An Epigenetically Derived Monoclonal Origin for Recurrent Respiratory Papillomatosis

ORIGINAL ARTICLE An Epigenetically Derived Monoclonal Origin for Recurrent Respiratory Papillomatosis

Josena Kunjoonju Stephen, MD; Lori E. Vaught, MD; Kang Mei Chen, MD; Veena Shah, MD; Vanessa G. Schweitzer, MD; Glendon Gardner, MD; Michael S. Benninger, MD; Maria J. Worsham, PhD

Objective: To investigate the contribution of pro- cer genes in the multigene panel had altered DNA meth- moter methylation-mediated epigenetic events in recur- ylation in at least 1 laryngeal papilloma biopsy speci- rent respiratory papillomatosis tumorigenesis. men. Identical abnormally methylated genes were found in 5 of 15 recurrent cases, of which the CDKN2B gene Design: Archival tissue DNA, extracted from microdis- was hypermethylated in all 5 cases. Dissimilar epige- sected papilloma lesions, was interrogated for methyl- netic events were noted in the remaining cases. ation status by means of the novel, multigene methylation- specific multiplex ligation-dependent probe amplification Conclusions: A clonal origin was derived for 5 of 15 assay. recurrent respiratory papillomatosis biopsy specimens based on identical epigenetic events. The high fre- Subjects: Fifteen subjects with recurrent respiratory pap- quency of epigenetic events, characterized by consis- illomatosis, 3 females and 12 males, all with adult onset tent promoter hypermethylation of multiple tumor of illness (age range, 23-73 years) except for 1 female pa- suppressor genes, points to the use of gene silencing tient with juvenile onset (1 year old). mechanisms in the pathogenesis of recurrent respiratory papillomatosis. Results: Promoter hypermethylation was recorded in 14 of 15 cases, and 19 of 22 unique methylation-prone can- Arch Otolaryngol Head Neck Surg. 2007;133(7):684-692

ECURRENT RESPIRATORY small percentage of RRP cases progress to (laryngeal) papillomatosis malignancy.9 (RRP), an extremely rare Laryngeal papillomas usually run a be- condition, is characterized nign but recurrent course. Spontaneous by benign neoplasms transformation of RRP to squamous cell car- withinR the respiratory tract and can be cinoma is not easily characterized by a his- potentially life threatening because of tologic progression through dysplasia over airway obstruction.1 Recurrent respiratory time, making these lesions difficult to di- papillomatosis presents primarily as tiny or agnose histologically and clinically early in larger warts on the vocal chords. Preva- the course of the transformation of the lence of RRP worldwide is approximately disease. 100 000, with 2300 new cases in the United Clonality, the property that the cells States each year.2-4 Juvenile-onset disease oc- within a tumor are derived from a single curs in patients from younger than 1 year parent cell, is often indicated by unifor- to 8 years old; shows no sex difference5,6; mity or relative uniformity of genetic ab- has a rapid but often unpredictable pattern errations contained within many or all cells of recurrence5; tends to be a long-term, of- of the tumor. Such aberrations are as- ten lifelong disease; and exhibits a con- sumed to confer or reflect biological dis- Author Affiliations: Research tinuum of severity and aggressiveness. The tinctions relevant to tumor behavior, and Division, Department of adult form of RRP has a variable age at on- thus to be relevant to tumor initiation and Otolaryngology–Head and Neck set (peak, approximately 20-30 years),6,7 clonal expansion.10-13 Surgery (Drs Stephen, Vaught, 6 Chen, Schweitzer, Gardner, with a higher incidence in males. The se- Epigenetics is the regulation of changes Benninger, and Worsham), and verity, aggressiveness, and recurrence of the in gene expression by mechanisms that do Department of Pathology adult form tend to be less than in the juve- not involve changes in DNA sequence. Es- (Dr Shah), Henry Ford nile form.6 Human papillomavirus types 6 tablishment and maintenance of epige- Hospital, Detroit, Michigan. and 11 account for 80% to 90% of RRP.8 A netic control (gene silencing) has several

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 684

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 MS-MLPA: Hhal digest Amplification of methylated target Antisense primer Antisense primer Labeled sense primer Labeled sense primer Ligation PCR Stuffer sequence Stuffer sequence with variable length Methylated with variable length Target DNA target CG CH Undigested by Methylated → no digestion → ligation 3 Antisense HhaI, binds to → Antisense primer probe PCR primer Labeled sense primer Labeled Hhal sense primer No PCR digestion

Unmethylated target Target DNA CG

Unmethylated → cut by Hhal → no ligation Digested by HhaI → does not bind to probe → no PCR

Figure 1. Methylation-specific multiplex ligation-dependent probe amplification (MS-MLPA).17 Probes designed to recognize HhaI sites within unmethylated regions will not generate a signal because these sequences have become cut by HhaI and cannot bind to the probe. Conversely, an MLPA probe will bind to an intact methylated site, spared by HhaI, and generate an amplification signal, producing 15 separate peaks (Figure 2) in a normal control DNA sample. Aberrant methylation is identified as the

appearance of a signal peak that is otherwise absent in normal DNA samples (Figures 2, 3, and 4). CH3 indicates methyl group; PCR, polymerase chain reaction.

