To the best of our knowledge, this is the first study that used rs-MRI to investigate the differences in brain network topological properties in breast cancer patients with different expressions of δ-catenin protein.
In the present study, rs-fMRI was used to investigate the differences of different δ-catenin expression status on the topological properties of small-world brain networks in breast cancer patients before treatment. Clinical factors such as Her-2 were added as covariates to equate the groups. The results showed that the EG index of the DL group was higher than HC group before treatment, while it was shorter in the LP group compared to the HC group, indicating that the functional integration ability of the network and the resting brain activity were slightly enhanced in the early stage of cancer compared with the HC group. In addition, the network metric results showed differences between the two breast cancer groups, suggesting that δ-catenin has a specific effect on the right supramarginal gyrus and triangular part of the right Inferior frontal gyrus. There was no cognitive decline in the two groups before treatment, which suggested that brain network changes induced by δ-catenin precede cognitive dysfunction, and early brain damage induced by high expression of δ-catenin may be involved in executive function and emotion regulation.
δ-catenin serial protein is an aneural platelet affinity protein-related armadillo repeat protein (neural plakophilin-related armadillo repeat protein, NPRAP) that was considered to be neurospecific and was first discovered by German scientists in 1997 through the sequencing of brain cDNA and was considered to be neurospecific [30]. It is a component of the cadherin-catenin cell adhesion complex. In recent years, more and more studies have shown that the expression of δ-catenin is up-regulated in some epithelial tumors [5–8]. Since δ-catenin contains multiple structural domains and is located at a key point of the signaling pathway at the molecular level, it is involved in and affects the occurrence and development of diseases in multiple ways [31–33]. It has been speculated that the increased expression of δ-catenin gene level in tumors may be related to multiple mechanisms [31–36]: (1) δ-catenin can regulate the expression levels of apoptosis-related proteins, such as Bcl-2L, surviving, cleaved caspase 3. It can also promote the proliferation of tumor cells and inhibit apoptosis; (2) δ-catenin protein may be related to its ability to stimulate the secretion of vasoactive factors in tumor tissues and participate in neovascularization. At the same time, this protein can also regulate the activity of small GTPases; (3) δ-catenin may be associated with overexpression of transcriptional activator E2F1 or co-expression of transcriptional repressor Hes1; (4) there are also data showing that the δ-Catenin mediated FAK/Src/Rac1 signaling pathway can promote tumor development. Her-2 is a transmembrane protein that has a key role in regulating cell growth, apoptosis, and differentiation, and its overactivity can lead to malignant biological behaviors of the mammary gland [37]. In this study, only the Her-2 index differed between DH and DL groups in clinical data, and the results also showed that Her-2 was a very important factor for breast cancer. Therefore, to exclude the influence of Her-2 on patients, Her-2 was taken as a covariable in statistical analysis to eliminate the influence of Her-2 index. The obtained results are indicative of the unique effect of δ-catenin protein on patients. Studies have suggested that the synergistic effect between δ-catenin and Her2 may have an important role in the progression of tumor invasion. By promoting the expression of cyclin CyclinD1 and Cdc34 and promoting the phosphorylation of Her-2, δ-catenin promotes cell mitosis, which may lead to the tumor-promoting effect of this protein [38]. In this study, the Her-2 index was significantly higher in the DH than in the DL group, which indirectly showed that high expression of δ-catenin in breast cancer patients was related to malignant cell behavior.
Lp and Eg are functional integration indicators, and the functional integration ability in the brain is measured by estimating the ease of communication between brain regions, which is usually based on the concept of path. Generally, the shorter Lp is, the stronger the functional integration ability is. Eg is the average value of node efficiency in the whole brain, which measures the information transfer ability between nodes in the network. Cp and Elog are used to measure the local information transmission capability of the network. In this study, the AUC values of Lp, Eg, σ, and γ were different between the three groups under sparsity conditions. Compared with the HC group, Lp was shorter, and Eg was higher in the DL group. These results indicated that the functional integration ability of the brain network in the early DL group was slightly stronger compared to the HC group. There was no significant difference between the DH and HC groups in Lp and Eg indexes. Only the values of σ and γ indexes in the DH group were slightly higher than those in the HC group.
