ORIGINAL ARTICLE—ALIMENTARY TRACT

Identification of two new HLA-A*0201-restricted cytotoxic Tlymphocyte epitopes from colorectal carcinoma-associatedantigen PLAC1/CP1

Fangfang Liu • Henghui Zhang • Danhua Shen •

Shan Wang • Yingjiang Ye • Hongsong Chen •

Xuewen Pang • Qiujing Song • Peiying He

Received: 11 September 2012 / Accepted: 2 April 2013

� Springer Japan 2013

Abstract

Background To explore the potential application of pla-

centa-specific PLAC1/Cancer Placenta (CP) 1 antigen for

immunotherapy in CRC patients, further identification of

the cytotoxic T lymphocyte epitopes from this antigen is

necessary.

Methods We assessed the protein expression of PLAC1/

CP1 using a tissue chip and immunochemistry staining in

CRC samples. Simultaneously, we predicted four PLAC1/

CP1-derived HLA-A*0201-restricted peptides by using

reverse immunology methods. Peptide-specific CD8? T

cell responses were assessed by an IFN-c release ELISPOT

assay. Effector CD8? T cells lyse HLA-A*0201 CRC cell

line SW620 was detected in a granzyme-B release ELI-

SPOT cytotoxicity assay.

Results Our results indicated that PLAC1/CP1 was highly

expressed in 56.7 % (55/97) of adenocarcinomas. PLAC1/

CP1 protein expression was associated with CRC tumor

differentiation, the tumor/node/metastasis stage, and lymph

node metastasis. Two of four peptides showed high affini-

ties in an HLA-A2 binding assay. In 66.7 % (6/9) of

peripheral blood mononuclear cells of CRC samples with

PLAC1/CP1 protein-positive expression, these two pep-

tides, PLAC1/CP1 p41-50 (FMLNNDVCV) and PLAC1/

CP1 p69-77 (HAYQFTYRV), were immunogenic in the

induction of peptide-specific CD8? T cell responses as

assessed by an IFN-c release ELISPOT assay. Furthermore,

the generated effector CD8? T cells could specifically lyse

the PLAC1/CP1 HLA-A*0201 CRC cell line SW620 in a

granzyme-B release ELISPOT cytotoxicity assay.

Conclusions These results show that the PLAC1/CP1

antigen is a possible prognostic marker of CRC and that

PLAC1/CP1 p41-50 and PLAC1/CP1 p69-77 are novel

HLA-A*0201-restricted CD8? T cell epitopes and potential

targets for peptide-based immunotherapy in CRC patients.

Keywords PLAC1/CP1 protein � Protein expression �Colorectal carcinoma � Peptides � Immunotherapy target �Tumor marker

Introduction

In recent decades, some human tumor-associated antigens

have been identified as vaccine candidates for tumor

immunotherapy. Immunotherapeutic approaches have been

attempted based on using CD8? cytotoxic T cells (CTLs)

to kill the targeted tumor cells [1, 2]. In some clinical trials,

a positive correlation between specific CD8? T cell

responses and clinical response has been observed in a

F. Liu and H. Zhang contributed equally to this paper.

F. Liu (&) � D. Shen � Q. Song

Department of Pathology, Peking University People’s Hospital,

Beijing 100044, People’s Republic of China

e-mail: liufangfang@pkuph.edu.cn

H. Zhang � H. Chen

Department of Hepatology Institute, Peking University People’s

Hospital, Beijing 100044, People’s Republic of China

S. Wang � Y. Ye

Department of Gastroenterological Surgery, Peking University

People’s Hospital, Beijing 100044, People’s Republic of China

X. Pang

Department of Immunology, Peking University Health Science

Center, Beijing 100083, People’s Republic of China

P. He

Department of Central Laboratoty, Peking University People’s

Hospital, Beijing 100044, People’s Republic of China

123

J Gastroenterol

DOI 10.1007/s00535-013-0811-4

subgroup of patients, resulting in partial or complete tumor

regression by using tumor antigenic peptide-based vaccines

[3–5]. Identifying more HLA class I restricted epitopes

from tumor antigens is necessary for the goal of designing

a multivalent vaccine.

Cancer-testis (CT) antigens have been regarded as attrac-

tive vaccine candidates because of their limited expression in

normal tissues (testis and placenta) and for being broadly

expressed in various types of cancers [6]. Especially, the CT

antigen NY-ESO-1 and MAGE family members, such as

MAGE-A1 and -A3, have been tested for antigen-specific

immunotherapy of cancer patients in many clinical trials

[7–9]. Another important role of tumor antigens is as tumor

markers, which can be useful factors for predicting the

occurrence or prognosis of cancer. Many tumor markers exist,

such as the prostate-specific antigen, a marker for prostate

cancer. However, most CT antigens are expressed only in a

minor proportion of CRC patients [10]. Currently, clinical

trials with CRC patients often focus on antigens overexpres-

sed in the tumor or on oncofetal proteins, such as mucin-1,

carcinoembryonic antigen, and the epithelial cell adhesion

molecule, but the objective clinical response rate (complete

plus partial responses) has been less than 1 % [11].

