
22031285.2
11
CA 02714880 2010-09-15
Monoclonal antibodies are obtained by screening the cells which can secret the
antibodies and then culturing the cells in vitro. The method is well known by
the people in
this field (Kohler G & Milstein C. (1975) Nature. 256, 495-7). The process for
Hsp90a-specific monoclonal antibody preparation is as follows: for the first
immunization,
recombinant human Hsp90a (rhHsp90a, 100 g) with Freund's complete adjuvant
was
injected subcutaneously on the back of BALB/c mice multi-pointly; 3 weeks
later, same dose
of rhHsp90a was injected intraperitoneal (i.p.) with Freund's incomplete
adjuvant; the third
immunization was administered by i.p. injection after 3 weeks with same dose
(5 to 7 days
later the blood titer was tested); the other 3 weeks later, 200 g rhHsp90a
was administered
for the booster immunization by intraperitoneal injection. 3 days later,
spleen cells were fused
with hybridoma SP2/0-Ag14 (SP2/0) (Source: ATCC, number: CRL-1581) using HAT
for
screening. Then hybridoma cells were limited diluted. immune blot and ELISA
methods were
used to identify and eventually to select cell lines that can secrete specific
Hsp90a antibodies.
According to the invention, the antibodies used for the preparation of the
method or kit
can specifically bind to Hsp90a, and preferably to Hsp90a in plasma. In an
preferred
embodiment. the antibodies specifically bind to Hsp90a with one or more
following amino
acid residues are phosphorylated: Thr90, Ser23 1, Ser263, Tyr309 and the
combination
thereof. In the preferred embodiment, antibodies described herein can
specifically bind to
Hsp90a with phosphorylated Thr90.
The invention also provides a method for detecting the level of plasma
polypeptide. The
level of plasma Hsp90a can be detected by any suitable methods. The methods
described here
include both direct and indirect means for detecting the polypeptide which can
be used for the
diagnosis of tumor development, malignancy and metastasis.
Direct measurements include methods of detecting the described polypeptide of
the
invention using its an agent capable of specifically binding to said
polypeptide, for example
using specific antibodies of the polypeptides by Western blot or ELISA.
The concentration of Hsp90a can also be indirectly determined by detecting the
activity
of Hsp90u. An example is the assay of thermal induced denaturation of
luciferase, which can
be used to detect the chaperone activity of Hsp90a (Johnson et al. (2000) J.
Biol. Chem., 275,
32499-32507).
Preferably, the level of plasma Hsp90a can be detected by ELISA or Western
blot,
22031285.2
12
CA 02714880 2010-09-15
comprising the steps as follows:
a) collecting the whole blood of cancer patients, and obtaining plasma or
serum by
centrifugation;
b) using ELISA or Western blot to detect the Hsp90a level of plasma or serum
obtained
from the step a), in which plasma of healthy people is used as the negative
control,
while plasma of patients with malignant tumors is used as the positive
control,
optionally generating an Hsp90a concentration standard curve;
c) determining the tumor malignancy and stage according to the level of plasma
Hsp90a, thereby determining tumor diagnosis, prognosis or efficacy of
treatment.
In step b) other methods can also be used, such as methods based on antigen-
antibody
reactions as well as methods based on other principles directly or indirectly
reflecting the
concentrations of Hsp90a, for example the examination of the concentration of
Hsp90a by
detecting the activity of Hsp90a.
Standard Hsp90a used for the standard curve of ELISA is purified from the
plasma of
cancer patients and can also be obtained by recombinant construction,
including the full
length Hsp90a, fragments, and other recombinant proteins or conjugates
containing the
sequence of the Hsp90a. "the standard curve of Hsp90a concentration" refers to
the
correlation curve between Hsp90a concentration and absorbance values detected
in the
ELISA using standard Hsp90a samples. "Hsp90a standard" refers to the plasma
Hsp90a
protein, recombinant Hsp90a protein, fragments and derivatives with the purity
of more than
95%.
"Determining the malignancy of tumor" refers to making a judgment of tumor
malignancy by examining the Hsp90a concentration in the plasma samples of
patients and
comparing this value with negative and positive controls.
Both sandwich and competitive ELISA can be used to detect the level of Hsp90a
in
plasma. The competitive ELISA with higher sensitivity is preferred.
The general steps of Sandwich ELISA include: a) immobilizing a specific
antibody onto
a solid support, and removing the unbound antibodies and impurities by one or
more washing
steps; b) adding samples to be tested and incubating for a period of time,
allowing the antigen
in the samples to bind to the antibody immobilized on the solid support and
form solid-phase
antigen-antibody immune complexes, then removing the unbound substances; c)
adding an
22031285.2
13
CA 02714880 2010-09-15
enzyme-conjugated-antibody which can bind to the antigen in the solid phase
immune
complexes, and then removing the unbound enzyme-conjugated-antibody by
thorough washes
(the amounts of enzyme-conjugated-antibody bound to the solid support are
positively
correlated to the amount of antigen); and d) adding substrate and quantifying
the amount of
antigen according to the color reaction.
