American Journal of Pathology, Vol. 152, No. 4, April 1998
Copyright ©) Amercan Societyfor Investigative Pathology
Methylation Alterations of the MyoDi Upstream Region Are Predictive of Subclassification of Human Rhabdomyosarcomas
Bin Chen,* Peter Dias,t Jesse J. Jenkins Van H. Savell,* and David M. Parham*
III,O
From the Department of Pathology,* University of Arkansas for Medical Sciences, and Arkansas Children's Hospital, Little Rock, Arkansas; PharMingen,t San Diego, California; and St. Jude Children's Research Hospital,t Memphis, Tennessee
MyoDl expression is a distinguishing characteristic of rhabdomyosarcoma. In this study, distinct methylation alterations were identified in the 5' flanking region of the MyoDI gene from the two major subtypes, ie, alveolar and embryonal rhabdomyosarcoma. The MyoDI methylation patterns of 26 rhabdomyosarcomas were compared with that of normal skeletal muscle and nonmuscle specimens by Southern blot analysis using methylation-sensitive restriction enzymes HhaI and HpaH. A 5-kb region immediately upstream of the MyoDI coding sequence was found to be methylated in adult muscle and all nonmuscle tissues tested. The MyoDI upstream region was unmethylated in the majority of the alveolar rhabdomyosarcomas (13 of 15, 87%) examined in this study. In contrast, 10 of 11 (91%) embryonal rhabdomyosarcomas showed a methylation pattern that was also observed in fetal muscle cells, in which the CpG sites in the MyoDI upstream region were partially methylated. Our data suggest that the methylation status of the MyoDI upstream CpG sites may be related to rhabdomyosarcoma tumorigenesis and may have valuable implications for its differential diagnosis. (Am JPathol 1998, 152:1071-1079)
Rhabdomyosarcoma is the most common soft-tissue malignancy in childhood, accounting for 4 to 8% of all pediatric cancers.1 Within the class of pediatric rhabdomyosarcomas, there are two major subtypes, embryonal and alveolar, which present with biologically and clinically distinct behaviors. Alveolar rhabdomyosarcomas usually have a more aggressive clinical behavior and are associated with a worse outcome than all embryonal variants.1 The molecular events that are involved in the development of a rhabdomyosarcoma remain sketchy; however, the discovery of the MyoD gene family has provided insight to the regulation of myogenesis and has
prompted the search for applicable markers that can assist the clinical diagnosis of rhabdomyosarcomas.24 The genes of the MyoD family create the nodal point of early myogenesis by up-regulating the expression of myogenic genes such as desmin, creatine kinase, and myosin.2 Members of this family, including MyoDl, myogenin, myf-5, and myf-6, are transcription factors characterized by a helix-loop-helix motif with an adjacent basic domain (bHLH).5- 8 Arbitrary expression of these genes causes initiation of myogenesis not only in primitive mesenchymal cells but also in cultured nonmuscle cells such as fibroblasts.2 In normal myogenesis, as occurring in growth and repair, MyoDl expression initiates a cascade of events leading to the formation of myotubes and is suppressed after terminal differentiation.9'10 Among the MyoD family genes, the expression of MyoDl has been found to be the most consistent molecular feature for rhabdomyosarcomas.3'4'11 As a result, the MyoDl protein has been recognized as a sensitive and specific marker for both childhood and adult rhabdomyosarcomas. 1,3 ,11
Abnormal patterns of DNA methylation are thought to play an important role in the development of cancer by altering gene expression and causing genomic instability.12 The human MyoDl gene is mapped at chromosome 11p15.4,13 adjacent to a number of imprinted genes including IGF2, H19, and p57KIP, all of which showed imprinting disturbances leading to aberrant gene expression associated with abnormal development and cancer.14-16 However, relatively little is known about the mechanisms that control the expression of the human MyoDl gene in normal myogenesis or in the development of rhabdomyosarcoma. Though sharing the common feature of expressing the MyoDl gene, embryonal and alveolar rhabdomyosarcomas have distinct etiologies. Loss of heterozygosity for chromosome 1 1p15, together with loss of imprinting for genes in this region, is found fre-
Supported in part by the Dean's CUMG Development Fund, University of Arkansas for Medical Sciences; by a National Cancer Institute core grant P30 CA217657; and by American Lebanese-Syrian Associated Charities. Accepted for publication January 30, 1998. Address reprint requests to Dr. David M. Parham, Department of Pathology, Slot 820, Arkansas Children's Hospital, 800 Marshall Street, Little Rock, AR 72202. E-mail:
[email protected]. P. Diaz's current address: Medinox, Inc., 11555 Sorrento Valley Road, Suite E, San Diego, CA 92121.
1071
1072
Chen et al
AJP April 1998, Vol. 152, No. 4
Table 1. Primers Used in the Current Study Primers
Sequences
Ml M2 MetF MetR Pax3/Pax7 FKHR reverse FKHR forward Pax3-specific Pax7-specific
5' ACT GCC AGC ACT TTG CTA TCT ACA 3' 5' CAC GGT CCT GGC ITI CGC CCA 3' 5' CCG AGT TTG GAG AGA TTG G 3' 5' GAC CCC GAG AGT TGA AGT G 3' 5' GAC AGC AGC TCT GCC TAC 3' 5' AT GAG CAT CCA CCA AGA AC 3' 5' GGT CM GAG CGT GCC CTA CT 3' 5' ACT GCC TCC CCA GCA CCA 3' 5' TTC TCC AGC TAC TCT GAC 3'
Materials and Methods Cloning and Sequencing of the Human MyoD1 5' Flanking Region
Figure 1. A: Alveolar rhabdomyosarcoma with typical staining pattern of diffuse, strong positivity (graded as + + +) for MyoDl. Negative cells represent lymphocytes and fibroblasts. B: The typical heterogeneous staining pattern of embryonal rhabdomyosarcoma, with tumor cells expressing variable amounts of MyoDl positivity (graded as +). Anti-human MyoDl, labeled streptavidin-biotin peroxidase stain, hematoxylin counterstain; original magnification, X200.
