Somatic Mitochondrial DNA Mutations in Neurofibromatosis Type 1-Associated Tumors

Andreas Kurtz,1 Maria Lueth,2 Lan Kluwe,3 Tingguo Zhang,1 Rosemary Foster,1 Victor-Felix Mautner,3 Melanie Hartmann,4 Duan-Jun Tan,2 Robert L. Martuza,1 Reinhard E. Friedrich,3 Pablo Herna´iz Driever,4 and Lee-Jun C. Wong2

1Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts; 2Institute for Molecular and Human Genetics, Georgetown University Medical Center, Washington, District of Columbia; 3University Hospital Eppendorf, Hamburg, Germany; and 4Charite´ Medical Center, Humboldt University, Berlin, Germany

Abstract Introduction Neurofibromatosis type 1 is an autosomal dominantly Neurofibromatosis type 1 (NF1) is an autosomal dominantly inherited disease predisposing to a multitude of tumors, inherited disorder with an estimated prevalence range from most characteristically benign plexiform neurofibromas 1:2,190 to 1:7,800 (1). There are no known ethnic groups in and diffuse cutaneous neurofibromas. We investigated which NF1 does not occur or is unusually common. NF1 is the presence and distribution of somatic mitochondrial caused by mutations in the NF1 tumor suppressor gene, DNA (mtDNA) mutations in neurofibromas and in resulting in reduced Ras-GTPase-activating protein activity of nontumor tissue of neurofibromatosis type 1 patients. neurofibromin (2-5). Clinical features of NF1 include cafe´ au MtDNA alterations in the entire mitochondrial genome lait spots, axillary freckling, iris hamartomas (Lisch nodules), were analyzed by temporal temperature gradient gel skeletal abnormalities, and learning deficiencies (4). The most electrophoresis followed by DNA sequencing. Somatic prevalent clinical manifestations are the development of benign mtDNA mutations in tumors were found in 7 of 19 plexiform and cutaneous neurofibromas. Plexiform neuro- individuals with cutaneous neurofibromas and in 9 of fibromas can progress at low frequency to highly malignant 18 patients with plexiform neurofibromas. A total of 34 peripheral nerve sheath tumors. Cutaneous neurofibromas somatic mtDNA mutations were found. All mutations typically start to develop around puberty, and the number of were located in the displacement loop region of the these tumors increases with age. Other tumor types less mitochondrial genome. Several plexiform neurofibromas frequently associated with NF1 include optical gliomas, from individual patients had multiple homoplasmic myeloid leukemias, phaeochromocytomas, and astrocytomas. mtDNA mutations. In cutaneous neurofibromas, the Loss of heterozygosity is thought to be the underlying cause same mtDNA mutations were always present in tumors for development of benign neurofibromas (6-10). Indeed, loss from different locations of the same individual. An of heterozygosity at the NF1 locus has been shown in Schwann increase in the proportion of the mutant mtDNA was cells of cutaneous and plexiform neurofibromas and in always found in the neurofibromas when compared with malignant peripheral nerve sheath tumors (11-14), whereas nontumor tissues. The somatic mtDNA mutations were other studies were unable to confirm loss of heterozygosity present in the Schwann cells of the analyzed multiple for NF1 in astrocytomas and neurofibromas of NF1 patients cutaneous neurofibromas of the same individual. The (15, 16). The variable expressivity, the plethora of mostly observed dominance of a single mtDNA mutation in benign tumors, and the variability in loss of heterozygosity for multiple cutaneous neurofibromas of individual patients NF1 suggest that the threshold for tumorigenesis is perhaps indicates a common tumor cell ancestry and suggests reduced by systemic modulators (17-19). a replicative advantage rather than random segregation The diffuse cutaneous neurofibromas typical for NF1 may for cells carrying these mutated mitochondria. arise at multiple sites due to the independent occurrence of (Mol Cancer Res 2004;2(8):433–41) tumorigenic mutations. Alternatively, an early mutation in a disseminating precursor cell during development may be responsible for the observed phenotype. Here, we use mutational analysis to identify somatic changes in mitochondria in benign neurofibromas and unaffected tissue of NF1 patients. Charac- Received 2/24/04; revised 6/17/04; accepted 6/17/04. Grant support: United States Medical Research Command Neurofibromatosis terization of somatic mitochondrial DNA (mtDNA) mutations in Research Program award DMAD17-00-1-0535 (A. Kurtz) and BCRP award multiple tumors of the same patient may illuminate the cellular DAMD 17-01-1-0258 (L-J.C. Wong). The costs of publication of this article were defrayed in part by the payment of origin of multiple tumors in NF1 from a common precursor. page charges. This article must therefore be hereby marked advertisement in The role of mitochondria in tumor development has gained accordance with 18 U.S.C. Section 1734 solely to indicate this fact. much attention with recent reports of somatic mtDNA Note: A. Kurtz and M. Lueth contributed equally to this study. Requests for reprints: Lee-Jun C. Wong, Institute for Molecular and Human mutations in brain, ovarian, esophageal, breast, and colorectal Genetics, Georgetown University Medical Center, 3800 Reservoir Road, human cancers (20-27). Mitochondria contain multiple copies M4000, Washington, DC 20007. Phone: 202-444-0760; Fax: 202-444-1770. E-mail: [email protected] of circular double-stranded DNA molecules that have a high Copyright D 2004 American Association for Cancer Research. degree of sequence variations among different individuals (28).

