Personalized evolutionary hypothesis of breast neoplasm based on differential hybridization signals of subtelomeric probes

Introduction

Breast cancer (BC) is the most common cause of cancer mortality among women worldwide. Therefore, providing reliable cellular targets for early detection of biological behavior and an appropriate clinical management is urgent. In the early 1970s, it was initially recognized that chromosomes (Chrs) are not able to replicate their terminal regions. Leonard Hayflick stated a limited somatic cell division theory, in which DNA sequences could be lost during cell replication. A critical level would be reached, thus triggering cell division termination (Olovnikov, 1971, 1973).

The terminal region of Chrs is formed by telomere repeats followed by subtelomere (ST) sequences, which are characterized by repeats sequences (Churikov and Price, 2008). Therefore, length of telomeres and complexity of STs could indicate the evolutionary history of the cell (Denayrolles et al 1997; Van Overveld et al., 2000; Linardopoulou et al., 2001). It has been shown that reduced telomere length (TL) in peripheral blood lymphocytes is correlated with an increased risk of developing cancer (Pooley et al., 2010). We have also reported that diverse TL in cancer cells play a critical role in cancer progression. Furthermore, the genomic and somatic evolution, TL diversity, and heterogeneity in BC and brain tumor have been previously published (Mehdipour et al., 2011). Breast cells are in close contact with the lymphatic channels, which pass through lymph nodes, acting as a filtering system against hazard substrates and cancer cells. The number and nature of the auxiliary lymph node (ALN) are usually used as a supportive and diagnostic tool in BC management (Fregnani and Macea, 2009).

Subtelomeres have also been assayed by fluorescence in situ hybridization (FISH) probes in individuals with mental retardation (Flint et al., 1995; Knight et al., 1999). Later, quantitative-FISH analysis was performed on individual Chrs (Zheng et al., 2009, 2011). Subtelomeres corresponds to about 0.1% of the human genome, being involved in more than 40% of interchromosomal duplications (Linardopoulou et al., 2003). This region is prone to DNA breaks and repair (Fagagna et al., 2003; Rudd et al., 2007). However, the precise role of STs in cancer development and progression requires further investigation, especially in BC.

Due to the high diversity, distinct evolution, tumor heterogeneity, and aging of patients, the importance of major and minor ST differences in cells is relevant. Such differences might affect tumor behavior, especially in breast neoplasm, including benign tumors and BC. In BC, for example, a severe heterogeneity could be traced within a specific cancer type. Therefore, measuring the global TL length as a sole diagnostic criterium seems inadequate. In this sense, ST analysis by FISH could be a complementary information regarding cancer predisposition, prognosis, and prediction for further clinical management. In addition, it is worth remembering that telomere, telomerase, and STs act as a triangle and three-edged sword, by which decision making on patient's aging and cancer development would be possible.

We have recently reported a different profile of FISH signal, which reflects a distinct copy number of ST sequences of buccal and ALN cells and provides a personalized signature for the progress of breast neoplasm (Mehdipour et al., 2013). Given this personalized insight, the aim of the present study was to analyze the ST profiles for distribution of hybridization signals and intensity of individual p and q Chr arms in non-affected (genomic) and ALN cells of a female patient previously affected by primary proliferative ductal carcinoma. Moreover, the ALN was diagnosed as having a benign lesion, in which the ST characteristics could pave the way through a personalized follow-up study for the possible development of any malignant lesion in this patient.

Materials and methods

Results

Signal copy numbers and signal intensities in ALN and G cells

Two cell populations, including normal and abnormal cells, were analyzed in order to assess their heterogeneity. The SCN and SI of FISH were used to characterize the STs present in each or all Chrs and distinguish the p and q arms.

Figure 1 shows, a typical labeling of ALN and relatively unaffected lymphocytes from the same patient, with specific probes for the p and q arms of Chr 1. Analysis with all probes showed that cells with positive signals correspond to 11.2% (4506/40.197) for ALN and 28.4% (2415/8494) for unaffected G cells. Highly significant differences were found in labeling each Chr (P < 0.001) when the status of the p/q ratio for SCNs in the p and q arms of ALN cells and lymphocytes was compared (Table 1). Moreover, distinct SIs, scored as weak-, medium-, and strong-intensity, also indicated significant mean differences in each Chr arm in signal-positive cells when comparing ALN and genomic cells (P < 0.001) (Table 2). However, lymphocytes without strong SI were found only in Chrs 6 and 12. A summary of differences between SCN and SI for the p and q arms of each Chr is shown in Table 3, emphasizing that the mean of ratio value for SCN and SI in genomic cells was significantly higher than in ALN cells. Specifically, the mean value for SI of p arms of all Chrs in G cells was also significantly higher than for weak and medium SIs in ALN cells, but not in cells with strong SI.