The general steps of competitive ELISA include: a) immobilizing a specific
antibody
onto a solid support and washing thoroughly; b) adding the mixture of the
sample to be tested
and a certain amount of enzyme-conjugated-antigen, and incubating for enough
time and
wash thoroughly. (If there is no antigen in the sample, the enzyme-conjugated-
antigen can
bind to the solid phase antibody without competition. If there are antigens in
the sample, then
the antigens in the sample compete with the enzyme-labeled-antigens to bind
the immobilized
antibody, and the amount of enzyme-labeled-antigen bound to immobilized
antibody decreases. As a control, the reference tubes only contain the enzyme-
labeled
antigen); c) adding substrate and obtaining the absorbance value for each
tube. The value of
the reference tube is highest, and the difference between the reference tube
and the sample
tube represents the amount of antigens in the sample. The lighter the color
is, the more
antigens the sample contains.
If Sandwich ELISA is used, the antibodies are two different species of plasma
Hsp90a
specific antibodies. The immobilized antibody can be a rabbit polyclonal
antibody with strong
binding capacities, while the detection antibody is a mouse monoclonal
antibody with high
specificity. These two antibodies should have no cross-reaction.
If competitive ELISA is used, the antibodies should be the specific antibodies
against
plasma Hsp90a, which should have strong binding capacities and high
specificity to both the
competitve substance and Hsp90a.
If competitive ELISA is used, the competitive substances can be a labled
plasma Hsp90a
standard protein. And the labeling does not interfere with the binding of
Hsp90a standard
protein to its antibody.
If Western blot method is used, the first antibody used is a specific antibody
against
plasma Hsp90a; the secondary antibody can be conjugated with horseradish
peroxidase, or
alkaline phosphotase; the substrate can be DAB or fluorescent substrate. The
fluorescence
substrate with high sensitivity is preferred.
22031285.2
14
CA 02714880 2010-09-15
The sensitivity of above methods should be 10 ng/ml or more sensitive.
The antibody specific for plasma Hsp90ct used in this present invention can be
a whole
antibody, or its fragments and derivatives.
The antibodies of this invention can also be replaced by other agents capable
of
specifically binding to Hsp90a, wherein the agents include small molecule
compounds,
peptides and their derivatives.
The competitive plasma Hsp90a standard proteins can be labeled by biotin and
various
fluorescent labeling reagents. Biotin is preferred.
The cancers which can be detected by the kit or method described in this
invention
include, but are not limited to, lung cancer, liver cancer, gastric cancer,
esophageal cancer,
osteosarcoma, pancreatic cancer, lymphoma, colon cancer, breast cancer,
prostate cancer, oral
cancer, nasopharyngeal cancer, cervical cancer, leukemia, malignant melanoma,
sarcoma,
renal cancer, cholangiocarcinoma. "Tumor and its malignancy" refers to whether
the tumor is
benign, malignant, or metastasis.
The inventors demonstrated that, the plasma Hsp90a level in non-cancer human
is 2-50
ng/ml, most in the range of 2-10 ng/ml. The level of plasma Hsp90a in patients
diagnosed
with cancer is higher than the normal level, while the plasma Hsp90a level in
the patients
having cancer metastasis is higher than 50 ng/ml, mostly higher than 200
ng/ml. This makes
the plasma Hsp90aa new tumor marker, which would be helpful for the diagnosis
of cancer,
especially metastasis.
Thus, in an embodiment, the kit or method of this invention can be used to
determine the
existence of tumor, especially malignant tumors and tumor metastasis. To this
end, the kit or
method of this invention can be used to measure the Hsp90a level of plasma
samples from
potential tumor patients or potential tumor metastasis patients, and
optionally compared with
normal controls, and then determine the possibility of patients with tumor or
tumor metastasis
according to the Hsp90a level in the sample. The elevated Hsp90a level in
plasma suggests
higher possibility that the patients have suffered from malignant tumor, and
for patients with
known cancer, an elevated Hsp90a level strongly suggests possibility of tumor
metastasis.
In another embodiment, the kit or method of the invention can be used to
detect the level
of plasma Hsp90a so as to screen high-risk populations. To this end, the kit
or method of the
invention can be used to detect plasma Hsp90ct level in samples from high-risk
populations,
22031285.2
CA 02714880 2010-09-15
and optionally compared with normal controls, and then one can determine
whether an
individual within the population may have tumor according to the level of the
sample.
Elevated Hsp90a level indicates the higher possibility of the presence of a
malignant tumor. It
is well-known for the people in this field how to determine different high-
risk groups
depending on the type of cancer to be screened, individual factors such as
age, family history,
lifestyle, work environment, history of exposure to harmful compounds and so
on. For
example, patients having chronic hepatitis B or hepatitis C are at high risk
of HCC.
In another embodiment, the kit or method of the invention can be used to
detect the level
of plasma Hsp90a to predict the prognosis of cancer patients. To this end, the
kit or method of
this invention can be used to detect plasma Hsp90a level in samples from
cancer patients, and
optionally compared with normal controls, and then one can determine the
prognosis of
cancer patients according to the level of Hsp90a in plasma samples. The
maintenance of
Hsp90a in a high level or further increase may relate to the poor prognosis.
Therefore, the
clinicians can be alerted to provide more closely observation to the patients,
and if necessary,
change the current treatment.