In addition to energy production by oxidative phosphoryl- mutations were identified. All mutations were found in the ation, mitochondria play a crucial role in programmed cell displacement loop (D-loop) region. Insertions or deletions in death (29-32) and generate DNA-damaging reactive oxygen the nucleotide positions (np) 303 to 309 poly(C) region were species as side products of normal function. MtDNA is an easy detected in multiple tumors. This region has been reported to be target for oxidative DNA damage due to the close proximity to the somatically unstable mutation hotspot of breast cancer reactive oxygen species production, the lack of protective (25, 39). All nucleotide substitutions were transitions between histone proteins, and the limited repair capabilities. The ac- T/C and A/G, consistent with oxidative DNA damage. Fourteen cumulation of reactive oxygen species might also contribute to of 27 of the somatic mtDNA mutations in the plexiform tumors increased nuclear gene mutagenesis (33). were alterations from homoplasmic state in blood to homo- The characteristics of multisystemic manifestation, variable plasmic state in tumor. A shift from homoplasmy in blood to expressivity, and somatic mosaicism of NF1 as well as the heteroplasmy in the corresponding tumor was found in six reported associations of neurofibromin with highly energy- cases, and a shift from heteroplasmy in blood to homoplasmy in dependent microtubules (34) and mitochondria (35) may in- tumor was detected in four cases. Three pairs of matched dicate the presence of specific somatic mtDNA mutations in samples had heteroplasmic nucleotide substitutions in both tumors of NF1 patients. We analyzed the mtDNA mutations in blood and tumor, but there were detectable differences in the neurofibromas and compared mutations between several degree of heteroplasmy as assessed by quantitative comparison distinct tumors of the same individual. The presence of somatic of the nucleotide peak amplitudes in the corresponding mtDNA mutations in nontumor cells and tumor cells was sequence profiles (Table 2). compared. One patient with plexiform tumor, T173, harbored nine somatic mtDNA mutations. To rule out an error in DNA sampling, identity analysis on tumor T173 and corresponding Results blood sample B174 samples was done. Identical alleles were Somatic mtDNA mutations in neurofibromas were detected detected at five polymorphic sites: the short tandem repeat in by parallel analysis of mtDNA from pairs of matched tumor and intron 3 of the PAH gene (chromosome 12), the CTG repeats of normal blood samples by temporal temperature gradient gel the myotonin protein kinase gene (disease gene for myotonic electrophoresis (TTGE). Multiple PCR products comprising the dystrophy, chromosome 19), the CAG repeats of the androgen entire mitochondrial genome were used. Somatic nucleotide receptor gene (X chromosome), and the SCA1 (chromosome 6) alterations are identified as differences in the banding patterns and SCA3 (chromosome 4) genes (data not shown). These between the DNA from matched normal and tumor tissues. results support that tumor T173 indeed had nine somatic These analyses were carried out on a total of 19 patients with mtDNA mutations, four of which are novel. The cause for this cutaneous neurofibromas and 18 patients with plexiform unusually large number of tumor-specific somatic mtDNA neurofibromas. Figure 1 illustrates the results from TTGE and mutations is not clear. Apparently, the point mutations in the sequencing analyses. origin of H-strand replication and the termination-associated When the analyzed mtDNA sequences from blood were sequence regulate the mtDNA synthesis and transcription in the compared with those of the published sequence (36), 71 germ tumor. A large number (>6) of somatic mtDNA mutations also line sequence variations were revealed (Table 1). These do not occurred in f5% to 10% of medulloblastomas and breast and represent all the variations because only the mtDNA PCR lung cancers (25, 40). products of regions that showed somatic mutations in the paired Cutaneous neurofibromas were analyzed from a total of 19 tumor sample were sequenced. Nine of the germ line variations patients. Seven of these tumors had somatic mtDNA mutations detected are novel. Many of the remaining reported poly- (Table 2), and all of them occurred in the hypervariable D-loop morphisms occurred in multiple individuals. Among them, region. NF1 patients usually have multiple cutaneous neuro- A73G and T16519C are common polymorphisms in various fibromas ranging from <10 to >1000. Surprisingly, multiple ethnic groups. The apparent high frequencies of A263G and cutaneous neurofibromas resected from distinct anatomic sites 303-309insC are because these are polymorphisms in the on an affected individual always shared identical somatic Cambridge reference sequence (37, 38). Although germ line mtDNA mutations (Tables 2 and 3): Analysis of tumor samples variations are generally considered silent, missense mutations T104 and T105 (Fig. 2C) revealed a change in a short poly(C) such as the novel A265V (C5263T) alteration in the sequence at np 303 to 309 in the conserved sequence block mitochondrial electron transport chain complex NADH dehy- (CSB), which shifted from C8/C9 heteroplasmy in the blood drogenase subunit 2 may have a functional effect. sample (B106) of the same patient to near homoplasmy of C8 in The complete results of our analysis of NF1-associated both tumor samples (Table 3). Analysis of mtDNA samples cutaneous and plexiform neurofibromas are listed in Table 2. from tumors T107 and T108 showed a shift from np 303 to 309 The overall percentage of neurofibromas with somatic mtDNA C7 homoplasmy in the matched blood sample B109 to C7/C8 mutations was similar to that found for glioblastoma and heteroplasmy in each of the tumors. Tumors T119 and T120 medulloblastoma but lower than those in lung, breast, and oral both harbor the same homoplasmic T16304C mutation when cancers (20-27). compared with the corresponding blood DNA B121, which is Somatic mtDNA mutations were detected in 9 of 18 homoplasmic for the wild-type T16304. Comparable changes plexiform neurofibromas (Table 2). Four of the plexiform were detected in tumors and blood of patients 4, 5, and 6 tumors with mtDNA mutations harbored a single alteration; five (Table 3). In patient 7, we detected a high proportion of np 303 cases had >1 mutation. A total of 27 different somatic mtDNA to 309 C8 in blood and in tumor samples, whereas no D-loop

