Chemokine receptors differentially expressed by race category and molecular subtype in the breast cancer TCGA cohort

Abstract

Racial disparities in mortality due to metastasis remain significant among breast cancer patients. Chemokine receptors contribute to breast tumors and metastatic outcome. We explored for significant differences in chemokine receptor expression in breast tumors from Black, Asian, and White patients in The Cancer Genome Atlas. We show that despite sharing the same molecular subtype, expression of the chemokine receptors ACKR1, CCR3, CCR6, CCRL1, CCRL2, CXCR1, CXCR2, CXCR4, CXCR6, and CXC3CR1 was significantly different depending on racial group. For patients with triple negative breast cancer, CCR3 was higher in Black versus White and CCRL2 was higher in Asian versus White. In luminal A tumors, ACKR1 was lower in Asian versus White, CCR3 was higher in Black versus White, and CCR6 and CXC3CR1 were lower in Black versus White. In luminal B tumors, CCRL2 was lower in Black versus White, CXCR1 and CXC3CR1 were lower in Asian versus White, and CXCR2 was lower in Black and Asian versus White. In HER2 enriched tumors, CCR3 was higher in Black versus White and CXCR4 lower in Asian versus White. CCR3, CCR6, and CXCR6 associated with worse patient survival. These findings can inform improved treatment strategies to decrease racial disparities in breast cancer burden.

Full text

1
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports
Chemokine receptors
dierentially expressed by race
category and molecular subtype
in the breast cancer TCGA cohort
Elissa D. Vazquez1,2, Xiangyi Fang1,2, Lauren A. Levesque1, Mike Huynh1, Citlali Venegas1,
Nhien Lu1 & Nicole Salazar1*
Racial disparities in mortality due to metastasis remain signicant among breast cancer patients.
Chemokine receptors contribute to breast tumors and metastatic outcome. We explored for signicant
dierences in chemokine receptor expression in breast tumors from Black, Asian, and White patients
in The Cancer Genome Atlas. We show that despite sharing the same molecular subtype, expression
of the chemokine receptors ACKR1, CCR3, CCR6, CCRL1, CCRL2, CXCR1, CXCR2, CXCR4, CXCR6, and
CXC3CR1 was signicantly dierent depending on racial group. For patients with triple negative breast
cancer, CCR3 was higher in Black versus White and CCRL2 was higher in Asian versus White. In luminal
A tumors, ACKR1 was lower in Asian versus White, CCR3 was higher in Black versus White, and CCR6
and CXC3CR1 were lower in Black versus White. In luminal B tumors, CCRL2 was lower in Black versus
White, CXCR1 and CXC3CR1 were lower in Asian versus White, and CXCR2 was lower in Black and Asian
versus White. In HER2 enriched tumors, CCR3 was higher in Black versus White and CXCR4 lower in
Asian versus White. CCR3, CCR6, and CXCR6 associated with worse patient survival. These ndings can
inform improved treatment strategies to decrease racial disparities in breast cancer burden.
An estimated 43,250 women and 530 men will die from breast cancer in the year 2022 in the United States and
approximately 685,000 women will die worldwide1,2. Racial disparities in mortality due to metastasis remain
signicant among breast cancer patients3–6. Black women are more likely to die from more aggressive and meta-
static breast cancer than other racial groups6,7. Despite having the lowest incidence of breast cancer rates across
racial groups, the rates have increased in the last decade for Asian and Pacic Islander women in the United
States6,8. e concurring summary from a robust emerging body of literature is that racial disparities observed in
cancer progression and outcome are a combination of social, environmental, ancestral and biological factors9–12.
Biological factors refer to the genetic, epigenetic, cellular, and molecular level changes associated with disease
development and progression. Biological factors are apparent when analyzing the cells within the context of
their unique tissue microenvironment.
Chemokines and their receptors are proteins that contribute significantly to the tumor
microenvironment10,13–15. Chemokine receptors are essential to the migratory pattern and positioning of cells
and are required for normal processes, including immune cell tracking and inammation16,17. Chemokine
receptor regulation plays an important role in the tumor microenvironment, being expressed not only by cancer
cells but also by cells that compose the tumor stroma, including immune cells, structural cells (i.e. broblasts
and pericytes), and endothelial cells. ey also play an important role in tumor development and metastatic
outcome18–20. Recent studies identied genes associated with breast cancer progression using weighted gene
co-expression network analysis (WGCNA) using e Cancer Genome Atlas (TCGA) database21–26. In other
studies, Keenan etal. and Huo etal. compared the genomic landscape between Black and White women in the
TCGA breast cancer cohort and demonstrated signicant dierences in tumor biology that can result in survival
disparities12,27. Here, we used breast cancer tissue RNA sequencing data from the TCGA breast cancer cohort to
assess the expression levels for the chemokine receptors as described by the IUPHAR17 in breast tumors across
molecular subtypes from patients of dierent racial categories. ese ndings provide a basis to study connec-
tions between chemokine receptors and breast cancer molecular subtype and their contribution to the breast
tumor microenvironment in patients from dierent racial categories.
OPEN
1Department of Biology, San Francisco State University, San Francisco, CA 94132, USA. 2These authors contributed
equally: Elissa D. Vazquez and Xiangyi Fang. *email: nsave@sfsu.edu
Content courtesy of Springer Nature, terms of use apply. Rights reserved