aspects, which include promoter region hypermethyl- lowed to cool to 60°C. Next, 6 µL of 20-mg/mL proteinase K ation, methyl-binding proteins, DNA methyltransfer- was added, mixed, overlaid with 3 drops of mineral oil, and ases, histone deacetylases, and chromatin state. Aber- spun for 5 seconds at 13 000g. This was followed by a 4- to 16- rant methylation of CpG islands is a hallmark of human hour (overnight) incubation at 60°C. The tube was heated for cancers and is found early during carcinogenesis.14 Genes 10 minutes at 90°C to denature the proteinase K and to dis- rupt nucleic acid formaldehyde adducts. On removal of the oil, in cellular pathways that are inactivated by promoter re- the tube was centrifuged for 15 minutes (at 13 000g) at room gion hypermethylation include MGMT (DNA repair), temperature. Next, 250 µL of the supernatant was transferred INK4a INK4b p16 , p15 (cell cycle), DAPK (apoptosis), and to a clean 1.5-mL tube. On addition of 10 µL of 5M sodium GSTP1 (detoxification).15 chloride and 1000 mL of ethanol to the 250-µL supernatant, We investigated alterations in DNA methylation in bi- the tube was incubated at −20°C for least 60 minutes. This was opsy specimens of recurrences from patients with RRP followed by centrifugation for 15 minutes at 13 000g at −4°C. to assess the contribution of promoter methylation- On removal of the supernatant, an additional centrifugation step mediated epigenetic events in RRP tumorigenesis. for 10 seconds ensured removal of the last traces of the super- Aberrant promoter methylation of 22 methylation- natant. Finally, the pellet was air dried and dissolved in 100 prone tumor suppressor genes was evaluated by means µL of double-distilled water. of a high-throughput multigene probe panel (41 gene probes, 35 unique genes, including control probes) in MS-MLPA ASSAY 15 RRP cases by using the methylation-specific multi- The MS-MLPA assay allows for the relative quantification of ap- plex ligation-dependent probe amplification (MS-MLPA) 16,17 proximately 41 different DNA sequences in a single reaction re- assay. quiring only 20 ng of human DNA. The standard use of the tech- nique to observe quantitative changes in copy number has been METHODS outlined in other studies.18-21 Adaptation of the MLPA to detect aberrant methylation (MS-MLPA) has been detailed elsewhere.16,17 The probe design is similar to that of ordinary MLPA probes. RRP COHORT For 26 of 41 probes, the recognition sequence detected by the MLPA probe is contained within a restriction site for the methyl- The RRP cohort comprised 15 subjects, 3 females and 12 males, sensitive enzyme HhaI(Figure 1). all with adult onset of respiratory papillomatosis (age range, The41-gene-probepanel(Table 1)interrogates35uniquegenes 23-73 years) except for 1 female patient with juvenile onset (1 implicated in cancer, including head and neck squamous cell car- year old). The number of biopsy specimens from patients with cinoma, for losses and gains in a separate reaction in the absence recurrences ranged from 1 to 6. Archival tissue DNA, ex- of the methyl-sensitive enzyme HhaI. Because there are 2 probes tracted from microdissected papilloma lesions, was interro- eachforVHL,CDKN2A,BRCA1,andBRCA2,and3probesforMLH1, gated for methylation status by means of the MS-MLPA assay. a normal control DNA sample will generate 41 individual peaks in the absence of HhaI(Figure 2). A concurrently run reaction DNA EXTRACTION with the 41-gene-probe set in the presence of HhaI is designed to detect aberrant promoter hypermethylation by taking advantage As a first step, 300 µL of P-buffer (50mM Tris hydrochloride, of an HhaI site in the promoter region of 22 of the 35 unique genes pH 8.5; 100mM sodium chloride; 1mM EDTA; 0.5% Triton (note that 1 of the 2 BRCA1 probes is designed to recognize a re- X100; 20mM dithiothreitol) was added to tubes containing whole gion outside the HhaI recognition site; Table 1). Fifteen of the 41 5-µm tissue sections or microdissected tissue. The tube was gene probes are designed outside an HhaI site and serve as undi- heated for 15 to 20 minutes at 90°C in a water bath and al- gested controls (Figure 2). On digestion of the sample DNA with