FIGURE 1. Detection of somatic mtDNA mutations in plexiform and cutaneous neurofibromas by TTGE and sequence analysis. A. Comparison of PCR-amplified mtDNA D-loop region from plexiform neurofibroma T173 and paired blood sample B174. Sequencing revealed multiple homoplasmic nucleotide substitutions in plexiform neurofibroma mtDNA. Direct sequencing of the blood and tumor mtDNA PCR products revealed five homoplasmic tumor-specific nucleotide substitutions in this region, including three novel ones (A16163G, C16186T, and C16221T). The detected T16189C and T16519C alterations have been reported previously (22, 25, 27). B. Comparison of mtDNA D-loop region from plexiform neurofibroma T191 and paired blood sample B192. Sequencing revealed two changes, T199C and G207A, from homoplasmic in normal to heteroplasmic in tumor and one heteroplasmic to heteroplasmic T204C change in the same region. Sequencing revealed two changes from homoplasmic T199 and G207 to heteroplasmic T199C and G207A. Interestingly, a heteroplasmic change in the proportion of T204C mutation between B192 and T191 was also detected, revealing a shift in the degree of heteroplasmy at T204C in the tumor sample. C. Comparison of mtDNA D-loop region from two cutaneous neurofibromas T104 and T105 from the same individual and paired blood sample B106. Sequencing revealed a gradual change from a heteroplasmic 303 to 309 C8/C9 to a homoplasmic C8 in both cutaneous neurofibromas. mutations were found in patient 8. Tumor tissues from differ- 8), no somatic mtDNA D-loop mutations were found in the ent parts of the same tumor also showed the identical tumor and skin samples (Table 3). The other two sets of tumors mtDNA mutations with similar degrees of plasmy (Table 3, displayed somatic mtDNA mutations (patients 4 and 5). A patients 4 and 5). progressive increase in mutant mtDNA content was showed These results suggest that the tumor-specific mtDNA among blood, skin, and the neurofibromas in both of these mutations may have already been present in nontumor cells patients (Table 3). In addition, the identical mutation was and accumulated in the neurofibromas. To examine this, we present in the skin, in different parts of the tumor, and analyzed mtDNA isolated from unaffected skin and matched in separate tumors from different locations of the same cutaneous neurofibromas from three NF1 patients. Mutation individual. analysis of these skin and tumor samples was focused on the Neurofibromas are mixed-cell tumors composed predomi- D-loop and its surrounding region, because the data obtained nantly of Schwann cells and fibroblasts, with a minority of from 37 neurofibromas revealed that all somatic mtDNA perineureal cells, neurons, mast cells, and endothelial cells. mutations occurred in the D-loop region. In one patient (patient To determine which cell type in an apparently homoplasmic

Table 1. Germ Line Sequence Variations cutaneous neurofibroma contains the mutated mitochondria, mutational analysis was done on laser capture microdissected Gene/Region Germ Line Mutation Frequency* Significance cells. S100-positive Schwann cells, S100-negative cells, and endothelial cells were dissected from neurofibromas 583 and A. Novel 584 (patient 4) and 592 and 593 (patient 5; Table 3). Fibroblasts D-loop T10C 1 7S DNA D-loop T55C 1 7S DNA and epithelial cells were also dissected from unaffected skin of D-loop T57C 1 Hypervariable segment 2 the same patients. MtDNA from the separate cell types was D-loop T408A 1 L-strand promoter analyzed for the presence of the previously identified D-loop 16S A2706G 1 16S RNA c ND2 C5263T 1 GCC-GTC, A265V mutation 303-309insC C7/C8. COI G6917A 1 GTG-GGG, V338V In tumors 583 and 584, S100-positive Schwann cells con- ND4 A11947G 1 ACA-ACG, T396T tained almost exclusively the C8 mtDNA mutation (Fig. 2A-C). C16465T 1 D-loop S100-negative tumor-derived cells, presumably fibroblasts, B. Reported were also highly enriched for the C8 mtDNA mutation (data D-loop T72C 4 Hypervariable segment 2 not shown). Although carefully selected, we cannot completely D-loop A73G 12 Hypervariable segment 2 D-loop T146C 2 H-strand origin exclude the possibility that the S100 cell fraction contains D-loop C150T 1 H-strand origin neoplastic Schwann cells, as it is notoriously difficult to D-loop T152C 4 H-strand origin distinguish between the different cell types in neurofibromas. D-loop A189G 1 H-strand origin D-loop C194T 2 H-strand origin Tumor-derived nonneoplastic endothelial cells contained only D-loop T195C 6 H-strand origin C7 mtDNA (Fig. 2D-F), as did epithelial cells obtained from D-loop T199C 1 H-strand origin the skin sample 586 of the same patient (Fig. 2G). Dermal D-loop T204C 2 H-strand origin fibroblasts from skin sample 586 contained C7 and C8 mtDNA D-loop G207A 2 H-strand origin f D-loop C242T 1 mtTF1 binding site at a ratio of 30:70 (Fig. 2H). Similar results were obtained for D-loop A263G 16 H-strand origin the different cell types of tumor samples 592 and 593 and D-loop C271T 1 H-strand origin matched skin sample 588. The results of the cell type–specific D-loop C295T 1 mtTF1 binding site D-loop 303-309insC 15 CSB II mtDNA analysis are summarized in Table 4. D-loop C462T 1 D-loop T489C 1 D-loop A508G 1 Discussion D-loop 514insCA 1 D-loop 514insCACA 1 In the analysis of NF1-associated tumors, one single somatic D-loop 568insCCC 1 mtDNA mutation was detected in each cutaneous neurofibro- 12s A663G 2 12S RNA ma, whereas an average of three mutations was detected 12s G709A 1 12S RNA ND2 G4580A 1 ATG-ATA, M37M per plexiform neurofibroma. This compares with an average ND2 A4769G 2 ATA-ATG, M100M of one to three somatic mtDNA mutations for other tumors COI T6776C 1 CAT-CAC, H291H (20, 22, 25, 27). It is noteworthy that 22 of 27 mutations in ND4 G11914A 1 ACG-ACA, T385T ND4 G12007A 1 TGG-TGA, W416W plexiform neurofibromas are nucleotide substitutions compared ND5 A12612G 1 GTA-GTG, V92V with only 2 of 7 mutations in cutaneous neurofibromas. One ND5 A12693G 1 AAA-AAG, K119K interesting observation is that mutations at np 204 and 207 ND5 C12705T 2 ATC-ATT, I123I ND6 A14233G 1 ATC-GTC, I29V occurred three times in three unrelated patients. This result CytB T14798C 1 TTC-CTC, F18L suggests that either the np 204 and 207 are mutation hotspots or D-loop G16145A 1 the mutant mitochondria have selective growth advantage. D-loop C16186T 1 Hypervariable segment 1 All of the mtDNA somatic mutations identified in our study D-loop C16188T 1 Hypervariable segment 1 D-loop T16189C 1 Hypervariable segment 1 occurred in the hypervariable D-loop region of the mitochon- D-loop T16192T 1 Hypervariable segment 1 drial genome. This is unique because numerous studies on D-loop C16193T 1 Hypervariable segment 1 lung, breast, ovarian, bladder, head and neck, glioblastoma, D-loop C16195T 1 Hypervariable segment 1 D-loop C16222T 1 Hypervariable segment 1 and oral cancers showed that 20% to 70% of somatic mtDNA D-loop C16223T 3 Hypervariable segment 1 mutations were found in coding regions (20-32, 40). The D-loop C16278T 4 Hypervariable segment 1 pathologic significance of mutations in noncoding regions of D-loop C16290T 1 Hypervariable segment 1 D-loop C16292T 1 Hypervariable segment 1 the mitochondrial genome is currently unknown. Apparently, D-loop C16294T 3 Hypervariable segment 1 mutations in the D-loop influence the origin of replication and D-loop C16296T 2 Hypervariable segment 1 promoter region and may affect the mitochondrial biogenesis, D-loop T16298C 3 Hypervariable segment 1 transcription, and protein expression (41, 42). D-loop T16304C 2 Hypervariable segment 1 D-loop A16309G 1 Hypervariable segment 1 The most common somatic mtDNA mutations identified in D-loop T16311C 2 Hypervariable segment 1 our study are insertions or deletions at np 303 to 309. Mutations D-loop T16362C 2 Hypervariable segment 1 at np 303 to 309 were not observed in mtDNA from 40 muscle D-loop G16390A 1 Hypervariable segment 1 5 D-loop T16519C 4 tissues of individuals ranging in age from 0 to 65 years. This suggests that the variability in the np 303 to 309 region is not NOTE: The total number of distinct germ line sequence variations is 71 (9 novel and 62 reported). Missense substitutions are in bold. *Number of tumors that carry germ line variation. c 5 ND2, NADH dehydrogenase subunit 2. L-J.C. Wong, unpublished observation.

