Tuberous Sclerosis Complex 2 Loss of Heterozygosity in Patients with Lung Disease and Cancer

To the Editor:

Lymphangioleiomyomatosis (LAM) is a rare multisystem disease affecting primarily women. It is characterized by cystic lung destruction, lymphatic infiltration, and renal angiomyolipomas, resulting from proliferation of smooth muscle-like LAM cells (1). LAM can occur sporadically (S-LAM) or be associated with tuberous sclerosis complex (TSC). TSC is an autosomal-dominant syndrome characterized by seizures, mental retardation, and hamartomatous tumors of multiple organs (e.g., skin, brain), resulting from mutations or epigenetic effects on one of two tumor suppressor genes: TSC1 on chromosome 9 (9q34) or TSC2 on chromosome 16 (16p13.3) (2). LAM cells may have mutations, altered expression, and/or loss of heterozygosity (LOH) of the TSC genes. These genetic alterations lead to hyperactivation of the mechanistic target of rapamycin pathway; rapamycin (sirolimus) is currently being used to treat patients with LAM (3).

LAM cells appear to spread by a metastatic process. In agreement, recipient LAM cells have been identified in a donor lung. Identical TSC2 mutations were found in angiomyolipomas and lung lesions (4, 5); LAM cells have been identified in blood, chylous effusions, urine, and bronchoalveolar lavage (6, 7). LAM lung lesions have been shown to have LOH and somatic mutations in the TSC genes (8). Previously, we reported that the frequency of TSC2 LOH seen in circulating LAM cells of S-LAM patients is approximately 90% (7), and we wanted to investigate the frequency of TSC2 LOH in cells isolated from blood and urine from patients with other lung diseases to determine the specificity of this finding for LAM. We have now determined the presence of TSC2 LOH in circulating cells in other lung diseases, and its association with different cancers, based on published reports and in silico studies (Tables 1–3 and Figure 1). Eighteen randomly selected patients with different lung diseases were enrolled between 2010 and 2014 at the National Institutes of Health Clinical Center in a protocol (96-H-0100) approved by the National Heart, Lung, and Blood Institutional Review Board. All patients’ samples were screened phenotypically (i.e., LAM cell markers) and genetically (i.e., TSC2 LOH), as described in Cai and colleagues (7). The diagnosis was based on clinical, radiographic, and histopathologic criteria. The study included patients with sarcoidosis (n = 9), pulmonary Langerhans cell histiocytosis (PLCH) (n = 3), benign metastasizing leiomyoma (BML) (n = 1), follicular bronchiolitis (n = 1), emphysema (n = 1), interstitial lung disease (n = 1), cystic lung disease (n = 1), and Mounier-Kuhn syndrome (n = 1) (Table 1).

Table 1.

Frequency of TSC2 LOH at Five Microsatellites in Lung Diseases

Patient	Type of Disease	Sex	Overall Allelic Status	TSC2 Microsatellite Markers
				D16S521	D16S3024	D16S3395	Kg8	D16S291
CN01	Sarcoidosis	F	ROH	NI	ROH	ROH	ROH	NI
CN02	Sarcoidosis	F	LOH	LOH	NI	LOH	ROH	NI
CN03	Sarcoidosis	F	ROH	NI	NI	ROH	ROH	NI
CN04	Sarcoidosis	M	ROH	NI	NI	NI	ROH	ROH
CN05	Sarcoidosis	M	ROH	NI	NI	ROH	NI	NI
CN06	Sarcoidosis	F	ROH	ROH	NI	NI	NI	ROH
CN07	Sarcoidosis	F	ROH	ROH	ROH	ROH	NI	NI
CN08	Sarcoidosis	F	ROH	NI	NI	ROH	ROH	NI
CN09	Sarcoidosis	F	ROH	NI	NI	NI	NI	ROH
P01	Benign metastasizing leiomyoma	F	ROH	NI	ROH	ROH	ROH	ROH
P02	Follicular bronchiolitis	F	ROH	ROH	NI	NI	ROH	NI
P03	Interstitial lung disease	M	ROH	NI	NI	NI	ROH	ROH
P04	Mounier–Kuhn syndrome	M	ROH	NI	ROH	NI	ROH	NA
P05	Cystic lung disease	F	ROH	ROH	NI	NI	ROH	NI
P06	Pulmonary Langerhans cell histiocytosis	F	ROH	ROH	ROH	NI	NI	ROH
P07	Pulmonary Langerhans cell histiocytosis	M	ROH	ROH	NI	ROH	ROH	NI
P08	Pulmonary Langerhans cell histiocytosis	F	LOH	NI	NA	NI	LOH	ROH
P09	Emphysema	F	ROH	NI	ROH	ROH	NI	NI

Definition of abbreviations: LOH = loss of heterozygosity; NA = not amplified; NI = noninformative; ROH = retention of heterozygosity.

