Journal of Cancer Research and Clinical Oncology

, Volume 132, Issue 1, pp 19–27

Identification of a candidate tumor suppressor gene RHOBTB1 located at a novel allelic loss region 10q21 in head and neck cancer

Authors

  • Levent B. Beder
    • Department of Otolaryngology, Graduate School of Medicine and DentistryOkayama University
  • Mehmet Gunduz
    • Department of Oral Pathology and Medicine, Graduate School of Medicine and DentistryOkayama University
  • Mamoru Ouchida
    • Department of Molecular Genetics, Graduate School of Medicine and DentistryOkayama University
  • Esra Gunduz
    • Department of Oral Pathology and Medicine, Graduate School of Medicine and DentistryOkayama University
  • Akiko Sakai
    • Department of Molecular Genetics, Graduate School of Medicine and DentistryOkayama University
  • Kunihiro Fukushima
    • Department of Otolaryngology, Graduate School of Medicine and DentistryOkayama University
  • Hitoshi Nagatsuka
    • Department of Oral Pathology and Medicine, Graduate School of Medicine and DentistryOkayama University
  • Sachio Ito
    • Department of Molecular Genetics, Graduate School of Medicine and DentistryOkayama University
  • Noriyasu Honjo
    • Department of Otolaryngology, Graduate School of Medicine and DentistryOkayama University
  • Kazunori Nishizaki
    • Department of Otolaryngology, Graduate School of Medicine and DentistryOkayama University
    • Department of Molecular Genetics, Graduate School of Medicine and DentistryOkayama University
Original Paper

DOI: 10.1007/s00432-005-0033-0

Cite this article as:
Beder, L.B., Gunduz, M., Ouchida, M. et al. J Cancer Res Clin Oncol (2006) 132: 19. doi:10.1007/s00432-005-0033-0

Abstract

Purpose: Aims of the study are to narrow-down the hotspot region on 10q21 defined by previous genome-wide loss of heterozygosity (LOH) analysis in head and neck squamous cell carcinomas (HNSCC) and to define candidate tumor suppressor genes (TSG) concerned with 10q21. Materials and methods: LOH analysis was carried out with ten polymorphic microsatellite markers. Expression analysis was performed by semi-quantitative RT-PCR, and mutation analysis by PCR and direct sequencing. Results: LOH analysis on 10q21 in 52 HNSCC indicated distinctive and frequent allelic loss at D10S589 (42%). Among flanking genes, we found the RHOBTB1 gene as a candidate TSG, since an intragenic marker demonstrated the highest LOH (44%). Expression analysis revealed down-regulation of RHOBTB1 mRNA in 37% of tumors. Interestingly, all the five tumors that showed decreased expression of RHOBTB1 were accompanied with LOH, supporting the haploinsufficiency and class 2 TSG characteristics of RHOBTB1. No pathogenic mutation of RHOBTB1 was found. Furthermore, another gene within the region, EGR2, was also taken under scope. LOH frequencies around the EGR2 gene were relatively low (23 and 33%). Albeit semi-quantitative expression analysis of EGR2 demonstrated downregulation in 45% of tumor samples, no relation was found between the expression levels and LOH status. Conclusion: Frequent allelic loss and decreased expression of RHOBTB1 suggested that this gene has a role in tumorigenesis of a subset of HNSCC.

Keywords

10q21RHOBTB1EGR2LOH analysisRT-PCR

Introduction

Loss of heterozygosity (LOH) analysis is a strong tool to define the location of putative tumor suppressor genes (TSG). LOH method reveals whether any of the alleles is retained or lost in a specific chromosomal region. Loss of one allele by deletion means one hit of the “two-hits theory” hypothesized by Knudson to explain the mechanism of the inactivation of the genes functioning as tumor suppressor, whereas mutation, as another “hit” destroys the remaining allele (Knudson 1971). Therefore, regions of genome showing frequent LOH are considered to carry TSGs.