due to a microsatellite instability. It is of interest to note that the that multiple mutations in tumors of the same patient were np 303 to 309 C8 mutation was always dominant in the tumors, almost always found in a homoplasmic state. In addition to whereas the C7 variant was dominant in nontumor tissue and random segregation and replicative advantage, mitochondrial was either diminished or not detected in tumors. segregation dynamics and maintenance may also be influenced MtDNA mutations may occur randomly, and multiple by the genetic background of cells (44). Because neurofibromin mutations may or may not occur simultaneously. To reach a is closely associated with microtubules and mitochondria homoplasmic state may require some mechanism for selection. (34, 35), haploinsufficiency of NF1 may perhaps affect the Homoplasmy for mtDNA mutations in tumors can also be stability and promote segregation of specifically mutated caused by random segregation after a sufficient number of mitochondria in neurofibromas. cell divisions (43). Thus, mutations in the origin of replication The present data indicate the presence of somatic mtDNA (D-loop region) may provide a replicative advantage of these mutations in a heteroplasmic state in cells of nontumor tissues mutant mtDNAs, or homoplasmy for the D-loop mutation in of NF1 patients. The mutated mitochondria, particularly the np tumors may be the result of random segregation. The last 303 to 309 C8 genotype, accumulate and ultimately dominate in mechanism is unlikely to explain the observed distribution of the multiple cutaneous neurofibromas of a patient. This specific mtDNA mutations in individual neurofibromas. The indepen- mitochondrial distribution pattern, in association with NF1 dent appearance of the same mtDNA mutation in distinct haploinsufficiency, may be indicative of a lowered tumorigenic tumors of the same patient in six analyzed individuals argues threshold in NF1. Furthermore, the mitochondrial signatures of against a random process. This is supported by the observation plexiform and cutaneous neurofibromas of NF1 patients vary,

Table 2. Somatic mtDNA Mutations in NF1-Associated Neurofibromas

Case No. Gene/Region Somatic Mutation Cambridge Sequence nl to tu Pattern* Functionc Previously Reported in Tumorsb

NF1-associated plexiform neurofibromas T159 D-loop A73G A Homo-homo Hypervariable segment 2 eso T159 D-loop C16193T C Homo-homo Hypervariable segment 1 Novel T159 D-loop C16278T T Homo-homo Hypervariable segment 1 ov T159 D-loop C16519T T Homo-homo Lung, glioblastoma T165 D-loop T64C C Hetero-hetero Hypervariable segment 2 Novel T171 D-loop 303-309delC, C9/C8-C8/C9 C7 Hetero-hetero CSB crc, gastric, eso, ov, brca T173 D-loop C64T C Homo-homo Hypervariable segment 2 Novel T173 D-loop A73G A Homo-homo Hypervariable segment 2 eso T173 D-loop T152C T Homo-homo H-strand origin ov T173 D-loop T195C T Homo-homo H-strand origin Lung, glioblastoma T173 D-loop A16163G A Homo-homo Termination-associated sequence Novel T173 D-loop C16186T C Homo-homo 7S DNA Novel T173 D-loop T16189C T Homo-homo 7S DNA brca T173 D-loop C16221T C Homo-homo Hypervariable segment 1 Novel T173 D-loop T16519C T Homo-homo Lung, glioblastoma T179 D-loop 303-309insC, C7/C8-C8 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T185 D-loop C204T T Homo-homo H-strand origin gastric, glioblastoma T185 D-loop A207G G Hetero-homo H-strand origin brca T187 D-loop 303-309delC, C7/C8-C7 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T189 D-loop T204C T Hetero-homo H-strand origin gastric, glioblastoma T189 D-loop G207A G Homo-hetero H-strand origin brca T189 D-loop 303-309insC, C7-C7/C8 C7 Homo-hetero CSB crc, gastric, eso, ov, brca T191 D-loop A73G A Homo-hetero Hypervariable segment 2 eso T191 D-loop T199C T Homo-hetero Hypervariable segment 2 ov T191 D-loop T204C T Hetero-hetero H-strand origin gastric, glioblastoma T191 D-loop G207A G Homo-hetero H-strand origin brca T191 D-loop 303-309insC, C7-C7/C8 C7 Homo-hetero CSB crc, gastric, eso, ov, brca