quently in embryonal rhabdomyosarcomas.4' 17 Alveolar rhabdomyosarcomas, on the other hand, are characterized by chromosomal translocations t(2;13) or t(1;13), which respectively generate the Pax3-FKHR or the Pax7FKHR fusion genes.18'19 Of interest to us has been our observation that patchy, heterogeneous expression of MyoDl is frequently observed in embryonal rhabdomyosarcoma, whereas strong diffuse positivity is usually observed in alveolar rhabdomyosarcomas (Figure 1). This phenomenon was also reflected by Northern blot analysis in a previous study by Scrable et al,4 which indicated higher levels of MyoDl mRNA transcripts in alveolar rhabdomyosarcomas as compared with tumors of the embryonal subtype. Based on these findings, we hypothesized that the expression of the MyoDl gene might be influenced by different regulatory mechanisms in these two tumor entities. The current study was aimed to investigate 1) whether altered MyoDl methylation is present in rhabdomyosarcomas, 2) whether differential methylation of the MyoDl CpG sites occurs in different rhabdomyosarcoma subtypes, and 3) whether assessment of the MyoDl CpG methylation has implications for differential diagnosis of rhabdomyosarcomas.
A human genomic EMBL3 SP6/T7 library (Clontech, Palo Alto, CA) was screened20 by using a 114-bp polymerase chain reaction (PCR)-amplified fragment that corresponds to nucleotides 7 to 120 of the published cDNA sequence of the human MyoDl gene.21 Primers Ml and M2 were used in the PCR amplification, and their sequences are shown in Table 1. Overlapping restriction fragments derived from the positive phage clones were then subcloned into the pGEM3Zf(+) vector (Promega, Madison, WI) according to the manufacturer's instructions. Sequencing analysis was performed on both strands of the subcloned DNA using a model 377 automated DNA sequencer (Perkin Elmer, Foster City, CA).
DNA Isolation from Frozen Tumor Specimens and Normal Tissues Frozen tumor material was obtained from St. Jude Children's Research Hospital (Memphis, TN) and from a bank of tumor tissues derived from a previous Intergroup Rhabdomyosarcoma Study.22 Diagnoses were obtained from previous Intergroup Rhabdomyosarcoma Study review or published retrospective analyses and were based on standard histological criteria.1 Normal tissue specimens were obtained from Departments of Pathology at Arkansas Children's Hospital and The University of Arkansas for Medical Sciences (Little Rock, AR). All tissue samples were snap-frozen in liquid-nitrogen-cooled isopentane immediately after surgical removal and stored at -80°C until DNA extraction. Genomic DNA was extracted from rhabdomyosarcoma and normal tissue specimens with the Puregene DNA extraction kit according to the manufacturer's instructions (Gentra Systems, Minneapolis, MN). All procedures performed for handling the specimens were in agreement with the ethical standards of the American College of Medical Genetics Storage of Genetics Materials Committee.23
Southern Blot Analysis The Southern blot analyses were performed as previously described.24 26 To analyze the methylation patterns of the MyoDl gene, each DNA sample was digested by
Altered MyoDl Methylation in Rhabdomyosarcomas 1073 AJP April 1998, Vol. 152, No. 4
1 kb -
MetF MetR
HhaI sites E
0-
B
A PB I
C D B E P
P
PP B II
A
A
A
B C
I
P A A
B I
Exon 1
D
MspIIHpaII sites PP2.0 PBO.5 - SSO.5 Figure 2. The 5' upstream region of the human MyoDl gene. Exon 1 is represented by the filled box. E, EcoRI; B, BamHI; P, PstIl. The upstream HpaII sites are shown by arrowheads, and the upstream HhaI sites are shown by long arrows. The relative positions of the primers MetF and MetR are indicated. EE4.0, BP2.0, PP2.0, PBO.5, and SSO.5 are subclones derived from the phage clones, and their positions relative to the HpaII and HhaI sites are indicated.
three groups of restriction enzymes: 1) Pstl only or Pstl plus Hhal, 2) Mspl or Hpall, and 3) Hhal, Hpall, or Hhal plus Hpall. Five micrograms each of normal and tumor DNA was separately digested with 40 U of Hhal, Hpall, Mspl, Pstl, 40 U each of Pstl and Hhal, or 40 U each of Hpall and Hhal (New England Biolabs, Beverly, MA) at 370C for 16 hours, followed by an additional 10 U of each enzyme for 8 hours. The digested DNA fragments were separated by electrophoresis in 1.2% agarose gels and blotted onto Hybond Plus nylon membrane (Amersham, Arlington Heights, IL) overnight. The subcloned DNA fragments, PP2.0, PBO.5, and SS0.5, were used as probes for hybridization. Approximately 100 ng of the plasmid insert DNA was 32P-labeled by random priming. All hybridizations were performed in 2X Denhardt's solution, 100 ,ug/ml denatured salmon sperm DNA, and 6X SSC at 650C for 24 hours. Washing was performed at 650C in 3X SSC for 1 hour, 2X SSC for 30 minutes, 1 X SSC for 30 minutes, 0.5X SSC for 30 minutes, and 0.1 X SSC for 30 minutes. The blots were then exposed to Hyperfilm-MP (Amersham) at -800C for 16 to 24 hours. The locations of the hybridization probes relative to the MyoDl gene are shown in Figure 2.
dithiothreitol, was incubated at 370C for 30 minutes and subsequently heated to 90°C to denature the reverse transcriptase. PCR amplification was performed with the same FKHR reverse primer as used in the reverse transcription and a consensus Pax3/7 primer that recognizes both Pax3 and Pax7. This primer pair amplifies a 219-bp fragment for Pax3-FKHR and a 206-bp fragment for Pax7-FKHR.28.29 A control amplification reaction was performed for each RNA sample using the FKHR reverse primer and the FKHR forward primer to amplify the normal FKHR transcript. The cycling parameters for both PCR reactions were 35 cycles of 940C for 30 seconds, 550C for 30 seconds, and 720C for 1 minute. Then the entire PCR reaction was electrophoresed through a 10% native polyacrylamide gel for 3 hours at 200 V. The gel was stained in a solution of 1X Tris-Boric acid-EDTA buffer containing 1 ,ug/ml ethidium bromide and photographed under UV light. To confirm the nature of the fusion genes, the samples that were positive for translocations were then amplified separately using the same FKHR reverse primer and a primer specific for Pax3 or a primer specific for Pax7. All primers were selected by the computer software Oligo 4.0 based on published FKHR, Pax3-FKHR, and Pax7-FKHR coding sequences.2829 The sequences of the RT-PCR primers are shown in Table 1.