PLAC1/CP 1, a new member of the CT family, encodes a

26-kDa protein primarily localized to both the nucleus and

cytoplasm compartments of trophoblastic cells. We identi-

fied its encoding gene from a human hepatocellular carci-

noma cDNA expression library in our lab, using a

suppression subtractive hybridization method, and its

sequence has been deposited in GenBank (GenBank

Accession NM_021796). Despite the apparent restriction of

PLAC1/CP1 to the placenta in normal tissues, we have

observed a wide expression of PLAC1/CP1 messenger

RNA in various types of human cancers including colo-

rectal carcinoma (CRC) and lung, liver, and gastric cancers.

In our previous study, we demonstrated that the PLAC1/

CP1 protein is a potent tumor antigen capable of inducing

both CD8? and CD4? T cell responses in vitro from

peripheral blood mononuclear cells (PBMCs) obtained

from CRC patients; the patients with a positive immune

response had a longer survival, suggesting that the PLAC1/

CP1 might be a useful target for CRC immunotherapy [12].

In this study, we explored the protein expression of

PLAC1/CP1 in CRC tumor tissues and identified new HLA-

A*0201-restricted T cell epitopes in PLAC1/CP1 protein.

Materials and methods

CRC patients and samples

The analyses were based on clinical samples obtained from

97 CRC patients who underwent surgical resection of the

tumor at Peking University People’s Hospital, in 2002–2007.

Among these patients, 56 were male, 41 were female. The

median age was 65.5 ± 13.0 (range 42–88 years). Most of

the cases (44) were stage TNM I, with 8, 30, and 15 cases of

stages II, III and IV, respectively. The median size of the

tumors was 4.9 ± 2.2 cm in diameter (they varied from 2 to

15 cm). Histopathologic examination indicated that 17 were

well differentiated, 61 moderately differentiated, and 19

poorly differentiated. Sample collection was approved by the

Hospital Ethics Review Committee and occurred after

obtaining informed consent.

Immunohistochemistry

Tissue sections were deparaffinized and rehydrated with

xylene and a series of grades of alcohol. After epitope retrieval

and inactivation of endogenous peroxidase, sections were

blocked with 10 % normal goat serum for 30 min and

sequentially incubated with the anti-CP1 Ab (afforded by

professor Michael Fant, Department of Pediatrics, University

of Texas Health Science Center, Houston, TX, USA) and

horseradish peroxidase-conjugated anti-rabbit IgG. After a

washing, slides were developed by adding the chromogen

diaminobenzidine tetrahydrochloride, followed by counter-

staining with hematoxylin. Sections of the normal human

placenta were used as a positive control. The rabbit serum

collected before immunization was used as a negative control.

Blood samples and cell lines

Blood samples from 17 HLA-A2? CRC patients were

collected with informed consent and peripheral blood

mononuclear cells (PBMCs) were isolated by standard

Ficoll-Hypaque density gradient centrifugation, then, fro-

zen in liquid nitrogen until use. The HLA-A2 subtypes of

each sample were determined by polymerase chain reac-

tion-sequence specific primer and sequence-based typing as

described previously [12].

Human TAP-deficient T2 cell line and human colon

cancer cell line SW620 were obtained from the American

Type Culture Collection and the Cell Institute (Shanghai,

China), respectively. All cell lines were maintained in

complete RPMI-1640 medium containing 10 % FCS,

2 mM glutamine and 100 units/mL penicillin/streptomycin.

Epitopes prediction and peptides synthesis

The combination of three computer programs available via

the internet were used to predict HLA-A*0201 epitopes from

PLAC1/CP1 protein. These were the predictive algorithm:

National Center for Biotechnology Information (http://bimas.

dcrt.nih.gov/molbio/hlabind/), The University of Tubingen

(http://www.uni-tuebingen.de/uni/kxi/), The University of

J Gastroenterol

123

Oklahoma Health Sciences Center (http://hlaligand.ouhsc.

edu/prediction.htm). Peptides, with more than 95 % purity,

were synthesized using solid phase techniques, purified by

reversed phase-high performance liquid chromatography

(HPLC) (Saibaisheng Genetech Company, Beijing, China).