2
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Results
WGCNA and module‑trait relationships. To identify if chemokine receptor genes in breast cancer are
associated with specic clinical traits, we performed WGCNA to construct modules of co-expressed genes from
breast invasive carcinoma bulk RNA sequencing data from the TCGA from all available patient samples (Sup-
plementary Fig.1A). We identied race as the trait with strongest clinical module-trait relationship (Supplemen-
tary Fig.1B), when comparing available clinical traits from the GDC TCGA data. Next, we dened a variable as
“sample type” to determine if the dierence between normal and tumor samples could explain major dierences
in clinical module-trait relationships. Supplementary Fig.1C shows the modied module-trait relationship from
the combined normal and tumor samples with clinical traits, where the trait that made the most signicant dif-
ference/strongly associated with modules was “sample type.” erefore, including both normal and tumor sam-
ples together in the same WGCNA analysis showed disease status as the major driver of observed dierences. As
the initial analysis indicated that race showed the strongest module-trait relationships (Supplementary Fig.1B),
we next asked whether these relationships could be explained not by race but rather due to molecular subtypes,
as racial groups dier in their breast tumor molecular subtype incidence8,28,29 and included the hormone recep-
tor status30 as part of the clinical trait relationship analysis, which showed that aer sample type, hormone recep-
tor status had the strongest module-trait relationships (Supplementary Fig.1C).
We repeated the WGCNA and module-trait relationship analysis, analyzing only tumor samples by remov-
ing the normal samples available from the breast TCGA cohort. We identied 23 co-expressed modules in the
tumor only WGCNA (Fig.1A,B). Using the identied modules from the WGCNA results, we generated the
module-trait relationship analysis to identify which modules/gene expression changes were associated with the
specic clinical traits of interest. Figure1C shows the module-trait relationship analysis between tumor samples
and clinical traits, where the traits most strongly associated with modules are molecular subtypes and PAM50
status, which are associated to breast tumor progression.
Dierential gene expression analysis. We performed dierential gene expression analysis comparing
Black versus White and Asian versus White because these three groups had the highest number of patient sam-
ples available from the TCGA. Hispanic ethnicity could not be associated with a race group and Native Ameri-
can only had 1 sample, therefore could not be used for group analyses. e dierentially expressed genes (DEGs)
between all the available Black and White patients in the TCGA cohort are in Supplementary-Table1. e DEGs
between all the available Asian and White patients in the TCGA cohort are in Supplementary-Table2). We
searched for known chemokine receptors to identify which were dierentially expressed between racial catego-
ries and labeled them in Table1, which shows ACKR1, CCRL2, and CXCR2 are the only common chemokine
Figure1. Network construction and module detection for Weighted Gene Co-expression Network Analysis
(WGCNA) of Breast Invasive Carcinoma RNA-seq samples from TCGA database. (A) Gene dendrogram of
clustered gene dissimilarity, based on consensus topological overlap, with corresponding gene module colors
of the TCGA breast invasive carcinoma RNA-seq dataset from n = 824 tumor patient samples, each with 20,253
genes analyzed in one block. (B) Number of genes in each module identied from WGCNA. e modules
containing the greatest number of genes are the blue, turquoise, yellow, and brown modules. (C) Module-trait
associations for tumor only samples (excluding normal samples) from the cohort. e corresponding p-value for
the correlation of the module with the clinical parameter is in parenthesis. e grid is color-coded by correlation
according to the color bar.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

3
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
receptor DEGs for both Black versus White and Asian versus White samples. CCR3, and ACKR4/CCRL1 were
also dierentially expressed between Black versus White samples. CCR6, CXCR1, CCRL2, CXCR4 and CXCR6
were dierentially expressed between Asian versus White samples.
e DEGs for tumor only samples between Black versus White patients in the TCGA cohort are in Supple-
mentary Fig.2A and Supplementary-Table3 and for Asian versus White patients in Supplementary Fig.2B and
Supplementary-Table4). e top dierentially expressed genes in tumors between Black and White patients aer
adjusting forcovariate eects, were RPL29P2 and CRYBB2. e top dierentially expressed genes in tumors
between Asian and White patients, aer adjusting for covariates, were CHGB and LEP (supplementary Tables2
and 4). Table1 shows ACKR1/DARC was the only common chemokine receptor DEG between Black versus
White and Asian versus White patient tumor samples. CCR3 and CX3CR1 were dierentially expressed between
Black versus White, and CCR6, CCRL2, and CXCR4 between Asian versus White breast tumors.
Dierences between normal and tumor tissues have been well described31,32. Signicant dierences in gene
expression in normal versus tumor samples within the breast cancer TCGA cohort have been reported before31.
However, when we queried for dierentially expressed genes in the limited available normal adjacent versus
tumor tissues (Supplementary-Tables5 and 6), we identied only the chemokine receptors CXCR2, CXCR6,
and CX3CR1 to be commonly dierentially expressed in normal versus tumor samples for both Black and White
patients. We were unable to do normal versus tumor analysis for the Asian group due to the lack of normal tissue
samples for Asian patients in the breast cancer TCGA.
Gene expression analysis. We plotted the tumor gene expression data to illustrate how the chemokine
receptor genes are expressed across race and tumor molecular subtype. Figure2 shows ACKR1, CCR3, CCR6,
CCRL2, CXCR2, and CX3CR1 expression is signicantly dierent based on racecategory (White, Black or
Asian), using the breast cancer TCGA data for tumor samples.
Since breast tumor molecular subtype (Luminal A or B, HER2 enriched, and triple negative), is a well-estab-
lished determinant of dierences in patient outcome33, we analyzed receptor expression within molecular sub-
types across racial categories. Figure3 shows ACKR1, CCR3, CCR6, ACKR4/CCRL1, CCRL2, CXCR1, CXCR2,
CXCR4, CXCR6, and CX3CR1 expression is signicantly dierent based on breast cancer molecular subtypes.
Figure4 shows that for patients with triple negative breast cancer, CCR3 is signicantly higher in Black versus
White patients while CCRL2 is higher in Asian versus White patients.
Figure5 shows that for patients with luminal A breast cancer, expression of ACKR1 is lower in Asian than
White, CCR3 is higher in Black versus White, and CCR6 and CXC3CR1 are lower in Black versus White.
Table 1. Signicant (p value < 0.05) DEG for Chemokine Receptors for two analysis, either all samples in
TCGA Breast cohort or for the tumor samples only. Bolded receptors names indicate those common in both
Black versus White and Asian versus White comparisons. Full list of DEG in Supplementary tables provided.
log2FoldChange padj
For all (normal and tumor) samples in TCGA breast cohort
Black versus White
ACKR1/DARC − 1.153 0.004
CCR3 0.687 0.016
CCRL1 − 0.641 0.014
CCRL2 − 0.412 0.023
CXCR2 − 0.413 0.007
Asian versus White
ACKR1/DARC − 1.594 0.000
CCR6 − 0.471 0.012
CCRL2 0.599 0.001
CXCR1 − 1.062 0.024
CXCR2 − 0.417 0.011
CXCR4 − 0.448 0.046
CXCR6 0.667 0.026
For tumor samples in TCGA breast cohort
Black versus White
ACKR1/DARC − 0.450 0.048
CCR3 0.661 0.006
CX3CR1 − 0.416 0.046
Asian versus White
ACKR1/DARC − 1.064 0.004
CCR6 − 0.476 0.005
CCRL2 0.617 0.000
CXCR4 − 0.406 0.026
Content courtesy of Springer Nature, terms of use apply. Rights reserved