NF1-associated cutaneous neurofibromas T104 D-loop 303-309delC, C9/C8-C8 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T105 D-loop 303-309delC, C9/C8-C8 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T107 D-loop 303-309insC, C7-C7/C8 C7 Homo-hetero CSB crc, gastric, eso, ov, brca T108 D-loop 303-309insC, C7-C7/C8 C7 Homo-hetero CSB crc, gastric, eso, ov, brca T119 D-loop T16304C T Homo-homo Hypervariable segment 1 ov T120 D-loop T16304C T Homo-homo Hypervariable segment 1 ov T590 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-hetero CSB crc, gastric, eso, ov, brca T591 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-hetero CSB crc, gastric, eso, ov, brca T583 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T584 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-homo CSB crc, gastric, eso, ov, brca T468 D-loop T196C T Hetero-hetero CSB Novel T469 D-loop T196C T Hetero-homo CSB Novel T554 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-hetero CSB crc, gastric, eso, ov, brca T555 D-loop 303-309insC, C7/C8-C8/C7 C7 Hetero-hetero CSB crc, gastric, eso, ov, brca

*nl, normal (blood); tu, tumor; homo, homoplasmic; hetero, heteroplasmic. c The letter before the np number indicates the nucleotide found in the normal tissue, and the letter after the np number indicates the nucleotide found in tumor tissue. The homoplasmic or heteroplasmic status of the mutation in normal and tumor tissue is indicated accordingly in the nl to tu Pattern column. b crc, colorectal cancer; eso, esophageal cancer; ov, ovarian cancer; brca, breast cancer.

Table 3. Somatic mtDNA Mutations in Two Separate Cutaneous Neurofibromas

Patient Sample Specimen Type Location Somatic Mutation % Heteroplasmy

1 B106 Blood 303-309insC C9/C8 C9 50%, C8 50% T105 Tumor 1 303-309delC C9/C8 C9 0%, C8 100% T104 Tumor 2 303-309delC C9/C8 C9 5%, C8 95% 2 B109 Blood 303-309insC C7/C8 C7 100%, C8 0% T108 Tumor 1 303-309insC C7/C8 C7 60%, C8 40% T107 Tumor 2 303-309insC C7/C8 C7 40%, C8 60% 3 B121 Blood T16304 T 100%, C 0% T120 Tumor 1 T16304C T 0%, C 100% T119 Tumor 2 T16304C T 0%, C 100% 4 B587 Blood 303-309insC C7/C8 C7 50%, C8 50% S586 Skin Distal from tumor 303-309insC C7/C8 C7 40%, C8 60% S585 Skin Overlaying the tumor 303-309insC C7/C8 C7 30%, C8 70% T583 Tumor 1 Thorax 303-309insC C7/C8 C7 10%, C8 90% T584 Tumor 2, part 1 Abdomen 303-309insC C7/C8 C7 10%, C8 90% T584 Tumor 2, part 2 Abdomen 303-309insC C7/C8 C7 5%, C8 95% 5 B594 Blood 303-309insC C7/C8 C7 50%, C8 50% S588 Skin Distal from tumor 303-309insC C7/C8 C7 40%, C8 60% S589 Skin Overlaying the tumor 303-309insC C7/C8 C7 40%, C8 60% T590 Tumor 1 Thorax right side 303-309insC C7/C8 C7 20%, C8 80% T591 Tumor 2, part 1 Abdomen 303-309insC C7/C8 C7 0%, C8 100% T592 Tumor 2, part 2 Abdomen 303-309insC C7/C8 C7 20%, C8 80% T593 Tumor 2, part 3 Abdomen 303-309insC C7/C8 C7 0%, C8 100% 6 B276 Blood T196C T 50%, C 50% T468 Tumor 1 T196C T 30%, C 70% T469 Tumor 2 T196C T 5%, C 95% 7 B9 Blood 303-309insC C8/C7 C8 70%, C7 30% T554 Tumor 303-309insC C8/C7 C8 95%, C7 5% T555 Tumor 303-309insC C8/C7 C8 70%, C7 30% 8 B598 Blood S597 Skin 2 cm away from 595 No mutation found T596 Tumor 1 Thorax right side No mutation found T595 Tumor 2 Thorax right side No mutation found

NOTE: Somatic mtDNA mutations in two separate cutaneous neurofibromas from each of eight patients and in paired skin samples of patients 4, 5, and 8. In all samples from patients 4 to 8, only D-loop region was analyzed. % Heteroplasmy was estimated from the sequencing chromatogram. They do not represent the actual proportion. However, the trend of progressive alteration was obvious (patients 4 and 5). For samples 104 and 105, TTGE gel chromatogram was used to estimate the percentage of mutant heteroplasmy, which was too low to be revealed by sequencing (Fig. 1C).

