It has been suggested that the expression of p53 isoforms codified by the WTp53 gene (Fig. 9A) identified in breast cancer patients may influence prognoses and survivability (Dimas-González et al. 2017; Gallardo-Alvarado et al. 2019). Additionally, the expression of p53 protein isoforms has been shown to be associated with clinical response and patient outcomes such as cancer cell invasion (Gadea et al. 2016), TNBC (Avery-Kiejda et al. 2014), recurrence free survival, apoptosis, and DNA damage (Steffens Reinhardt et al. 2022b). Despite such critical information, to our knowledge this study is the first to identify and analyze the protein expression patterns of p53 isoforms and their relationship with clinical and pathological features, mitochondrial haplogroups, and mtDNAcn in Mexican breast cancer patients.
We first identified the frequency and distribution of mtDNA haplogroups in 91 tissue samples, showing that haplogroup A was the most frequent (37.36%), followed by B (28.57%), C (15.39%), and D (9.89%), K1a (3.3%), L2d (2.2%), H1ba (1.1%), J1 (1.1%), and U4 (1.1%). Similarly, in a previous study, haplogroup frequencies in 82 breast cancer patients in a Mexican population were reported as follows: A (41.7%), B (32%), C (18%), and (D 8.3%) (Domínguez-de-la-Cruz et al. 2020). In a second previous study, haplogroups were evaluated in 92 breast cancer tissues samples of Mexicans, showing haplogroup A with the highest frequency (44.6%) followed by B (22.8%), D (12%), C (11.9%), L (5.4%), H (2.2%) and J (1.1%) (Pérez-Amado et al. 2020).
The relationship between haplogroups and clinical features of cancer patients in this study showed that those in haplogroup A have a significantly higher average body weight (73.24±13.4 kg) compared with other haplogroups (67.52±10.7 kg), with a P= 0.03, while patients in haplogroup C had an average weight of 62.01±0.1 kg and a BMI of 26±3.6 kg/m2 significantly lower than the other haplogroups (71.05±11.9 kg, P= 0.007; 30±4.7 kg/m2, P= 0.006). This suggests that mitochondrial haplogroups and weight may contribute to the progression of cancer, since obesity has been reported as a risk factor for this disease. To our knowledge, there are no previous studies that linked Native American haplogroups with BMI in samples of breast cancer, suggesting a role for the mtDNA haplogroups and weight in this disease. Similarly, previous studies have shown an association between haplogroup A and obesity in a Mexican population with osteoarthritis (Ramos-Louro et al. 2022), while in European populations, haplogroup T was associated with obesity and with a decreased OXPHOS compared to individuals with haplogroup H (Ruiz-Pesini et al. 2000; Ebner et al. 2015), suggesting that together mitochondrial haplogroups and weight may contribute to development of different diseases.
When evaluating the relationship of haplogroups with pathological characteristics, we observed that breast cancer stages I-IV were distributed in A and B haplogroups, breast cancer stage IV was absent in haplogroup C, and patients in stages II and III exclusively displayed haplogroup D. Some data indicated that patients in stages II and III have worse prognoses, particularly stage III (Pascual et al. 2022). Furthermore, in our study we observed that the most frequent breast cancer stages were IIA (21.98%) and IIIA (25.27%). In a previous study including Mexican breast cancer samples from 2005-2014, stages IIB and IIIC were the most frequent (45.2%) (Maffuz-Aziz et al. 2017).
The histological subtype IDC was the most frequent in our analyzed samples (79.12%), while ILC was detected in 18.68% of samples, slightly higher than the 5-15%, which was previously reported in patients with invasive breast cancer, showing a marked global increase in the last two decades (Wilson et al., 2021). In haplogroups C and D, ILC displayed a frequency of 28.57% and 44.44%, respectively, compared with previous reports (Wilson et al. 2021). Some studies indicate an increase in ILC incidence in postmenopausal women, which may be related to postmenopausal hormone replacement treatment (Makki 2015). On the other hand, luminal A cancer was identified in all haplogroups, except for haplogroup D, that displayed exclusively luminal A cancer, although, the number of patients with this haplogroup was very low (n= 9).