4
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Figure6 shows that in patients with luminal B breast cancer, expression of CCRL2 is lower in Black versus
White, CXCR1 and CX3CR1 are lower in Asian versus White, and CXCR2 is lower in both Black and Asian
versus White.
Figure2. Dierentially expressed chemokine receptor genes based on race using the breast cancer TCGA data
for tumor samples only. Box and whisker plots represent minimum expression at the bottom whisker, maximum
at the top whisker and median at the middle line with a Log10 axis scale showing geneexpression of the
chemokine receptors. Statistical signicance was determined with global signicance for ANOVA with Kruskal
Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

5
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Figure7 shows that for patients with HER2 enriched tumors there is higher CCR3 in Black versus White, and
lower CXCR4 in Asian versus White breast cancer patients.
Figure3. Dierentially expressed chemokine receptorgene expression based on molecular subtype using the
breast cancer TCGA data for tumor samples only. Box and whisker plots represent minimum expression at the
bottom whisker, maximum at the top whisker and median at the middle line with a Log10 axis scale showing
gene expression of the chemokine receptors. Statistical signicance was determined with global signicance for
ANOVA with Kruskal Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

6
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Survival analysis. Ten chemokine receptors (CXCR1, CXCR2, CXCR4, CXCR6, CCR3, CCR6, ACKR4/
CCRL1, CCLR2, CX3CR1, and ACKR1/DARC) were identied as dierentially expressed from the DEG analy-
Figure4. Dierentially expressed chemokine receptorgeneexpression based on race within triple negative
subtype using the breast cancer TCGA data for tumor samples only. Box and whisker plots represent minimum
expression at the bottom whisker, maximum at the top whisker and median at the middle line with a Log10 axis
scale showing gene expression of the chemokine receptors. Statistical signicance was determined with global
signicance for ANOVA with Kruskal Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