Immunohistochemistry The expression of MyoDl protein in rhabdomyosarcoma specimens was examined as previously described.3 In brief, sections (6 ,um thick) were obtained from frozen blocks that had been stored at -800C and warmed to -200C. The sections were then fixed, permeabilized, and stained for MyoDl with a mouse monoclonal antibody (MAb) 5.8A, using a labeled avidin-biotin complex procedure.3
Methylation-Sensitive PCR RNA Extraction and Detection of Pax3-FKHR and Pax7-FKHR Fusion Genes in Alveolar
Rhabdomyosarcomas by Reverse Transcription
(RT)-PCR Total RNA from normal muscle, rhabdomyosarcoma, and non-rhabdomyosarcoma specimens was extracted with the Purescript RNA extraction kit according to the manufacturer's instructions (Gentra Systems). The concentration of the extracted RNA was determined by spectrophotometry. For detection of the Pax3-FKHR or Pax7FKHR fusion mRNA transcripts that are characteristic for alveolar rhabdomyosarcomas, the first strand of cDNA was synthesized as described27 with the following modifications. The 20-,lI mixture, containing 1 ,ug of RNA, 200 U of M-MLV reverse transcriptase (Promega), 20 pmol of FKHR reverse primer, 40 U of RNasin (Promega), and 1 mmol/L deoxyribonucleotide triphosphates (dATP, dCTP, dGTP, and TTP) in a buffer containing 50 mmol/L Tris/HCI (pH 8.3), 75 mmol/L KCI, 3 mmol/L MgCl2, and 10 mmol/L
One microgram each of normal or tumor DNA was digested with 10 U of Hhal (New England Biolabs) at 370C for 16 hours, followed by an additional 1 U of the enzyme for 4 hours. The undigested DNA used as controls in the PCR amplification was prepared under the same conditions, except that no Hhal was added to the reaction. Each reaction mixture consisted of 100 ng of DNA, 250 mmol/L of each dNTP, 1.5 mmol/L MgCO2, 50 mmol/L KCI, 5 pmol each of primers MetF and MetR, and 2 U of Taq DNA polymerase (Perkin Elmer) in a total volume of 25 ,l. Amplification was performed for all samples using 25 cycles of 940C for 30 seconds, 550C for 30 seconds, and 720C for 1 minute. After amplification in a Perkin Elmer 9600 thermal cycler, the PCR products were electrophoresed through a 2% agarose gel containing 0.5 ,ug/ml ethidium bromide and photographed under UV light. All samples were analyzed at least twice in duplicate. The positions of primers MetF and MetR in the MyoDl upstream region are indicated in Figure 2 and their sequences are shown in Table 1.
Chen et al
1074
AJP April 1998, Vol. 152, No. 4
A i
A
I A..
kb 2
-
-T
r
,
-
I!+ kb 2-
I =
i-
-
A
A
a
mI
kb 5
I
0.5 -
-
3-
0.5
0.25'
Figure 3. Methylation of the MyoDl upstream region from nornal tissues, represented by adult muscle, liver, and lung. A: PstI only or PstI/HbaI double digestion, hybridized to probe PP2.0. B: The same blot was stripped and hybridized to probe PBO.5. Notice the presence of the methylated 1-kb band in both hybridizations. C: MspI or HpaII digestion, hybridized to probe PP2.0. A 5-kb band is present in all HpaII digestions, indicating the hypermethylated MyoDl upstream region. P, PstI digestion; P+H, PstI and HhaI double digestion; M, MsI digestion.
Results Cloning of the 5' Upstream Region of the Human MyoDl Gene Of approximately 3 x 105 phage clones screened, three overlapping clones were identified that contained the exon 1 sequence of the human MyoDl gene. A restriction map was constructed, and the 5' region of the cloned MyoDl gene is shown in Figure 2. The sequence of the MyoDl upstream region was submitted to GenBank (Accession number AF027148). The HpallIMspl and Hhal sites in the 5' flanking region are shown in Figure 2.