Although head and neck squamous cell carcinomas (HNSCC) are one of the most frequent malignancies worldwide, pathogenesis of the disease and underlying molecular genetic events that contribute to its development are largely undefined. To unveil new tumor suppressor loci, our group has previously performed a comprehensive genome-wide LOH analysis in HNSCC and demonstrated 20 hotspot areas with a high frequency of allelic loss and five of them were novel loci in HNSCC. Results of that study also revealed that 17p13 and 9p21 regions exerted the highest level of LOH frequency and confirmed the roles of p53 and p16 genes in the carcinogenesis of HNSCC (Beder et al. 2003).

Objectives of the present study are to analyze the hotspot regions of previous work in detail to narrow down the target areas. At first, we selected and analyzed 10q21, 5q13 and 5q31 hot loci. In the previous work, 5q13 and 5q31 bands represented by D5S1501 and D5S1480 showed 45 and 54% allelic loss, respectively (Beder et al. 2003). We examined LOH with six and four new microsatellite markers within 5q13 and 5q31, respectively. LOH analysis of those markers showed extensive losses (35–45%), however minimally deleted focus could not be revealed for further analysis (data not shown). For that reason, we abandoned the analysis of chromosome 5 at this step. However, detailed analysis on 10q defined minimally deleted region at 10q21.3.

In the second step, after reviewing flanking genes considering TSG potential, the RHOBTB1 gene located at 10q21.3 was selected for further analysis regarding close relation with minimally deleted focus of 10q21.3 region. We assessed for the first time, alterations in sequence and expression of RHOBTB1, as well as correlation between the expression level and LOH status in HNSCC. Moreover, LOH status and expression of EGR2 (early growth factor 2), a gene located close to RHOBTB1, was also evaluated, since it has been suggested to have tumor suppressive function and induces apoptosis in human tumors and cancer cell lines (Unoki and Nakamura 2001, 2003).

Materials and methods

Tissue samples

Paired normal and tumor samples were obtained from 52 patients with primary HNSCCs at the Department of Otolaryngology, Okayama University Hospital after acquisition of informed consent from each patient. All tissues were frozen in liquid nitrogen immediately after surgery and stored at −80°C until the extraction of DNA and RNA. Tumor samples are not microdissected. However, all tumor samples were confirmed as squamous cell carcinoma and the tumor cell ratio was over 70% by histological examinations performed at the Department of Pathology.

DNA and RNA extraction

Genomic DNAs were isolated from frozen tissues by SDS/proteinase K treatment, phenol-chloroform extraction, and ethanol precipitation. Total RNAs were prepared by using a modified acid guanidium phenol chloroform method (ISOGEN; Nippon Gene Co., Tokyo, Japan).

LOH analysis

Loss of heterozygosity analysis was performed with highly polymorphic microsatellite markers. After sense primers were labeled with 5-iodoacatamidefluorescein, PCR was carried out in 20 μl of reaction mixture containing 10 pmol of each primer, 100 ng of genomic DNA, 1× PCR buffer, 200 μM of each deoxynucleoside triphosphate, and 0.5 U of Taq DNA polymerase (Takara, Kyoto, Japan). The PCR products were applied on ABI Prism 3100 DNA sequencer (Applied Biosystems, Foster City, CA, USA) and analyzed by the Genescan analysis software version 3.7 (Applied Biosystems).

Mutation analysis of RHOBTB1

Exon/intron boundaries were established from the NCBI sequence database and Genetyx software (Wisconsin Package Version 9.0 of the Genetics Computer Group, Madison, WI, USA) and PCR amplicons were designed to span nine coding exons of RHOBTB1, starting with exon 3. Nine coding exons of the RHOBTB1 gene were amplified with primers shown in Table 1. PCR was performed as above. All PCR products were treated with ExoSAP (Amersham Bioscience, Tokyo, Japan), labeled with the Big Dye sequencing kit (Applied Biosystems) and analyzed on ABI PRISM 3100. Nucleotide changes observed were confirmed by independent PCR amplification and sequencing.
Table 1

Primers used for mutation analysis

Gene

Sequences (5′–3′)