7
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
sis between Black, White and Asian patient normal and tumor samples. We used overall survival (OS) for breast
cancer tumors in TCGA and in the GEO repository to determine if receptor expression correlated with survival.
Figure8 shows Kaplan–Meier plots for chemokine receptors ACKR1, CXCR6, CCR6, and CX3CR1, which were
the only receptors that signicantly correlated with the overall survival of patients with breast cancer.
We also assessed dierences in chemokine receptors expression by race.With a p value of 0.05, higher expres-
sion of CXCR6 correlated with better overall survival for Black patients only. No signicant receptor correlations
with survival were observed for White or Asian breast cancer patients in the TCGA breast cancer cohort. We also
looked at the Relapse-free survival rate from patients that survived with no symptoms of cancer, and found no
signicant dierences in survival rate based on these receptor genes. Next, we assessed dierences in chemokine
receptors expression by molecular subtypes.Figure9 shows the chemokine receptors that signicantly correlated
with breast cancer overall survival in basal tumors. Figure10 shows the receptors that signicantly correlated
with overall survival rate for breast cancer tumor samples of HER2 enriched, luminal A, and luminal B subtypes.
Chemokine receptor specic ndings. ACKR1/DARC . Atypical Chemokine Receptor 1 or Duy An-
tigen Receptor for Chemokines (ACKR1/DARC) is an internalizing receptor or chemokine-scavenging/decoy
receptor whose ligand binding results in chemokine sequestration, degradation, or transcytosis because of its
inability to signal via classic G-protein-mediated pathways. ACKR1 can regulate chemokine bioavailability and
leukocyte recruitment when expressed on endothelial cells. It is also expressed by erythrocytes where it serves
as a blood reservoir of cognate chemokines but also as a chemokine sink, buering potential surges in plasma
chemokine levels. ACKR1 can inhibit breast cancer growth and progression by sequestration of angiogenic
chemokines and subsequent inhibition of tumor neovascularity34. DEG analysis showed that ACKR1 is signi-
cantly downregulated in Black versus White patients (Table1), consistent with a recent study showing ACKR1
tumor expression in breast cancer is lower in African American patient’s breast cancer tumors than in European
American patient tumors22. RNA count expression plotting for the TCGA breast cancer cohort only conrmed a
signicant decrease in ACKR1 expression in Asian compared to White and to Black patients (Fig.2). In addition,
there was no signicant dierence within the triple negative molecular subtype across the dierent racial catego-
ries (Fig.4). All our dierent types of analyses did however, support that ACKR1 is signicantly downregulated
in Asian versus White and specically in luminal A patient samples. ACKR1 expression varies signicantly
based on molecular subtype, with higher expression in luminal A than luminal B, HER2 enriched and triple
negative breast tumors. ACKR1 associated with better overallsurvival (Fig.8A).
CCR3. e CC Motif Chemokine Receptor 3, CCR3, is expressed in eosinophils, basophils, T, NKT, and airway
epithelial cells. It may contribute to accumulation and activation of eosinophils and other inammatory cells35.
CCR3 has been associated with improved relapse-free survival in breast cancer with high expression of CCR3
in luminal-like rather than triple negative or HER2 enriched tumors36. Our results for the breast TCGA cohort
showthat CCR3 is higher in the triple negative subtype than in luminal A and B (Fig.3). In addition, we found
CCR3 was upregulated in luminal A, HER2 enriched, and triple negative tumors of only Black versus White
patients.
CCR6. e CC Motif Chemokine Receptor 6, CCR6, is expressed by immature dendritic cells and memory T
cells. e ligand of this receptor is macrophage inammatory protein 3 alpha (MIP-3 alpha)35. CCR6 is highly
expressed in pro-tumorigenic macrophages within the mammary gland microenvironment, promoting breast
cancer tumors37. CCR6 was dierentially expressed, being down regulated in Asian versus White (Table1) and in
Black versus White patients (by expression count data) (Fig.2). High expression of CCR6 signicantly associated
with worse overall breast cancer patient survival (Fig.8D), but correlated with better survival in basal and HER
enriched patients (Fig.9C, 10C). We found CCR6 is higher in luminal A and in triple negative than in luminal B
breast cancers. We also found higher CCR6 expression in Black versus White patients with luminal A molecular
subtype breast tumors.
CCRL1. e atypical chemokine receptor 4/CC‐chemokine receptor‐like 1, ACKR4/CCRL1, or CCX-CKR,
is expressed by cancer cells, thymic epithelial cells, bronchial cells, and keratinocytes. ACKR4/CCRL1 down
regulation correlates with worse outcome in breast cancer. CCRL1 is a negative regulator of growth and metas-
tasis in breast cancer by sequestering chemokines and inhibiting intratumoral neovascularity38,39. DEG analysis
shows CCRL1 was downregulated across the Black versus White patient samples in the TCGA breast cohort
(Table1). Figure3 shows CCRL1 is signicantly lower in luminal B compared to triple negative breast cancer
samples.High ACKR4/CCRL1expression correlated with worse survival in patients with triple negative or basal
tumors (Fig.9D).
CCRL2. e CC Motif Chemokine Receptor Like 2, CCRL2, is expressed at high levels in endothelial cells, acti-
vated macrophages, and myeloid derived leukocytes35,40,41. CCRL2 is a receptor for CCL19 and the chemokine/
adipokine chemerin (RARRES2). CCRL2 modulates chemokine-triggered immune responses by capturing and
internalizing CCL19 or presenting RARRES2 ligand to the receptor CMKLR1. CCRL2 on activated endothelial
cells acts in concert with CMKLR1 to coordinate chemerin-dependent leukocyte adhesion invitro and recruit-
ment invivo41. CCRL2 expressing adipose tissue progenitor cells cooperate to induce epithelial-to-mesenchymal
transition (EMT) gene expression in luminal breast cancer cells to enhance tumor progression and metastatic
dissemination42. is may be important in the obesity-driven inammatory microenvironment known to cor-
relate with tumor development and progression in post-menopausal women42. Our dierentially expressed gene
analysis shows that CCRL2 is downregulated in Black versus White patients and is upregulated in Asian versus
Content courtesy of Springer Nature, terms of use apply. Rights reserved

8
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
White samples (Table1). Figure2 RNA count data indicates that CCRL2 is signicantly upregulated in Asian
versus White and Black patients. We found that CCRL2 is signicantly higher in HER2 enriched than in luminal
Figure5. Dierentially expressed chemokine receptorgene expression based on race within Luminal A
subtype using the breast cancer TCGA data for tumor samples only. Box and whisker plots represent minimum
expression at the bottom whisker, maximum at the top whisker and median at the middle line with a Log10 axis
scale showing geneexpression of the chemokine receptors. Statistical signicance was determined with global
signicance for ANOVA with Kruskal Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

9
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Figure6. Dierentially expressed chemokine receptor gene expression based on race within Luminal B
subtype using the breast cancer TCGA data for tumor samples only. Box and whisker plots represent minimum
expression at the bottom whisker, maximum at the top whisker and median at the middle line with a Log10 axis
scale showing gene expression of the chemokine receptors. Statistical signicance was determined with global
signicance for ANOVA with Kruskal Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

10
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
Figure7. Dierentially expressed chemokine receptorgene expression based on race within HER2 subtype
using the breast cancer TCGA data for tumor samples only. Box and whisker plots represent minimum
expression at the bottom whisker, maximum at the top whisker and median at the middle line with a Log10 axis
scale showing geneexpression of the chemokine receptors. Statistical signicance was determined with global
signicance for ANOVA with Kruskal Wallis.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