In another embodiment, the kit or method of the invention can be used to
detect the
plasma level of Hsp90a, which is thus used to determine whether the treatment
such as the
surgery, radiotherapy or chemotherapy on cancer patients is effective. To this
end, the kit or
method of this invention can be used to detect plasma Hsp90a level in samples
from cancer
patients, and optionally compared with normal controls, and then one can
determine the
prognosis of cancer patients according to the level of Hsp90a in plasma
samples. According
to the level of the Hsp90a, one can determine whether the surgery,
radiotherapy or
chemotherapy on cancer patients is effective and/or whether the treatment
should be
continued.
The inventors further demonstrated that the Hsp90a in plasma is secreted by
tumor cells
and is different from that within the tumor cells. Therefore, it is assumed
that the tumor
development and metastasis may be suppressed by inhibiting the secretion of
Hsp90a into
plasma. This point was demonstrated by the mouse tumor metastasis model using
the specific
antibody against plasma Hsp90a. Therefore, the secretion of Hsp90a can be used
as a new
target for screening new anticancer drugs.
In another aspect, the invention provides a method for the preventing or
treating tumor
22031285.2
16
CA 02714880 2010-09-15
metastasis comprising administering to a cancer patient an inhibitor of the
polypeptide of the
invention.
According to one embodiment of this invention, the aforementioned inhibitor is
a
specific antibody of the polypeptide. Preferably, the antibody is a humanized
antibody or its
fragments. In one embodiment, these antibodies specifically bind to a
phosphorylated form of
the polypeptide of the invention, said phosphorylated form of the polypeptide
contains one or
more phosphorylated amino acid residues in the amino acid sequence of SEQ ID
No.1
selected from the group consisting of Thr90, Ser231, Ser263, Tyr309 and a
combination
thereof. In a preferred embodiment, the antibody can bind to the polypeptide
phosphorylated
at Thr90. In one specific embodiment, the antibody is MAb E9 or D10 which is
produced by
the cell line deposited under CGMCC No. 2903 or 2904, respectively. As
demonstrated
herein, these antibodies can completely inhibit the lymph node metastasis of
tumors in mouse
model and can inhibit the lung metastasis to an extent up to 56%.
So, in another aspect, this invention also relates to an antibody specifically
binds to the
Hsp90a in blood. In a specific embodiment, the antibody is MAb E9 or D 10
which is
produced by the cell line deposited under CGMCC No. 2903 or 2904,
respectively.
Preferably, the antibody is a humanized antibody or an antigen binding
fragment thereof. In a
specific embodiment, the antibody specifically binds to a phosphorylated form
of the
polypeptide of the invention, said phosphorylated form of the polypeptide
contains one or
more phosphorylated amino acid residues in the amino acid sequence of SEQ ID
No.1
selected from the group consisting of Thr90, Ser231, Ser263, Tyr309 and a
combination
thereof. In one preferred embodiment, the antibody specifically binds to the
polypeptide
phosphorylated at Thr90. Preferably, the antibody inhibits tumor growth,
especially
metastasis. This invention also related to a conjugate comprising the antibody
and a diagnosis
or treatment moiety. For example, the diagnosis moiety is a fluorophore, while
the treatment
moiety is a chemotherapeutic agent.
The inventor also found the amount of Hsp90a secreted by tumor cells is
related to the
level of Protein phosphatase 5 (PP5). In benign tumor, secreted Hsp90a is low
but PP5
expression level is high. In malignant tumor, secreted Hsp90a is high but PP5
expression
level is low. Therefore, the level of secreted Hsp90a is negatively
interrelated with the level of
PP5. As a result, by detecting the expression level of PP5, it is possible to
predict the amount
22031285.2
17
CA 02714880 2010-09-15
of secreted Hsp90a.
The inventors also found that the secretion of Hsp90a can be inhibited by
cellular PP5.
Over-expression of a nucleic acid molecule encoding PP5 in the cell can
inhibit Hsp90a
secretion, while a decrease in the expression level of PP5 will result in an
increase in Hsp90a
secretion. So by over expressing PP5, Hsp90a secretion can be inhibited and
the tumor
progression and metastasis may be inhibited. According to our experiment, we
found that over
expression of PP5 inhibits the metastasis of MCF-7. So PP5 can be a new
therapeutic target
for tumor treatment.
So, in a further aspect, this invention provides a method for inhibiting tumor
invasiveness
and metastasis, wherein a step of inhibiting the phosphorylation of Hsp90a
within tumor cells.
In one embodiment, this method comprises a step of inhibiting Thr90
phosphorylation of
Hsp90a in tumor cells. In one specific embodiment, this method comprises a
step of
over-expressing a nucleic acid molecule encoding PP5 in tumor cells.
Preferably, the
over-expression of PP5 is achieved by the means of gene introduction. In one
embodiment,
this method comprises a step of over-expressing a nucleic acid molecule
encoding PP5 having
an amino acid sequence of SEQ ID No.5. In one specific embodiment, the nucleic
acid
molecule comprises the nucleotide sequence of SEQ ID No.6. This invention also
provides a
method of inhibiting the phosphorylation of Hsp90a in tumor cells comprising a
step of
over-expressing PP5 in tumor cells using a vector carrying a polynucleotide
encoding PP5
operably linked to a promoter. The method can be used to inhibit tumor
invasiveness and
metastasis.
This invention also provides methods and models for screening for anti-tumor
drugs by
using plasma Hsp90a and its derivatives. Such anti-tumor drugs includes, but
not limited to,
plasma Hsp90a binding proteins, small peptides, and small compounds aswell as
inhibitors
which can suppress the activity of plasma Hsp90a.
