Invest Clin 62(3): 236 - 246, 2021 https://doi.org/10.22209/IC.v62n3a05

Corresponding Author: Qi Xing, Department of Neurological Rehabilitation, Hubei Provincial Maternal and Child

Health Hospital, Tongji Medical College, Huazhong University of Science and Technology, No.745 Wuluo Road

Hongshan District Wuhan, Hubei, China. Tel: 86018963942319. Email: 56052706@qq.com

Hepatocyte Nuclear Factor -1α stimulates

cervical cancer cells to migrate and invade

through regulating pyruvate kinase L/R.

Xiao-Ling Tao

, Wei-Chang Yu

, De-Jun Chen

, Li-Ming Wang

, Lu Liu

and Qi Xing

Department of Gynecology, Hubei Provincial Maternal and Child Health Hospital,

Tongji Medical College, Huazhong University of Science and Technology, Wuhan,

China.

Department of Neurological Rehabilitation, Hubei Provincial Maternal and Child

Health Hospital, Tongji Medical College, Huazhong University of Science and

Technology, Wuhan, China.

Key words: cervical cancer; hepatocyte nuclear factor -1α (HNF-1α); pyruvate kinase

L/R(PKLR).

Abstract. This study was aimed to analyze the role of hepatocyte nuclear

factor -1α (HNF-1α) in regulating migrative and invasive potentials in cervical

cancer via the involvement of pyruvate kinase L/R (PKLR). The expression of

HNF-1α and PKLR in cervical cancer tissues classified by tumor size and FIGO

(Federation International of Gynecology and Obstetrics) stage were detected

by qRT-PCR. The expression correlation between HNF-1α and PKLR in cervi-

cal cancer tissues was analyzed by Pearson correlation test. After intervening

HNF-1α and PKLR levels in SiHa and Hela cells, their migratory and invasive

abilities were examined by the Transwell assay. HNF-1α was upregulated in cer-

vical cancer tissues, particularly those with large tumor size or advanced FIGO

stage. PKLR was highly expressed in cervical cancer tissues as well, presenting

a positive correlation with the HNF-1α level. Knockdown of HNF-1α attenuated

migratory and invasive abilities in SiHa cells, whereas overexpression of HNF-1α

enhanced migratory and invasive abilities in SiHa cells. PKLR was able to abol-

ish the regulatory effects of HNF-1α on cervical cancer metastasis. HNF-1α and

PKLR synergistically promote cervical cancer to migrate and invade.

HNF-1α stimulates migration and invasion in cervical cancer 237

Vol. 62(3): 236 - 246, 2021

El factor nuclear del hepatocito-1α estimula las células de

cáncer cervical para migrar e invadir a través de la regulación

de la piruvato cinasa L/R.

Invest Clin 2021; 62 (3): 236-246

Palabras clave: cáncer de cuello uterino; factor nuclear del hepatocito -1α (HNF-1α);

piruvato quinasa L/R (PKLR).

Resumen. Este estudio tuvo como objetivo analizar el papel del factor

nuclear del hepatocito -1α (HNF-1α) en la regulación de los potenciales migra-

torios e invasivos en el cáncer de cuello uterino a través de la participación de

la piruvato quinasa L/R (PKLR). Se detectó mediante qRT-PCR, la expresión de

HNF-1α y PKLR en tejidos de cáncer de cuello uterino, clasificados de acuerdo

al tamaño del tumor y su estadio FIGO (Federación Internacional de Ginecolo-

gía y Obstetricia). La correlación de expresión entre HNF-1α y PKLR en tejidos

de cáncer de cuello uterino se analizó mediante la prueba de correlación de

Pearson. Después de intervenir los niveles de HNF-1α y PKLR en las células SiHa

y Hela, se examinaron las capacidades migratorias e invasivas mediante el en-

sayo Transwell. HNF-1α se reguló positivamente en tejidos de cáncer de cuello

uterino, en particular aquellos con tumores de gran tamaño o estadio FIGO

avanzado. PKLR también se expresó en gran medida en los tejidos del cáncer de

cuello uterino, presentando una correlación positiva con el nivel de HNF-1α. La

eliminación de HNF-1α atenuó las capacidades migratorias e invasivas en las cé-

lulas SiHa, mientras que la sobreexpresión de HNF-1α aumentó las capacidades

migratorias e invasivas en las células SiHa. La PKLR pudo abolir los efectos re-

guladores del HNF-1α sobre la metástasis del cáncer de cuello uterino. HNF-1α

y PKLR promueven sinérgicamente la invasión y migración del cáncer de cuello.