11
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
A, B, and triple negative breast tumors. CCRL2 was higher in Asian than White patients with triple negative
breast cancer. CCRL2 was also higher in White than in Black patients with luminal B breast tumors.High
CCRL2 correlated with better survival in triple negative or basal tumors (Fig.9E).
CXCR1 and CXCR2. Chemokine receptors CXCR1 and CXCR2 are G-protein-coupled receptors whose bio-
logical eects are mediated by the inammatory chemokine CXCL843. While not abundant in cancer cells,
CXCR1 was identied on breast cancer stem-like cells (CSCs) where it plays a major role in regulating the breast
CSCs and surrounding cancer cell survival via Fas-ligand mediated pathways44,45. Similarly, CXCR2 expression
is not consistently found to be highly expressed in cancer cells within breast tumors, but rather this receptor is
most highly expressed in the stromal cells of breast tumors, particularly neutrophils. High levels of CXCR2 are
also associated with higher inltration of T and B lymphocytes in breast tumors46. We showed that both CXCR1
and CXCR2 are downregulated in Black versus White patients based on RNA counts (Fig.2). DEG analysis
indicates downregulation of both CXCR1 and CXCR2 in the Asian versus White breast cohort (Table1). Only
CXCR2 was downregulated for both Black versus White and Asian versus White patients in the TCGA cohort
(Table1). However, DEG results for tumor only samples did not indicate signicant dierences for these two
receptors based on race category(Table1). Molecular subtype analysis revealed CXCR1 is signicantly lower in
triple negative than in luminal A, B and HER2 enriched breast cancers. CXCR2 is signicantly lower in luminal
B and HER2 enriched than in luminal A tumors (Fig.3). When analyzing race categorywithin the luminal B
subtype, we found that CXCR1 was lower in Asian versus White breast tumors. Similarly, within the luminal
B subtype, CXCR2 expression is lower in Black and Asian compared to White patient breast tumors (Fig.6).
CXCR2expression correlated with better survival in HER2 enriched tumors (Fig.10B).
CXCR4. CXCR4 is a G-protein-coupled receptor highly expressed in breast cancer cells. CXCR4 plays a role
in the growth and metastasis of breast cancer via its ligand CXCL12/SDF-1, which activates mTOR to promote
the epithelial–mesenchymal transition (EMT), and is therefore important in metastasis47. CXCR4 inhibitors
can impair tumor growth and metastatic dissemination in HER2 enriched breast cancer cells but do not reduce
Figure8. Overall survival rate of breast cancer patients tumor samples. Kaplan–Meier curves of overall survival
based on KMplotter mRNA gene chip expression of chemokine receptors in breast cancer patients. Red line
represents higher expression and black line represents lower expression. Patients with a high expression of the
genes for (A) ACKR1, (B) CXCR6, and (C) CX3CR1 have a higher survival rate than those who have a low
expression of the gene. (D) CCR6 low expression had better survival only at later time points.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

12
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
tumor growth, and can increase metastatic spread of triple negative breast cancer. CXCR4 inhibitors also reduce
myobroblast content in all breast cancer subtypes, but only decrease angiogenesis in HER2 enriched breast
cancer. is is surprising considering that we would expect the more aggressive triple negative breast cancer
would harbor a more angiogenic and metastatic microenvironment with increased CXCR4 expression. DEG
analysis showed CXCR4 was dierentially downregulated only for the Asian versus White breast cohort and
samples (Table1). Molecular subtype analysis showed CXCR4 expression was higher in triple negative samples
than in luminal A, B, and HER2 enriched. CXCR4 was also higher in HER2 enriched than in luminal A and B
(Fig.3). We only found a signicant dierence for Asian patients with HER2 enriched, who have higher CXCR4
expression than White patients with HER2 enriched tumors (Fig.7). CXCR4expression correlated with better
overall survival of patients with basal tumors (Fig.9B).
CXCR6. CXCR6 is a G-protein-coupled receptor enriched in inamed tissue lymphoid cells, and also
expressed in some epithelial and nonepithelial cancer cells. CXCR6 mediates tumor promoting inammation
via its ligand CXCL16 by inducing macrophage polarization toward a pro-tumoral phenotype in solid tumors18.
Reduction of CXCR6 expression in breast cancer mouse models decreases metastasis of those tumors48. DEG
analysis indicated CXCR6 was upregulated in the Asian versus White TCGA cohort (Table1). Molecular sub-
type analysis showed CXCR6 is signicantly higher in triple negative and in HER2 enriched than in luminal A
and B breast tumors (Fig.3). Higher expression of CXCR6 correlated with slightly worse overall patient survival
(Fig.8B) while CXCR6 correlated with slightly better survival for Black patients (not shown)and better survival
for patients with basal tumors (Fig.9C). is suggests that CXCR6 is potentially one of the chemokine receptors
mediating pro-inammatory microenvironment in Asian and Black breast tumors.
CX3CR1. e chemokine receptor CX3CR1 is over-expressed in both primary and metastatic breast tumors.
CX3CR1 determines the arrest and initial lodging of breast cancer cells to transmigrate through endothelial cells
and extravasate to lodge on to the skeleton in response to its ligand, Fractalkine (FKN/CX3CL1)49,50. Target-
ing CX3CR1 delays the egress of circulating tumor cells from the blood circulation, induces cancer cell apop-
tosis, and reduces metastasis51. DEG analysis shows CX3CR1 is signicantly downregulated in Black versus
Figure9. Overall survival rate of breast cancer patients tumor samples of basal subtype. Kaplan–Meier curves
of overall survival based on KMplotter mRNA gene chip expression of chemokine receptors in breast cancer
patients. Red line represents higher expression and black line represents lower expression. Patients with a high
expression of the genes for (A) CCR6, (B) CXCR4, and (C) CXCR6, (D) ACKR4 and (E) CCRL2.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