The HLA-A*0201-restricted flu matrix peptide p58-66

(GILGFVFTL) and HBV peptide (p18–27, FLPSDFFPSV)

were used as positive and negative control, respectively.

T2 binding assays

The binding affinity of each PLAC1/CP1 peptide and posi-

tive control peptide to HLA-A*0201 molecules was deter-

mined as described previously [10]. Briefly, T2 cells were

incubated overnight with peptide (50 lg/mL) in serum-free

medium supplemented with human b2-microglobulin (b2-

M; 5 mg/mL; Sigma). Cells were washed, stained with anti-

HLA-A2 mAb for 30 min at 4 �C, and then analyzed by flow

cytometry (FACS Calibur, Becton-Dickinson). The fluo-

rescence index (FI) was calculated by the following formula:

FI = [mean fluorescence intensity (MFI) sample - MFI

background]/MFI background, where MFI background

represents the value without peptide. Samples were mea-

sured in duplicate and mean FI was calculated.

HLA-A typing

Genomic DNA was extracted from peripheral blood cells

with the PUREGENE Genomic DNA Purification Kit

(Gentra Systems, Minneapolis, MN, USA) according to the

manufacturer’s instructions. DNA typing of the HLA-A

locus was first performed using the polymerase chain

reaction (PCR)-sequence specific primer (SSP) method

(Biotest HLA SSP Reagents Kit, Frankfurt am Main,

Germany), and HLA-A*02 subtyping involved the PCR-

sequence based typing (SBT) method and the detailed

procedure of HLA-A*02 subtyping involved the PCR-SBT

method was as previously described [12].

In vitro induction of CTLs with peptides from PBMCs

of CRC patients

For antigen presentation, the autologous DCs were incu-

bated with a mixture of peptides (10 lg/mL of each pep-

tide) for 2 h. CD8? T cells were isolated by positive

selection with anti-CD8-beads (Stem cell technologies,

Hangzhou, China). After washing, CD8? T cells were co-

cultured with PLAC1/CP1 peptides-pulsed Daces at a ratio

of 5:1 with the same medium and procedure as described

above. Seven days later, the cultured T cells were reticu-

lated with freshly prepared peptide-pulsed DCs and cul-

tured for another 7 days. After two consecutive rounds of

stimulation, the peptide-specific CTLs in cultures were

assessed by functional assays.

Interferon-c enzyme-linked immunospot assay

The IFN-c release ELISPOT assay was performed as described

previously [10] using a commercial kit (BD company, Amer-

ica). Briefly, plates (BD Company, America) were coated

overnight at 4 �C with 100 lL/well anti-IFN-c capture Ab and

washed six times with PBS. After blocking with 10 % human

AB serum RPMI 1640, stimulator cells (peptides-pulsed T2

cells) (2.5 9 104 cells per well) were added together with

responder T cells (2.5 9 104 cells per well) to each well and

incubated for 20 h at 37 �C, the cells were then removed, and

the plates were incubated with a second (biotinylated) Ab to

human IFN-c and streptavidin–alkaline phosphatase. The

plates were developed with substrate at room temperature in the

dark and the reaction was stopped by rinsing the plates with

distilled water. The membranes were air dried and the spots

were counted using the BD ELISOT reader system. In the

ELISPOT assay, the response was scored as positive where the

number of spots in the presence of antigen was at least 2-fold

higher than that in the negative control [13].

Cytotoxicity assay

Granzyme-B release ELISPOT assay

Effectors CD8? T cells (2.5 9 104/well) were added to trip-

licate wells precoated with anti-human granzyme-B (GrB)

capture Abs (GB-10), together with SW620 or SW480 cells

(2.5 9 104/well). After coculture at 37 �C for 4 h, plates were

developed with a biotinylated anti-human GrB detecting Ab

(GB-11) and streptavidin–alkaline phosphatase phosphatase.

Spots were visualized and enumerated using the Champ Spot

II ELISOT reader system (Sage Creation, Beijing, China).

Statistical analysis

The correlation of CP1 protein expression and the clinical

pathologic factors of the CRC patients were analyzed by

the v2 test and logistic analysis with SPSS 16.0 software

(SPSS, Inc, Chicago, IL, USA). A two-tailed P value\0.05

was considered significant.

Results

The protein expression of PLAC1/CP1 in CRC tissue

samples

The protein expression of PLAC1/CP1 was studied using

immunochemistry staining methods in a CRC tissue chip.

J Gastroenterol

123

This protein was expressed in 56.7 % (55/97 cases) of

colorectal adenocarcinoma tissue samples. The protein

expression of PLAC1/CP1 in poorly differentiated colo-

rectal primary adenocarcinomas was significantly higher at

(78.9 %) than in well-differentiated (35.3 %) and moder-

ately differentiated (55.7 %) adenocarcinomas (P \ 0.05).