RHOBTB1

 Exon 3

  F

TGGTGTCCTTACTGTGGTGTTC

  R

CTAGGGATGCTGAAGACACTC

 Exon 4

  F

CTTCTGACTGACACCGAGCA

  R

GCATTCAAGCTCAACCACGC

 Exon 5

  F

ACCTGTTAAAAGGCAATACAAGC

  R

AACACCAGGCTACAGAGACAC

 Exon 6

  F-a1

CATTAGGTGCTTAGATAGGCCC

  F-a2

CGACACTCAATATCCGCCCCT

  R-a

CCCACCTAAAGAAAGTCCAGA

  F-b1

AGCCTGTGAAGAAAGAGAAGCAG

  R-b1

CTGCTCATGTCTGGCTAACGT

  R-b2

ACTGAAGCGTCCATCCTGA

 Exon 7

  F

GTTGGCTCATGGTTGTGTTTCC

  R

CACACATACATTCATAAAGACACC

 Exon 8

  F

GCCCACTCCCTGCTTTCTGC

  R

AGACACTGCATGGCCCTGAGT

 Exon 9

  F

GTGGTTTAATGTTCTCCTGCCC

  R

TTAGACAGATGCAACAATGGAATG

 Exon 10

  F

AAATAGTCTTACCCTCCTTTTGC

  R

AGGTGAGGAGAGCTGGATGAA

 Exon 11

  F

GGCATCAATGAACCGCAGGTC

  R

CCTGAACTATGTCGTATCTCTAAC

p53

 Exon 4

  F

ATCTACAGTCCCCCTTGCCG

  R

GCAACTGACCGTGCAAGTCA

 Exon 5

  F

GACTTTCAACTCTGTCTCCTTC

  R

AACCAGCCCTGTCGTCTCTC

 Exon 6

  F

ACCATGAGCGCTGCTCAGAT

  R

AGTTGCAAACCAGACCTCAG

 Exon 7

  F

CTTGGGCCTGTGTTATCTCCT

  R

AGGGTGGCAAGTGGCTCCTG

 Exon 8

  F

CCTTACTGCCTCTTGCTTCTC

  R

TGAATCTGAGGCATAACTGCAC

 Exon 9

  F

AGTTATGCCTCAGATTCACTTTT

  R

GATAAGAGGTCCCAAGACTTAG

Semi-quantitative RT-PCR

Total RNA was reverse-transcribed with the Toyobo pre-amplification system (Toyobo, Osaka, Japan) starting with 2 μg of total RNA from each sample, according to the procedures provided by the supplier. One microliter of each RT reaction was amplified in 50 μl mixture containing 1.2 mM MgCl2, 1× PCR buffer, 200 μM of each deoxynucleoside triphosphate, 20 pmol of each primer, and 1 unit of rTth DNA polymerase XL (Applied Biosystems). Amplification was performed using specific primers for RHOBTB1, RS1 (5′-GCT CTC TTA CTT GGA ATT GGC T) and RAS1 (5′-CGC TGG TAG TGA TCT TCT TCC), for EGR2, RS1 (5′-CCG TAG ACA AAA TCC CAG TAA C) and RAS1 (5′-TTG CCC ATG TAA GTG AAG GTC T) and for GAPDH, S1 (5′-AGA CCA CAG TCC ATG CCA TCA C) and AS1 (5′-GGT CCA CCA CCC TGT TGC TGT). Appropriate cycling number for non-saturating measure was determined empirically by a quantitative PCR system (ABI GeneAmp 5700, Applied Biosystems). PCR amplification was performed as described in microsatellite analysis except that the annealing steps were at 60°C for 1 min, 32 PCR cycles for RHOBTB1 and EGR2, and 25 cycles for GAPDH.

Quantitation of the RT-PCR products

RT-PCR products were separated through 2% agarose gel and stained with ethidium bromide. The sizes of the RT-PCR products were 204, 268, and 456 bps for RHOBTB1, EGR2 and GAPDH, respectively. The intensity of ethidium bromide staining of each band was measured by a CCD image sensor (GelPrint 2000/VGA, Toyobo, Osaka, Japan), and analyzed by a computer program for band quantification (Quantity One, Toyobo). The values of tumor specific RHOBTB1 and EGR2 expressions were determined by calculating the ratio of the expression level in the tumor and that in the matched normal sample, each of which was normalized for the corresponding GAPDH expression level. Decreased and increased expression levels were defined when this ratio was ≤0.50 (50% or more decrease) and ≥1.50 (50% or more increase), respectively. Reproducibility was confirmed by independent PCR repeated twice.