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 685

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 and 1 minute at 72°C; and 30 cycles of 30 seconds at 95°C, 30 Table 1. Methylation MLPA Probe Panel seconds at 60°C, and 1 minute at 72°C. All (ligated) probes were amplified by the same primer pair, one of which was tagged No. Probe Size Gene Probe Chromosome Locus with a fluorescent dye. The relative amounts of polymerase chain 1 M-238a TP73 01p36 reaction product obtained reflect the relative amounts of li- 472 NRAS 01p13.2 gated probes at the start of the polymerase chain reaction. Am- 364 MSH6 02p11 plification products were analyzed on a DNA sequencer (ABI 2 M-265a FANCD2 03p26 310/ABI 3100; Applied Biosystems Inc, Foster City, Califor- 3 M-220a VHLb 03p26 a b nia), and the products were normalized and quantified. 4 M-274 VHL 03p26 Aberrant methylation was identified as the appearance of a 5 M-166a MLH1b 03p21.3 6 M-292a MLH1b 03p21.3 signal peak that was otherwise absent in normal DNA samples 7 M-463a MLH1b 03p21.3 (Figures 2, 3, and 4). To quantify whether 1, 2, or more copies 8 M-328a RASSF1 03p21.3 of a specific gene locus became aberrantly hypermethylated, a 9 M-409a FHIT 03p14.2 mathematical algorithm was used.17 10 M-148a APC 05q21 154 IL4 05q31.1 11 M-373a ESR1 06q25.1 RESULTS 481 MFHAS1 08p23.1 12 M-160a CDKN2Ab 09p21 13 M-427a CDKN2Ab 09p21 Promoter hypermethylation was recorded in 14 of 15 14 M-211a CDKN2B 09p21 cases. Of the 22 unique methylation-prone cancer genes 15 M-346a DAPK1 09q34.1 in the multigene panel, 19 had altered DNA methyl- 136 CREM 10p12.1 16 M-193a MEN1 11q13 ation in at least 1 laryngeal papilloma biopsy specimen 17 M-454a GSTP1 11q13 (Table 2). 18 M-319a IGSF4 11q23 Of the 15 recurrent cases, 5 cases had at least 1 ab- 175 TNFRSF1A 12p13 normally methylated gene in a recurrent biopsy, of which 445 TNFRSF7 12p13 19 M-382a CDKN1B 12q13.1 the CDKN2B gene showed consistent hypermethylation 20 M-301a BRCA2b 13q12 in all 5 cases (Table 3; Figures 2, 3, and 4). All 3 bi- 418 BRCA2b 13q12.3 opsy specimens in case 4 indicated hypermethylation of 310 RB1 13q14.3 CDKN2B,withgainofaTP73 epigenetic event in the sub- 202 MLH3 14q24.3 283 TSC2 16p13.3 sequent 2 recurrences. APC and GSTP1 were hypermeth- 21 M-400a ASC 16p12 ylated in 1 recurrent biopsy in case 7. In case 13, all 3 22 M-247a CDH13 16q24.2 specimens showed hypermethylation of the APC and the a 23 M-355 HIC1 17p13.3 VHL promoter regions (Table 3). 337 BRCA1b 17q21 24 M-436a BRCA1b 17q21 Dissimilar epigenetic events were noted in the remain- 256 BCL2 18q21.3 ing cases (Table 2). Aberrant methylation of CDKN2B and 25 M-184a KLK10 19q13.3 APC genes was most frequent, occurring in 8 of 14 cases, 391 KLK3 19q13 followed by CDKN2A in 6, TIMP3 and VHL in 5, and 229 NF2 22q12 26 M142a TIMP3 22q12.3 DAPK1, HIC1, and GSTP1 in 4 of 14 cases (Table 2).

Abbreviation: MLPA, multiplex ligation-dependent probe amplification. COMMENT a Probes with HhaI sites in CpG-rich promoter regions b Genes represented by more than 1 gene probe. The scientific understanding of RRP has been a slow pro- cess, and RRP remains an enigmatic disease that contrib- HhaI, probes that recognize the unmethylated regions will not gen- utes a substantial human and financial cost to the public. erate a signal because these sequences have become cut by HhaI Originally called juvenile laryngeal papillomatosis, the dis- andcannotbindtotheprobe(Figure1).Conversely,anMLPAprobe ease has been increasingly recognized in adults and is now will bind to an intact methylated site, spared by HhaI, and gener- generally referred to as recurrent respiratory papillomato- ate an amplification signal (Figures 2, 3, and 4). sis. Histologically, RRP is a benign disease characterized In this study, approximately 20 ng of DNA was used for each of the 2 reactions, one without HhaI and the other with HhaI. by unregulated growth of wartlike neoplasms on the lar- Briefly, DNA from normal controls (male and female) and RRP ynx, trachea, and bronchi. Human papillomavirus has been 8 DNA for each of the 2 MLPA reactions was diluted with water shown to be the cause. Even though a small percentage to a total volume of 5 µL. The DNA was denatured and frag- of RRP cases progress to malignancy,9 spontaneous trans- mented by heating for 5 minutes at 98°C in a thermocycler. Bi- formation of RRP to squamous cell carcinoma does not ap- nary MLPA probes were added and allowed to hybridize to their pear to be characterized by a histologic progression through targets during a 16-hour incubation at 60°C. Dilution buffer dysplasia over time,22 making these cancers difficult to di- and a special ligase 65 enzyme (MRC Holland, Amsterdam, the agnose histologically and clinically early in the course of Netherlands) were added to the vial. During 15-minute incu- the transformation of the disease. bation at 60°C, the 2 parts of a probe could be ligated to each The molecular pathogenesis of RRP implicates dys- other and become an amplifiable molecule provided that the complementary sequence was present in the sample (Figure 1). regulation of apoptosis as determined by abnormal ex- This was followed by the addition of polymerase chain reac- pression of antiapoptotic factors like survivin and XIAP 22 tion primers, deoxynucleotide triphosphates, and Taq poly- as contributing to papilloma growth and survival. Hu- merase, followed by the following cycles: 1 cycle of 1 minute man papillomavirus 11 infection is a likely early event at 95°C; 10 cycles of 30 seconds at 95°C, 30 seconds at 70°C, associated with progression of RRP to carcinoma, with

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 686

0 120 160 200 240 280 320 360 400 440 480 520

B 3000 2000 Control DNA 1000 0 120 160 200 240 280 320 360 400 440 480 520

C 2000 Case 12: biopsy 1 Fluorescence Intensity 1000

0 120 160 200 240 280 320 360 400 440 480 520

D 2000 CDKN2B Case 12: biopsy 1 1000

0 120 160 200 240 280 320 360 400 440 480 520 PCR Product Size, bp

Figure 2. Methylation-specific multiplex ligation-dependent probe amplification probe mix without (A, C) and with (B, D) HhaI enzyme (DNA sequencer, ABI 3130). Note 15 peaks in the control DNA sample (B). Presence of a peak (D) in case 12, biopsy 1, not present in the control DNA (B) is that of aberrantly methylated CDKN2B gene. bp indicates base pairs; PCR, polymerase chain reaction.