Examples
Example 1: The collection and preparation of mouse plasma samples, and the
detection of
plasma Hsp90a.
Balb/c mice with average body weight of 20 gram (purchased from Beijing Vital
River
Laboratory Animal Technology Co., Ltd.) were randomized divided into two
groups, 3 for
22031285.2
18
CA 02714880 2010-09-15
each. The mice of experiment group were inoculated with 106 H22 (CCTCC ID:
GDC091)
cells, while the mice in control group were not inoculated. When the diameter
of tumor
reached about 2 cm (about 20 days), blood was collected from eye down
veniplex.
Anticoagulant was added and hemolysis was prevented. If hemolysis occurred,
the blood
would be recollected. The blood sample was centrifuged at 4 C with 6000g for
twice, and
supernatant was saved. The amount of plasma Hsp90a was detected by Western
blotting using
Rabbit anti-human Hsp90a pAb (Labvasion). BCA method was used to determine the
total
amount of proteins in samples, which ensured that the loading amounts of
different samples
were equal. The result is shown in Fig. 1. Compared to control mice, the
plasma Hsp90a was
elevated in tumor bearing mice.
Example 2: The collection and preparation of plasma samples from normal people
and tumor
patients, and the detection of plasma Hsp90a.
The blood of normal people and cancer patients were collected and delivered to
lab
within 24 hours at 4C. If hemolysis occurred, the samples should be
recollected. The samples
were centrifuged at 4 C, 6000g twice and supernatant was saved. The amount of
Hsp90a was
determined by Western blotting. If the samples could not be examined
immediately, the
samples should be stored at -80 C. By comparing the results with clinical
diagnosis, the
correlation between Hsp90a and tumor malignancy was confirmed.
The protocol for Western blotting: the plasma samples were mixed 1:1 with
loading
buffer. 1-2 l sample is loaded for SDS-PAGE. Primary antibody is the one
which can
specifically recognize plasma Hsp90a(rat mAb SPA-840, Stressgen). The
aecondary antibody
is goat anti-rat antibody conjugated with HRP (purchased from
Zhongshanjingiao). As the
result shows in Fig.2, the amount of Hsp90a in HCC patients is 10-fold higher
than that of
normal people (A), while it is elevated by 2-fold in benign galactocele and
hysteromyoma
patients compared with that of normal people.
Example 3:Preparation of rabbit pAb and mouse mAb against Hsp90a
Primers with the following sequences were used to clone the gene of Hsp90a:
Hsp90a-Sall-Re: ACGCGTCGACTTAGTCTACTTCTTCCATGC (SEQ ID No.8) and
Hsp90a-Sphl-For: ACATGCATGCATGCCTGAGGAAAC CCAGACC (SEQ ID No.9).
22031285.2
19
CA 02714880 2010-09-15
Primers were synthesized by Invitrogen. Pfu DNA polymerases (NEB) were used to
amplify
Hsp90a from human liver cDNA library (Stratagene). Sphl and Sall (NEB) were
used to
digest the amplified PCR products and pQE80L vector (Qiagen). Then T4 ligases
were used
for ligation. The products were transformed into Top 10 cells (Transgen) for
amplification and
validation. The validated plasmids were further transformed into BL21DE3 cells
(Transgen)
for expression. The recombinant Hsp90a proteins were purified by ion-exchange
chromatography SP HP, pH6.8, collecting 10 ms/ml peak and Q HP, pH7.8,
collecting 19
ms/ml peak.
Recombinant human Hsp90a with a purity higher than 95% was used to immune
adult
male New Zealand Rabbit by dorsal subcutaneous multi-point injection at 100 gg
for each
time. Two weeks later, the secondary immunization was conducted according to
the same
method, except that the amount of Hsp90a injection was reduced to 50 g. Boost
injections
were conducted every 1 week after the secondary immunization for twice. The
titer of
antibody in serum was determined 7-10 days after the boost immunizations.
Eight days after
the last immunization, serum was collected and stored at -20 C. Affinity
chromatography
conjugated with antigen was used to purify the antibody from the serum.
Purified rabbit pAb
was named as S2.
BALB/C mice were immunized by recombinant human Hsp90a. Primary immunity: 100
g antigen and Freund's complete adjuvant was injected dorsal subcutaneously
multipoint.
The second immunity was conducted 3 weeks later with Freund's incomplete
adjuvant using
the same dose and i.p. injection. The third immunization was operated 3 more
weeks later
without adjuvant (blood was collected for test after 5-7 days). Three more
weeks later, 200 g
antigen was injected i.p. as the boost immunization. 3 days later, spleen
cells were collected
and fused with SP2/0-Ag14(SP2/0) hybridomas (ATCC: CRL-1581). HAT was used for
the
screening. Limited dilution was used to colonize hybridomas. Western blotting
and ELISA
were used for identification. Finally, E9 and D 10, which secrete specific
antibody against
Hsp90a, were obtained and stored as CGMCC No.2903 and 2904 on February 24,
2009.
Indirect ELISA was used to determine the titers of E9 and D10. Shown as Fig.3,
the
average titers of E9 and D10 respectively reaches 500,000, which is qualified
to be used in the
detection of plasma Hsp90a. Indirect ELISA: recombinant human Hsp90a was
plated at 4 C
overnight with the concentration of 10 g/ml. Then the plate was blocked at 37
C for 1 hour.