Table 1. Comparable analysis of signal copy numbers (SCNs) and signal intensities (Sis) by subtelomere fluorescence in situ hybridization labeling of chromosomes (Chrs) in the genomic (G) and auxiliary lymph node (ALN) cells and lymphocytes

		SCNs per Chr				SIs per Chr
Target tissues	Arms	Zero	One	Two	Three	Weak	Medium	Strong
G>ALN	p	–	All Chrs, except 2, 6, 8, 11	All Chrs	3, 5, 6, 7, 10, 11, 18, X	All Chrs	All Chrs, except 6, 17	All Chrs, except 6, 8, 11, 20
G>ALN	q	–	1, 4, 5, 7, 9, 10, 11, 12, 13, 14, 17, 18, 19, 20	1–11, 13, 14, 18-22, X	2, 3, 10, 20, 21, 22	1–10, 12–15, 18-22, X	1–4, 7–15, 18–22	1–4, 7, 9–11, 13–15, 17–22, X
ALN>G	p	All Chrs	–	–	4, 8, 9, 16, 17	–	All Chrs, except 6, 8, 11, 20	–
ALN>G	q	All Chrs, except 13, 16, 18	–	12, 15, 17	-	11, 16, 17	5, 6, 16, 17, X	5, 6, 8, 12, 16
G = ALN	p	–	–	–	20	–	–	–
G = ALN	q	–	–	–	–	–	–	–

• >, The value in G cells is greater than in ALN cells or vice versa; =, The values are equal in both G and ALN cells.

Table 2. Comparative analysis of the p and q arm signal intensities (SIs) in the auxiliary lymph node (ALN) and genomic (G) cells by fluorescence in situ hybridization

	SIs per Chr in the p arm			SIs per Chr in the q arm
Target tissues	Weak	Medium	Strong	Weak	Medium	Strong
G>ALN	2, 3, 5, 6, 9, 11, 12, 16, 17, 18, 19	1, 3, 5, 6, 7, 10, 11, 16, 18, 20, X	2, 3, 7, 10, 12, 18, X	1, 2, 3, 4, 6, 8, 9, 10, 14, 15, 20, 21, X	1, 2, 3, 9, 10, 11, 12, 13, 22	1, 2, 3, 7, 9, 10, 11, 13, 15, 17, 19, 20, 22, X
ALN>G	1, 4, 7, 8, 10, 20, X	2, 4, 8, 9, 12, 17, 19	1, 4, 5, 6, 8, 9, 11,	5, 7, 11, 12, 13, 16, 17, 18, 19, 22	4, 5, 6, 7, 8, 14, 15, 16, 17,18, 19,20,	4, 5, 6, 8, 12, 14
			16, 17, 19, 20		21, X	16, 18, 21
G = ALN	–	–	–	–	–	–

• >, The value in G cells is greater than in ALN cells or vice versa; =, The values are equal in both G and ALN cells.

Table 3. Summary of characteristics in subtelomere labeling of chromosomes (Chrs) in the auxiliary lymph node (ALN) and genomic (G) cells

		From total cells		From positive cells
Chr-arms	Factors	ALN	G	ALN	G	P values
p	Zero	0.90 ± 0.04	0.73 ± 0.13	_	_	_
	One	0.083 ± 0.031	0.11 ± 0.05	_	_	_
	Two	0.012 ± 0.006	0.15 ± 0.09	_	_	_
	Three	0.0017 ± 0.0018	0.0068 ± 0.0078	_	_	_
	Weak SI	0.056 ± 0.025	0.16 ± 0.09	0.46 ± 0.13	0.48 ± 0.15	**
	Medium SI	0.022 ± 0.011	0.067 ± 0.042	0.19 ± 0.09	0.19 ± 0.08	**
	Strong SI	0.019 ± 0.007	0.041 ± 0.026	0.17 ± 0.07	0.13 ± 0.06	**
q	Zero	0.92 ± 0.18	0.88 ± 0.07	_	_	_
	One	0.042 ± 0.024	0.053 ± 0.030	_	_	_
	Two	0.0031 ± 0.0028	0.065 ± 0.052	_	_	_
	Three	0.0005 ± 0.0007	0.0016 ± 0.0042	_	_	_
	Weak SI	0.022 ± 0.013	0.068 ± 0.047	0.23 ± 0.15	0.24 ± 0.19	**
	Medium SI	0.013 ± 0.009	0.028 ± 0.022	0.15 ± 0.11	0.10 ± 0.10	**
	Strong SI	0.0095 ± 0.0065	0.025 ± 0.019	0.11 ± 0.10	0.10 ± 0.09	**

• SI, signal intensity.