Mutation analysis of p53

Each of the coding regions of exon 4–11 was amplified by PCR with intron spanning primers (Table 1). The PCR mixture contained 100 ng of genomic DNA, 1.2 mM MgCl2, 1× PCR buffer, 200 μM of each deoxynucleoside triphosphate, 20 pmol of each primer, and 1 U of rTth DNA polymerase in 50 μl volume. Initial denaturation at 94°C for 3 min was followed by 33 cycles of a denaturation step at 94°C for 30 s, an annealing step at 58°C for 1 min, and an extension step at 72°C for 1 min. A final extension step at 72°C for 7 min was added. The resultant PCR products were purified using GeneClean III kit (BIO 101 systems, Qbiogene, Carlsbad, CA, USA). Purified PCR products were re-amplified with BigDye terminator sequencing kit, ethanol precipitated and directly sequenced on an automated sequencer using the primers above (ABI prism 3100).

Statistical analysis

Fisher’s exact test (two-sided) was used for statistical analysis. P values less than 0.05 were considered as statistically significant.

Results

LOH analysis on chromosome 10q21

The loci of 10q21 represented by D10S1221 microsatellite marker displayed 40% LOH frequency in the previous work (Beder et al. 2003). In this work, seven new microsatellite markers with high heterozygosity located around D10S1221 (D10S196, D10S1790, D10S1756, D10S589, D10S1225, D10S581, D10S210) were analyzed in a total of 52 pairs of HNSCC matched with adjacent normal tissue for allelic loss. An average distance between these markers is about 2 Mbp. Among these, D10S589 showed a prominent result: 14 out of 33 (42%) informative samples displayed allelic loss as the highest LOH frequency of the region. Moreover, five samples have distinctive loss compared with flanking markers (Fig. 1a). Figure 1b shows representative examples with LOH and retention cases. The genes located around D10S589 were defined and reviewed in detail regarding TSG potential by both PubMed search and map view homepage of the National Center for Biotechnology Information (NCBI: http://www.ncbi.nlm.nih.gov/mapview). Figure 1a shows locations of genes and primers of the region. The RHOBTB1 gene as a TSG candidate was selected for further analysis. To confirm high LOH frequency in regard to the RHOBTB1 gene, two additional new microsatellite markers, one (MS-1) located in intron 7 of the gene, the other (MS-2) 50 kbp telomeric side of the gene, were designed by using sequence at map view web site of the NCBI and RepeatMasker program of Institute for Systems Biology homepage (http://www.repeatmasker.org). LOH analysis of the markers D10S589 and MS-1 showed higher LOH frequencies (42 and 44%, respectively) among all markers whereas flanking microsatellites displayed only 17 and 28% deletions (Fig. 1a).
https://static-content.springer.com/image/art%3A10.1007%2Fs00432-005-0033-0/MediaObjects/432_2005_33_Fig1_HTML.gif
Fig. 1

Loss of heterozygosity analysis on chromosome 10q21 in HNSCC. aRight, Schematic representation of LOH distribution. Case numbers are shown at the top. Only cases with at least one LOH were shown. Microsatellite markers are shown to the left and LOH frequencies to the right. Filled box LOH; open box retention of heterozygosity; shaded box not informative. Left, A physical map of the D10S196-D10S210 genomic interval. The location of the markers and genes are based on the latest mapping information derived from the National Center for Biotechnology Information (NCBI) homepage (http://www.ncbi.nlm.nih.gov/mapview). b Representative electropherograms of LOH analysis on the RHOBTB1 gene locus (MS-1) by microcapillary electrophoresis and data analysis by Applied Biosystems 3100 genetic analyzer. LOH was scored by comparing the peak heights of alleles of the matched tumor and normal samples. The arrows mark the lost allele in samples 20–11 and 20–16, while sample 8–19 represents retention

Another gene of the region, EGR2, was also taken into scope because of its suggested tumor suppressive role. To clarify allelic loss status of EGR2 region, two flanking microsatellite markers, MS-3 and MS-4 were designed, located 60 kbp centomeric and 30 kbp telomeric to the gene, respectively. LOH results were 33 and 23% for MS-3 and MS-4, respectively.