underlying their separate developmental origins. Although the The age of patients with cutaneous tumors ranges from 16 to present findings cannot exclude a mechanism of frequent muta- 50 years with a mean age of 37.6 years. The age of patients with tional events in the NF1 gene, leading to multiple cutaneous plexiform tumors ranges from 8 to 73 years with a mean age of neurofibromas, they support the notion that multiple cutaneous 28.8 years. neurofibromas in NF1 patients may arise from a widely distri- buted cell carrying an early somatic mitochondrial mutation. DNA Isolation DNA was isolated from frozen tissues using proteinase K Materials and Methods and phenol/chloroform extraction method. DNA was extracted Tissue Samples from peripheral blood lymphocytes using a modified nonenzy- Patients with NF1 were recruited through the Departments of matic method (46). Total DNA was quantified using fluorescent Neurosurgery and Neurogenetics, Massachusetts General Hos- Hoechst dye H33258 (Sigma, St. Louis, MO) with DyNA pital, Harvard University and the Department of Neurology, Quant 200 (Amersham Biosciences, Uppsala, Sweden) accord- Klinikum Nord Ochsenzoll (Hamburg, Germany) according to ing to the manufacturer’s protocol and diluted to 5 ng/ALtobe institutional review board–approved protocols. Patients were used in PCR reactions (47). phenotypically characterized for features of NF1, including number, location, and size of cutaneous neurofibromas. Only patients with a clear diagnosis of NF1 according to NIH criteria Mutational Analysis of the Entire Mitochondrial Genome (45) were included in this study. Cutaneous and plexiform DNA isolated from 19 pairs of matched blood and cutaneous neurofibromas were removed during routine surgery, dissected neurofibroma samples and 18 pairs of matched blood and into multiple aliquots, and frozen immediately. Two or more plexiform neurofibroma samples was used for mutational cutaneous neurofibromas resected from different anatomic sites analysis of the mitochondrial genome by TTGE. on each individual were obtained from 19 patients. For three Thirty-two pairs of overlapping primers were used to of these patients, skin samples were biopsied from an area amplify the entire mitochondrial genome by PCR (47). The overlaying the resected cutaneous neurofibroma and from an area PCR-amplified DNA fragments vary from 306 to 805 bp long. distal to the tumor (Table 2, patients 4, 5, and 6). A single plexi- The positions and sequence of the PCR primers and the PCR form neurofibroma sample was obtained from each of 18 patients. and TTGE conditions were as described recently (47). Briefly,

FIGURE 2. Cell specificity of mtDNA mutations in cutaneous neurofibromas and skin of the same patient. A to C. Cutaneous neurofibroma 583 with selected Schwann cells (A), microdissected cells (B), and sequence results of np 303 to 309 C7/C8 D-loop region (C). Arrow, C8. D to F. The same neurofibroma as in A with selected endothelial cells (D), microdissected cells (E), and sequencing results (F). Arrow, C7. Magnification is 200Â in all images. G. Se- quencing results of epithelial cells from a skin sample of the same patients as in A. Arrow, C7. H. Sequencing results of dermal fibroblasts from a skin sample of the same patient as in A. Arrow, C8 (70%)/C7(30%). the DNA template, after the initial denaturation at 94jCfor homoduplexes and heteroduplexes were maintained at 4jC 5 minutes, was amplified with 35 cycles of 94jC for 45 seconds, until TTGE analysis was done on a D-Code apparatus (Bio-Rad 55jC for 45 seconds, and 72jC for 45 seconds and completed Laboratories, Hercules, CA). Five microliters of denatured and by 4 minutes of extension at 72jC. The PCR products were reannealed PCR products were loaded onto a polyacrylamide denatured at 95jC for 30 seconds and slowly cooled to 45jC gel (acrylamide/bis 37.5:1) prepared in 1.25Â Tris-acetate- for 45 minutes at a rate of 1.1jC/min. The reannealed EDTA buffer containing 6 mol/L urea. Electrophoresis was

Table 4. Somatic mtDNA Mutations in Cells Dissected from Cutaneous Neurofibromas and Paired Skin of Patients 4 and 5 of Table 2

Patient Sample Specimen Type Cell Type Somatic Mutation % Heteroplasmy

4 586 Skin Epithelial 303-309insC C7/C8 C7 100%, C8 0% 586 Skin Dermal fibroblast 303-309insC C7/C8 C7 30%, C8 70% 583 Tumor 1 S100-positive Schwann cell 303-309insC C7/C8 C7 0%, C8 100% 583 Tumor 1 S100-negative cells 303-309insC C7/C8 C7 5%, C8 95% 583 Tumor 1 Endothelial cell 303-309insC C7/C8 C7 100%, C8 0% 5 588 Skin Epithelial 303-309insC C7/C8 C7 100%, C8 0% 588 Skin Dermal fibroblast 303-309insC C7/C8 C7 35%, C8 65% 592 Tumor 2, part 2 S100-positive Schwann cell 303-309insC C7/C8 C7 5%, C8 95% 592 Tumor 2, part 2 S100-negative cells 303-309insC C7/C8 C7 10%, C8 90% 592 Tumor 2, part 2 Endothelial cell 303-309insC C7/C8 C7 100%, C8 0%

NOTE: Percentage of heteroplasmy was estimated from the sequencing chromatogram.