In cells with either one or two SCNs, the mean value in genomic cells was also higher than in ALN cells. In contrast, the mean values for three SCNs in genomic cells were higher than in ALN cells for half of the chromosomal set. As an example, we show the image of Chr 20 (Figure 2). For two and three SCNs, the mean value in genomic cells was higher than in ALN cells in each Chr.

The mean difference between SCNs for the p/q ratio of ALN and genomic cells per individual Chrs from total cells was also evaluated (data not shown). In the case of labeling without SCN or presenting one signal, the mean value of the p/q ratio reflected significant differences for all Chrs (P < 0.001). Similar significance levels were obtained for two SCNs, except in Chrs 2, 11, 12, and 17, which showed a lower significance level (P < 0.05). In the cases of three SCNs, differences were highly significant for Chrs 3 and 20 (P < 0.001), and also significant for Chrs 2, 5, 7, 8, 10, 16, 18, and X (P < 0.05). For more information, the p/q ratio is also shown for all Chrs (Table 3).

Telomere length analysis

In the present investigation, we evaluated the status of TL in lymphocytes and ALN cells. The ALN showed a decrease in TL as detected by Southern blot and Q-FISH analysis (Figure 3A).

Protein expression of the key cell cycle markers

We also assayed the key targets involved in the cell cycle control of ALN cells to evaluate their degree of proliferation. In this regard, we analyzed the Ki67, cyclin D1, and cyclin E expression profiles (Figure 4) and we found a labeling diversity in the cell population.

Discussion

The human telomeric structure contains the simple tandem repeat (TTAGGG)_n, which is accompanied by an extraordinary and complementary complex ST domain (Moyzis et al., 1988; Brown et al. 1990). In a previous publication (Mehdipour et al., 2011), we provided some insights comparing TL in tumor and other tissues. In addition, the ST of Chr 3q has been reported to be related to 35 other chromosomal ends (Trask, 2005; Brown et al., 1990), and the ST of Chr 7q was found to have a simple structure (Riethman et al., 1993). Eichler suggested that “exceptional duplicated regions underlie exceptional biology” (Eichler, 2001). The ST domain is a surprising region of the human genome, with an unusual evolutionary history, structure, and function. Therefore, STs are dynamic structures that may play a role in diversifying gene families, creating phenotypic diversity, and promoting disease-causing rearrangements.

However, no publication on human ST signals and SI for solid tumors, including BC, was available. Therefore, the aim of the present investigation was to explore the ST region by a systematic follow-up study of a patient with primary BC at the G and ALN levels. This study highlights the diversity and nature of ST signals and SI in the p and q Chrs’ arms of lymphocytes and ALN cells in a patient with BC.

Neoplastic tissues can act as proliferative reservoirs, supplying cells for cysts within the breast tissue or ALN. The formation of neoplastic tissue may be complicated by its progression. In this case, early detection, prediction, and prognosis is possible.

Regarding ST neighborhood, the telomeric diversity and evolution was discussed at the genomic and somatic levels (Mehdipour, 2013). The present results support the idea that remarkable cooperation between telomere and ST territories seems to exist, thus highlighting a further heterogeneity in human neoplasm. Although telomeres protect the Chr ends, SIs have extraordinary characteristics, being able to initiate and promote diversity.

Furthermore, there is a chain of events in the ST organization, including evolution, variation, and functional machinery, which may play a role in structural aberrations. STs are sensitive to macro- and micro-environmental factors, being prone to cause neoplastic initiation and progression.

Despite the limited publications on STs, this region is highlighted as “the gene-rich nature” of the ST region (Saccone et al., 1992). In addition, the role of ST rearrangements was reported in congenital abnormalities and mental retardation (Flint et al., 1995). Thus, here we tried to provide information to draw an individual map of signal distribution and intensity through the p and q arms of each Chr in different tissues of a patient with primary BC. This could provide a follow-up in different stages of the patient's life, allowing planning and management in the case patients with BC or even healthy individuals who have both been affected by hazard of environmental factors and inappropriate nutritional habits.