Genomic structure of the human RHOBTB1 gene

The RHOBTB1 gene (KIAA0740) with two alternative transcripts (Genebank no: NM_014836 and NM_198225) was mapped 1 Mbp telomeric to D10S589 at 10q21.3 band. The RHOBTB subfamily includes three members (RHOBTB1, RHOBTB2 and RHOBTB3) and belongs to the Rho GTPase family. RHOBTB1 sequence shares high homology (77% identity) with RHOBTB2, but no identity with RHOBTB3 when the open reading frames were compared with each other (Ramos et al. 2002). The human RHOBTB1 gene has 11 exons encoding 696 amino acid residues and spans over 131 Kbp genomic region. RHOBTB proteins are characterized by modular organization consisting GTPase region immediately followed by a short proline-rich region, a tandem of two BTB domains and a carboxyl-terminal region. The precise function of the RHOBTB proteins is unknown at present.

Mutation analysis of RHOBTB1 in HNSCC

All coding exons and exon-intron junctions of RHOBTB1 were screened for mutation by PCR-direct sequencing in 52 primary HNSCC samples. Sequence analysis revealed silent nucleotide changes at codon 506 (GGG–GGT) in exon 7 for sample of 7–33, at codon 514 (AGT–AGC) in exon 7 for 10 samples (7–33, 7–35, 8–12, 8–20, 9–1, 9–9, 20–7, 20–8, 20–11, 20–12) and codon 584 (GAC–GAT) in exon 9 for two samples (8–20, 9–3). Thus, no presumably pathogenic mutations of RHOBTB1 were detected in HNSCC.

Expression analysis of RHOBTB1 in normal and tumor tissues

RHOBTB1 expressions were analyzed by semi-quantitative RT-PCR (using primer set covering exons 9–10–11) in RNA available 46 HNSCC samples and corresponding normal tissues of the same patient (Fig. 2a). RHOBTB1 expression was detected in all investigated HNSCC samples and normal tissues at different levels (Table 2). After quantitation of RT-PCR products were normalized for GAPDH expression, comparison of expression between normal and tumor tissues were made according to 50–150% cutoff levels. The comparison for RHOBTB1 revealed decreased expression in 37% of tumor specimens (17 of 46), increased expression in 35% of tumor specimens (16 of 46) and no change in 28% of specimens (13 of 46). When the relationship between mRNA expression and LOH analysis of the RHOBTB1 gene was examined, a significant result appeared that all low-expression samples displayed allelic loss of the gene for informative cases (Table 3, P=0.01), although number of the non-informative cases for the marker was fairly high.
https://static-content.springer.com/image/art%3A10.1007%2Fs00432-005-0033-0/MediaObjects/432_2005_33_Fig2_HTML.gif
Fig. 2

Expression levels of the RHOBTB1a and EGR2b genes in HNSCC and adjacent normal tissues. Gene expression was examined by semi-quantitative RT-PCR analysis. Lower panel represent the expression of a housekeeping gene, GAPDH

Table 2

Expression analysis quantitations of RHOBTB1 and EGR2 and comparison with LOH and p53 mutation status

Sample no

RHOBTB1 Exp

LOH

p53 mut

EGR2 exp

LOH*

5–15

4.04

H

+

0.30

L

+

5–22

0.50

L

NI

0.40

L

7–24

55.00

H

NI

+

ND

 

+

7–25

0.27

L

NI

+

0.18

L

7–30

1.02

N

NI

1.00

N

7–31

ND

 

+

0.30

L

+

7–33

2.50

H

+

ND

 

7–35

10.09

H

NI

+

4.80

H

7–38

1.99

H

NI

+

1.96

H

+

7–40

0.09

L

NI

ND

 

NI

8–18

ND

 

ND

 

NI

8–19

2.07

H

1.51

H

8–21

8.24

H

 

1.06

N

8–26

0.13

L

NI

+

0.50

L

+

9–7

7.24

H

ND

 

20–2

0.50

L

NI

0.64

N

20–11

0.50

L

+

+

0.35

L

+

20–12

3.79

H

0.71

N

NI

20–16

0.50

L

+

+

0.31

L

NI

20–22

0.86

N

NI

0.15

L

20–23

0.77

N

+

+

0.50

L

+

20–24

0.50

L

NI

0.40

L

7–29

ND

 