HER2-positivity was identified in samples with haplogroups A (66.66%) and C (33.34%), while luminal B was recognized in samples with haplogroups A (66.66%) and B (33.34%). TNBC was present in all Native American and non-Native American haplogroups except for D, J1 and L2d, with an “n” of 9, 1, and 2, respectively. Similarly, in a previous study in a Mexican population, luminal A and B subtypes were identified in haplogroups B and C, and TNBC was found in samples classified as haplogroups A and B (Pérez-Amado et al. 2020). Interestingly, we detected a total of 4 BCT samples classified as HER3-positive in haplogroups A, B, C, and L2. HER3-positivity is an emerging biomarker of breast cancer and it has been shown that breast cancer patients with this receptor are resistant to therapies used for patients with other HER receptors, or for chemotherapy. Recently, the current strategy against HER3-positive patients is to block the receptor by using antibodies that recognize the amino acids of extracellular region of the HER3 protein (Gandullo-Sánchez et al. 2022).
When levels of p53 protein expression were analyzed, we showed that Δ40p53α, Δ133p53α, and p53β/γ isoforms were differentially expressed in BCT compared to NAT among patients with different clinical and pathological features, and different mitochondrial haplogroups, which may play a role in cancer development and progression. Overexpression of Δ40p53α in NAT compared with BCT in our study is in agreeance with previous reports showing that an increase in Δ40p53α mRNA expression led to favorable low-grade tumor and improved recurrence-free survival in serous ovarian and mucinous ovarian cancers (Hofstetter et al. 2011, 2012). Moreover, in hepatocellular carcinoma cells it has been observed that overexpression of Δ40p53α induces an increase of WTp53 protein expression in cancerous cells, leading to elevated levels of apoptosis (Ota et al. 2017). In melanoma cells, Δ40p53α suppresses cell growth and enhances senescence and G1 arrest (Takahashi et al. 2014). These antitumoral effects of Δ40p53α, reported previously, may be due to the lack of amino-terminal transactivation domain I, where the Mdm2 protein binds to WTp53 protein inducing proteasomal degradation, and therefore, Δ40p53α forms heterotetramers with WTp53, escaping Mdm2 binding and degradation of these tetramers (Fig. 8B). Thus Δ40p53α-WTp53 can regulate the transcriptional activity of WTp53 target genes that depend on the protein expression levels of Δ40p53α subunits contained in the heterotetramers (Hafsi et al. 2013). This data suggests that the high expression of Δ40p53α in NAT found in this study could have an antitumor effect to reduce the transformation of normal cells into cancer cells.
Our data showed that the Δ133p53α protein isoform was associated with BCT samples (OR= 2.02, 95% CI= 1.12–3.64, P= 0.02), and that the expression level was significantly higher in BCT compared to NAT samples (P= 0.01). These results agree with previous studies that detected mRNA expression of the Δ133p53α isoform in breast cancer tumors, although, the relative expression was not associated with tested clinical features (Avery-Kiejda et al. 2014; Gadea et al. 2016). Additionally, it is known that the expression of Δ133p53α inhibits WTp53-dependent apoptosis and G1 arrest without inhibiting WTp53-dependent G2 arrest, as well as promotes angiogenesis, metastasis, and the migration of endothelial cells by regulating the expression of angiogenic genes independently of WTp53 (Aoubala et al. 2011; Bernard et al. 2013).
Our results showed a higher expression of Δ133p53α in BCT samples from patients >50 years old (P= 0.0002), obese (P= 0.01), or overweight (P= 0.007) compared with NAT. It is known that senescent cells increase with age (Mylonas and O’Loghlen 2022) and with the accumulation of adipose tissue, which has been associated with an inflammatory process (Fujita et al. 2009; Singh et al. 2019). Furthermore, because it has been observed that overexpression of Δ133p53α can promote senescence escape and therefore cell proliferation, this can result in cell transformation from adenoma to carcinoma by repressing WTp53 that induce senescence genes, as occurs in colon cancer (Horikawa et al. 2017). Therefore, it is possible that this mechanism can be occurring in our study in BCT samples from patients >50 years old, and in samples classified as overweight and/or obese.