Collectively, higher and lower SCNs at ALN cells could be considered as a heterogeneous alteration at the neoplastic level; differently, lower SCNs in lymphocytes could be addressed as a reliable warning sign at the genomic level. Higher SCNs at the genomic level may suggest two possibilities: (i) an inherited positive prognostic trait or (ii) an acquired predisposing alteration.

Regarding the mean difference between SI values for the p arm, we found that all Chrs are predisposed to be affected by loss of genetic material (Table 1). Furthermore, the scenario in the q arm was found to be different regarding the mean difference between SCNs in the ALN and genomic cells. By focusing on cells without SCN, an exception was found in the q arm of Chrs 13, 16, and 18. By comparing the p and q arms for no signal, both arms of Chrs 1, 12, and 19 were involved.

However, more Chrs seem to have a higher mean value at the genomic level for the p arm. By comparing the p and q arms for the presence of three SCNs, Chrs 3 and 10 were found to be involved. By comparing the p and q arms for SI, all p arms showed higher mean values for weak SI in G cells on the basis of total cell count. But q-arm with weak SI was less frequent at the genomic level than in ALN, whereas Chrs 11, 16, and 17 showed lower SI value in the ALN than in G cells. However, reliable predisposing targets include Chrs 5, 6, 11, and X. Nevertheless, most Chrs are also trustable early detective targets at the genomic level (Tables 1 and 2). The use of the ST DNA probe only for Chr 17q was considered for breast tumor (Rashid-Kolvear et al., 2007).

Regarding the SI of the p arm, almost all Chrs showed higher intensity for somatic cells. In the ALN, a higher signal value was found in most Chrs, most characteristically in Chrs 6 and 8. In addition, weak SI at the genomic level was found only in Chrs 6 and 12. Collectively, the spectrum of diversity for SI at the genomic or ALN levels could reflect the ST heterogeneous behavior of the p and q arms in two different cells that show significantly different behavior. These findings could characterize different Chr arms with diverse behavior.

In addition, exploration of heterogeneity within a cell population with positive signals could suggest more reliable and specific estimations. Regarding the mean difference between different SI degrees in the p and q arms, we concluded that: (i) Chrs 2, 3, and 10 are modified in both arms, indicating that most Chrs behave differently in the p and q arms of G cells, even more than in ALN cells; (ii) in the ALN, most Chrs are differently affected by diverse SI categories, with values significantly higher than in G cells.

By evaluating both arms as a unique target, these findings show that there is no difference between the p and q arms. This is just to demonstrate that the p and q arms of individual Chrs with distinct regions and behaviors are influential targets in cancer cell biology. Such findings not only strengthen the analysis of individual Chrs, but also highlight the application of such a strategy in personalized cancer management.

By signal-positive cell-based analysis of the p arm in the whole Chr set for weak and medium intensity, the mean value for different SI categories in G cells was significantly higher than in ALN cells. Regarding q arm, only mean for cells with weak SI in lymphocytes were higher than in the ALN. Moreover, this also highlights that the p and q arms behave differently. However, these distinct results indicate the heterogeneous ST behavior at either genomic or ALN levels, and this could be considered as a predisposing factor at both genomic and neoplastic levels.

We also compared the global characteristics of SCN for the p and q arms as a common territory between ALN and G cells. For zero SCN, the mean in G cells was higher than in ALN cells and for all Chrs, except for Chrs 2 and 4, which may be involved in the pathogenic process at the somatic level. Regarding one SCN, no similar Chr involvement between two categories could be traced, including when the SCN in G cells was higher than in ALN and vice versa. This finding reflects the different behavior of G cells, which seems entirely different from that of ALN cells. However, the noticeable point is that losing genetic material is the key role in this category. Regarding two SCNs, the mean in G cells was higher than in ALN cells for all Chrs. This indicates the severity of the event in cancer cells, where down-regulated and heterogeneous ST alteration occurred at the neoplastic level with a benign histopathologic diagnosis. Such findings are warning signs for early management at the ALN level and probably at a further progressive level. Regarding three SCN, the mean in G cells was higher than in ALN cells. Interestingly, only Chr 7 showed similar mean values in the G and ALN cells. Finally, by comparing medium and strong intensities, it seems that Chrs 2, 3, 7, 10, and 13 were affected. Thus, the predisposition of both Chrs 2, 3, and 10 for different SI degrees in G cells is greater than in ALN cells.