 

ND

 

7–37

2.17

H

NI

+

2.15

H

7–41

ND

 

NI

 

ND

 

8–12

0.50

L

NI

0.60

N

8–17

1.41

N

NI

5.79

H

8–20

0.03

L

+

+

0.13

L

+

9–1

34.00

H

ND

 

9–4

1.26

N

2.16

H

9–6

0.35

L

NI

0.25

L

9–9

0.15

L

NI

ND

 

NI

9–12

0.77

N

NI

0.70

N

NI

20–3

1.55

H

NI

0.41

L

20–4

0.90

N

NI

+

0.30

L

20–6

0.86

N

0.11

L

NI

20–7

0.50

L

NI

3.45

H

20–8

0.50

L

+

0.73

N

20–10

1.18

N

NI

0.48

L

NI

20–14

0.88

N

+

+

1.57

H

+

20–15

0.82

N

+

+

1.23

N

20–18

0.50

L

+

1.42

N

NI

20–20

1.22

N

NI

+

1.47

N

01–04

0.57

L

NI

+

0.34

L

NI

01–06

2.02

H

3.87

H

01–07

0.90

N

+

1.29

N

NI

4–19

3.01

H

+

1.24

N

5–36

0.45

L

NI

0.17

L

6–13

ND

 

+

ND

 

NI

8–24

5.57

H

3.22

H

9–3

ND

 

+

ND

 

NI

9–10

16.00

H

+

ND

 

NI

ND Not done, NI not informative, L low, H high, N normal expression

− and + indicate presence or absence of either p53 mutation or LOH of related gene

*MS-3 and MS-4 results were combined

Table 3

Correlation of RHOBTB1 and EGR2 gene expression with LOH status and p53 mutation

RHOBTB1 expression

EGR2 expression

LOH status

Total (n)

Low

Normal

High

P*

Total (n)

Low

Normal

High

P*

 Retention

11

0

3

8

0.01

23

11

7

5

0.08

 LOH

10

5

3

2

 

7

4

0

3

 

P53 mutation

 p53 mutant

18

6

6

6

0.94

17

10

3

4

0.30

 p53 wild

28

10

8

10

 

23

8

9

6

 

*Fisher’s exact test-extended

When we examined the promoter region of RHOBTB1, the first 300 bp of upstream sequence of exon 1 showed a 68% GC content with 34 CpG motifs. When we included exon1 (non-coding region), GC content increased to 72% with additional 35 CpG motifs. Those results suggest that up-stream sequence of RHOBTB1 transcription start site as well as exon1 itself could be methylated which may result in decreased expression of RHOBTB1.

Expression analysis of EGR2 in tumor vs normal tissues

EGR2 expression was analyzed by semi-quantitative RT-PCR using primer set covering exons 1–2 (Fig. 2b). Tissue samples, quantitation methods and cut-off levels were identical as RHOBTB1 expression analysis. When mRNA expression levels of tumor samples were compared with adjacent normal tissues, low expression in 45%, high expression in 25% and normal expression in 30% of tumor specimens were displayed (Table 2). Results of the expression analysis were compared with LOH analysis of flanking markers, MS-3 and MS-4. Contrary to the results of RHOBTB1, relationship between expression levels and LOH status of EGR2 flanking markers was not significant (Table 3).

Association of RHOBTB1 and EGR2 expressions with p53 mutation status

Previous reports have revealed that the p53 gene functions as TSG in HNSCC. According to the results of mutation analysis, 19 out of 50 tumor samples (38%) harbored mutations in the p53 gene. To find out whether any causative relation-interaction exists, results of RHOBTB1 and EGR2 expression were compared with p53 mutation status. However we did not find significant relationship between p53 mutation status and either RHOBTB1 or EGR2 expression levels (Table 3).

Association of RHOBTB1 and EGR2 expressions with clinicopathological features

First we analyzed the relationship between LOH result of the RHOBTB1 gene locus and various clinical characteristics of the patients including age, gender, smoking status, tumor localization, nodal status, TNM stage, histological grade and mutation status of p53. Second, same clinical features were compared with scores of the semi-quantitative mRNA expression analysis. We did not find significant relation between genetic-epigenetic modifications of both genes and clinicopathological markers in tumor samples.