The Upstream Region of the Human MyoDl Gene Is Methylated in Differentiated Muscle and Nonmuscle Tissues The methylation pattern of the MyoDl gene was examined in normal skeletal muscle (n = 6) and nonmuscle tissues, including liver (n = 4), lung (n = 4), pancreas (n = 3), peripheral lymphocytes (n = 20), heart (n = 3), kidney (n = 3), and spleen (n = 2), by Southern blot analysis using the methylation-sensitive restriction enzymes Hhal and Hpall. Methylation of the Hhal recognition site GCGC at the internal cytosine residue inhibits the enzyme digestion. Similarly, the presence of an internal 5-methylcytosine residue in the Hpall site CCGG prevents Hpall digestion. Mspl is the methylation-insensitive isozyme of Hpall and digests DNA whether the internal CpG is methylated or not. Therefore, Mspl digestion was used as a control for Hpall digestion, as the difference would reflect CpG methylation of the DNA being analyzed. Methylation of the MyoDl gene was detected in all nonmuscle and adult muscle tissues, as evidenced by a 1-kb fragment on PstlIHhal digestion and a 5-kb fragment on Hpall digestion (Figure 3). The 1-kb band on PstlIHhal digestion was present whether the 2-kb Pstl fragment (PP2.0) or the 556-bp Pstl-BamHl fragment (PBO.5) was
used as a hybridization probe (Figure 3, A and B). When the same blots were hybridized to a 565-bp Smal fragment (SSO.5, nucleotides -229 through +336), no fragment of increased size was observed (not shown). Therefore, the 1-kb PstlIHhal fragment was generated due to CpG methylation of Hhal site C (Figure 2). A 580-bp and a 420-bp band, as well as fragments smaller than 200 bp, were observed for all normal tissues in hybridization to probe PP2.0 (Figure 3A). When PB0.5 was used as a probe to hybridize the same blots, the 1-kb and the 580-bp fragments remained and the smaller fragments disappeared (Figure 3B). This indicates that the Hhal site C is not completely methylated in the cells from nonmuscle or adult muscle tissues and that the 580-bp fragment results from digestion of the DNA unmethylated at Hhal site C. The 27 additional Hhal sites throughout the 2-kb Pstl fragment, which contains a 1.4-kb 5' noncoding region and the first 602 bases of the MyoDl exon 1, did not show CpG methylation. Methylation of the MyoDl upstream region from normal tissues was detected also with the methylation-sensitive enzyme Hpall, as evidenced by a 5-kb band present in all Hpall-digested normal tissue DNAs (Figure 3C). A 1.2-kb fragment was also observed in the Hpall digestions and is the only fragment present in the Mspl digestions. This hybridization pattern remained unchanged whether PP2.0 or PB0.5 was used as a hybridization probe. However, when SS0.5 was used to probe the same blots, neither the 5-kb nor the 1.2-kb band was detectable (not shown). These findings indicate that the 5-kb fragment resulted from CpG methylation of the upstream Hpall sites B and C and that the 1.2-kb fragment was the product of Hpall digestion of the DNA unmethylated at sites C and D. Twelve additional Hpall sites are present in the region extending from Hpall site D through the end of exon 1, but these sites were not found to have CpG methylation in the tissues examined. To further study the distribution of methylated CpGs in the MyoDl upstream region, an HhalIHpall double digestion was performed to test whether the methylated Hpall sites were located on the same DNA strand as the methylated Hhal site C. In all nonmuscle tissues as well as in the six adult muscle specimens, the methylated Hpall sites were on the same DNA strand as the methylated Hhal site C, as indicated by the persistence of the 5-kb fragment harboring both methylated Hhal and methylated Hpall sites and by a 667-bp fragment generated by digestion of a DNA strand unmethylated at both Hpall site C and Hhal site C (Figure 4, represented by a muscle DNA sample from a 16-year-old individual and a liver DNA sample).
Partial Methylation of the MyoD1 Upstream CpG Sites in Embryonal Rhabdomyosarcomas and in Normal Fetal Muscles MyoDl methylation of 11 embryonal rhabdomyosarcomas was studied using the same Southern blot method as described in the previous paragraph. One embryonal rhabdomyosarcoma (E4) was found to have a deletion in
1075
Altered MyoDl Methylation in Rhabdomyosarcomas
AJP April 1998, Vol. 152, No. 4
16 wk Muscle ~3t+
30 wk
16 yo
Muscle
Muscle
t3
+
~3
+
Normal Liver t3
A 2
Embryonal RMS #8
B
mI :M51
3+
Z
+
1
1
3
=I,
kb 2 -
kb 5-
3-
10-
2
4
3
I
A.
:.
CA.
A.1
A+.
AL.
kb 2-
I
-
10-
0.5
-
0.5-
0.25-
0.25-
m +t
0.5D
C
Figure 4. Partial methylation of the MyoDl upstream region in embryonal rhabdomyosarcomas and in fetal muscle tissues. Lanes 1, 4, 7, 10, and 13, HpaII digestion; lanes 2, 5, 8, 11, and 14, HhaI digestion; lanes 3, 6, 9, 12, and 15, HpaII/HbaI double digestion. The blot was hybridized to probe PBO.5. Notice that the 1.1-kb band on HpaII/HhaI double digestion was present in the embryonal tumor DNA and in the fetal muscle DNAs but not in the 16-year-old normal muscle or liver DNA samples.
1
1
2
3
m mlm
4
1
m--
-
9
-
kb 5
I10.5
3
-
0.25-
1
2 kb 2-
2
r--
1
2
l
it
kb 5
3,
0.5. 0.25
I
Figure 6. Hypomethylation of the MyoDl upstream region in alveolar rhabdomyosarcomas. A to C: PstI or PstI/HhaI digestion. D: MspI or HpaII digestion. A: Hybridization to probe PP2.0. Lane 1, peripheral blood sample; lane 2, normal skeletal muscle sample; lane 3, alveolar rhabdomyosarcoma (Al). B: A comparison between hybridizations with probe PP2.0 and probe PBO.5. Lane 1, embryonal tumor E8; lane 2, alveolar tumor A14. The blot (lanes 1 and 2) was hybridized to probe PP2.0. The same blot was rehybridized to probe PBO.5, and the result is shown in lanes 3 (E8) and 4 (A14). C: Alveolar rhabdomyosarcoma A13 (lanes 1 and 3) and embryonal tumor E8 (lanes 2 and 4), hybridized to PP2.0 (lanes 1 and 2) or PBO.5 (3 and 4). D: MspI or HpaII digestion of DNA from A14 (lane 1) and E8 (lane 2), hybridized to probe PP2.0. Notice that the 5-kb methylated band is absent in the alveolar tumor A14 but present in the embryonal rhabdomyosarcoma E8. P, PstI digestion; P+H, double digestion by PstI and HhaI; M, MspI digestion.
2
m
+w +1+w kb 2
4
both embryonal rhabdomyosarcomas and fetal myocytes.