increased expression of p53 and topoisomerase ␣-II and mens, marking this epigenetic event as an initiating clonal a reduced expression of p21(WAF1) protein gene as mark- alteration in the recurrence continuum in RRP. ers of transformation to invasive papillomatosis and squa- The cyclin-dependent kinase 2A (CDKN2A) and mous cell carcinoma.23,24 Additional markers of an ag- CDKN2B genes map to 9p21 and are in tandem, with gressive clinical course include high S-phase fraction, CDKN2B located 25 kilobases centromeric to proliferative index, and Ki-67 expression.25 CDKN2A.28,29 The CDKN2A locus controls the Rb path- There are few reported studies examining the mo- way (which regulates G1/ S-phase transition) and the lecular genetics of RRP, especially from the genomic and p53 pathway (which induces growth arrest or apoptosis the epigenetics perspectives, and much remains to be in response to either DNA damage or inappropriate learned. This study should serve as a forerunner in the mitogenic stimuli) by generating 2 gene products, p16 delineation of genes that succumb to promoter hyper- and p14.30,31 Mutations in CDKN2A/p16 inactivate the methylation silencing as underlying events in the epige- Rb pathway, whereas deletion of the CDKN2A locus netic pathogenesis of RRP. In addition, it provides evi- (CDKN2A/p16 and CDKN2A/p14) alter both the Rb and dence of a clonal origin for RRP and permits the tracing p53 pathways, which are important in many cancers. of an epigenetic continuum implicating key tumor sup- Inactivation of the CDKN2B/p15, CDKN2A/p14, and pressor genes in RRP. CDKN2A/p16 genes is a frequent event in human oral Epigenetic alterations produce heritable changes in squamous cell carcinomas.17,19,32,33 The presence of aber- gene expression without a change in the DNA coding se- rant methylation of p15 and p16 in precancerous oral quence itself. Promoter region hypermethylation is known tissues32 implicates methylation of p15 and p16 as early to be an early event in carcinogenesis.26,27 The conse- events in the pathogenesis of oral lesions. quence of CpG island hypermethylation, especially for Hypermethylation of the APC gene was another con- those islands associated with tumor suppressor gene pro- sistent epigenetic event in RRP, occurring in 8 of 14 cases, moters is the loss of tumor suppressor gene function, including biopsy specimens of recurrences in cases 7 and which contributes to tumorigenesis.17 13. APC (adenomatosis polyposis coli) is a tumor sup- The most commonly methylated genes in this RRP co- pressor gene originally implicated in colon cancer. It has hort were CDKN2B and APC (8 of 14 cases), CDKN2A an important role in the Wnt signaling pathway, which (6 cases), TIMP3 and VHL (5 cases), and DAPK1, HIC1, is involved in the development of colorectal carcino- and GSTP1 (4 cases; Table 2). In addition, in the 5 RRPs mas. Genetic and epigenetic alterations in this gene have with at least 1 commonly methylated gene (Table 3), since been recognized in other malignant neoplasms, in- CDKN2B was identified in all subsequent biopsy speci- cluding oral squamous cell carcinomas, gastric cancers,

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 687

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 A Without HhaI With HhaI 4500 4000 Control DNA 3500 3000 2500 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480

3500 Case 4: biopsy 1 3000 2500 CDKN2B 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480

4500 4000 TP73 Case 4: biopsy 2 Fluorescence Intensity 3500 CDKN2B 3000 2500 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480

4500 Case 4: biopsy 3 4000 3500 TP73 3000 CDKN2B 2500 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480 PCR Product Size, bp

Figure 3. Methylation-specific multiplex ligation-dependent probe amplification probe mix without and with HhaI enzyme (DNA sequencer, ABI 3100) in a control DNA specimen (A) and 3 biopsy specimens from case 4 (B-D). Note methylation of CDKN2B in all 3 biopsy specimens and methylation of TP73 in specimens 2 and 3. bp indicates base pairs; PCR, polymerase chain reaction.

and esophageal adenocarcinomas. Uesugi et al34 previ- nign. A study of clear-cell renal carcinomas showed that ously reported APC as being mutated and/or deleted in hypermethylation of VHL promoter region was associ- primary oral squamous cell carcinoma tissues and sug- ated with absence of transcript expression. It was also gested that loss of APC function contributes to carcino- found that treatment of these methylated VHL tumors with genesis in the oral region. Promoter hypermethylation a demethylating agent resulted in reexpression of the VHL is also an important mechanism of APC inactivation in transcripts.36 oral cancers, occurring in 25% of oral squamous carci- TIMP3 induces apoptosis,37 inhibits angiogenesis,38 noma cells.34 impedes cell migration,39 and is a physiologic regulator In Barrett metaplasia and dysplasia,35 hypermethyl- of inflammation.40 Promoter methylation of TIMP3 ation of APC, CDKN2A, and ESR1 were usually found in has been observed in many tumor types41,42 and is in- a large contiguous field, suggesting either a concerted volved in the genesis of esophageal adenocarcinoma methylation change associated with metaplasia or a clonal notably during the progression from dysplasia to expansion of cells with abnormal hypermethylation. carcinoma.43,44 Hypermethylation of the TIMP3 and VHL genes oc- Hypermethylation of DAPK1, GSTP1, and HIC1 was curred in 5 of 14 RRP cases, and of the VHL gene in all 3 less frequent, occurring in 4 of 14 cases as well as in biopsy specimens of case 13. The VHL gene is a tumor biopsy specimens of recurrence in case 7. Death- suppressor gene located at 3p26-p25 and is responsible associated protein kinase 1, DAPK1, located at 9q34.1, for Von Hippel–Lindau syndrome, which is an inher- encodes a 160-kDa cytoskeletal-associated calcium/ ited familial cancer syndrome that makes patients sus- calmodulin-dependent serine/threonine kinase that was ceptible to a variety of neoplasms, malignant and be- initially identified as a positive mediator of interferon