We also found that expression of Δ133p53α was higher in samples with the follow pathological features: stage IA (P= 0.01); histological subtypes IDC I (P= 0.0001) or IDC II (P= 0.002); or receptor status ER-positive (P= 0.01), PR-positive (P= 0.0001), HER2-negative (P= 0.0001), HER3-negative (P= 0.0002), or luminal A (P= 0.0002) in BCT compared with NAT. A related study (Nutthasirikul et al., 2013) showed high expression of Δ133p53α to be associated with stages III or IV and a significant negative association between Δ133p53α expression and survivability in patients with intrahepatic cholangiocarcinoma (Nutthasirikul et al. 2013). Furthermore, Δ133p53α expression was found to be significantly high in mucinous and serous ovarian cancers, while high expression of Δ133p53 was an independent prognostic marker for overall and recurrence-free survival in patients with p53 mutant ovarian cancer (Hofstetter et al. 2011, 2012). Also, Δ133p53α induced an invasive and migratory phenotype by activating the JAK-STAT and RhoA-ROCK pathways and a shorter disease-free survival in colorectal cancer cells (Campbell et al. 2018).
It is important to mention that the expression of Δ133p53α induced by gamma irradiation promotes double-stranded DNA breakage repair pathways in cancer cells, which may suggest that Δ133p53α confers tolerance to radiotherapy in cancer patients (Liu et al. 2015). The tumor effects of the Δ133p53α isoform acts in a dominant negative manner towards WTp53 due to the absence of the two transactivation domains of Δ133p53α but retains its binding domain that allows binding to WTp53 target promoters. This, in turn, inhibits the tumor suppression function of WTp53 via competition, as shown previously (Horikawa et al. 2017; Vieler and Sanyal 2018).
We did not find significant differences in p53β/γ expression levels between BCT and NAT, and there was no association between p53β/γ with any clinical feature. Previous studies have shown that increased p53γ expression is associated with reduced progression-free survival for patients with uterine serous carcinomas (Bischof et al. 2018). To the contrary, in breast cancer patients expressing mutant versions of p53 and p53γ had low cancer recurrence and an overall survival as good as patients expressing WTp53 breast cancer (Bourdon et al. 2011). In breast cancer, p53β expression has been shown to be negatively associated with tumor size and positively associated with disease-free survival. High expression of p53β was protective in patients with p53 mutations, suggesting that p53β may counteract the carcinogenic effects of p53 mutants (Avery-Kiejda et al. 2014). A recent study, however, found that high expression levels of p53β was associated with worse disease-free survival in tumors with WTp53 (Steffens Reinhardt et al. 2022a). To determine the differential expression between p53β and p53γ in our study it would be necessary to distinguish between mRNA expression of these isoforms.
In this study, the mean expression of p53 isoform protein was evaluated between haplogroups, showing higher expression of Δ133p53α in BCT compared to NAT in patients samples with haplogroups A or B (P= 0.0002 and P= 0.02, respectively). Therefore, because the highest expression of Δ133p53α in BCT showed poor survival due to cancer invasiveness activity and resistance to radiotherapy treatments, patients containing haplogroups A or B may have a risk factor for early death (Liu et al. 2015). High levels of expression of p53β/γ proteins in samples in haplogroup C in NAT (when compared to other haplogroups; P= 0.01) needs further study to differentiate between expression of p53β and p53γ in NAT to know which isoform is producing the higher expression, and producing a possible protective effect. Previous studies have shown that a high level of p53β or p53γ expression is associated with a reduction in tumor size and an improved prognosis for the patient (Bourdon et al. 2011; Avery-Kiejda et al. 2014). It should be noted, however, that there are also studies that have shown that high expression of p53β, is associated with poor prognosis in patients with breast cancer (Steffens Reinhardt et al. 2022a).