PLAC1/CP1 protein expression was also significantly

greater in stage TNM III ? IV samples (71.2 %) than in

stage TNM I ? II samples (40 %) (P \ 0.05). In addition,

expression of PLAC1/CP1 was more common in tumors

with lymph node metastasis (69.6 %) than in tumors

without lymph nodes metastasis (45.1 %) (P \ 0.05), and

the expression level was higher in the primary lesion with

increasing numbers of lymph nodes involved. No correla-

tion was found between a positive PLAC1/CP1 protein

expression and clinical-pathologic parameters such as age,

sex, and tumor size. Protein expression was separately

localized in the nucleus or cytoplasm or occurred in both

(Fig. 1a–c). Separately, 27.8 % (27/97) of PLAC1/CP1

protein was expressed in the nucleus and 43.3 % (42/97) in

the cytoplasm. In further studies, these significant results

depended specifically on nuclear expression (Table 1). Of

interest, the positive rate of nuclear protein expression of

PLAC1/CP1 in women (41.5 %) was greater than in men

(17.9 %) (P \ 0.05) (Table 1). Protein expression in the

cytoplasm was not related to these clinical-pathologic

factors (data not shown).

Epitopes and T2 binding assay

To screen for the HLA-A*0201-restricted epitopes from

the PLAC1/CP1 protein, we used three computer programs

Fig. 1 Immunohistochemistry staining of CP1 proteins expressed in

CRC tumor tissues. Typical results are shown. PLAC1/CP1 protein

immunohistochemistry reveals strong nuclear and cytoplasmic stain-

ing in various differentiated colorectal adenocarcinoma tissues.

a Arrow indicates nuclear expression in poorly differentiated CRC.

b Arrow indicates both nuclear and cytoplasmic expression in

moderately differentiated CRC. c Arrow indicates cytoplasmic

expression in well-differentiated CRC (SP 9200)

Table 1 Relationship between

the nuclear expression of

PLAC1/CP1 protein and

clinico-pathologic

characteristics of 97 CRC

patients

Characteristics No. of patients No. of CP1

positive cases (%)

v2 value P value P value

(logistic analysis)

Gender

Male 56 10 (17.9) 6.567 0.010

Female 41 17 (41.5)

Histological differentiation

Well 17 2 (11.8) 8.321 0.016 0.013

Moderately 61 15 (24.6)

Poorly 19 10 (52.6)

TNM classification

Stage I ? II 45 3 (6.7) 18.726 0.000 0.000

Stage III ? IV 52 24 (46.2)

Lymph node metastasis

No 51 5 (9.8) 17.407 0.000 0.000

Yes 46 22 (47.8)

Numbers of lymph node metastasis

N0 51 5 (9.8) 22.632 0.000 0.000

N1 28 10 (35.7)

N2 18 12 (66.7)

J Gastroenterol

123

(BIMAS, SYFPEITHI, and HLALIGAND) to predict

binding affinity and stability. This analysis revealed eight

potential epitopes with relatively higher predicted binding

scores among the different programs. Of these, we had

previously studied four peptides [12]. The remaining four

peptides were synthesized and assessed for HLA-A*0201

binding capacity using T2 binding assays. Flu matrix

peptide p58–66 (GILGFVFTL) served as a positive control

in this assay. The results showed that PLAC1/CP1 peptides

p1 (FMLNNDVCV, p41–50) and p2 (HAYQFTYRV,

p69–77) had high binding affinities for HLA-A*0201

molecules, whereas the other two peptides showed low

affinities (data not shown).

HLA-A typing

We collected 33 case samples of paired cancer tissues and

PBMC samples. Of these, 18 were PLAC1/CP1 protein?,

17 were HLA-A2?, and 9 were PLAC1/CP1 protein? and

HLA-A2?. The clinical-pathological factors of 17 HLA-

A2? CRC patients were shown in Table 2.

Specific CD8? T cell responses to PLAC1/CP1-derived

peptides in patients with CRC

The specific CD8? T cell responses to the PLAC1/CP1-

derived peptides were assessed in 11 HLA-A2? CRC

patients (9 bearing PLAC1/CP1 protein? and 2 bearing

PLAC1/CP1 protein- tumors). These 11 patients with

CRC were stimulated with a mixture of p1

(FMLNNDVCV, p41–50) and p2 (HAYQFTYRV,

p69–77) peptides presented by the respective autologous

DCs. After one round of peptide stimulation, CD8? T

cells were purified with anti-CD8 MACS beads from

PBMC samples and tested for their ability to recognize T2

cells loaded with p1 (FMLNNDVCV, p41–50) and p2

(HAYQFTYRV, p69–77) in the IFN-c release ELISpot

assay. The peptide-specific CD8? T cell responses could

be detected in 6 PBMC samples obtained from PLAC1/

CP1 protein? patients (Fig. 2) and could not be detected

in 2 PBMC samples obtained from PLAC1/CP1 protein-

patients (Table 3).