Received: 02-04-2021 Accepted: 03-06-2021

INTRODUCTION

Cervical cancer is a common malignancy

of the female reproductive system (1). It is re-

ported that in 2021, the diagnostic rate and

mortality rate of cervical cancer in our coun-

try was 16.25/100,000, and 5.08/100,000,

respectively (2). It is necessary to deeply un-

derstand the molecular mechanisms of cer-

vical cancer. Seeking for effective cervical

cancer biomarkers is conductive to stop the

malignant phenotypes of cancer cells.

Hepatocyte Nuclear Factors (HNFs)

are transcription factors predominately ex-

pressed in the liver, including HNF-1, 3, 4, 6,

CCAAT / enhancer binding-proteins and D-

binding proteins. HNF-1 contains two mem-

bers, that is, HNF-1α (also known as LFB1/

TCF1) and HNF-1β. HNF-1α is a transcrip-

tion factor containing a variant homology

domain. It is located on human chromosome

12q24.2, and expressed in liver, kidney, intes-

tinal and pancreatic tissues. HNF-1α is able

to regulate expressions of multiple tissue-

specific genes (3). A previous study has iden-

tified the close relationship between HNF-1α

and hepatocellular carcinoma (HCC) pro-

gression, and HNF-1α is capable of maintain-

ing hepatocyte differentiation phenotypes

(3,4). In addition, HNF-1α mutation causes

238 Tao et al.

Investigación Clínica 62(3): 2021

maturity onset diabetes of the young type

3 (MODY3) (5). Some low-frequency muta-

tions in HNF-1α (e.g. a single rare missense

mutation) are associated with type 2 diabe-

tes (6,7). The potential influence of HNF-1α

on cervical cancer is rarely reported.

Pyruvate Kinase L/R (PKLR) is located

on human chromosome 1q22, encoding py-

ruvate kinases (PKs). PKs are involved in an-

aerobic glycolysis and provide 50% of ATP to

mature red blood cells. Dysfunctional PKLR

can lead to PKs deficiency (PKD) and ATP de-

ficiency, shortening the lifespan of red blood

cells (8). The Warburg effect (aerobic glycol-

ysis) is a well-defined metabolic change that

is relevant to cancerous phenotypes, includ-

ing the accelerated proliferative, invasive

and migratory potentials (9). Nie et al. (10)

pointed out that mineralocorticoid recep-

tors (MRs) are able to alleviate the Warburg

effect and cancer progression in HCC by

targeting the miR-338-3p/PKLR axis. There-

fore, PKLR is considered as a key regulator

of glycolytic reprogramming and tumor cell

functions. A recent research showed that

HNF-1α drives the growth and anti-apoptosis

capacity in pancreatic cancer via its target

gene PKLR (11). This study aims to eluci-

date the role of HNF-1α and PKLR in the pro-

gression of cervical cancer, and to provide a

novel guidance in clinical treatment.

MATERIALS AND METHODS

Collection of pathological tissues

A total of 50 cervical cancer tissues,

confirmed by pathological examination and

resected by surgical procedures, were col-

lected. During the same period, 50 adjacent

normal tissues that were 2 cm away from the

cancer focus were collected as well. Tissues

were immediately frozen in liquid nitrogen

for RNA extraction. All patients did not re-

ceive neoadjuvant chemotherapy or radio-

therapy before operation, and there was no

surgical contraindication. The study was ap-

proved by the Ethics Committee of Hubei

Provincial Maternal and Child Health Hos-

pital, and informed consent was signed by

patients and their families before operation.

Cell culture

HcerEpic, Hela and SiHa cell lines were

provided by Fudan University (Shanghai, Chi-

na). Cells were cultivated in Dulbecco’s modi-

fied eagle medium (DMEM) (Gibco, Rock-

ville, MD, USA) containing 10% fetal bovine

serum (FBS) (Gibco, Rockville, MD, USA) in

a humidified incubator with 5% CO

at 37°C.

For cell culture passages, cells were isolated

and purified by pancreatin digestion and then

were cultured and subculfured in DMEM with

20 % fetal bovine serum. Cells in the logarith-

mic growth phase were used for experiments.

Cell transfection

HNF-1α siRNAs were constructed by Am-

bion (Austin, TX, USA) based on the HNF-1α

sequences in GeneBank (Accession No. NM

000545). Primer sequences were as follows:

siHNF-1α-#1 forward primer, 5′-CCGGT-

GCTAGTGGAGGAGTGCAATTTCAAGAGA-

ATTGCACTCCTCCACTAGCTTTTTG-3′ and

reverse primer, 5′-AATTCAAAAAGCTAGT

GGAGGAGTGCAATTCTCTTGAAATTG-

CACTCCTCCACTAGCA-3′; siHNF-1α-#2

forward primer, 5′-CCGGTGCAGAAGT

ACCCTCAAGCATTCAAGAGATGCTT-

GAGGG TACTTCTGCTTTTTG-3′ and re-

verse primer, 5′-AATTCAAAAAGCAGAAGTAC

CCTCAAGCATCTCTTGAATGCTTGAGGG-

TACTTCTGCA-3′. Cells were prepared to

suspension after 0.25% trypsin digestion,

and inoculated in the 6-well plate with 1.0 ×

cells/well. Once cell confluence reached

about 70%, transfection was conducted using

Lipofectamine

2000 (Invitrogen, Carlsbad,

CA, USA). Fresh medium was replaced at 6-8

h, and medium containing 2 μg/mL puromy-

cin was applied at 48 h. After 72 h cell cul-

ture, cells were passaged to a new 6-well plate

and cultivated for 1-2 weeks. Visible colonies

were picked up and inoculated in a 96-well

plate for extended culture. Their growth was

regularly observed and passaged to 6-well

plates and culture bottles two days later.

HNF-1α stimulates migration and invasion in cervical cancer 239

Vol. 62(3): 236 - 246, 2021

Quantitative real-time polymerase chain

reaction (qRT-PCR)

Total RNA isolated using TRIzol (Invi-

trogen, Carlsbad, CA, USA) was reversely

transcribed to complementary deoxyribose

nucleic acid (cDNA) using the First Chain

cDNA Synthesis Kit in a system contain-

ing 1 μL of Oligo dT Primer, 1 μL of dNTP

Mixture, 5 μL of RNA template and 3 μL

of ddH

O. The system was incubated at

65°C for 5 min and immediately frozen on

ice. Reverse transcription solution was pre-

pared as follows: 4 μL of 5×Prime Script II

Buffer, 0.5 μL of RNase Inhibitor, 1 μL of

Prime Script II RTase and 4.5 μL of RNase

free ddH

O. The reaction conditions were

45 min at 42℃, 5 min at 95℃ and cooling

on ice. PCR primers were: HNF-1α forward

primer: 5′-TCTACAGCCACAAGCCCGAG-3′

and reverse primer: 5′-GAGGTGAAGACCT-

GCTTGGT-3′; PKLR forward primer:

5′-TGGGAAAACTGGGTGGGATGGATG-3′

and reverse primer: 5′-GAAGGAAGCAGCC-

GGGGATTTGAC-3′; β-actin forward primer:

5′-TTGGCCTTAGGGTTCAGAGGGG-3′, and

reverse primer: 5′-CGTGGGCCGCCCTAG-

GCACCA-3′. PCR system in a total of 25

μL was then prepared, including 12.5 μL

of SYBR Fast qPCR Mix (2×), 1 μL of PCR

Forward Primer (10 μmol/L), 1 μL of PCR

Reverse Primer (10 μmol/L), 0.5 μL of ROX

Reference Dye (50×), 1 μL of cDNA and 8 μL

of ddH

O. The system was amplified at 95°C

for 30 s, 40 cycles at 95°C for 5 s and 60°C

for 32 s, followed by 95°C for 15 s, 60°C for

1 min, 95°C for 15 s and 60°C for 15 s. The

expression level was calculated by 2

-ΔΔCt

Western blot

Cells were digested in 0.25% trypsin,

washed in phosphate buffered saline (PBS)

for three times and lysed in radioimmu-

noprecipitation assay (RIPA) (Beyotime,

Shanghai, China). The isolated protein

samples were separated by sodium dodecyl

sulphate-polyacrylamide gel electrophoresis

(SDS-PAGE) and loaded on polyvinylidene

fluoride (PVDF) membranes (Millipore, Bil-

lerica, MA, USA). After blockage of non-spe-

cific antigens in 5% skim milk for 1 h, mem-

branes were reacted with primary antibodies

(1: 1000) at 4°C overnight, and horseradish

peroxidase (HRP)-labeled sheep anti-mouse

antibody (1:5000) for 1 h. Band exposure

and grey value analyses were finally conduct-

ed by Image-Pro Plus.

Transwell assay

Until cells were cultured to 75% con-

fluence, serum-free medium was replaced.

100 μL of serum-free suspension (1.0 × 10

cells/mL) and 600 μL of serum-containing

medium were applied to the top and bottom

Transwell chamber, respectively, and cultured

overnight. Cells in the bottom were subjected

to methanol fixation for 15 min, and crystal

violet staining for 20 min. Migratory cells

were counted in five randomly selected fields

per sample. Transwell invasion assay was con-

ducted in chambers pre-coated with 50 μL of

Matrigel diluted in serum-free medium.

Statistical analysis

Data processing was conducted by Sta-

tistical Product and Service Solutions (SPSS)

20.0 (IBM, Armonk, NY, USA). Linear relation-

ship of genes was analyzed by Pearson corre-

lation test. Differences between groups was

compared by the Student’s t-test. Compari-

son between multiple groups was done using

One-way ANOVA test followed by Post Hoc Test

(Least Significant Difference). Percentage (%)

was used to express the enumeration data and

chi-square test was used for data analysis. A sig-

nificant difference was set at p<0.05.