22031285.2
CA 02714880 2010-09-15
E9 or D10 with 1:400, 1:1600, 1:6400, 1:25600, 1:102400, 1:509600 dilutions
was added
and incubated for 2 hours at RT. Then goat anti-mouse antibody conjugated with
HRP was
added and incubated for 1 hour at room temperature, then o-phenylendiamine was
added and
absorbance at OD 490 nm was detected.
Example 4: Measurement of the concentration of Hsp90a by E9, S2 and sandwich
ELISA
In the method of sandwich ELISA to test the concentration of plasma Hsp90a,
two
antibodies from distinct species were used. S2 with high binding capacity (the
preparation of
S2 was described in example 3) was used as the coating antibody and E9 (the
preparation of
E9 was described in example 3) with high binding specificity was used as the
detecting
antibody. There is no cross reaction between these two antibodies. The method
is repeatable
and sensitive. Fig.4 shows that the sensitivity of this method is 5 ng/ml.
Example 5: The collection and preparation of human plasma, the detection of
plasma Hsp90a
and determination of tumor malignancy (sandwich ELISA).
The blood of normal persons, cancer patients and inflammation patients was
delivered to
lab at low , temperature within 24 hours. If hemolysis occurrs, the blood
should be
re-collected. The samples were centrifuged at 4 C, 6000g twice and supernatant
was saved.
The amount of plasma Hsp90a was determined by ELISA. The samples were stored
at -80 C
if they were not examined immediately. By comparing the results with clinical
diagnosis, the
correlation between Hsp90a and tumor malignancy was confirmed.
Two different Hsp90a antibodies were used in sandwich ELISA. S2 was coated on
the
plate and incubated overnight at 4C. The plate was blocked for 1 hour at 37 C.
The samples
were diluted by 10 folds and were added to the plate (100 p1/well). After 2
hours incubation at
37 C, E9 was added and incubated at 37 C for 2 more hours. Goat anti-mouse
antibody
conjugated with HRP was added to incubate for another 1 hour. O-
phenylendiamine was
added and absorbance at OD490 nm was detected. The results were shown as
Fig.5, 6 and 7.
The standard curve comprises a serial of samples containing the gradient
concentration of
standard Hsp90a proteins and 10% of negative plasma in each sample, which was
used to
exclude the background of plasma.
As shown in Fig.5, the amount of plasma Hsp90a in benign tumor patients
(including
22031285.2
21
CA 02714880 2010-09-15
benign galactocele and hysterromyoma patients, seven cases) is between 2-10
ng/ml, mainly
in 2-5 ng/ml. 69% (20/29) liver cancer patients have plasma Hsp90a above 50
ng/ml, which is
in average 10 folds higher than that of benign tumor patients, p=0.00263,
student t test (A);
64% (9/14) lung cancer patients have plasma Hsp90a above 50 ng/ml. The average
concentration is 10 folds higher than that of benign tumor patients, p=0.0497,
student t test
(B). 78% (25/32) breast cancer patients have plasma Hsp90a above 50 ng/ml. The
average is
folds higher than that of benign tumor patients, p<0.00I; student t test (C).
100% (10/10)
pancreatic cancer patients have plasma Hsp90a above 50 ng/ml. The average is
10 folds
higher than that of benign tumor patients, p<0.05, student t test (D).
10 As shown in Fig.6, in liver cancer (17 cases, 7 metastasis) (A), lung
cancer (10 cases, 2
metastasis) (B) and breast cancer (21 cases, 10 metastasis) (C) patients, the
levels of Hsp90a
in metastasis patients were significantly higher than those in patients
without metastasis.
Liver cancer p=0.003, breast cancer p=0.002, student t test.
As shown in Fig.7, inflammation patients, including 10 pneumonia cases (A), 10
hepatitis cases (type A 5 cases, type B 5 cases) (B) have plasma Hsp90a
between 2-10 ng/ml,
which shows no significant difference when compared with that of normal
persons (3 cases),
p=0.2988, 0.5177, 0.138 by student t test, respectively.
The normal people's samples were collected from healthy volunteers, while
tumor
patients and inflammation patients' samples were collected from Beijing Cancer
Hospital and
Xiamen First Hospital.
Example 6: The plasma Hsp90a was secreted by tumor cells
Nude mice (purchased from Beijing Vital River Laboratory Animal Technology
Co.,
Ltd.) with average body weight of 20 g were divided into two groups, 6 mice
for each group.
Each mouse was injected with 106 Hela cells (ATCC: CCL-2). Control group were
normal
mice without tumor. When the diameter of tumor reached 2 cm (about 20 days),
blood was
collected from eye veniplex. Hsp90a antibody (rat mAb, Stressgen) that
specifically
recognizes human Hsp90a but not mouse Hsp90a was used to detect the plasma
Hsp90a. As
shown in Fig.8, the plasma Hsp90a in the tumor-bearing mice is specifically
recognized by
human but not mouse Hsp90a antibody, which indicates that the plasma Hsp90a
was secreted
by the xenograft tumor cells.
2203128.2
22
CA 02714880 2010-09-15
Example 7: The Hsp90a secreted by rumor cells was C-terminally truncated.