Cytotoxicity of PLAC1/CP1 peptide-specific effector

CD8? T cells to CRC target cells

PLAC1/CP1 peptide-sensitized CD8? T cells from 3 (3/6,

50 %) patients (RM05, RM12 RM15) were further tested

for their ability to lyse HLA-A2? PLAC1/CP1? CRC cell

lines. The bulk effector CTLs generated from priming with

mixed peptides (p1 and p2) exhibited a high level of tox-

icity against the HLA-A2? PLAC1/CP1 protein? CRC cell

line SW620 but were unable to kill cells of the HLA-A2?

PLAC1/CP1 protein- line SW480 (Table 3; Fig. 3).

Table 2 The clinical–

pathological factors associated

with 17 HLA-A2? CRC patients

Note: Patients with HLA-

A*0201 subtype and A2

supertype (A*0203, A*0206

and A*0207) have been

highlighted in bold

Patients HLA

subtyping

Gender Age

(years)

Histological

differentiation

Size of

tumors (cm)

TNM

classification

Lymph node

metastasis

RM01 A*0204 M 69 Well 5.0 310 ?

RM02 A*0201 M 24 Poorly 3.5 420 ?

RM03 A*0201,

A*0206

F 52 Poorly 7.0 321 ?

RM04 A*0201 F 48 Moderate 5.5 320 ?

RM05 A*0201,

A*0203

F 64 Poorly 5.0 311 ?

RM06 A*0210 M 72 Moderate 5.0 300 -

RM07 A*0211 M 54 Poorly 14.0 320 ?

RM08 A*0201,

A*0207

M 75 Well 5.0 300 -

RM09 A*0208 M 72 Poorly 3.5 320 -

RM10 A*0201 F 55 Well 7.5 411 ?

RM11 A*0207 F 79 Moderate 4.8 410 ?

RM12 A*0201 F 51 Poorly 4.0 320 ?

RM13 A*0211 M 41 Well 7.0 300 -

RM14 A*0203 M 52 Moderate 6.0 300 -

RM15 A*0201 M 49 Poorly 4.0 320 ?

RM16 A*0201,

A*0211

M 79 Moderate 4.8 410 ?

RM17 A*0204 M 44 Moderate 6.0 400 -

J Gastroenterol

123

Discussion

PLAC1/CP1 is a recently described X-linked gene with

expression restricted primarily to cells derived from the

trophoblast lineage during embryonic development. Stud-

ies of the PLAC1 promoter regions indicate that its

expression in both normal placenta and cancer cells is

driven by specific interactions involving a combination of

transcription factors. Using a mutant mouse model, Fant

and her colleagues [14] have sought to demonstrate that

PLAC1/CP1 is necessary for placental and embryonic

development. Liu identified [15] a novel HLA-A2-restric-

ted cytotoxic T lymphocyte epitope from PLAC1 in breast

cancer. Although the exact function of PLAC1/CP1

remains to be defined, results of a recent study suggest that

PLAC1 may participate in the regulation of cell prolifer-

ation and chemotaxis, and preliminary studies indicate that

cancer-derived PLAC1 has the potential to promote tumor

growth and function [16].

In our previous work, we identified PLAC1/CP1 as a novel

tumor antigen associated with CRC and reported a positive

correlation between PLAC1/CP1-specific T cell responses

and patient survival. Here we mainly analyzed the correlation

of PLAC1/CP1 expression and clinical features in a relatively

large cohort of CRC patients. Moreover, we report two new

HLA-A2-restricted CTL epitopes derived from PLAC1/CP1,

which may be useful for the design of PLAC1/CP1-based

peptide vaccines for CRC immunotherapy.

Our results showed that the expression of PLAC1/CP1

protein was associated with CRC tumor differentiation,

tumor/node/metastasis (TNM) stage, and lymph node

metastasis. TNM stage and lymph node metastasis are very

important prognostic factors for tumor patients. We pro-

pose that PLAC1/CP1 will be a promising marker for

Fig. 2 Specific CD8? T cell responses to PLAC1/CP1-derived pep-

tides in CRC patients. CD8? T cells from CRC patients were stimulated

with autologous DCs pulsed with PLAC1/CP1 peptides. IFN-csecretion by these cells was determined using the ELISPOT assay.