13
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
White patient tumors (Table1). CX3CR1 was lower in Asian patient tumor samples based on RNA gene counts
(Fig.2). Molecular subtype analysis showed CX3CR1 is signicantly lower in luminal B, HER2 enriched, and tri-
ple negative breast tumors than in luminal A (Fig.3). CX3CR1 expression associated withbetter overall survival
(Fig.8C) and worse patient survival for luminal A patients(Fig.10D).
Discussion
We veried that expression dierences observed across racial category were not confounded by molecular sub-
type prevalence in the dierent racial groups, as it is well established that minorities, such as Black women
present more oen with triple negative breast cancer and Asian and Latinas with HER2 enriched breast cancers
than White women8,28,29. e expression of distinct chemokine receptors in the bulk breast cancer tissue could
account for a dierential recruitment of inammatory factors such as lymphocyte inltrate, stromal cells, and
cytokine concentration gradients in the tumor microenvironment. e dierential expression of these chemokine
receptors across racial group category with the same molecular subtype may explain the variability in tumor
microenvironment across racial groups and their potential response to future immune therapies. In addition,
our analysis highlights the importance of stratifying patient racial dierences within tumor molecular subtypes
to better understand tumor microenvironment features. Dierential gene expression showed that while there are
many genes that are dierentially expressed, only a few chemokine receptor genes (Table1) were signicantly
dierent when comparing samples from White versus Black or Asian breast cancer patients. Our results warrant
further studies of gene expression assessment in breast tumors across diverse patient samples to harness targeting
these receptors and their distribution within cell subsets as potential therapeutics to reduce cancer disparities. A
major limitation of this study is that we did not access DNA sequencing data to correlate ancestry to the patient
samples’ race category, and large variations could result from incorrect or mixed self-reported racial group clas-
sication. Other limitations include the limited number of tumor samples and metastatic samples within this
Figure10. Chemokine receptors identied as signicantly correlated with overall survival rate for breast cancer
tumor samples of HER2 and luminal subtypes. Kaplan–Meier curves of overall survival based on KMplotter
mRNA gene chip expression of chemokine receptors in breast cancer patients. Red line represents higher
expression and black line represents lower expression. e receptors in (A) ACKR1, (B) CXCR2, and (C)
CCR6, were signicant for HER2 enrichedsubtype (D) CX3CR1 was correlated with the luminal A subtype. (E)
ACKR1 and (F) CCR3 expression correlated with luminal B subtype.
Content courtesy of Springer Nature, terms of use apply. Rights reserved

14
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
breast cancer cohort, as we could not associate metastatic status to expression level of our genes of interest. In
addition, low sample size for minority racial groups in the TCGA makes the survival analysis by race stratication
unreliable. While the TCGA is not a population-based study sample, it provides access to hundreds of patient
samples with tumor genomics and clinical features making it an important dataset to form the basis of improved
tumor targeted therapies, including normal tissue adjacent to the tumor samples31.
Despite these limitations, the chemokine receptor genes we described have been consistently identied as
being involved in metastatic processes and the epithelial to mesenchymal transition, as mentioned above. In
addition, here we used bulk RNA-seq ndings to hypothesize about the tumor microenvironment’s chemokine
receptor composition of tumor samples across race, when resolution at the single cell level will be necessary to
allow us to elucidate the role of these receptor’s cellular distribution across dierent patient’s microenvironments.
Methods
Data collection. We downloaded Breast Invasive Carcinoma RNA-seq count and clinical data from e
Cancer Genome Atlas (TCGA) database from 623 White, 58 Black, 52 Asian, 27 Latina and 1 American and
Indian/Alaska Native patients, that had a primary tumor sample via the Genomic Data Commons (GDC) Portal
on or before July 2020. Level 3 RNA-Seq data was used for this study, which is de-identied and publicly avail-
able through TCGA. e study was carried out in accordance with relevant guidelines and regulations.
Weighted gene co‑expression network analysis (WGCNA). WGCNA was applied to the normal-
ized, ltered, nal expression matrix to identify gene groups (modules). Modules are clusters of highly intercon-
nected genes. Only the top 33% of genes with appreciable expression levels (FPKM > 1) in more than half of the
breast cancer patients were subjected to analysis. We proceeded to cut out 14 sample outliers, which resulted in
810 tumor samples for nal analysis. We used the WGCNA R package52,53 with so-thresholding power 6, to
build a weighted gene co-expression network that contained 20,564 nodes (genes) analyzed in one block.
Module‑trait relationships. e identied modules were correlated with the available clinical traits avail-
able in the TCGA, and also the tumor molecular subtypes and the PAM50 RNA status for the available samples
from the TCGA study30. Signicant modules were plotted against clinical traits. Each row corresponds to a
module eigengene, and each column corresponds to a clinical parameter, i.e. sample type, ethnicity, gender,
molecular subtype, and PAM50 RNA. e module eigengene is dened as the rst principal component of a
given module and considered a representative of the gene expression proles in a module. Each grid contains the
correlation value, calculated based on eigengene expression and clinical traits. Listed in the heatmap are bicor
correlation rho values and p-values for the correlation (in parentheses), dening relationships between overall
weighted expression proles of modules across samples and clinical traits. e module colors are shown on the
le side of each row. e modules and number of genes each module contains are identied by a color, except
for the “grey” module, which is reserved for unassigned genes with transcripts with lower correlations across
samples not considered strongly co-expressed52,53.
Dierential expression analysis. We used DESeq2, an R package that utilizes negative binomial general-
ized linear models (GLM) to test for dierential expression54,55. We analyzed the breast cancer gene expression
data from the TCGA to identify genes dierentially expressed in Black and Asian patients compared to White
patients. We used sample expression and the associated patient metadata. e GLM used to model the eect of
cancer diagnosis, stage, race, as well as other covariates such as age, on gene expression levels was: radiation_
therapy + ethnicity + race. Genes dierentially expressed due to race were identied by dropping the race factor
from the model. Genes with false discovery rate (FDR)-corrected p value < 0.05 were considered signicant.
Common genes for normal adjacent versus tumor gene lists were identied using Venny56.
Gene expression based on clinical parameters. We used RNA-seq HTSeq-counts and clinical sample
data from the TCGA breast cancer patients, accessed through the GDC portal. Box and whisker plots represent
the interquartile range and median line with a Log10 axis scale showing RNA-seq HTSeq-counts with global
signicance for ANOVA with Kruskal Wallis.
Survival analysis. For analysis of TCGA breast cancer patients from dierent racial categories, we used
RNA-seq HTSeq-counts and clinical sample data from the TCGA, accessed through the GDC portal. Age and
race adjusted Cox KM plots were made with the survt.coxph function in R. Expression data was divided at the
median and plotted against survival time in days. For Figs.8, 9 and 10, we used KMPlotter57 (http:// kmplot. com/
analy sis/) results with logrank p-value < 0.05.
Received: 23 December 2021; Accepted: 13 June 2022
References
1. Siegel, R. L., Miller, K. D., Fuchs, H. E. & Jemal, A. Cancer statistics, 2022. CA: Cancer J. Clin. 72, 7–33 (2022).
2. Sung, H. et al. Global cancer statistics 2020: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185
countries. CA Cancer J. Clin. 71, 209–249 (2021).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