Using Hsp90a-pc3. I -Nhe 1-For-Myc: GCTAGCTAGCGCCACCATGGA
ACAAAAACTCATCTCAGAAGAGGATCTGCCTGAGGAAACCCAGACCCAAGAC
(SEQ ID No.10) and Hsp90a-pc3.I-Xhol-Re-nostop: CCCGCTCGA
GTGTCTACTTCTTCCATGCGTGATG (SEQ ID No.12) as primers (synthesized by
Invitrogen) and Pfu DNA polymerase (NEB) to amplify Hsp90a from the template
of
pQE80L-Hsp9Oa (obtained in example 3). The products were digested and inserted
into
pcDNA3.l/Myc-His(-) (Invitrogen) to obtain Hsp90a with Myc tag at the N
terminus, which
was named as Myc-H. Similarly, His-Myc-H with N-terminal tandem His-Myc tags
was
amplified by Hsp90a-pc3. I -Nhe I -For-His-Myc:
GCTAGCTAGCGCCACCATGCATCATCATCATCATCATGAACAAAAACTCATCTCAG
AAGAGGAI'CTGCCTGAGGAAACCCAGACCCAAGAC (SEQ ID No.11) and SEQ ID
No.12 primers. These two plasmids were transiently transfected into MCF-7 to
observe the
secretion of exogenous Hsp90a. Antibodies against Hsp90a, Myc and His were
used to detect
the secreted Hsp90a. The result shows that the secreted Hsp90a was C-
terminally truncated
(Fig.9A).
Using Myc-His-H as a template, the mutations of the final four amino acids
(EEVD) of
Hsp90a C-terminus was constructed. EE-> AA represents the mutation of two EE
to two Ala
(using SEQ ID No.11 and Hsp90a-EE- AA:
GGCCGCTCGAGTGTCTACTGCTGCCATGCGTGATGTG (SEQ ID No.13) as primers),
VD-> AA means that VD were mutated to two Ala (using SEQ ID Noll and
Hsp90a-VD-AA: GGCCGCTCGA GTTGCTGCTTCTTCCATGCGTGATGTG (SEQ ID
No.14) as primers). All Ala represents the mutation of EEVD to four Ala (using
SEQ ID
No.11 and Hsp90a-EEVD-AAAA: GGCCGCTCGAGTTGCTGCTGCTGCCATGCG
TGATGTG (SEQ ID No .15) as primers), CM4 represents the deletion of last EEVD
four
amino acids (using SEQ ID No.I1 and Hsp90a-CA4-Xho:
CCGCTCGAGTCATGCGTGATGTGTCGTCATCTC (SEQ ID No.16) as primers). Human
breast cancer cell line MCF-7 was transiently transfected with these types of
mutants to
observe the secretion of exogenous Hsp90a (over-expressed Hsp90a, which is
different from
the endogenous one). The antibody against Hsp90a was used to detect the
changes of secreted
22031285.2
23
CA 02714880 2010-09-15
Hsp90ain the extracellular medium. The results show that four C-terminal amino
acid
residues regulate the secretion of Hsp90a. Any site-directed mutation or
deletion of these four
amino acid residues can lead to the secretion of Hsp90a without C-terminal
truncation, which
suggests that four C-terminal amino acid residues EEVD are deleted in the
secreted
extracellular Hsp90a (Figure 9B).
Example 8: Examination of the existing form of Hsp90a in human plasma
We collected whole blood samples of liver cancer patients, which were
centrifuged twice
within 24 hours after collection, then detected Hsp90a in plasma using the
method of
immunoprecipitation and immunoblotting. First, specific rabbit polyclonal
antibody against
Hsp90a (Source: Labvision) was used for immunoprecipitation, and then a rabbit
polyclonal
antibody which can specifically recognize the Hsp90a C-terminal four amino
acid residues
EEVD (lab stock, antigen used for immunizationused is a carrier protein
coupled with
three-repeated peptides of EEVD, synthesized from the SBS Genetech Co,. Ltd.)
was used to
detect Hsp90a in plasma. As shown in Figure 10, the EEVD antibody specifically
recognize
Hsp90a from whole cell lysate, but does not recognize Hsp90a in plasma, which
indicates
that intracellular Hsp90a is different from plasma Hsp90a, which lacks the
four C-terminal
amino acid residues EEVD. (Figure 10).
Example 9: Detection of the phosphorylated form of Hsp90a in plasma
Whole blood samples of the liver cancer patients were centrifuged twice to
extract the
plasma within 24 hours after collection. Then Hsp90a in plasma was detected
using the
method of immunoprecipitation and immunoblotting. First, specific rabbit
polyclonal
antibody against Hsp90a (Source: Labvision) was used to immunprecipitate
plasma Hsp90a,
then an antibody (Rabbit anti-phospho-(Ser/Thr) PKA substrate pAb, Cell
signaling) which
can specifically recognize the Thr9O phosphorylated Hsp90 was used to detect
the Thr90
phosphorylation status of plasma Hsp90. As shown in Figure 11, the Hsp90a in
plasma is
Thr90 phosphorylated.