Discussion

We found for the first time high and distinctive LOH focus of the chromosomal 10q21 band in HNSCC. High LOH frequencies and several hotspot areas including 10q23, 10q24 and 10q26 in chromosome 10 have been reported previously for a variety of human neoplasms (Cappellen et al. 1997; Rao et al. 2003). Conversely, only a few studies exist evaluating the LOH status of chromosome 10q in HNSCC. Gasparotto et al. showed two separate hotspots at 10q22-26 interval in HNSCC (Gasparotto et al. 1999). It seems likely that the authors had selected the interval considering localization of the PTEN/MMAC1 gene, however remaining parts of the chromosome including q21 band were not taken into evaluation. The other study also focused on 10q26 band where the DMBT1 gene located, although allelotyping of the markers together with the DMBT1 gene locus could not reveal significant allelic loss (Petersen et al. 2000).

Our previous genome-wide LOH screening study revealed 10q21 band as hot spot. As continuation of the previous work, LOH analysis with eight new microsatellite markers surrounding 10q21 was performed in 52 HNSCC cases to confirm and define the minimally deleted region. Thus, current analysis revealed D10S589 marker (10q21.3) with high and distinctive LOH results (42%) in HNSCC as a novel minimally deleted region.

In the second step of the study, genes flanking D10S589 were reviewed in detail with respect to TSG potential. As to the putative TSGs studied on chromosome 10, PTEN located on 10q23 is well-characterized tumor suppressor gene (Cantley and Neel 1999). Inactivation of PTEN by mutation was also shown in HNSCC (Poetsch et al. 2002). ANX7, suggested as TSG especially for prostate carcinomas (Srivastava et al. 2001), is located 14 Mbp telomeric to our hotspot region and the microsatellite marker (D10S1432) proximal to this region showed only 24% of LOH. Thus, we suspected existence of another tumor suppressor gene in 10q21.3 location.

CDC2 and H4 genes of 10q21.3 were shown to be involved in cell cycle. CDC2 is a kinase controlling entry into mitosis (Nurse 1990). p53, APC and Rb genes were shown to be targets for human p34(cdc2)-cyclin B1 (Bischoff et al. 1990; Lin et al. 1991; Trzepacz et al. 1997). CDC2 involved in the cell cycle-dependent phosphorylation of those three TSGs. p53 also represses the transcription of CDC2 (Yun et al. 1999). On the other hand, H4 of the cell-cycle dependent histone genes is transcriptionally and coordinately activated at the G1/S phase transition (Mitra et al. 2003). The tumor suppressor pRB was shown to repress the H4 gene promoter (Gupta et al. 2003). Those evidences suggest both genes have oncogenic properties rather than tumor suppressive role.

We have selected the RHOBTB1 gene closed to D10S589 as a target for further molecular analysis considering two factors. First, BTB/POZ (Broad complex, Tramtrack, and Bric a brac) domain proteins are found in transcriptional regulators and many of those are transcriptional repressors and influence cellular development (Collins et al. 2001). Moreover, other putative TSGs carrying BTB domain such as HIC1, BPOZ and APM1 have been reported (Pinte et al. 2004; Reuter et al. 1998; Unoki and Nakamura 2001). RHOBTB1 is a member of the Rho GTPase family. Recently, increasing amount of evidence indicates that several members of the Rho GTPase are important players in tumor biology. These GTP-hydrolyzing enzymes relay on intracellular signaling (Boettner and Van Aelst 2002). However, there is still little information available on the clinical significance of Rho GTPase expression in human cancer specimens. RHOBTB subgroup (RHOBTB1 and RHOBTB2) was suggested to have a function distinct from classical Rho GTPases, without obvious role in organizing the actin filament system (Aspenstrom et al. 2004). RHOBTB proteins also have a different modular organization than Rho and other small GTPases. On the other hand, depending upon the BTB domains, RHOBTBs may have a role in transcriptional regulation (Collins et al. 2001).