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 688

4500 4000 Control DNA 3500 3000 2500 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480

1500 Case 7: biopsy 1

1000 CDKN2B APC GSTP1 500

0 120 160 200 240 280 320 360 400 440 480

Fluorescence Intensity 3000 CDKN2B Case 7: biopsy 2 2500 2000 1500 1000 500 0 120 160 200 240 280 320 360 400 440 480

3000 CDKN2B Case 7: biopsy 3 2500 2000 1500 APC GSTP1 1000 500 0 120 160 200 240 280 320 360 400 440 480 PCR Product Size, bp

Figure 4. Methylation-specific multiplex ligation-dependent probe amplification probe mix without and with HhaI enzyme (DNA sequencer, ABI 3100) in a control DNA specimen (A) and 3 biopsy specimens from case 7 (B-D). Note methylation of CDKN2B in all 3 biopsy specimens and methylation of APC and GSTP1 in specimens 1 and 3. bp indicates base pairs; PCR, polymerase chain reaction.

␥–induced programmed cell death in HeLa cells.45 Glutathione S-transferase ␲ (GSTP1) encodes for the DAPK1 expression is commonly lost in urinary bladder, glutathione S-transferase ␲ enzyme, which plays an im- breast, and B-cell neoplasms and renal cell carcinoma portant role in detoxification. It maps to 11q1354 and also cell lines because of promoter hypermethylation. Aber- has a role in susceptibility to cancer and other diseases. rant promoter methylation of DAPK1 has been shown to Inactivation of GSTP1 by promoter hypermethylation is frequently occur in human head and neck cancers,17,46 characteristic of corticosteroid-related neoplasms such non–small-cell lung carcinomas,47 gastric and colorectal as breast, liver, and prostate cancers.46,55 The ␲ class of carcinomas,48,49 and uterine cervical carcinomas.50 In glutathione S-transferase enzymes has been associated head and neck squamous cell carcinoma, DAPK1 pro- with preneoplastic and neoplastic changes.55 Promoter moter hypermethylation has been associated with metas- hypermethylation pattern of the p16, MGMT, GSTP1, and tasis to lymph nodes as well as advanced disease stage.46 DAPK genes have been used as molecular markers for can- HIC1 is a tumor suppressor gene that encodes a cer cell detection in paired serum DNA, and almost half transcriptional repressor with 5 Kruppel-like C2H2 of the patients with head and neck squamous cell carci- zinc finger motifs and an N-terminal BTB/POZ noma with methylated tumors were found to display these domain. Epigenetic silencing of HIC1 has been shown epigenetic changes in paired serum.46 to significantly influence tumorigenesis.51,52 Loss or Persistence of the same aberrantly methylated gene in reduced HIC1 messenger RNA in pediatric tumor cell 36% of multiple recurrent biopsy specimens (5 of 14 cases) lines with aberrantly methylated HIC1 became reex- in this study cohort supports a monoclonal origin for RRP. pressed in all cell lines by treatment with the demeth- Neoplasia typically develops as a clonal expansion from ylating agent 5-aza 2Јdeoxycytidine.53 a single cell of origin. There is ample experimental and

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 689

Patient No.

1a 2a 3a 4b 5a 6c 7b 8a 9a 10a 11b 12b 13b 14a 15a No. of biopsies 762642422333522 TIMP3d M MMMM APCd MMM MM MMM CDKN2Ad MM MMM MLH1 MM KLK10 MM M MEN1 M CDKN2Bd M MM M MMMM VHLd MM MMM TP73 MMM FANCD2 M BRCA2 MM IGSF4 MM RASSF1 M DAPK1d MMMM HIC1d MMMM ESR1 M M CDKN1B M BRCA1 MM GSTP1d MMMM

Abbreviation: M, methylated. a Cases with dissimilar epigenetic events in multiple biopsy specimens. b Cases with similar epigenetic events in multiple biopsy specimens. c Case with absence of M genes. d Commonly M genes (present in Ͼ3 cases).