15
Vol.:(0123456789)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
3. Akinyemiju, T., Sakhuja, S., Waterbor, J., Pisu, M. & Altekruse, S. F. R acial/ethnic disparities in de novo metastases sites and survival
outcomes for patients with primary breast, colorectal, and prostate cancer. Cancer Med. 7, 1183–1193 (2018).
4. Ren, J.-X., Gong, Y., Ling, H., Hu, X. & Shao, Z.-M. Racial/ethnic dierences in the outcomes of patients with metastatic breast
cancer: Contributions of demographic, socioeconomic, tumor and metastatic characteristics. Breast Cancer Res. Treat. https:// doi.
org/ 10. 1007/ s10549- 018- 4956-y (2018).
5. Vaz-Luis, I. et al. Racial dierences in outcomes for patients with metastatic breast cancer by disease subtype. Breast Cancer Res
Treat 151, 697–707 (2015).
6. DeSantis, C. E. et al. Breast cancer statistics, 2019. CA: Cancer J. Clin. 69, 438–451 (2019).
7. Yedjou, C. G. et al. Health and racial disparity in breast cancer. Adv. Exp. Med. Biol. 1152, 31–49 (2019).
8. Gomez, S. L. et al. Breast cancer in Asian Americans in California, 1988–2013: Increasing incidence trends and recent data on
breast cancer subtypes. Breast Cancer Res. Treat. 164, 139–147 (2017).
9. Zavala, V. A. et al. Cancer health disparities in racial/ethnic minorities in the United States. Br. J. Cancer 124, 315–332 (2021).
10. Allinen, M. et al. Molecular characterization of the tumor microenvironment in breast cancer. Cancer Cell 6, 17–32 (2004).
11. Davis, M. B. & Newman, L. A. Breast cancer disparities: How can we leverage genomics to improve outcomes?. Surg. Oncol. Clin.
N. Am. 27, 217–234 (2018).
12. Huo, D. et al. Comparison of breast cancer molecular features and survival by African and European ancestry in the cancer genome
atlas. JAMA Oncol. 3, 1654–1662 (2017).
13. Orimo, A. et al. Stromal broblasts present in invasive human breast carcinomas promote tumor growth and angiogenesis through
elevated SDF-1/CXCL12 secretion. Cell 121, 335–348 (2005).
14. Burger, J. A. & Kipps, T. J. CXCR4: A key receptor in the crosstalk between tumor cells and their microenvironment. Blood 107,
1761–1767 (2006).
15. Charan, M. et al. Molecular and cellular factors associated with racial disparity in breast cancer. Int. J. Mol. Sci. 21, 5936 (2020).
16. Grith, J. W., Sokol, C. L. & Luster, A. D. Chemokines and chemokine receptors: Positioning cells for host defense and immunity.
Annu. Rev. Immunol. 32(32), 659–702 (2014).
17. Bachelerie, F. et al. Chemokine receptors (version 2020.5) in the IUPHAR/bps guide to pharmacology database. IUPHAR/BPS
Guide to Pharmacology CITE 2020, (2020).
18. Poeta, V. M., Massara, M., Capucetti, A. & Bonecchi, R. Chemokines and chemokine receptors: New targets for cancer immuno-
therapy. Front. Immunol. 10, 379 (2019).
19. Morein, D., Erlichman, N. & Ben-Baruch, A. Beyond cell motility: e expanding roles of chemokines and their receptors in
malignancy. Front. Immunol. 11, 952 (2020).
20. Sarvaiya, P. J., Guo, D., Ulasov, I., Gabikian, P. & Lesniak, M. S. Chemokines in tumor progression and metastasis. Oncotarget 4,
2171–2185 (2013).
21. omas, J. K., Mir, H., Kapur, N., Bae, S. & Singh, S. CC chemokines are dierentially expressed in breast cancer and are associated
with disparity in overall survival. Sci. Rep. 9, 1–12 (2019).
22. Jenkins, B. D. et al. Atypical chemokine receptor 1 (DARC/ACKR1) in breast tumors is associated with survival, circulating
chemokines, tumor-inltrating immune cells, and african ancestry. Cancer Epidemiol. Biomark. Prev. 28, 690–700 (2019).
23. Quan, L. et al. Cytokine and cytokine receptor genes of the adaptive immune response are dierentially associated with breast
cancer risk in American women of African and European ancestry. Int. J. Cancer 134, 1408–1421 (2014).
24. Ohandjo, A. Q. et al. Transcriptome network analysis identies CXCL13-CXCR5 signaling modules in the prostate tumor immune
microenvironment. Sci. Rep. 9, 1–13 (2019).
25. Tang, J. et al. Prognostic genes of breast cancer identied by gene co-expression network analysis. Front. Oncol. 8, 374 (2018).
26. Lan, L., Xu, B., Chen, Q., Jiang, J. & Shen, Y. Weighted correlation network analysis of triple-negative breast cancer progression:
Identifying specic modules and hub genes based on the GEO and TCGA database. Oncol. Lett. 18, 1207–1217 (2019).