Example 10: Detect the concentration of Thr90 phosphorylated Hsp90 in plasma
Whole blood samples of both liver cancer patients and normal people were
centrifuged
22031285.2
24
CA 02714880 2010-09-15
twice to extract the plasma within 24 hours after collection . The relative
level of Hsp90a in
the plasma was detected using the method of sandwich ELISA. Protocol: firstly,
using
self-made rabbit polycolonal antibody S2 to coat the plate overnight at 4^,
then added
10-fold diluted plasma samples at 100 pl per well. After incubation at 37 ^
for 2 hours,
antibody which can specifically recognize the Thr90 phosphorylated Hsp90a
(cell signal) was
added. The plate is incubated at 37 ^ for 2 hours; and then horseradish
peroxidase conjugated
goat anti-rabbit secondary antibodies were added and incubated for 1 hour.
Finally
o-phenylenediamine was added to detect the absorption at OD490nm. The results
show that
the levels of Hsp90a in liver cancer patients are higher than those of normal
people, P=0.003,
Student t test, which indicates that the level of Thr90 phosphorylated Hsp90
is increased in
liver cancer patients (Figure 12).
Example 11: Detection of the consistency of the Thr90 phosphorylated Hsp90
level in plasma
and the total Hsp90a content.
The levels of the Thr90 phosphorylated Hsp90 and the total amount of Hsp90a in
liver
cancer patients plasma (8 cases) were detected. The method to detect the total
amount of
plasma Hsp90a was the same as that used in example 5; the method to detect the
level of
Thr90 phosphorylated Hsp90 was the same as that used in example 10. The
results show that
the level of Thr90 phosphorylated plasma Hsp90 is consistent with the total
amount of plasma
Hsp90a, which further indicates that the phosphorylation of plasma Hsp90a is
on Thr90, and
the increment of the total amount of Hsp90a can represent the increment of
Thr90
phosphorylated Hsp90a (Figure 13).
Example 12: The phosphorylation of Thr90 is necessary for the secretion of
Hsp90a
Using pcDNA3.l-Myc-His-Hsp90a plasmid as the template (Also known as wild-type
Hsp90a (WT Hsp90ct)), mutant Hsp90o. (T90A, threonine mutated to alanine) was
constructed by quickchange PCR using the following primers:
Hsp90a-T89A-Sense:GATCGAACTCTTGCAATTGTGGATACTGGAATTGGAATG(SEQI
DNo.17) and Hsp90a-T89-AntiSense: CATTCCAATTCCAGTATCCACAATTGCAAGAGT
TCGATC (SEQ ID No.18). T90A Hsp90a mutant can not be phosphorylated at Thr90.
Human
breast cancer cell line MCF-7 (purchased from ATCC, No. HTB-22) was
transfected with
22031285.2
CA 02714880 2010-09-15
wild-type Hsp90a (WT) or mutant Hsp90a (T90A). The medium was then collected
and the
secretion of exogenous Hsp90a was detected by anti-Hsp90a antibody.
The results show that overexpressed exogenous wild-type Hsp90a can be detected
in the
extracellular medium, while the T90A mutant can not be detected, which
indicates that
phosphorylation at residue Thr9O is a prerequisite for Hsp90a
secretion.(Figure 14).
Example 13: PP5 is responsible for the dephosphorylation of pT90-Hsp9Oa
The preparation of the Thr90 phosphorylated Hsp90a (pT90-Hsp9Oa): The
recombinant
human Hsp90a protein and recombinant protein kinase A (Promega Corporation
USA) were
incubated in a reaction buffer (NEB UK company) at 30 C for 1 h, then pT90-
Hsp9Oa protein
was purified. After removing free phosphate by dialysis , the purified pT90-
Hsp9Oa protein
mixed with recombinant human PP5 protein were incubated at 30 C, the free
phosphate
released from the pT90-Hsp90a was detected using the non-radioactive
serine/threonine
phosphatase assay kit (Promega Corporation USA). Peptide substrate is a
composition of the
kit and was used as the positive control. As shown in Figure 15A, when PP5 was
incubated
with the peptide substrate, the release of free phosphate was significantly
increased, P value
<0.005, Student t test, indicating that PP5 can directly dephosphorylate the
peptide substrate
(positive control). However, when PP5 was incubated with pT90-Hsp9Oa protein,
the release
of free phosphate was also significantly increased, P value <0.005, Student t
test, indicating
that PP5 can directly dephosphorylate pT90-Hsp9Oa.
The nucleotide sequence of PP5 (SEQ ID No.6) was amplified from human liver
eDNA
library and constructed into the pcDNA3.1/Myc-His (-) (vector source:
Invitrogen). The PP5
vector was transfected into and overecpressed in human breast cancer cell line
MCF-7. On the
other side, using the RNA interference technology, the expression of PP5 can
also be knocked
down using the siRNA with the sequence of 5'-ACTCGAACACCTCGCTAAAGAGCTC-3
'(SEQ ID No.7) (synthesized by Invitrogen). Then the status of Hsp90a Thr90
phosphorylation was examined with the overexpression or knock down of PP5. As
shown in
Figure 15B, with overexpression of human PP5, the Thr90 phosphorylated Hsp90a
(pT90-Hsp9Oa) was significantly reduced (0.55 of the control). When the
expression of
human PP5 was suppressed, the Thr90 phosphorylated Hsp90a (pT90-Hsp9Oa) was
significantly increased (1.58 of the control).