behavior. Knowledge of whether a neoplasm has a Table 3. Epigenetically Linked Recurrent Laryngeal single or multiple cell origin may provide important Papilloma Cases information about its etiology and pathogenesis. The high frequency of epigenetic events characterized Patient No. Biopsy APC CDKN2B VHL TP73 GSTP1 by consistent aberrant promoter hypermethylation of multiple tumor suppressor genes points to the 4 1 [Reference] M use of gene silencing mechanisms as one of the driving 2 (10 mo) M M 3 (30 mo) M M forces behind the growth of recurrent laryngeal 7 1 [Reference] M M M papillomas. Additional studies to further confirm and 2 (3 mo) M validate the results of this study in a larger sample are 3 (6 mo) M M M in progress. 11 1 [Reference] M Recurrent genomic aberrations are good indicators of 2 (15 mo) M genes that are causally associated with cancer develop- 12 1 [Reference] M 2 (14 mo) M ment or progression. Because promoter hypermethyl- 13 1 [Reference] M M ation is potentially reversible, the molecules that regulate 2(1mo) M M M the methylation status of DNA are considered promising 3(3mo) M M M targets for new cancer therapies. Identifying epigenetic alterations in a precancerous lesion may lead to the discov- Abbreviation: M, methylated. ery of biomarkers that add to the knowledge of risk as- sessment and early detection, and may provide molecular targets for chemopreventive interventions. clinical evidence favoring a monoclonal origin of cancer, and some of the strongest arguments are derived from Submitted for Publication: August 1, 2006; final revi- cytogenetic investigations.10,56-58 sion received January 22, 2007; accepted February 27, More recent approaches have used a combination of 2007. fluorescence in situ hybridization and karyotyping of tu- Correspondence: Maria J. Worsham, PhD, Department mors with chromosome rearrangements as clonal mark- of Otolaryngology–Head and Neck Surgery, Henry Ford ers.11 Loss of heterozygosity patterns at different loci have Hospital, 1 Ford Pl, 1D, Detroit, MI 48202 (mworsha1 also been useful as clonal markers,59-61 and p53 muta- @hfhs.org). tions have been used as clonal markers.62-64 Author Contributions: Drs Stephen and Worsham had Clonal epigenetic alterations in precancerous lesions full access to all the data in the study and take respon- may reflect biological peculiarities pertinent to tumor sibility for the integrity of the data and the accuracy of