Six of nine samples from PLAC1/CP1 protein? patients demonstrated a

positive response for peptide 1, and four of nine samples from PLAC1/

CP1 protein? patients demonstrated a positive response for peptide 2;

DCs pulsed with hepatitis B virus (HBV) peptide were a negative

control while DCs pulsed with Flu peptide were a positive control.

Result shown is from a typical PLAC1/CP1-responsive patient (RM05).

The error bars indicate that the results shown are the standard

deviations of the spot number from duplicate wells

Table 3 Immune response to the CP1 peptides in HLA-A2? CRC patients

Patients CP1 protein HLA-A2 subtyping T-cells response to CP1 peptidesa

IFN-c ELISPOT GrB ELISPOT

Peptide 1 Peptide 2 Peptide 1 Peptide 2

RM02 ? 0201 46 (9) 37 (9) NDb ND

RM03 ? 0201, 0206 11 (18) 18 (18) ND ND

RM04 ? 0201 120 (20) 21 (20) ND ND

RM05 ? 0201, 0203 123 (29) 88 (29) 144 (30) 132 (30)

RM08 - 0201, 0207 16 (11) 19 (11) ND ND

RM10 ? 0201 61 (52) 36 (52) ND ND

RM11 ? 0207 57 (23) 26 (23) ND ND

RM12 ? 0201 116 (15) 79 (15) 145 (22) 112 (22)

RM14 - 0203 53 (33) 47 (33) ND ND

RM15 ? 0201 86 (27) 116 (27) 95 (19) 84 (19)

RM16 ? 0201, 0211 108 (25) 57 (25) ND ND

a CP1-sensitized CD8? T cells were tested for IFN-c or GrB release in coculture with CP1 peptide-pulsed autologous DCs or CP1? SW620

cells. The results shown are the average of the spot number from duplicate wells. Values in parentheses indicate the spot number from controls

without CP1 peptide or CP1- SW480. The response was scored as positive when the number of spots in the presence of antigen was at least

2-fold higher than in the negative control. The positive response is highlighted in boldb Not determined either because there were insufficient cells or the test was not applicable

J Gastroenterol

123

evaluating the prognosis of CRC patients. In this study, we

detected 56.7 % PLAC1/CP1 protein expression in CRC

tissues, including 27.8 % (27/97) in the nucleus and

43.3 % (42/97) in the cytoplasm. This result was similar to

that of our previous study in which PLAC1/CP1 protein

was detected in the nucleus in 28.6 % of CRC tissues. We

also found a small amount of PLAC1/CP1 cytoplasmic

expression in the previous work, but most of it was of weak

staining intensity. After enlarging samples, we found more

cases of cytoplasmic expression. For this reason, we

decided to investigate which expression pattern was asso-

ciated with the clinical–pathologic factors. Although the

detailed biological function of nuclear expression vs

cytoplasmic expression requires investigation, we at least

demonstrate that these significant results depend specifi-

cally on nuclear expression. The findings show (Table 1)

that nuclear PLAC1/CP1 protein expression mainly occurs

in poorly differentiated colorectal adenocarcinomas.

Compared to cytoplasmic expression of PLAC1/CP1,

nuclear expression of PLAC1/CP1 appears to be rather

disadvantageous for antigen presenting and eliciting the

host immune response, which might contribute to the

escape of poorly differentiated colorectal adenocarcinomas

from the anti-tumor immune response. Indeed, peptide

vaccines show potential in the immunotherapy of cancer

for this very reason. The positive rate of nuclear protein

expression of PLAC1/CP1 was higher in women than in

men, suggesting that PLAC1/CP1 may be an especially

important marker for female CRC patients.

Peptide vaccines are very useful in the immunotherapy

of cancer, and the identification of HLA-restricted CTL

peptides is essential for immunotherapy of CRC patients.

Potential CD8? T cell epitopes for tumor vaccines have

been identified using a number of methods, including

eluting peptides from purified HLA of tumor cells [17, 18]

and using expression cloning [19, 20]. An alternative

would be to generate a series of overlapping peptides and

determine either in vitro or in vivo which peptide can

stimulate CTL responses [21]. However, these methods are

relatively laborious and inefficient. Although the poly-

peptide elution method may directly identify the nature of

the antigen peptide, because of the technical difficulty

involved at the instrument level, quantitative analysis of

the distribution of the antigen peptide in the tumor tissue

has so far served to identify only a minority of tumor

antigen peptides in the study group. The specific CTL

epitope clone screening method is mature and allows for

direct identification of CTL recognition of tumor antigen;

the disadvantage, however, is the need to clone CTL that

can kill tumor cells and the tumor cell lines that cannot be

killed by CTL. In addition, the conversion efficiency of

tumor cell-deficient strains is not high enough. This method

is complicated and technically difficult, which is an

obstacle to its widespread use.