Second, recent findings of homozygous deletion, somatic missense mutations and extinguished expression of the RHOBTB2 gene of the same subfamily (or DBC2 gene: deleted in breast cancer) are evidences supporting tumor suppressive roles of the RHOBTB subfamily members. Growth-inhibitory activity of the RHOBTB2 was also displayed, when it was reintroduced to RHOBTB2-deficient cell (Hamaguchi et al. 2002). Furthermore, comparison of amino acid sequences among the members of the RHOBTB family showed that RHOBTB1 and RHOBTB2 are closely related to each other (79% identity). However, insert region of the first BTB domain is highly divergent both in length and amino acid composition between RHOBTB1 and RHOBTB2 (Ramos et al. 2002). Collectively those findings implicate RHOBTB1 as a potential TSG at 10q21 in HNSCC. We confirmed high LOH results(44%) for the gene locus by designating a specific microsatellite marker located in intron 7 of RHOBTB1. It is also noteworthy that 4 tumor samples showed LOH exclusively for RHOBTB1 locus (Fig. 1a).

Semi-quantitative RT-PCR analysis was performed to evaluate expression pattern of RHOBTB1, because one of the mechanisms for gene inactivation is suppression of expression. Overall, 17/46 of the samples (37%) showed tumor specific reduction of RHOBTB1 expression. When expression levels were compared with LOH patterns, all 5 informative cases with the low expression (<50%) showed LOH. That association is a significant result defining the role of allelic loss in down-regulation of the gene product. Expression levels were also compared with p53 mutation status, however no relation was found.

To determine whether inactivating mutations of RHOBTB1 exist, we carried out a comprehensive screening including all nine exons and splicing junctions of the gene for somatic mutations in the same samples. No somatic mutation of RHOBTB1 was detected in the series. When we examined promoter region of RHOBTB1, we detected a high GC content with a lot of CpG islands, suggesting the possible methylation of the promoter region which may result in decreased expression of RHOBTB1. However, when downregulated samples with LOH are taken into consideration, loss of one allele may result in decrease in gene product level less than half of the normal two-allele condition. The variance model in gene expression system showed also that gene expression levels of one-allele system is fluctuating and can approach to 10% of wild-type levels which mimic a null mutation (Cook et al. 1998).

The opinion which states loss of only one allele of a TSG might contribute to tumor process, is supported by considerable evidences appeared recently. Although two-hit model of Knudson, which stipulated inactivation of both alleles has been held for many years to explain tumor progression, recently it is denoted that Knudson’s hypothesis must be modified. Haploinsufficiency is one of the potential mechanisms to explain that kind of heterozygous effect and it means reduction in gene dosage leading a phenotypic change and contributes tumorigenesis. Such an affect was demonstrated for a group of TSGs including ANX7, PTEN and p53 (Kwabi-Addo et al. 2001; Srivastava et al. 2003; Venkatachalam et al. 1998). Effects of haploinsufficiency may be partial or strong especially when exposed to carcinogens such as u.v. or smoking. Modest impairments in DNA repair caused by haploinsufficiency can result in increasing mutation rate instead of directly increasing target cell population (Santarosa and Ashworth 2004).

EGR2, another candidate gene of the region and the gene, studied extensively by Unoki and Nakamura, has been suggested to have tumor suppressive role (Unoki and Nakamura 2001, 2003). We checked LOH status of the gene region by two markers but LOH frequency was not significantly high (23–33%). When expression of the gene was evaluated in tumor samples, 45% of the samples showed low expression. However, no relation was found between expression level and LOH status. Similarly, the expression levels showed no association with p53 mutation status and clinical characteristics of the patients. Those results make EGR2 less prominent compared with RHOBTB1 regarding tumor suppressor potentiality.

In conclusion, our study demonstrated high and distinctive LOH at the RHOBTB1 locus and attenuated expression levels of the gene in about 40% of HNSCC samples. All cases with low expression had LOH indicating the haploinsufficiency effect of the lost allele of the gene. Further functional analysis is warranted to define the mechanisms for loss of the gene function and fine roles in controlling cell proliferation.

Acknowledgements

Thanks are due to all the surgeons in the Department of Otolaryngology of our University Hospital for cooperation. This work was supported by a Grant-in-Aid (No.12213084) from the Ministry of Education, Culture, Sports, Science and Technology of Japan to K.S.

Copyright information

© Springer-Verlag 2005