Table 3.

TSC1 and TSC2 Loss of Heterozygosity in Cancer and LAM

Diseases	TSC1 Loss of Heterozygosity			TSC2 Loss of Heterozygosity
	Patient Loss of Heterozygosity Number [n (%)]	Markers		Patient Loss of Heterozygosity [n (%)]	Markers
Lung adenocarcinoma	7/9 (77.8)	D9S149, D9S150, DBH (18)		5/6 (83.3)	D16S291, D16S525, D16S300 (18)
Glial and glioneuronal tumors	9/89 (10.1)	PM4, PM2, PM1, exon 14, intron 21, exon 22 of TSC1, PM5 (19)		6/98 (6.1)	Kg8, exon 40 of TSC2, IVS8, LP7, LP10 (19)
Renal cell carcinoma	7/38 (18.4)	PM4, PM2, PM1, exon 14, intron 21, exon 22 of TSC1, PM5 (20)		1 /42(2.4)	LP1, IVS8, LP10, exon 40 of TSC2, Kg8 (20)
Oral squamous cell carcinoma	17/46 (37.0)	D9S179, D9S1830 and D9S915 (21)		18/46 (39.1)	D16S3024, D16S3395, D16S475 (21)
Bladder cancer	40/70 (57.1)	24 markers between D9S61 and D9S66 (22)
Ovarian cancer	26/33 (78.8)	11 markers on 9p, 12 markers on 9q (24)
Angiomyolipomas				3/29 (10.3)	D16S283, D16S291, Kg8, D16S525 (23)
Papillary carcinomas				2/6 (33.3)	D16S663, D16S665 (25)
Cardiac rhabdomyomas				2/6 (33.3)	D16S283, D16S291, Kg8, D16S525 (26)
S-LAM				47/52 (90.4)	D16S521, D16S3024, D16S3395, Kg8, D16S291 (7)

Definition of abbreviations: LAM = lymphangioleiomyomatosis; S-LAM = sporadic LAM.

Figure 1.

Relative chromosomal location of TSC gene–associated microsatellite markers. Left: TSC1. Right: TSC2.

Cells from blood and urine were sorted on the basis of cell surface markers (CD235a, CD45, CD9, and CD44v6) that have been shown to identify LAM cells in body fluids and cultured lung cells from patients with LAM (6, 7). CD235a⁺CD45⁻ and CD235a⁻CD45⁻ cells were isolated from blood, and CD44v6⁺CD9⁺ and CD44v6⁻CD9⁻ cells were isolated from urine. DNA was extracted from the different cell populations, and five microsatellite repeats spanning the TSC2 locus were tested (D16S521, D16S3024, D16S3395, Kg8, and D16S291) (6, 7).

Here, TSC2 LOH was detected in two patients (Table 1). In LAM, TSC2 LOH was observed in the CD235a⁺CD45⁻, CD235a⁻CD45⁻, and CD9⁺CD44v6⁺ cell populations. One case of sarcoidosis (CN-02) did not show TSC2 LOH in either subpopulation from blood but did show TSC2 LOH in the CD9⁻CD44v6⁻ cell population from urine at markers D16S3395 and D16S521. This cell population is different from the CD9⁺CD44v6⁺ cell population from urine that contains LAM cells. TSC2 LOH was also detected in PLCH (P08) at marker Kg8, but only in the unsorted blood cells, after density separation on Oncoquick columns (6), and not in the two subpopulations in which LAM cells are typically found (CD235a⁺CD45⁻ and CD235a⁻CD45⁻). In fact, cells presenting with TSC2 LOH in these different diseases or conditions appear to differ phenotypically (i.e., express different cell surface markers) from LAM cells. Thus, these data support the idea that circulating cells with TSC2 LOH are not unique to patients with LAM. We also investigated the presence of circulating cells expressing the glycoprotein CD1a, a marker for Langerhans cells, which are involved in PLCH lesions. Interestingly, one of the patients with PLCH (P06) showed TSC2 LOH in CD1a⁺ cells. Therefore, this work warrants further study to determine the role of TSC2 LOH in sarcoidal and PLCH lesions and in CD1a⁺ cells.