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 690

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 the data analysis. Study concept and design: Stephen, ease progression in head and neck squamous cell carcinoma. Arch Otolaryngol Benninger, and Worsham. Acquisition of data: Vaught, Head Neck Surg. 2003;129(7):702-708. 20. Kunjoonju JP, Raitanen M, Grenman S, Tiwari N, Worsham MJ. Identification of Chen, Shah, Schweitzer, Gardner, and Worsham. Analy- individual genes altered in squamous cell carcinoma of the vulva. Genes Chro- sis and interpretation of data: Stephen and Worsham. Draft- mosomes Cancer. 2005;44(2):185-193. ing of the manuscript: Stephen, Chen, Benninger, and 21. Worsham MJ, Pals G, Schouten JP, et al. High-resolution mapping of molecular Worsham. Critical revision of the manuscript for impor- events associated with immortalization, transformation, and progression to breast tant intellectual content: Stephen, Shah, Schweitzer, cancer in the MCF10 model. Breast Cancer Res Treat. 2006;96(2):177-186. 22. Go C, Schwartz MR, Donovan DT. Molecular transformation of recurrent respi- Gardner, Benninger, and Worsham. Statistical analysis: ratory papillomatosis: viral typing and p53 overexpression. Ann Otol Rhinol Worsham. Obtained funding: Worsham. Administrative, Laryngol. 2003;112(4):298-302. technical, and material support: Chen, Shah, Schweitzer, 23. Gupta D, Holden J, Layfield L. Topoisomerase alpha II, retinoblastoma gene prod- Benninger, and Worsham. Study supervision: Stephen, uct, and p53: potential relationships with aggressive behavior and malignant transformation in recurrent respiratory papillomatosis. Appl Immunohistochem Mol Gardner, Benninger, and Worsham. Morphol. 2001;9(1):86-91. Financial Disclosure: None reported. 24. Lele SM, Pou AM, Ventura K, Gatalica Z, Payne D. Molecular events in the pro- Funding/Support: This study was supported by R01 NIH gression of recurrent respiratory papillomatosis to carcinoma. Arch Pathol Lab DE 15990 from the National Institutes of Health. Med. 2002;126(10):1184-1188. Previous Presentation: This study was presented at the 25. Stern Y, Hurtubise PE, Cotton RT. Significance of DNA ploidy and cell proliferation in juvenile respiratory papillomatosis. Ann Otol Rhinol Laryngol. 1998; American Head and Neck Society 2006 Annual Meeting 107(10, pt 1):815-819. and Research Workshop; August 19, 2006; Chicago, 26. Gasco M, Sullivan A, Repellin C, et al. Coincident inactivation of 14-3-3␴ and Illinois. p16INK4a is an early event in vulval squamous neoplasia. Oncogene. 2002;21 (12):1876-1881. 27. Nuovo GJ, Plaia TW, Belinsky SA, Baylin SB, Herman JG. In situ detection of the REFERENCES hypermethylation-induced inactivation of the p16 gene as an early event in oncogenesis. Proc Natl Acad Sci U S A. 1999;96(22):12754-12759. 28. Sherr CJ. Cancer cell cycles. Science. 1996;274(5293):1672-1677. 1. Bauman NM, Smith RJ. Recurrent respiratory papillomatosis. Pediatr Clin North 29. Johnson DG, Walker CL. Cyclins and cell cycle checkpoints. Annu Rev Pharma- Am. 1996;43(6):1385-1401. col Toxicol. 1999;39:295-312. 2. Derkay CS. Task force on recurrent respiratory papillomas: a preliminary report. 30. Haber DA. Splicing into senescence: the curious case of p16 and p19ARF. Cell. Arch Otolaryngol Head Neck Surg. 1995;121(12):1386-1391. 1997;91(5):555-558. 3. Kashima H, Mounts P, Leventhal B, Hruban RH. Sites of predilection in recur- 31. Xing EP, Nie Y, Song Y, et al. Mechanisms of inactivation of p14ARF, p15INK4b, and rent respiratory papillomatosis. Ann Otol Rhinol Laryngol. 1993;102(8, pt 1): p16INK4a genes in human esophageal squamous cell carcinoma. Clin Cancer Res. 580-583. 1999;5(10):2704-2713. 4. Shah KV, Stern WF, Shah FK, Bishai D, Kashima HK. Risk factors for juvenile onset recurrent respiratory papillomatosis. Pediatr Infect Dis J. 1998;17(5): 32. Shintani S, Nakahara Y, Mihara M, Ueyama Y, Matsumura T. Inactivation of the 372-376. p14(ARF), p15(INK4B) and p16(INK4A) genes is a frequent event in human oral 5. Schultz S, Trudeau M. Juvenile laryngeal papilloma: a review of etiology and squamous cell carcinomas. Oral Oncol. 2001;37(6):498-504. treatment. Hearsay. 1988;2:30-32. 33. Yeh KT, Chang JG, Lin TH, et al. Epigenetic changes of tumor suppressor genes, 6. Armstrong L, Langston R. RRP Registry Update. Atlanta, GA: Centers for Dis- P15, P16, VHL and P53 in oral cancer. Oncol Rep. 2003;10(3):659-663. ease Control and Prevention; 1998. 34. Uesugi H, Uzawa K, Kawasaki K, et al. Status of reduced expression and hyper- 7. Erisen L, Fagan JJ, Myers EN. Late recurrences of laryngeal papillomatosis. Arch methylation of the APC tumor suppressor gene in human oral squamous cell Otolaryngol Head Neck Surg. 1996;122(9):942-944. carcinoma. Int J Mol Med. 2005;15(4):597-602. 8. Duggan MA, Lim M, Gill MJ, Inoue M. HPV DNA typing of adult-onset respira- 35. Eads CA, Lord RV, Kurumboor SK, et al. Fields of aberrant CpG island hyper- tory papillomatosis. Laryngoscope. 1990;100(6):639-642. methylation in Barrett’s esophagus and associated adenocarcinoma. Cancer Res. 9. Doyle DJ, Henderson LA, LeJeune FE Jr, Miller RH. Changes in human papillo- 2000;60(18):5021-5026. mavirus typing of recurrent respiratory papillomatosis progressing to malig- 36. Herman JG, Latif F, Weng Y, et al. Silencing of the VHL tumor-suppressor gene nant neoplasm. Arch Otolaryngol Head Neck Surg. 1994;120(11): by DNA methylation in renal carcinoma. Proc Natl Acad SciUSA. 1994;91 1273-1276. (21):9700-9704. 10. Worsham MJ, Wolman SR, Zarbo RJ. Molecular genetic fingerprints: clues to 37. Baker AH, Zaltsman AB, George SJ, Newby AC. Divergent effects of tissue in- monoclonal origin of multifocal disease. Mod Pathol. 1996;9(3): hibitor of metalloproteinase-1, -2, or -3 overexpression on rat vascular smooth 163-165. muscle cell invasion, proliferation, and death in vitro: TIMP-3 promotes apoptosis. 11. Worsham MJ, Wolman SR, Carey TE, Zarbo RJ, Benninger MS, Van Dyke DL. J Clin Invest. 1998;101(6):1478-1487. Common clonal origin of synchronous primary head and neck squamous cell 38. Qi JH, Ebrahem Q, Moore N, et al. A novel function for tissue inhibitor of metal- carcinomas: analysis by tumor karyotypes and fluorescence in situ hybridization. loproteinases-3 (TIMP3): inhibition of angiogenesis by blockage of VEGF bind- Hum Pathol. 1995;26(3):251-261. ing to VEGF receptor-2. Nat Med. 2003;9(4):407-415. 12. Garcia SB, Park HS, Novelli M, Wright NA. Field cancerization, clonality, and epi- 39. van der Laan WH, Quax PH, Seemayer CA, et al. Cartilage degradation and in- thelial stem cells: the spread of mutated clones in epithelial sheets. J Pathol. 1999; vasion by rheumatoid synovial fibroblasts is inhibited by gene transfer of TIMP-1 187(1):61-81. and TIMP-3. Gene Ther. 2003;10(3):234-242. 13. Ha PK, Califano JA. The molecular biology of mucosal field cancerization of the 40. Mohammed FF, Smookler DS, Taylor SE, et al. Abnormal TNF activity in Timp3. head and neck. Crit Rev Oral Biol Med. 2003;14(5):363-369. Nat Genet. 2004;36(9):969-977. 14. Egger G, Liang G, Aparicio A, Jones PA. Epigenetics in human disease and pros- 41. Momparler RL, Bovenzi V. DNA methylation and cancer. J Cell Physiol. 2000;183 pects for epigenetic therapy. Nature. 2004;429(6990):457-463. (2):145-154. 15. Esteller M. CpG island hypermethylation and tumor suppressor genes: a boom- 42. Esteller M, Corn PG, Baylin SB, Herman JG. A gene hypermethylation profile of ing present, a brighter future. Oncogene. 2002;21(35):5427-5440. human cancer. Cancer Res. 2001;61(8):3225-3229. 16. Nygren AO, Ameziane N, Duarte HM, et al. Methylation-specific MLPA (MS- 43. Bian YS, Osterheld MC, Fontolliet C, Bosman FT, Benhattar J. p16 inactivation MLPA): simultaneous detection of CpG methylation and copy number changes by methylation of the CDKN2A promoter occurs early during neoplastic of up to 40 sequences. Nucleic Acids Res. 2005;33(14):e128. progression in Barrett’s esophagus. Gastroenterology. 2002;122(4): 17. Worsham MJ, Chen M, Muduri V, et al. Epigenetic events of disease progres- 1113-1121. sion in head and neck squamous cell carcinoma. Arch Otolaryngol Head Neck 44. Eads CA, Lord RV, Wickramasinghe K, et al. Epigenetic patterns in the Surg. 2006;132(6):668-677. progression of esophageal adenocarcinoma. Cancer Res. 2001;61(8): 18. Schouten JP, McElgunn CJ, Waaijer R, Zwijnenburg D, Diepvens F, Pals G. 3410-3418. Relative quantification of 40 nucleic acid sequences by multiplex ligation- 45. Cohen O, Feinstein E, Kimchi A. DAP-kinase is a Ca2ϩ/calmodulin-dependent, cy- dependent probe amplification. Nucleic Acids Res. 2002;30(12):e57. toskeletal-associated protein kinase, with cell death–inducing functions that de- 19. Worsham MJ, Pals G, Schouten JP, et al. Delineating genetic pathways of dis- pend on its catalytic activity. EMBO J. 1997;16(5):998-1008.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 691