carried out at 145 V for 4 to 5 hours at a constant 1jC/h to Acknowledgments 2jC/h temperature increment (47). The temperature range was We thank Dr. Mia MacCollin for valuable discussions and continuing support of determined by computer simulation from the melting curve of this project. the analyzed DNA fragment (MacMelt software, Bio-Rad Laboratories). The gels were stained with 2 mg/L ethidium References bromide for 5 minutes and imaged with a digital CCD gel 1. Friedman JM. Epidemiology of neurofibromatosis type 1. Am J Med Genet 1999;89:1-6. documentation system (high-performance ultraviolet transillu- 2. Whittinghofer A. Signal transduction via Ras. Biol Chem 1998;379:933-7. minator, Ultraviolet Products, Upland, CA). 3. Bernards A. Neurofibromatosis type 1 and Ras-mediated signaling: filling the On TTGE analysis, a single-band shift represents a homo- GAPs. Biochim Biophys Acta 1995;1242:43-59. plasmic DNA alteration, and a multiple-band pattern represents 4. Gutmann DH, Collins FS. Neurofibromatosis 1. 8th ed. Vol. 1. New York: a heteroplasmic mutation (48). Any DNA fragments showing McGraw-Hill; 2001. p. 877-96. different banding patterns between matched blood and tumor 5. Weiss B, Bollag G, Shannon K. Hyperactive Ras as a therapeutic target in sample pairs were sequenced to identify the exact mutations. neurofibromatosis type 1. Am J Med Genet 1999;89:14-22. To rule out the possibility that some mutations may not 6. Carey JC, Viskochil DH. Neurofibromatosis type 1: a model condition for the study of the molecular basis of variable expressivity in human disorders. Am J be detectable by TTGE, we randomly chose six samples that Med Genet 1999;89:7-13. did not show TTGE-positive banding patterns and sequenced 7. Carey JC, Lant JM, Hall BD. Penetrance and variability in neurofibromatosis: 10 coding regions containing stretches of six to eight homo- a genetic study in 60 families. Birth Defects 1979;15:271-81. polynucleotides. No mutations were found. 8. Tonsgard JH, Yelaavarthi KK, Cushner S, Shon MP, Lindgren V. Do NF1 gene deletions result in a characteristic phenotype? Am J Med Genet 1997;73:80-6. 9. Heim RA, Kam-Morgan LNW, Binnie CG, et al. Distribution of 13 truncating Sequence Analysis mutations in the neurofibromatosis 1 gene. Hum Mol Genet 1995;4:975-81. Confirmation of the nucleotide alteration was carried out by 10. Upadhyaya M, Ruggieri M, Maynard J, et al. Gross deletions of the neurofibromatosis type 1 (NF1) gene are predominantly of maternal origin and direct DNA sequencing of the purified PCR product using the commonly associated with a learning disability, dysmorphic features and original PCR primers and a BigDye terminator cycle sequenc- developmental delay. Hum Genet 1998;102:591-7. ing kit (Perkin-Elmer, Foster City, CA) and analyzed on 11. Knudsen A. Mutation and cancer: statistical study of retinoblastoma. Proc ABI 377 (Applied Biosystems, Foster City, CA) automated Natl Acad Sci USA 1971;68:820-3. sequencer. All mutations identified were confirmed by repeat 12. Serra E, Puig S, Otero D, et al. Confirmation of a double-hit model for the NF1 gene in benign neurofibromas. Am J Hum Genet 1997;61:512-9. PCR and sequencing. The results of DNA sequence analysis 13. Kluwe L, Friedrich R, Mautner VF. Loss of NF1 allele in Schwann cells but were compared with the published Cambridge sequence (37) not in fibroblasts derived from an NF1-associated neurofibroma. Genes Chro- using MacVector 7.0 (Oxford Molecular Ltd., Oxford, United mosomes Cancer 1999;24:283-5. Kingdom) software. Sequence variations found in both tumor 14. Stark M, Assum G, Krone W. Single-cell PCR performed with neurofibroma and blood mtDNA were scored as germ line variations. Each Schwann cells reveals the presence of both alleles of the neurofibromatosis type 1 (NF1) gene. Hum Genet 1995;96:619-23. was then checked against the MITOMAP database (36). 15. Wimmer K, Eckart M, Meyer-Puttlitz B, Fonatsch C, Pietsch T. Mutational Variations not recorded in the database were categorized as and expression analysis of the NF1 gene argues against a role as tumor suppressor novel mtDNA variations, and those appearing in the database in sporadic pilocytic astrocytomas. J Neuropathol Exp Neurol 2002;61:896-902. were reported as polymorphisms. Any mtDNA sequence 16. Upadhyaya M, Han S, Consoli C, et al. Characterization of the somatic differences found between a tumor sample and its corres- mutational spectrum of the neurofibromatosis type 1 (NF1) gene in neurofibromatosis patients with benign and malignant tumors. Hum Mutat 2004;23:134-46. ponding blood sample were scored as somatic mtDNA 17. Easton DF, Ponder MA, Huson SM, Ponder BA. An analysis of variation in mutations specific to the tumor. expression of neurofibromatosis (NF) type 1 (NF1): evidence for modifying genes. Am J Hum Genet 1993;53:305-13. 18. McLaughlin ME, Jacks T. Progesterone receptor expression in neuro- Laser Capture Microdissection fibromas. Cancer Res 2003;63:752-5. Microdissection was done using a PixCell II (Arcturus, 19. Wiest V, Eisenbarth I, Schmegner C, Krone W, Assum G. Somatic NF1 Mountain View, CA) laser capture microscope according to mutation spectra in a family with neurofibromatosis type 1: toward a theory of genetic modifiers. Hum Mutat 2003;22:423-7. the manufacturer’s instructions. Frozen 10 Am sections were 20. Liu VWS, Shi HH, Cheung ANY, et al. High incidence of somatic prepared on uncoated slides from cutaneous neurofibromas and mitochondrial DNA mutations in human ovarian carcinomas. Cancer Res 2001; unaffected skin tissues from the same patients (Table 2, patients 61:5998-6001. 4 and 5). The sections were stained in H&E immediately before 21. Alonso A, Martin P, Albarran C, et al. Detection of somatic mutations in microdissection. Adjacent control slides were stained with the mitochondrial DNA control region of colorectal and gastric tumors by heteroduplex and single-strand conformation analysis. Electrophoresis 1997; antibody to S100 protein to confirm histologic and morphologic 18:682-5. differentiation of Schwann cells within the neurofibromas. 22. Fliss MS, Usadel H, Caballero OL, et al. Facile detection of mitochondrial S100-negative cells, S100-positive Schwann cells, and endo- DNA mutations in tumors and bodily fluids. Science 2000;287:2017-9. thelial cells were dissected from neurofibromas. Epithelial skin 23. Copeland WC, Wachsman JT, Johnson FM, Penta JS. Mitochondrial DNA cells and dermal fibroblasts were dissected from skin samples. alterations in cancer. Cancer Invest 2002;20:557-69. About 200 to 300 S100-positive Schwann cells, S100 negative 24. Hibi K, Nakayama H, Yamazaki T, et al. Mitochondrial DNA alteration in esophageal cancer. Int J Cancer 2001;92:319-21. cells, and morphologically distinguishable endothelial cells 25. Tan D-J, Bai R, Wong L-JC. Comprehensive scanning of somatic from neurofibromas, epidermal epithelial cells, and dermal mitochondrial DNA mutations in breast cancer. Cancer Res 2002;62:972-6. fibroblast from normal skin were dissected for DNA extraction 26. Polyak K, Li Y, Zhu H, et al. Somatic mutations of the mitochondrial genome (PicoPure extraction kit, Arcturus). The D-loop region of the in human colorectal tumors. Nat Genet 1998;20:291-3. mtDNA was amplified by PCR and directly sequenced. 27. Kirches E, Krause G, Warich-Kirches M, et al. High frequency of