B
A
3
I
I-
the MyoDl upstream region (Figure 5). In 10 of 11 (91%) tumors, PstlIHhal or Hpall digestions did not show significant methylation differences from the normal tissues (not shown). However, a distinct HpallIHhal digestion pattern was observed for the 10 nondeleted embryonal rhabdomyosarcomas and for 5 skeletal muscle specimens from aborted fetuses (gestational age, 20 to 33 weeks). As shown in Figure 4, in addition to the 666-bp Hpall-Hhal fragment, a novel band, approximately 1.1 kb, was present in the Hpall/Hhal-digested DNA from fetal muscles and embryonal rhabdomyosarcomas. This extra band could be accounted for only by the 1082-bp fragment from Hpall site C to Hhal site D, a result of digestion at an unmethylated Hpall site C on a DNA strand that carries a methylated Hhal site C (Figure 4). Therefore, partially methylated MyoDl upstream DNA is present in
2
I
Figure 5. A deletion at the MyoDl upstream region in an embryonal rhabdomyosarcoma (E4). A: PstI or PstI/HbaI double digestion, hybridized to probe PP2.0. In lane 1, the 0.6-kb band is replaced by a smaller fragment in E4. 2. Another embryonal rhabdomyosarcoma (E7) shows the methylated 1-kb band. In lanes 3 and 4, the same blot was stripped and rehybridized to probe PBO.5. B: MspI or HpaII digestion of DNA from E7 (lane 1) or E4 (lane 2 ), hybridized to probe PP2.0. Notice the absence of the 5-kb band and the replacement of the 1.2-kb band with a novel 1-kb fragment in E4. P, PstI digestion; P+H, PstI and HhaI double digestion; M, MspI digestion.
Hypomethylation of the MyoD 1 Upstream Region in Alveolar Rhabdomyosarcomas The methylation status of the upstream CpG sites in 15 alveolar rhabdomyosarcomas was analyzed with the same Southern blot strategies. Hypomethylation of the MyoDl upstream region was found in 13 of 15 (87%) alveolar rhabdomyosarcomas examined in this study, revealed both by the absence of the 1-kb band on the PstlIHhal digestion (Figure 6, A-C) and by the absence of the 5-kb band on the Hpall digestion (Figure 6D). Therefore, the upstream CpG sites that showed methylation in both normal tissues and embryonal rhabdomyosarcomas were completely unmethylated in the majority of the alveolar rhabdomyosarcomas. The remaining two tumors (A8 and A9) presented with a MyoDl methylation pattern similar to that of the partially methylated embryonal rhabdomyosarcomas by all three groups of restriction digestions (not shown). A8 arose from the buttock of an 8-year-
1076 Chen et al AJP April 1998, Vol. 152, No. 4
Table 2. Methylation Status of the MyoDl Gene and RT-PCR Analysis of Pax3/7-FKHR Fusion Genes in Alveolar and Embryonal
Rhabdomyosarcomas
Pax3/7-FKHR Tumors
MyoDl protein
Diagnosis
RT-PCR*
expressiont
Methylation of MyoDl 5' region
Al A2 A3 A4 A5 A6 A7 A8 A9 A10 Al1 A12 A13 A14 A15 El E2 E3 E4 E5 E6 E7 E8 E9 E10 Ell
Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Alv rms Emb rms Emb rms Emb rms Emb rms Emb rms Emb rms Emb rms Emb rms Emb rms Emb rms B rms
+ + NT + +
++ NP +++ +++ NP NP NP NP NP NP NP
Unmethylated Unmethylated Unmethylated Unmethylated Unmethylated Unmethylated Unmethylated Partially methylated Partially methylated Unmethylated Unmethylated Unmethylated Unmethylated Unmethylated Unmethylated Partially methylated Partially methylated Partially methylated
+
NT + + +
+++ +++ +++ +++ + + + +++
NT NT NT NT -
NP NP
-
+ + +
-
NP NP
-
Unmethylatedt Partially methylated Partially methylated Partially methylated Partially methylated Partially methylated Partially methylated Partially methylated
Alv, alveolar; rms, rhabdomyosarcoma; Emb, embryonal; B, botryoid. *Results of simultaneous amplification of Pax3-FKHR and Pax7-FKHR using the Pax3/7 consensus primer. Subsequent PCR analysis using Pax3specific or Pax7-specific primers indicated the presence of Pax7-FKHR in tumor A13 and Pax3-FKHR in the rest of the translocation positive tumors. NT, not tested because RNA was not available. tResults of immunohistochemistry using MAb 5.8A. +++, positive-staining cells >50%; ++, positive-staining cells between 25 and 50%; +, positive-staining cells between 10 and 25%. NP, not performed. $This tumor has a short deletion in the 5' flanking region of the MyoDl gene (see text and Figure 5).
old child and A9 from the hand of a 17-year-old patient. Both tumors had alveolar characteristics by histological examination, but neither could be confirmed as having a t(2;13) or a t(1;13) translocation by RT-PCR analysis (Table 2).
Correlation of Chromosomal Translocations, MyoDl Expression, and Hypomethylation of the MyoD1 Upstream Region in Alveolar Rhabdomyosarcomas To confirm the histological diagnosis of the alveolar rhabdomyosarcomas, RT-PCR analysis was performed to detect Pax3-FKHR or Pax7-FKHR mRNA transcripts in the rhabdomyosarcomas. Total RNA was successfully isolated from 1 1 of the 15 alveolar rhabdomyosarcomas and 9 of the 11 embryonal rhabdomyosarcomas. The Pax3FKHR fusion transcripts were observed in 7 tumors and the Pax7-FKHR transcripts in 1 tumor, all of which had a hypomethylated MyoDl upstream region (Figure 7). MyoDl immunohistochemistry was performed for 7 of the 15 alveolar rhabdomyosarcomas and 6 of the 11 embryonal rhabdomyosarcomas. Strong MyoDl expression (++ or +++) was detected in all seven alveolar tumors (Table 2). Of the six embryonal tumors tested, five had a patchy and heterogeneous MyoDl immunostaining
pattern (graded as +), whereas the remaining tumor (E4) showed strong diffuse MyoDl positivity (graded as + + +). Interestingly, the MyoDl upstream region in tumor E4 was found unmethylated and a deletion was detected in this region (Figure 5). Table 2 summarizes the results of chromosomal translocations, MyoDl expression, and methylation state of the MyoDl upstream region from alveolar and embryonal rhabdomyosarcomas.