©2007 American Medical Association. All rights reserved. Downloaded From: https://jamanetwork.com/ on 09/27/2021 46. Sanchez-Cespedes M, Esteller M, Wu L, et al. Gene promoter hypermethylation quences near the ␲-class glutathione S-transferase gene accompanies human in tumors and serum of head and neck cancer patients. Cancer Res. 2000; prostatic carcinogenesis. Proc Natl Acad Sci U S A. 1994;91(24): 60(4):892-895. 11733-11737. 47. Esteller M, Sanchez-Cespedes M, Rosell R, Sidransky D, Baylin SB, Herman JG. 56. Worsham MJ, Carey TE, Benninger MS, et al. Clonal cytogenetic evolution in a Detection of aberrant promoter hypermethylation of tumor suppressor genes in squamous cell carcinoma of the skin from a xeroderma pigmentosum patient. serum DNA from non–small cell lung cancer patients. Cancer Res. 1999;59 Genes Chromosomes Cancer. 1993;7(3):158-164. (1):67-70. 57. Carey TE, Worsham MJ, Van Dyke DL. Chromosomal biomarkers in the clonal 48. Lee TL, Leung WK, Chan MW, et al. Detection of gene promoter hypermethyl- evolution of head and neck squamous neoplasia. J Cell Biochem Suppl. 1993; ation in the tumor and serum of patients with gastric carcinoma. Clin Cancer Res. 17F:213-222. 2002;8(6):1761-1766. 58. Carey TE, Van Dyke DL, Worsham MJ. Nonrandom chromosome aberrations and 49. Satoh A, Toyota M, Itoh F, et al. DNA methylation and histone deacetylation as- clonal populations in head and neck cancer. Anticancer Res. 1993;13(6B):2561- sociated with silencing DAP kinase gene expression in colorectal and gastric 2567. cancers. Br J Cancer. 2002;86(11):1817-1823. 59. Califano J, van der Riet P, Westra W, et al. Genetic progression model for head 50. Dong SM, Kim HS, Rha SH, Sidransky D. Promoter hypermethylation of mul- and neck cancer: implications for field cancerization. Cancer Res. 1996;56(11): tiple genes in carcinoma of the uterine cervix. Clin Cancer Res. 2001;7(7):1982- 2488-2492. 1986. 60. Scholes AG, Woolgar JA, Boyle MA, et al. Synchronous oral carcinomas: inde- 51. Chopin V, Leprince D. Chromosome arm 17p13.3: could HIC1 be the one [in pendent or common clonal origin? Cancer Res. 1998;58(9):2003-2006. French]? Med Sci (Paris). 2006;22(1):54-61. 61. Partridge M, Emilion G, Pateromichelakis S, Phillips E, Langdon J. Field canceri- 52. Rathi A, Virmani AK, Harada K, et al. Aberrant methylation of the HIC1 promoter sation of the oral cavity: comparison of the spectrum of molecular alterations in is a frequent event in specific pediatric neoplasms. Clin Cancer Res. 2003; cases presenting with both dysplastic and malignant lesions. Oral Oncol. 1997; 9(10, pt 1):3674-3678. 33(5):332-337. 53. Moscow JA, Townsend AJ, Goldsmith ME, et al. Isolation of the human anionic 62. el-Naggar AK, Lai S, Luna MA, et al. Sequential p53 mutation analysis of pre- glutathione S-transferase cDNA and the relation of its gene expression to estrogen- invasive and invasive head and neck squamous carcinoma. Int J Cancer. 1995; receptor content in primary breast cancer. Proc Natl Acad SciUSA. 1988; 64(3):196-201. 85(17):6518-6522. 63. Ribeiro U, Safatle-Ribeiro AV, Posner MC, et al. Comparative p53 mutational analy- 54. Esteller M, Corn PG, Urena JM, Gabrielson E, Baylin SB, Herman JG. Inactiva- sis of multiple primary cancers of the upper aerodigestive tract. Surgery. 1996; tion of glutathione S-transferase P1 gene by promoter hypermethylation in hu- 120(1):45-53. man neoplasia. Cancer Res. 1998;58(20):4515-4518. 64. Lydiatt WM, Anderson PE, Bazzana T, et al. Molecular support for field cancer- 55. Lee WH, Morton RA, Epstein JI, et al. Cytidine methylation of regulatory se- ization in the head and neck. Cancer. 1998;82(7):1376-1380.

(REPRINTED) ARCH OTOLARYNGOL HEAD NECK SURG/ VOL 133 (NO. 7), JULY 2007 WWW.ARCHOTO.COM 692