mitochondrial DNA mutations in glioblastoma multiforme identified by direct 39. Parrella P, Xiao Y, Fliss MS, et al. Detection of mitochondrial DNA mutations sequence comparison to blood samples. Int J Cancer 2001;93:534-8. in primary breast cancer and fine-needle aspirates. Cancer Res 2001;61. 28. Wallace DC, Lott MT, Brown MD, Kerstann K. Mitochondrial and 40. Wong LJ, Lueth M, Li XN, Lau CC, Vogel H. Detection of mitochondrial neuroophthalmologic diseases. 8th ed. Vol. 2. New York: McGraw-Hill; 2001. DNA mutations in the tumor and cerebrospinal fluid of medulloblastoma patients. p. 2425-509. Cancer Res 2003;63:3866-71. 29. Yang J, Liu X, Bhalla K, et al. Prevention of apoptosis by bcl-2: release of 41. Barthelemy C, de Baulny HO, Lombes A. D-loop mutations in mito- cytochrome c from mitochondrial blocked. Science 1997;275:1129-32. chondrial DNA: link with mitochondrial DNA depletion? Hum Genet 2002; 110:479-87. 30. Torroni A, Stepien G, Hodge JA, et al. Neoplastic transformation is associated with coordinate induction of nuclear and cytoplasmic oxidative 42. Tang Y, Schon EA, Wilichowski E, Vazquez-Memije ME, Davidson E, phosphorylation genes. J Biol Chem 1990;265:20589-93. King MP. Rearrangements of human mitochondrial DNA (mtDNA): new in- sights into the regulation of mtDNA copy number and gene expression. Mol Biol 31. Liang BC. Evidence for association of mitochondrial DNA sequence Cell 2000;11:1471-85. amplification and nuclear localization in human low-grade gliomas. Mutat Res 1996;354:27-33. 43. Coller HA, Khrapko K, Bodyak ND, Nekhaeva E, Herrero-Jimenez P, Thilly WG. High frequency of homoplasmic mitochondrial DNA mutations 32. Boultwood C, Fidler C, Mills KI, et al. Amplification of mitochondrial DNA in human tumors can be explained without selection. Nat Genet 2001;28: in acute myeloid leukemia. Br J Haematol 1996;95:426-31. 147-50. 33. Penta JS, Johnson FM, Wachsman T, Copeland WC. Mitochondrial DNA in 44. Holt IJ, Dunbar DR, Jacobs HT. Behavior of a population of partially human malignancy. Mutat Res 2001;488:119-33. duplicated mitochondrial DNA molecules in cell culture: segregation, mainte- 34. Gregory PE, Gutmann DH, Mitchell A, et al. Neurofibromatosis type 1 nance and recombination dependent upon nuclear background. Hum Mol Genet gene product (neurofibromin) associates with microtubules. Somat Cell Mol 1997;6:1251-60. Genet 1993;19:265-74. 45. Korf BR. Diagnosis and management of neurofibromatosis type 1. Curr 35. Roudebush M, Slabe T, Sundaram V, Hoppel CL, Golubic M, Stacey Neurol Neurosci Rep 2001;1:162-7. DW. Neurofibromin colocalizes with mitochondria in cultured cells. Exp Cell Res 46. Lahiri D, Nurnberger J Jr. A rapid non-enzymatic method for the prepa- 1997;236:161-72. ration of HMW DNA from blood for RFLP studies. Nucleic Acids Res 1991; 36. MITOMAP: a human mitochondrial genome database; 2004. Available from: 19:5444. http://www.mitomap.org. 47. Wong L-JC, Liang M-H, Kwon H, Park J, Bai R, Tan D. Comprehensive 37. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organization of the scanning of the whole mitochondrial genome for mutations. Clin Chem 2002; human mitochondrial genome. Nature 1981;290:457-65. 48:1901-12. 38. Andrews RM, Kubacka I, Chinnery PF, Lightowlers RN, Turnbull DM, 48. Chen TJ, Boles R, Wong L-JC. Detection of mitochondrial DNA muta- Howell N. Reanalysis and revision of the Cambridge reference sequence for tions by temporal temperature gradient gel electrophoresis. Clin Chem 1999; human mitochondrial DNA. Nat Genet 1999;23:147-7. 45:1162-7.

Mol Cancer Res 2004;2(8). August 2004 Downloaded from mcr.aacrjournals.org on September 27, 2021. © 2004 American Association for Cancer Research. Somatic Mitochondrial DNA Mutations in Neurofibromatosis Type 1-Associated Tumors 11United States Medical Research Command Neurofibromatosis Research Program award DMAD17-00-1-0535 (A. Kurtz) and BCRP award DAMD 17-01-1-0258 (L-J.C. Wong). Note: A. Kurtz and M. Lueth contributed equally to this study.

Andreas Kurtz, Maria Lueth, Lan Kluwe, et al.

Mol Cancer Res 2004;2:433-441.

Updated version Access the most recent version of this article at: http://mcr.aacrjournals.org/content/2/8/433

Cited articles This article cites 43 articles, 11 of which you can access for free at: http://mcr.aacrjournals.org/content/2/8/433.full#ref-list-1

Citing articles This article has been cited by 3 HighWire-hosted articles. Access the articles at: http://mcr.aacrjournals.org/content/2/8/433.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/2/8/433. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.