In this study, we observed elevated NE, BC, and DC in some brain regions in DH and DL groups, and the elevated NE was concentrated in the right superior frontal gyrus, inferior frontal gyrus, bilateral caudate nucleus, bilateral thalamus, and right putamen. The BC value of the right caudate nucleus in the DH and DL groups was higher than that in the HC group. The DC value of the bilateral caudate nucleus and bilateral thalamus in the DL group was higher than that in the HC group. These results showed that some brain regions in the DH group and DL group had a compensatory increase in resting state activity, which primarily occurred in the frontal lobe [39, 40], cingulate gyrus, caudate nucleus [41, 42], thalamus[43, 44] and other brain areas related to language and high-level cognitive function. As these brain areas are also important functional areas of the default mode network (DMN) [45, 46], these results suggest that more neuronal synergy is needed to maintain as much as possible of subjective cognitive function. Areas in the DMN generally have been shown to have high resting cerebral blood flow, and multiple studies [47–50] have shown that brain regions with high resting cerebral blood flow tend to have increased resting state indexes (such as ReHo/Alff, etc.). Moreover, the weakened function caused by neuronal injury can be compensated by increasing neuronal recruitment and activation [51–54], which is regarded as a manifestation of neuronal compensatory response. The present study speculated that after the patients were informed of their cancer, they feared that treatment would bring more negative effects in the early stage so that the brain increased the intensity of spontaneous activity by recruiting neurons and increased the activity of the resting state brain, so as to achieve the inhibitory excitation balance of the early brain network.
Although there was a certain equilibrium compensation mechanism, we found several differences in the small-world attribute indexes of brain networks between the two groups of patients with different expressions of δ-catenin. The NE and DC indexes showed differences between the DH and DL group in the right supramarginal gyrus, a part of the parietal lobule and an important part of the frontoparietal network, which mainly brings together the nerve tracts of the parietal, temporal, and occipital lobes [55, 56] that are involved in cognitive control and processing of working memory. The BC indices showed differences in the triangular part of the right Inferior frontal gyrus. The triangular inferior frontal gyrus belongs to the prefrontal lobe's sub-region, the most highly developed and complex part of the brain involved in many cognitive activities such as memory, attention, emotion, and language. It is also considered the higher emotional center. Pathological anxiety may be related to its abnormal function [57, 58]. These results indicated that the expression of δ-catenin had a unique effect on these two brain regions, which belong to the prefrontal cortex, and are associated with the frontal lobe-striatum frontal, parietal network. Both of these are involved in cognitive and memory processing. This is consistent with our results detecting differences in local efficiency in neuronal processing as well as a statistical difference in memory performance between the two groups.
Considering the BC and DC indexes, the two groups showed decreased functional areas in different brain regions. The BC value of the left Anterior cingulate gyrus Para cingulate gyri, and the left paracentral lobule in the DL group were lower than the HC group. The DC value of the left inferior parietal supramarginal and angular gyri, left parietal lobule and superior temporal gyrus in the DL group were lower than in the HC group. The reduced brain areas in the DL group were mostly distributed in the left hemisphere. In the DH group, the decreased brain area was mainly located in the right hemisphere compared with the HC group: the BC value of the right median cingulate and paracingulate gyri was decreased, and the DC value of the right caudate nucleus and the left Heschel gyrus was decreased. The results of this study indicated asymmetric brain distribution differences. In the absence of chemotherapy drugs and other indicators except for Her-2, we speculated that different expressions of δ-catenin had different effects on the function of bilateral cerebral hemispheres in patients with early breast cancer. Although there is a certain compensation mechanism in patients, the effects of different brain regions caused by different states of δ-catenin in patients exceeded the compensable range, and the system balance could not be maintained, which led to the weakening of the information processing ability of the corresponding damaged brain regions. However, it remains unclear to what extent the damage-compensatory effect of δ-catenin on the right brain ceases to have an effect after the disease progresses, which should be further addressed by a longitudinal study.
There are some limitations in the present study. First, although the sample size of this study met the requirements for the analysis, it was still not large enough. In addition, this was a single-center study. Increasing the sample size and conducting multi-center studies in the future may further improve the accuracy of verifying the influence of δ-catenin protein. Second, the patient's disease course is an important variable that may affect certain cognitive, emotional, and brain network changes. However, we could not accurately control this variable, so further longitudinal analysis is needed to improve the understanding of the changes in the neural mechanism related to the patient's hidden executive function under the impact of δ-catenin protein. Third, this study only discussed the small-world attributes of brain networks, and the next step is to conduct the resting state indicators of rs-fMRI signs and other sequence analysis methods, such as the volume of white matter, white matter fiber bundle connections, and other indicators. Fourth, the current comparison index was only based on the baseline group data before chemotherapy, and the first, second and third scan data after short-term chemotherapy is being supplemented to conduct a longitudinal study of patients with different δ-catenin groups of breast cancer before and after, so as to further explore the physiological mechanism of the influence of δ-catenin protein on breast cancer patients.
In summary, the present study adopted the topological attributes of the resting small-world brain network to study the change of δ-catenin in breast cancer patients. All the results in this group showed that different states of δ-catenin protein had a significantly different effect on the attributes of the patients' brain network and had characteristic effects on some brain regions. They could also be involved in executive function-related cognitive functions and changes in regulating emotions. It has also been demonstrated that δ-catenin is an important cancer protein. Future self-controlled studies in patients before and after chemotherapy are needed to prove that high expression of δ-catenin is associated with cell malignancy-related behaviors in breast cancer patients so as to use δ-catenin protein as a potential biomarker for breast cancer patients. This could also have important clinical significance for the early prevention of cognitive impairment in patients.