The next step was to compare the total mean for SCN and SIs in the p and q arms in ALN and G cells as a Multi-Target Platform (Tables 1 and 2). It is worth to emphasize the importance of individual Chrs and each arm. With this regard, a comparison between the total mean for the p arm in ALN and G cells revealed a significantly higher value for one and three SCNs in G cells. However, such values were significantly higher for one and two SCNs of the q arm in the total cell count. These findings indicate that the p arm was significantly predisposed for signal loss and gain at the genomic level, highlighting the p arm as a reliable and prognostic marker candidate for early detection of neoplasm in this patient and probably in her relatives. However, the q arm was predisposed only for loss of the SCN.

Regarding different SI degrees, they were significantly higher in G cells than in ALN cells for both p and q arms (Tables 1 and 2). Such different intensities provide a wide spectrum for early qualification of both p and q arms, which also indicates molecular alterations at the ST region. In addition, the mean for SCN in ALN cells was higher than in genomic cells, indicating an evolutionary heterogeneity at the ALN level. By evaluating the total mean differences between SIs in the signal-positive cell population, a higher mean of weak and medium intensity was found for the p arm. However, weak intensity for the q arm was found only at the genomic level (Table 3). These results indicate that SI at the genomic level for the p and q arms is reliable.

Briefly, only a few cells exhibited a correlation between the high SCN and SI for the p and q arms of Chr 1 in both cellular environments. Furthermore, the heterogeneous distribution of SCN and SI of the p and q arms in both target tissues was remarkable. The strong SI in the q arm of Chr 1, even as a minor clone, could be a key element for a further developmental process, such as metastasis in ALN cells. In addition, the merged image of the p and q arms indicates the different ST characteristics of the p and q arms in G and ALN cells.

In addition, based on the comparative profile of the TLs, telomere signal pattern, and ST signal in the whole set of Chrs, we assumed that the global TL in the left breast of the same patient with primary infiltrating BC was initially low, either at the genomic or tumor level by the age of 36. Furthermore, telomerase activity was found to be positive in this patient (Mehdipour et al., 2011).

Most publications have focused on the molecular, and not on the classification of cellular patterns. The sequences of STs in Chrs 4p, 16p, and 22q were reported to have similar structures (Flint et al., 1997, 1997; Daniels et al., 2001). However, we have observed that these Chrs have different characteristics regarding SCN and SI. In this regard, it was previously stated that Chrs 20p, 4q, and 18p have a dual-domain structure (Chute et al., 1997). Our data allowed us to characterize individual Chrs according to the status of SCN and SI, which reflect reliable interactions in different cell populations and two different chromosomal arms. Moreover, interaction between frequent repeats leading to exchange has been reported only for Chrs 4q and 10q (Van Overveld et al., 2000). Such common characteristics could be specifically traced through different investigation channels, including SCN and SI.

Conclusion

Human subtelomeric sequences are variable, therefore we aimed to consider these regions as the hot spots marker to unmask the personalized evolutionary alterations in the breast neoplasm. The provided periodic charts allowed us to generate ST signal profiles characteristic of the alterations in SCN and SI in the patient's chromosomes (Figures 5 and 6). This resource can be also applied for other patients and establish correlations between evolution and the nature of BC and/or ALN lesions. The vertical and horizontal arrays show ST variations through individual chromosomes at the genomic and lymph node levels. Regarding the various categories of the p/q ratio for SCN, the most harmonic behavior was found in Chr 18, in its vertical profile. By focusing on the qualitative aspect, SI profiling mostly shows high intensity in almost all chromosomes, except numbers 1, 5, and 18. However, a non-uniformity in all SCN categories and both normal and medium SI was observed for Chr 18. In contrast, most characteristics for the absence of SCN and week SI show a regular mode. Our data, together with karyotyping, may lead to unmask the new fragile sites in breast neoplasm. The charts facilitate translation of the reliable personalized information, from lymphocytes to ALN or vice versa, for appropriate management. Furthermore, the periodic charts can be used as complementary single cell-based array systems in the cell domain, including tumor biology. Given the high significant correlation between the ST co-factors, stability and instability of the ST behavior (including quantities and qualities of genetic material), non-invasive samples (including genomic cells) are assayable and can be further followed up at the neoplastic level for each individual. In addition, analysis of ST profiling in the peripheral blood is a key gate for early prediction through bridging cancer cell biology to the personalized and translational insights. The cascade of evolution relies on tumor heterogeneity through the patients’ (and apparently healthy individuals’) life in different tissues, and these are the fundamental events of the provided Hypothesis, which facilitates an early detection within the ‘Prediction - Prevention - Prognosis’ frame.