The epitope prediction method is usually referred to as

the ‘‘reverse immunology approach’’ for the identification

of whether or not a specifically overexpressed protein in

tumor cells is a tumor antigen. The advantage of this

method is its use as a screening tool of T cells derived from

healthy individuals rather than tumor patients, which could

lead to identification of new CTL epitopes. With the

deepening of cancer genomics research, the epitope pre-

diction method in the screening of tumor antigens has come

into increasingly wide use. We identified the potential

HLA-A*0201-restricted CTL epitopes from the PLAC1/

CP1 antigen using the strategy of reverse immunology. In

our previous study, we characterized the PLAC1/CP1

immune response based on two rounds of protein stimu-

lation by the ELISPOT assay method. Some researchers

have suggested the possibility of producing a naıve T cell

immune response by stimulating more than two rounds. In

this study, we induced these immune responses to pep-

tides based on only one round of stimulation, which

demonstrated directly the immune function of peptides

p1 (FMLNNDVCV, p41–50) and p2 (HAYQFTYRV,

p69–77).

Because HLA-A*0201 is the most frequent HLA-A

allele in Caucasian and Asian individuals, we used epitope

prediction programs to select four PLAC1/CP1-derived

peptides that bind to HLA-A*0201 molecules. These four

peptides were tested for their actual affinity to HLA-

A*0201 molecules according to T2-binding assays, and

two of them (p1 and p2) showed relatively high affinity.

Then, we used a mixture of these peptides to stimulate

Fig. 3 Cytotoxicity assay of PLAC1/CP1 peptide specific effector

CD8? T cells to CRC target cells. PLAC1/CP1 peptides specific

CD8? cells were cocultured with CP1? SW620 or CP1- SW480

cells. GrB release by PLAC1/CP1 peptide specific CD8? T cells was

measured by ELISPOT assay. Of the 3 selected samples, 3 showed

response to SW620 but not SW480; a typical one (RM05) is shown

J Gastroenterol

123

CD8? T cells in vitro to determine whether the peptides

could be naturally processed and presented by antigen-

presenting cells and induced PLAC1/CP1-specific CTLs

in vitro from nine CRC PLAC1/CP1 protein? patients.

ELISpot results showed that T2 cells loaded with p1 and p2

peptide could trigger CD8? T cells to release IFN-c in six

of nine PLAC1/CP1 protein? patients, while no IFN-crelease was observed in PLAC1/CP1 protein- patients.

Furthermore, in an analysis of the specific CD8? T cell

response to PLAC1/CP1-derived peptides p1 and p2 in

CRC patients, three of them showed strong cytotoxic

activity against the HLA-A2? PLAC1/CP1 protein? CRC

cell line SW620 in the granzyme B cytotoxicity assay. This

study showed that the induced CTLs in CRC patients could

kill CRC cell lines in a PLAC1/CP1-antigen-specific way.

In conclusion, our results suggest that the PLAC1/CP1

protein could be a useful CRC marker and that two

novel PLAC1/CP1-derived epitopes, PLAC1/CP1 p41–50

(FMLNNDVCV) and PLAC1/CP1 p69–77 (HAYQF-

TYRV), could serve as good vaccine candidates for pep-

tide-based immunotherapeutic strategies in the treatment of

CRC patients bearing PLAC1/CP1 protein? tumors.

Acknowledgments This article supported by grants from the

National Natural Science Fund of China (No. 30801093).

Conflict of interest All the authors have read and approved the final

manuscript. No potential conflicts do exist.

References

1. Platsoucas CD, Fincke JE, Pappas J, Jung WJ, Heckel M, Sch-

warting R, et al. Immune responses to human tumors: develop-

ment of tumor vaccines. Anticancer Res. 2003;23:1969–96.

2. Boon T, Coulie PG, Van den Eynde BJ, van der Bruggen P.

Human T cell responses against melanoma. Annu Rev Immunol.

2006;24:175–208.

3. Noguchi M, Kobayashi K, Suetsugu N, Tomiyasu K, Suekane S,

Yamada A, et al. Induction of cellular and humoral immune

responses to tumor cells and peptides in HLA-A24 positive

hormone-refractory prostate cancer patients by peptide vaccina-

tion. Prostate. 2003;57:80–92.

4. Uemura H, Fujimoto K, Tanaka M, Yoshikawa M, Hirao Y,

Uejima S, et al. A phase I trial of vaccination of CA9-derived

peptides for HLA-A24-positive patients with cytokine-refractory

metastatic renal cell carcinoma. Clin Cancer Res. 2006;12:

1768–75.