22031285.2
26
CA 02714880 2010-09-15
Example 14: Regulation of the secretion of Hsp90a by promoting or inhibiting
the expression
of PP5
PP5 can dephosphorylate Thr90 phosphorylated Hsp90a. Primers as follows were
used
to amplify PP5 from human liver cDNA library: PP5-Nhel-For:
CTAGCTAGCATGTACCCATACGACGTCCCAGACTACGCT (SEQ ID No.19) and
PP5-XhoI-Re: CCGCTCGAGTTAATGATGATGATGATG ATGCACGTGTACC (SEQ ID
No.20). The full length human PP5 was cloned into pcDNA3.1/Myc-His (-) (vector
source:
Invitrogen). Human breast cancer cell line MCF-7 was transfected with PP5
vector, then the
secretion of Hsp90a from the cells was examined. The results showed that after
overexpression of human PP5, the secretion of Hsp90a was significantly
decreased (Figure
16A).
On the other side, when the expression of human PP5 in MCF-7 cells was knocked
down
by RNA interference (against human PP5, Invitrogen), the secretion of Hsp90a
was
significantly increase (Figure 16B).
Example 15: The level of PP5 and the tumor malignancy
The relationship between the expression level of PP5 and the secretion of
Hsp90a was
examined in human breast cancer cell lines MCF-7, SKBR3, MDA-MB-453, 435s and
231
(ATCC, number, respectively HTB-22, -30, -131, -129, and HTB-26) using the
method of
Western blotting. MCF-7, SKBR3 breast cancer cell lines are less malignant
cell lines. In the
nude mice tumor model, these two cell lines can form primary tumors, but do
not metastasize.
MDA-MB-453, 435s and 231 are more malignant,They can not only form primary
tumors,
but also metastasize to distant organs in the nude mice tumor models. Thus MDA-
MB-435s
and 231 are often used to establish tumor metastasis models. In Figure 17,
secretion levels of
Hsp90a by these five breast cancer cell lines correlate with their malignancy.
The results show that the cells, which express high level of PP5, secrete low
level of
Hsp90a; whereas cells with low expression of PP5 can secrete more Hsp90a
(Figure 17).
Meanwhile, the results also show that the level of secreted Hsp90a is
positively correlated,
whereas the expression level of PP5 is negtively correlated with the tumor
malignancy (Figure
17). The level of secreted Hsp90a and its regulatory factors such as PP5 can
be used to
22031285.2
27
CA 02714880 2010-09-15
determine the tumor malignancy.
Example 16: The expression level of PP5 is linked to tumor invasiveness
The wound healing model was employed to examine the relationship between the
expression level of PP5 and tumor cell migration.
Human breast cancer cell line MCF-7 was transfected with human PP5 vector or
PP5
siRNA. Then the cells were inoculated into 12-well plate. When the cells grew
to confluence,
pippete tips were used to scrape cells to form a "wound". The scraped cells
were washed
away, and the rest part of cells were cultured in fresh DMEM medium (GIBCO) at
37 ^ in an
incubator with 5% of CO2. The images of the "wound" at the time of 0 h, 12 h,
and 24 h were
captured (Figure 18A). The effect of PP5 expression on cell migration was
examined by
analyzing the "wound" healing rate. The results show that overexpression of
PP5 inhibits
MCF-7 cell migration, while PP5 siRNA can promote MCF-7 cell migration (Figure
18B).
Example 17: The effect of plasma Hsp90a specific antibodies on tumor cell
migration
The wound healing model was used to detect the effect of plasma Hsp90a
specific
antibody on tumor cell migration.
MCF-7 or MDA-MB-231 cells (ATCC, No. HTB-22, respectively, and HTB-26) were
inoculated into 12-well plate. When the cells grow to confluence, pippete tips
were used to
scrape cells to form a "wound". The scraped cells were washed away, and the
rest part of cells
were moved into fresh DMEM medium (GIBCO). Meanwhile, E9 (20 g/ml), the
specific
mouse monoclonal antibody against Hsp90a, or control IgG (20 gg/ml ) was
added. Then the
plate was incubated at 37 ^, with 5% CO2. Images of the "wound" after 0 h, 6
h, 12 h, 24 h,
48 h, and 72 h of incubation were captured (Figure 18A). The effect of plasma
Hsp90 specific
antibodies on tumor cells migration was examined by monitoring the "wound"
healing rate.
As shown in Figure 19, the specific antibody against plasma Hsp90a can
significantly inhibit
the migration of both MDA-MB-231 (Figure 19A) and MCF-7 (Figure 19B)
(Inhibitory
activity>40%).
Example 18: Detect the effect of plasma Hsp90a specific antibody on tumor
metastasis
Nude mice with an average body weight of 20 g were purchased from Beijing
Vital River
22031285.2
28
CA 02714880 2010-09-15
Laboratory Animal Technology Co., Ltd. B16/F10 mouse melanoma cells (ATCC,
number:
CRL-6475) (2 10 5) were injected into the mice via tail vein. Next day the
mice were
randomly divided into two groups (n = 8): the control group (IgG was
administered) and the
Hsp90a Ab (mouse monoclonal antibody E9) treated group. The antibodies were
administered
once every other day with a dosage of 40 g/mouse/time. Metastasis was
detected 15 days
after inoculation. As shown in Figure 20, the specific antibody against plasma
Hsp90a can
completely inhibit the lymph node metastasis of B16/F10 cells (A), and 56% of
the lung
metastasis can be inhibited (B).
22031285.2
29