Al A2 A4 A5 A7 A13 E3 E4 E5
219
bp-
151
bp-
Pax3/7-FKHR
FKHR
Figure 7. RT-PCR analysis of the Pax3-FKHR or Pax7-FKHR mRNA transcripts in alveolar rhabdomyosarcomas. Six alveolar (Al, A2, A4, A5, A7, and A13) and three embryonal (E3, E4, and E5) rhabdomyosarcomas were analyzed in this run. Tumors Al, A2, A4, A5, and A7 yield a 219-bp amplification product indicating the Pax3-FKHR fusion, and A13 shows a smaller PCR product consistent with the Pax7-FKHRfusion. The normal FKHRrmRNA was amplified for each tumor in a control PCR reaction and is shown below.
Altered MyoDl Methylation in Rhabdomyosarcomas 1077 AJP April 1998, Vol. 152, No. 4
2
1
3
n n U D U D
m
bp 300200100-
U
D
4
5
6
7
8 m
U D U DU D U DU D
Figure 8. Detection of MyoDl CpG hypomethylation in alveolar rhabdomyby a methylation-sensitive PCR assay. Lanes 1 to 6, six alveolar rhabdomyosarcomas; lane 7, an embryonal rhabdomyosarcoma; lane 8, a normal skeletal muscle sample. U, amplified with undigested DNA; D, amplified with HhaI-digested DNA. After PCR amplification, the reactions were electrophoresed through a 2% agarose gel and visualized under UV light. osarcomas
Detection of the MyoDl Hypomethylation in Alveolar Rhabdomyosarcomas by a Methylation-Sensitive PCR Assay In this study, Southern blot analyses indicated that methylation of Hhal site C in the MyoDl upstream region, which led to the formation of the 1-kb fragment on PstlI Hhal digestion, occurred in embryonal rhabdomyosarcomas and in all normal tissues. In contrast, complete unmethylation of this Hhal site was found only in alveolar rhabdomyosarcomas. To explore the diagnostic utility of this finding, a PCR assay was developed based on the methylation status of the upstream Hhal site C. When the internal CpG of this enzyme site is methylated, the DNA template would be protected from Hhal digestion and would lead to PCR amplification by extension of the two primers, MetF and MetR, which flank the Hhal site. Because Hhal site C is completely unmethylated in alveolar rhabdomyosarcoma cells, the DNA templates would be disrupted by Hhal digestion and no amplification would result. Under the experimental conditions described in this report, none of the 13 tumors with MyoDl CpG hypomethylation showed detectable amplification, whereas all normal tissues and 10 of the 11 embryonal rhabdomyosarcoma DNA samples demonstrated PCR products (Figure 8).
Discussion The present study indicates that methylation alterations occur in the 5' flanking region of the human MyoDl gene in rhabdomyosarcomas and that the two major pediatric subtypes, alveolar and embryonal rhabdomyosarcoma, have distinct MyoDl methylation patterns. Partially methylated MyoDl upstream DNA is present in embryonal rhabdomyosarcomas as well as in skeletal muscle cells from normal fetuses. In contrast, the same CpG sites in the MyoDl upstream region are completely unmethylated in the majority of alveolar rhabdomyosarcomas. Our previous data and those of others have suggested that alveolar rhabdomyosarcomas usually show higher MyoDl expression than tumors of the embryonal subtype, which resemble fetal muscle tissue with regard to both histology and expression of MyoDl.3,4,30 Therefore, the distinct MyoDl methylation patterns in these two malignancies are consistent with their divergent levels of MyoDl ex-
pression. Additional studies, including reporter gene assays and genomic sequencing, are currently underway in our laboratory to determine the role of promoter methylation in MyoDl transcription regulation both in rhabdomyosarcoma cells and during normal myogenesis. Our discovery that hypomethylation of the MyoDl 5' region is associated with a human malignancy appears to represent a novel finding. More importantly, the high level of MyoDl expression in alveolar rhabdomyosarcoma potentially could be explained by the hypomethylated status of the promoter region. Hypermethylation of certain intragenic CpG sites in the MyoDl gene has been reported in several forms of human malignancies, including colorectal cancer,31 breast carcinomas,32 and ovarian carcinomas.33 However, as these tumors are of nonmyogenic cell origin, the observed MyoDl hypermethylation was suggested to be a marker of neoplasia in general rather than of cellular proliferation. Also, as cDNA probes of either mouse or human origin were used in these previous studies, only those methylation alterations involving the MyoDl coding region were detected. Therefore, the cloning of the MyoDl upstream region, together with the finding that this region is methylated in normal nonmuscle tissues, should provide a new tool for reassessing the significance of the MyoDl hypermethylation in non-rhabdomyosarcoma tumors. Methylation plays an important role in genomic imprinting, a phenomenon that distinguishes alleles of different parental origin and results in their differential expression.12 The human MyoDl gene is located at 1 1p15.4,13 adjacent to a number of imprinted genes.34 36 Previous studies indicated that disturbance of normal imprinting at 11p15 led to bi-allelic IGF2 expression and suppressed H19 expression in a proportion of rhabdomyosarcomas.37'38 Taken together with our findings, it is likely that the altered MyoDl methylation in rhabdomyosarcomas is the result of a generalized imprinting disturbance in the 11p15 region, and MyoD1 expression might simply be a concomitant event during tumorigenesis. Alternatively, MyoDl might have a role in the development of rhabdomyosarcoma other than its classic role of promoting cellular differentiation. This hypothesis awaits future testing, although previous studies with the RD rhabdomyosarcoma cell line indicated that MyoDl expression did not affect tumorigenicity.39 Results of the present study indicate an interesting correlation of chromosomal translocations, hypomethylation of the MyoDl upstream region, and MyoDl expression in alveolar rhabdomyosarcomas. However, as summarized in Table 2, this correlation is not strict. Three of the fifteen alveolar tumors, ie, A6, A8, and A9, had no detectable Pax3-FKHR or Pax7-FKHR fusions by RT-PCR. Tumor A6 arose from the nose of a 16-year-old patient and was diagnosed as an alveolar rhabdomyosarcoma based on typical histology. Though negative for the characteristic fusion transcripts, A6 had a hypomethylated MyoDl upstream region as the majority of alveolar tumors examined in our study. As shown by previous studies, neither molecular assays nor cytogenetic investigation could detect the Pax3/Pax7-FKHR fusions in a small percentage of alveolar rhabdomyosarcoma cases.40'41