5. Yajima N, Yamanaka R, Mine T, Tsuchiya N, Homma J, Sano M,

et al. Immunologic evaluation of personalized peptide vaccina-

tion for patients with advanced malignant glioma. Clin Cancer

Res. 2005;11:5900–11.

6. Scanlan MJ, Gure AO, Jungbluth AA, Old LJ, Chen YT. Cancer/

testis antigens: an expanding family of targets for cancer

immunotherapy. Immunol Rev. 2002;188:22–32.

7. Chen Q, Jackson H, Parente P, Luke T, Rizkalla M, Tai TY, et al.

Immunodominant CD4? responses identified in a patient vacci-

nated with full-length NY-ESO-1 formulated with ISCOMA-

TRIX adjuvant. Proc Natl Acad Sci USA. 2004;101:9363–8.

8. Davis ID, Chen W, Jackson H, Parente P, Shackleton M, Hopkins

W, et al. Recombinant NY-ESO-1 protein with ISCOMATRIX

adjuvant induces broad integrated antibody and CD4(?) and

CD8(?) T cell responses in humans. Proc Natl Acad Sci USA.

2004;101:10697–702.

9. Van Baren N, Bonnet MC, Dreno B, Khammari A, Dorval T,

Piperno-Neumann S, et al. Tumoral and immunologic response

after vaccination of melanoma patients with an ALVAC virus

encoding MAGE antigens recognized by T cells. J Clin Oncol.

2005;23:9008–21.

10. Scanlan MJ, Simpson AJ, Old LJ. The cancer/testis genes:

review, standardization, and commentary. Cancer Immunity.

2004;4:1.

11. Nagorsen D, Thiel E. Clinical and immunologic responses to

active specific cancer vaccines in human colorectal cancer. Clin

Cancer Res. 2006;12:3064–9.

12. Liu FF, Dong XY, Pang XW, Xing Q, Wang HC, Zhang HG,

et al. The specific immune response to tumor antigen CP1 and its

correlation with improved survival in colon cancer patients.

Gastroenterology. 2008;134:998–1006.

13. Mizukoshi E, Nakamoto Y, Marukawa Y, Arai K, Yamashita T,

Tsuji H, et al. Cytotoxic T cell responses to human telomerase

reverse transcriptase in patients with hepatocellular carcinoma.

Hepatology. 2006;43:1284–94.

14. Jackman SM, Kong X, Fant ME. Plac1 (placenta-specific 1) is

essential for normal placental and embryonic development. Mol

Reprod Dev. 2012;79:564–72.

15. Liu W, Zhai M, Wu Z, Qi Y, Wu Y, Dai C, et al. Identification of

a novel HLA-A2-restricted cytotoxic T lymphocyte epitope from

cancer-testis antigen PLAC1 in breast cancer. Amino Acids.

2012;42:2257–65.

16. Koslowske M, Sahin U, Mitnacht-Kraus R, Seitz G, Huber C,

Tureci O. A placenta specific gene ectopically activated in many

human cancers is essentially involved in malignant cell processes.

Cancer Res. 2007;67:9528–34.

17. Maeurer MJ, Martin D, Elder E, Storkus WJ, Lotze MT. Detec-

tion of naturally processed and HLA-A1-presented melanoma

T-cell epitopes defined by CD8 (?) T-cells’ release of granulo-

cyte-macrophage colony-stimulating factor but not by cytolysis.

Clin Cancer Res. 1996;2:87–95.

18. Nakatsuka K, Sugiyama H, Nakagawa Y, Takahashi H. Purification

of antigenic peptide from murine hepatoma cells recognized by

Class-I major histocompatibility complex molecule-restricted cyto-

toxic T-lymphocytes induced with B7-1-gene-transfected hepatoma

cells. J Hepatol. 1999;30:1119–29.

19. Brichard V, Van Pel A, Wolfel T, Wolfel C, De Plaen E, Lethe B,

et al. The tyrosinase gene codes for an antigen recognized by

autologous cytolytic T lymphocytes on HLA-A2 melanomas.

J Exp Med. 1993;178:489–95.

20. Kawakami Y, Eliyahu S, Delgado CH, Robbins PF, Rivoltini L,

Topalian SL, et al. Cloning of the gene coding for a shared human

melanoma antigen recognized by autologous T cells infiltrating

into tumor. Proc Natl Acad Sci USA. 1994;91:3515–9.

21. Beattie T, Kaul R, Rostron T, Dong T, Easterbrook P, Jaoko W,

et al. Screening for HIV-specific T-cell responses using over-

lapping 15-mer peptide pools or optimized epitopes. AIDS.

2004;18:1595–8.

J Gastroenterol

123