RESULTS

Correlation between HNF-1α and clinical

features in cervical cancer

Based on the median level of HNF-1α ex-

pression in recruited cervical cancer patients,

they were classified into high HNF-1α level

group (n=25) and low HNF-1α level group

(n=25), respectively. No significant differ-

ences in age and histology were observed be-

240 Tao et al.

Investigación Clínica 62(3): 2021

tween groups (p>0.05). Notably, tumor size,

FIGO stage and lymphatic metastasis inci-

dence were significantly different (p<0.05)

(Table I). It is indicated that HNF-1α was

closely linked to tumor size, FIGO grade and

lymphatic metastasis in cervical cancer pa-

tients, and might be a potential biomarker.

Upregulation of HNF-1α in cervical cancer

Compared with adjacent normal tissues,

HNF-1α was upregulated in cervical cancer

ones (Fig. 1A). In addition, collected cervical

cancer tissues were classified into two groups

based on the tumor size (4 cm as the cut-off

value). Higher level of HNF-1α was detected in

cervical cancer tissues with ≥ 4 cm in tumor

size compared with those < 4 cm (Fig. 1B).

In particular, we detected higher abundance

of HNF-1α in FIGO stage III-IV cervical cancer

patients than those FIGO stage I-II patients

(Fig. 1C). It is suggested that HNF-1α may

function as an oncogene driving the progres-

sion of cervical cancer.

HNF-1α overexpression and knockdown

models

HNF-1α level was first detected in cervi-

cal cancer cell lines (Fig. 2A). Based on the

differential level of HNF-1α, we constructed

HNF-1α knockdown and overexpression mod-

els in SiHa and Hela cells, respectively. Trans-

fection of either si-HNF-1α 1# or si-HNF-1α

2# could effectively decrease HNF-1α level in

SiHa cells (Fig. 2B, 2C). Besides, HNF-1α level

was markedly upregulated in Hela cells trans-

fected with pcDNA-HNF-1α (Fig. 2D, 2E).

HNF-1α promoted cervical cancer cells

to migrate and invade

Cell migration and invasion was a vital

process of tumor metastasis. Transwell as-

say was conducted to assess migratory and

TABLE I

CORRELATION BETWEEN HNF-1Α AND CLINICAL FEATURES IN CERVICAL

CANCER PATIENTS (N=50).

Clinico-pathologic

features

Number

of cases

HNF-1α expression

Low (n=25) High (n=25)

Age (years)

<50 31 16 15 0.7708

≥50 19 9 10

Histology

Squamous 23 12 11 0.7766

Adenocarcinoma 27 13 14

Tumor size

<4CM 26 19 7 0.0007*

≥4CM 24 6 18

FIGO stage

I~II 22 16 6 0.0044*

III~IV 28 9 19

Lymph node metastasis

No 26 17 9 0.0235*

Ye s 24 8 16

The high and low expression was classified based on the median level of HNF-1α expression in recruited cervical

cancer patients, *p<0.05.

HNF-1α stimulates migration and invasion in cervical cancer 241

Vol. 62(3): 236 - 246, 2021

Fig. 1. Upregulation of HNF-1α in cervical cancer tissues. (A) Higher level of HNF-1α was detected in cervical

cancer tissues (Tumor) (n=50) than adjacent normal ones (Peritumor) (n=50); (B) Higher level of

HNF-1α was detected in cervical cancer tissues larger than 4 cm in tumor size (≥ 4 cm) than those

smaller than 4 cm (<4 cm); (C) Higher level of HNF-1α was detected in FIGO grade III-IV(III-IV) cer-

vical cancer patients(III-IV) than those FIGO grade I-II patients(I-II). *p<0.05.

Fig. 2. HNF-1α overexpression and knockdown models. (A) HNF-1α levels in HcerEpic, Hela and SiHa cells

detected by qRT-PCR. The comparison was conducted between HcerEpic and Hela, and another com-

parison was conducted between HcerEpic and SiHa; (B) The mRNA level of HNF-1α was downregula-

ted by transfection of either si-HNF-1α 1# or si-HNF-1α 2#. The comparison was conducted between

si-NC and si-HNF-1α 1#, and another comparison was conducted between si-NC and si-HNF-1α 2#

1α 2#; The comparison was conducted between si-NC and si-HNF-1α 1#, and another comparison

was conducted between si-NC and si-HNF-1α 2# (D) The mRNA level of HNF-1α was upregulated by

transfection of pcDNA-HNF-1α in Hela cells; The comparison was conducted between pcDNA-NC and

pcDNA-HNF-1α.(E) The protein level of HNF-1α was upregulated by transfection of pcDNA-HNF-1α in

Hela cells. The comparison was conducted between pcDNA-NC and pcDNA-HNF-1α.*p<0.05.