It is known that mtDNAcn decrease in patients with cancer including breast cancer (Mambo et al. 2005; Bai et al. 2011a). Importantly, we detected a negative correlation between mtDNAcn in NAT samples and the age of breast cancer patients. Previous studies have reported that mtDNAcn decreases with age after 50 years (Picard 2021). Furthermore, we detected significantly lower mtDNAcn in BCT samples classified as histological subtype ILC I and stage IIIA compared with their matched NAT. Similarly, samples classified as haplogroup C had lower mtDNAcn in BCT compared to other haplogroups, which may be due to A249del, A290del, and A291del variants detected in these samples (Domínguez-de-la-Cruz et al. 2020). These deletions, reported previously, may possibly affect the binding of mitochondrial transcription and replication elements that affect mtDNAcn, as has been observed with other variants identified in the D-loop region (Shadel 1999; Lee 2004). Related results have been reported in a U.S. population where decrease in mtDNAcn was detected in BCT compared with their matched NAT samples (Mambo et al. 2005; Bai et al. 2011b).
Decrease in mtDNAcn has been associated with a decoupling of OXPHOS and increase of ROS production to activate carcinogenic signals (Abd Radzak et al. 2022b). We discovered a positive correlation between haplogroup D and mtDNAcn both in NAT (P < 0.01) and BCT (P< 0.05), although the “n” is very low (9 samples). Haplogroup D has previously been associated with predisposition to breast cancer in a Chinese population, and experimental data indicated that the D5 haplogroup promotes tumorigenesis through over-activation of AKT, mediated by increased ROS, as well as a reduced mitochondrial OXPHOS function caused by a decrease in the activity of complexes I and III. No differences were detected in mtDNAcn between haplogroup D patients and control subjects (Ma et al. 2017).
It has been observed that the decrease of mtDNAcn can cause the decline of mitochondrial function to contribute to development of the Warburg effect, in which there is a decrease in mitochondrial respiration and an increase in glycolysis. This metabolism change promotes the development, progression, and chemoresistance of cancer cells (Vaupel and Multhoff 2021; Liu et al. 2021). Therefore, the reduction of mtDNAcn detected in BCT in our study may play a significant role in breast tumor progression. In summary, our data suggests that:
(1) Expression of WTp53 was detected in all BCT and NAT samples, potentially due to the activation of anti-tumoral genes to reduce cellular stress of cancer cells such as DNA damage, aberrant growth, hypoxia or other oncogenic signals (Pietsch et al. 2008; Chen et al. 2022b). This result, therefore, may be an indicator of tissue oncogenic environment (Fig. 9B, dark green color).
(2) Expression of Δ40p53α was significantly higher in NAT compared to BCT matched samples classified as histological subtype IDC I. This may be due to tetramer formation among Δ40p53α and WTp53 that slow tumor progression with anti-oncogenic capacity (Fig. 9B and C, light green color) (Hafsi et al. 2013).
(3) Expression of Δ133p53α (Fig. 9 B and D, yellow color) was higher in BCT samples in patients classified as >50 years old, overweight (BMI= 25–29.9 kg/m2), and/or obese (BMI= ≥30 kg/m2).
(4) Expression of Δ133p53α was higher in BCT samples with the following pathological features: breast cancer stage IA; histological subtypes IDC I and IDC II; receptor status ER-positive, PR-positive, HER2-negative, HER3-negative; breast cancer subtype luminal A; mtDNA haplogroups A or B, high expression of Δ133p53α observed in this study possibly prevents the anti-tumoral activity of WTp53 via competition to favor progression of breast cancer (Fig. 9B, D).
(5) p53β/γ isoform (Fig 9 B and E, orange color), expression in NAT samples were higher in patients classified as haplogroup C compared to other haplogroups (Fig. 9E), therefore, these patients may have a better prognosis, as has been previously reported for patients who express p53β or γ isoforms in breast cancer (Thompson et al. 2007; Bourdon et al. 2011).
(6) Breast cancer patients with a higher body weight were classified as haplogroup A more frequently than other haplogroups. To the contrary, patients in haplogroup C displayed lower bodyweight and BMI, suggesting that those patients with high weight (haplogroup A) may have a decrease in disease-free survival and overall survival, while those patients with lower weight (haplogroup C) may have an improved prognosis.
7) mtDNAcn was lower in BCT samples compared to NAT samples, and specifically in those classified as possessing histological subtype ILC I, breast cancer stage IIIA, or haplogroup C (Fig. 9F), which may be due to a decrease in the efficiency of electron transport chain complexes and OXPHOS, resulting in an increase of glycolysis (Warburg effect) to promote tumor progression.