1078 Chen et al AJP April 1998, Vol.
152, No. 4
Therefore, identification of the MyoDl methylation status may be helpful in some fusion-negative alveolar rhabdomyosarcomas, as with our case A6. The remaining two fusion-negative tumors, A8 and A9, presented with a partial methylation pattern in the MyoDl 5' region more similar to the embryonal tumors, despite their alveolar histology. Clarification of the nature of these tumors awaits further study. The methylation-sensitive PCR assay based on the methylation status of Hhal site C in the MyoDl upstream region provides a sensitive and specific diagnostic test that could be used to assist the diagnosis of alveolar rhabdomyosarcoma. Immunohistochemistry detects MyoDl expression present in both alveolar and embryonal subtypes, but the results can be difficult to interpret. Fluorescence in situ hybridization and RT-PCR both identify the Pax3- or Pax7-FKHR translocations characteristic for alveolar rhabdomyosarcomas but are also technically challenging. Compared with the above, the PCR assay presented in this study allows a quick and reliable assessment of the methylation status of the MyoDl upstream region and might be useful in future diagnostic and biological studies. Finally, although our results are intriguing, the significance of MyoDl methylation in the tumorigenesis of rhabdomyosarcoma and the mechanisms by which this phenomenon occurs await future elucidation.
11. 12. 13.
14.
15. 16.
17.
18. 19. 20. 21.
Acknowledgments
22.
We thank Ms. Helen 0. A. Alexander for excellent secretarial assistance. 23.
References 1. Parham DM: The molecular biology of childhood rhabdomyosarcoma. Semin Diagn Pathol 1994, 11:39-46 2. Weintraub H, Davis R, Tapscott S, Thayer M, Krause M, Benezra R, Blackwell TK, Turner D, Rupp R, Hollenberg S, Zhuang Y, Lassar A: The myoD gene family: nodal point during specification of the muscle cell lineage. Science 1991, 251:761-766 3. Dias P, Parham DM, Shapiro DN, Tapscott SJ, Houghton PJ: Monoclonal antibodies to the myogenic regulatory protein MyoDl: epitope mapping and diagnostic utility. Cancer Res 1992, 52:6431-6439 4. Scrable H, Witte D, Shimada H, Seemayer T, Wang-Wuu S, Soukup S, Koufos A, Houghton P, Lampkin B, Cavenee W: Molecular differential pathology in rhabdomyosarcoma. Gene Chromosomes & Cancer 1989, 1:23-35 5. Davis RL, Weintraub H, Lassar AB: Expression of a single transfected cDNA converts fibroblasts to myoblasts. Cell 1987, 51:987-1000 6. Braun T, Bober E, Winter B, Rosenthal N, Arnold HH: Myf-6, a new member of the human gene family of myogenic determination factors: evidence for a gene cluster on chromosome 12. EMBO J 1990, 9:821-831 7. Rhodes SJ, Konieczny SF: Identification of MRF4: a new member of the muscle regulatory factor gene family. Genes Dev 1989, 3:20502061 8. Wright WE, Sassoon DA, Lin VK: Myogenin, a factor regulating myogenesis, has a domain homologous to MyoD. Cell 1989, 56:607-617 9. Montarras D, Chelly J, Bober E, Arnold H, Ott MO, Gros F, Pinset C: Developmental patterns in the expression of Myf5, MyoD, myogenin, and MRF4 during myogenesis. New Biol 1991, 3:592-600 10. Ott MO, Bober E, Lyons G, Arnold H, Buckingham M: Early expression of the myogenic regulatory gene, myf-5, in precursor cells of
24.
25. 26. 27.
28.
29.
30.
31. 32.
33.