Chromosomal alterations have been reported in sarcoidosis, PLCH, BML, and chronic obstructive pulmonary disease. Sputum cytological specimens from patients with sarcoidosis showed genetic alterations in chromosome 9p and chromosome 17q (9). Lung from patients with PLCH showed LOH in chromosomes 9p and 22q (10). Loss of chromosome 1p in sputum cells occurred in patients with BML (11), and induced sputum cells from patients with severe chronic obstructive pulmonary disease showed LOH of chromosome 17 (12). LOH involving chromosome 14q was found in sputum samples of patients with asthma (13). Sarcoidosis and PLCH were not known to have a circulating disease-related cell, so it is noteworthy that circulating cells with TSC2 LOH were detected in these cases. It is unknown whether the circulating cells exhibiting TSC2 LOH in these diseases are relevant to disease pathogenesis or progression. It would be interesting to determine the activation of the mechanistic target of rapamycin pathway in these cells. Although TSC2 LOH may not be diagnostic for LAM in general, the presence of TSC2 LOH in cells isolated from either the CD235a⁺CD45⁻, CD235a⁻CD45⁻ populations from blood or the CD9⁺CD44V6⁺ cell population from urine seems to be characteristic of patients with LAM.

The TSC2 gene is mutated or associated with LOH in cancers in different organs and with different frequencies (Tables 2 and 3 and Figure 1). Some of the cancer studies tested the same microsatellite markers used to identify circulating LAM cells (Table 3 and Figure 1) (14). Among cancers, TSC LOH is frequently observed in lung carcinoma and LAM (Tables 2 and 3 and Figure 1). TSC1 or TSC2 LOH was detected in 139 or 92, respectively, of 476 lung cancer cases; 38 cases had both TSC1 and TSC2 LOH (14). Interestingly, LAM and lung adenocarcinoma are more frequently found in women (1, 15). Because the lung is exposed to environmental factors (e.g., pollution, airborne oxidants), these data suggest lung cells could be a source of metastatic cells with TSC2 gene mutations in cancers and LAM. Because LOH is an important event in the “two-hit” hypothesis of cancer development (16) and also represents a frequent event in tumorigenesis, we hypothesize that TSC2 LOH could be a common event in different cancerous processes, and specific patients may benefit from rapamycin treatment. There are more than 900 clinical trials currently looking at the use of rapamycin as anticancer agent (17).

Table 2.

TSC1 and TSC2 Loss of Heterozygosity and Mutations in Cancer

Cancer Tissue*	TSC1 Loss of Heterozygosity		TSC1 Mutation		TSC2 Loss of Heterozygosity		TSC2 Mutation
	Samples (n)	Loss of Heterozygosity [n (%)]	Samples (n)	Mutated Samples [n (%)]	Samples (n)	Loss of Heterozygosity [n (%)]	Samples (n)	Mutated Samples (%)
Lung	476	139 (29.2)	1,184	19 (1.6)	476	92 (19.3)	1,183	20 (1.7)
Breast	852	138 (16.2)	1,174	8 (0.7)	852	56 (6.6)	1,174	6 (0.5)
CNS	638	45 (7.1)	1,231	4 (0.3)	638	49 (7.7)	1,232	14 (1.1)
Endometrium	246	35 (14.2)	271	14 (5.2)	246	19 (7.7)	271	17 (6.3)
Kidney	435	87 (20.0)	911	11 (1.2)	435	29 (6.7)	847	20 (2.4)
Large intestine	540	36 (6.7)	723	30 (4.1)	540	20 (3.7)	723	31 (4.3)
Ovary	462	159 (34.4)	577	5 (0.9)	462	159 (34.4)	610	5 (0.8)

Definition of abbreviation: CNS = central nervous system.

^*Data are collected from catalog of somatic mutations in cancer database (14).