skeletal muscle in the mouse embryo. Development 1991, 111:10971107 Tapscott SJ, Thayer MJ, Weintraub H: Deficiency in rhabdomyosarcomas of a factor required for MyoD activity and myogenesis. Science 1993, 259:1450-1453 Bird AP: The relationship of DNA methylation to cancer. Cancer Surv 1996, 28:87-101 Scrable HJ, Johnson DK, Rinchik EM, Cavenee WK: Rhabdomyosarcoma-associated locus and MYOD1 are syntenic but separate loci on the short arm of human chromosome 11. Proc Natl Acad Sci USA 1990, 87:2182-2186 Ogawa 0, Eccles MR, Szeto J, McNoe LA, Yun K, Maw MA, Smith PJ, Reeve AE: Relaxation of insulin-like growth factor 11 gene imprinting implicated in Wilms' tumour. Nature 1993, 362:749-751 Taniguchi T, Okamoto K, Reeve AE: Human p57(KIP2) defines a new imprinted domain on chromosome 11p but is not a tumour suppressor gene in Wilms tumour. Oncogene 1997, 14:1201-1206 Taniguchi T, Sullivan MJ, Ogawa 0, Reeve AE: Epigenetic changes encompassing the IGF2/H19 locus associated with relaxation of IGF2 imprinting and silencing of H19 in Wilms tumor. Proc Natl Acad Sci USA 1995, 92:2159-2163 Ohlsson R, Nystrom A, Pfeifer-Ohlsson S, Tohonen V, Hedborg F, Schofield P, Flam F, Ekstrom TJ: IGF2 is parentally imprinted during human embryogenesis and in the Beckwith-Wiedemann syndrome. Nature Genet 1993, 4:94-97 Barr FG, Galili N, Holick J, Biegel JA, Rovera G, Emanuel BS: Rearrangement of the PAX3 paired box gene in the paediatric solid tumour alveolar rhabdomyosarcoma. Nature Genet 1993, 3:113-117 Davis RJ, D'Cruz CM, Lovell MA, Biegel JA, Barr FG: Fusion of PAX7 to FKHR by the variant t(1;13)(p36;q14) translocation in alveolar rhabdomyosarcoma. Cancer Res 1994, 54:2869-2872 Maniatis T, Fritsch EF, Sambrook J (Eds): Molecular Cloning: A Laboratory Manual, ed 2. Cold Spring Harbor, NY, Cold Spring Harbor Laboratory Press, 1987 Pearson-White SH: Human MyoD: cDNA and deduced amino acid sequence. Nucleic Acids Res 1991, 19:1148 Parham DM, Webber B, Holt H, Williams WK, Maurer H: Immunohistochemical study of childhood rhabdomyosarcomas and related neoplasms: results of an Intergroup Rhabdomyosarcoma Study Project. Cancer 1991, 67:3072-3080 American College of Medical Genetics Storage of Genetic Materials Committee: ACMG statement: statement on storage and use of genetic materials. Am J Hum Genet 1995, 57:1499-1500 Rideout WM ll, Eversole-Cire P, Spruck CH l1l, Hustad CM, Coetzee GA, Gonzales FA, Jones PA: Progressive increases in the methylation status and heterochromatinization of the myoD CpG island during oncogenic transformation. Mol Cell Biol 1994, 14:6143-6152 Southern EM: Detection of specific sequences among DNA fragments separated by gel electrophoresis. J Mol Biol 1975, 98:503-517 Tornaletti S, Pfeifer GP: Complete and tissue-independent methylation of CpG sites in the p53 gene: implications for mutations in human cancers. Oncogene 1995, 10:1493-1499 Friedman EY, Rosbash M: The synthesis of high yields of full-length reverse transcripts of globin mRNA. Nucleic Acids Res 1977, 4:34553471 Galili N, Davis RJ, Fredericks WJ, Mukhopadhyay S, Rauscher FJ ll, Emanuel BS, Rovera G, Barr FG: Fusion of a fork head domain gene to PAX3 in the solid tumour alveolar rhabdomyosarcoma. Nature Genet 1993, 5:230-235 Schafer BW, Czerny T, Bernasconi M, Genini M, Busslinger M: Molecular cloning and characterization of a human PAX-7 cDNA expressed in normal and neoplastic myocytes. Nucleic Acids Res 1994, 22:4574-4582 Dias P, Parham DM, Shapiro DN, Webber BL, Houghton PJ: Myogenic regulatory protein (MyoDl) expression in childhood solid tumors: diagnostic utility in rhabdomyosarcoma. Am J Pathol 1990, 137:1283-1291 lacopetta BJ, Harmon D, Spagnolo DV, House AK, Kay PH: Hypermethylation of the Myf-3 gene in human colorectal cancer. Anticancer Res 1997, 17:429-432 Hahnel R, Harvey J, Kay P: Hypermethylation of the myogenic gene Myf-3 in human breast carcinomas. Anticancer Res 1996, 16:211 12115 Cheng P, Schmutte C, Cofer KF, Yu MC, Dubeau L: Alterations in
Altered MyoDl Methylation in Rhabdomyosarcomas 1079 AJP ApIl 1998, Vol. 152, No. 4
34.
35.
36. 37.
38.
DNA methylation are early, but not initial, events in ovarian tumorigenesis. Br J Cancer 1997, 75:396-402 Zhan S, Shapiro DN, Helman LJ: Loss of imprinting of IGF2 in Ewing's sarcoma. Oncogene 1995, 11:2503-2507 Pedone PV, Tirabosco R, Cavazzana AO, Ungaro P, Basso G, Luksch R, Carli M, Bruni CB, Frunzio R, Riccio A: Mono- and bi-allelic expression of insulin-like growth factor 11 gene in human muscle tumors. Hum Mol Genet 1994, 3:1117-1121 Issa JP, Baylin SB, Herman JG: DNA methylation changes in hematologic malignancies: biologic and clinical implications. Leukemia 1997, 11(Suppl 1):S7-S11 Zhan S, Shapiro DN, Helman LJ: Activation of an imprinted allele of the insulin-like growth factor 11 gene implicated in rhabdomyosarcoma. J Clin Invest 1994, 94:445-448 Casola S, Pedone PV, Cavazzana AO, Basso G, Luksch R, d'Amore
ES, Carli M, Bruni CB, Riccio A: Expression and parental imprinting of the H19 gene in human rhabdomyosarcoma. Oncogene 1997, 14: 1503-1510 39. Hiti AL, Bogenmann E, Gonzales F, Jones PA: Expression of the MyoDl muscle determination gene defines differentiation capability but not tumorigenicity of human rhabdomyosarcomas. Mol Cell Biol 1989, 9:4722-4730 40. Downing JR, Khandekar A, Shurtleff SA, Head DR, Parham DM, Webber BL, Pappo AS, Hulshof MG, Conn WP, Shapiro DN: Multiplex RT-PCR assay for the differential diagnosis of alveolar rhabdomyosarcoma and Ewing's sarcoma. Am J Pathol 1995, 146:626-634 41. Barr FG, Chatten J, D'Cruz CM, Wilson AE, Nauta LE, Nycum LM, Biegel JA, Womer RB: Molecular assays for chromosomal translocations in the diagnosis of pediatric soft tissue sarcomas. J Am Med Assoc 1995, 273:553-557