27. Keenan, T. et al. Comparison of the genomic landscape between primary breast cancer in African American versus white women
and the association of racial dierences with tumor recurrence. J. Clin. Oncol. 33, 3621–3627 (2015).
28. Carey, L. A. et al. Race, breast cancer subtypes, and survival in the carolina breast cancer study. JAMA 295, 2492–2502 (2006).
29. Serrano-Gomez, S. J., Fejerman, L. & Zabaleta, J. Breast cancer in Latinas: A focus on intrinsic subtypes distribution. Cancer
Epidemiol. Biomark. Prev.: Publ. Am. Assoc. Cancer Res. Cosponsored Am. Soc. Prev. Oncol. 27, 3–10 (2018).
30. Koboldt, D. C. et al. Comprehensive molecular portraits of human breast tumours. Nature https:// doi. org/ 10. 1038/ natur e11412
(2012).
31. Aran, D. et al. Comprehensive analysis of normal adjacent to tumor transcriptomes. Nat. Commun. 8, 1077 (2017).
32. Hanahan, D. & Weinberg, R. A. Hallmarks of cancer: e next generation. Cell 144, 646–674 (2011).
33. Sorlie, T. et al. Gene expression patterns of breast carcinomas distinguish tumor subclasses with clinical implications. Proc. Natl.
Acad. Sci. U.S.A. 98, 10869–10874 (2001).
34. Wang, J. et al. Enhanced expression of Duy antigen receptor for chemokines by breast cancer cells attenuates growth and metastasis
potential. Oncogene 25, 7201–7211 (2006).
35. e GeneCards Suite: From gene data mining to disease genome sequence analyses: Stelzer—2016—Current Protocols in Bioin-
formatics—Wiley Online Library. (2020) https:// doi. org/ 10. 1002/ cpbi.5.
36. Gong, D. H., Fan, L., Chen, H. Y., Ding, K. F. & Yu, K. D. Intratumoral expression of CCR3 in breast cancer is associated with
improved relapse-free survival in luminal-like disease. Oncotarget 7, 28570–28578 (2016).
37. Boyle, S. T., Faulkner, J. W., McColl, S. R. & Kochetkova, M. e chemokine receptor CCR6 facilitates the onset of mammary
neoplasia in the MMTV-PyMT mouse model via recruitment of tumor-promoting macrophages. Mol. Cancer 14, 1–14 (2015).
38. Massara, M., Bonavita, O., Mantovani, A., Locati, M. & Bonecchi, R. Atypical chemokine receptors in cancer: Friends or foes?. J.
Leukoc. Biol. 99, 927–933 (2016).
39. Feng, L. Y., Ou, Z. L., Wu, F. Y., Shen, Z. Z. & Shao, Z. M. Involvement of a novel chemokine decoy receptor CCX-CKR in breast
cancer growth, metastasis and patient survival. Clin. Cancer Res. 15, 2962–2970 (2009).
40. Zabel, B. A. et al. Mast cell-expressed orphan receptor CCRL2 binds chemerin and is required for optimal induction of IgE-
mediated passive cutaneous anaphylaxis. J. Exp. Med. 205, 2207–2220 (2008).
41. Monnier, J. et al. Expression, regulation, and function of atypical chemerin receptor CCRL2 on endothelial cells. J. Immunol. 189,
956–967 (2012).
42. Orecchioni, S. et al. Complementary populations of human adipose CD34(+) progenitor cells promote growth, angiogenesis, and
metastasis of breast cancer. Can. Res. 73, 5880–5891 (2013).
43. Runi, P. A. e CXCL8-CXCR1/2 axis as a therapeutic target in breast cancer stem-like cells. Front. Oncol. 9, 40 (2019).
44. Goldstein, L. J. et al. A window-of-opportunity trial of the CXCR1/2 inhibitor reparixin in operable HER-2-negative breast cancer.
Breast Cancer Res. 22, 1–9 (2020).
45. Ginestier, C. et al. CXCR1 blockade selectively targets human breast cancer stem cells invitro and in xenogras. J. Clin. Investig.
120, 485–497 (2010).
46. Boissiere-Michot, F. et al. Prognostic value of CXCR2 in breast cancer. Cancers 12, 2076 (2020).
47. Yang, F. et al. Inhibition of dipeptidyl peptidase-4 accelerates epithelial-mesenchymal transition and breast cancer metastasis via
the CXCL12/CXCR4/mTOR axis. Can. Res. 79, 735–746 (2019).
Content courtesy of Springer Nature, terms of use apply. Rights reserved

16
Vol:.(1234567890)
Scientic Reports | (2022) 12:10825 | https://doi.org/10.1038/s41598-022-14734-5
www.nature.com/scientificreports/
48. Xiao, G. et al. CXCL16/CXCR6 chemokine signaling mediates breast cancer progression by pERK1/2-dependent mechanisms.
Oncotarget 6, 14165–14178 (2015).
49. Jamieson-Gladney, W. L., Zhang, Y., Fong, A. M., Meucci, O. & Fatatis, A. e chemokine receptor CX3CR1 is directly involved
in the arrest of breast cancer cells to the skeleton. Breast Cancer Res. 13, R91 (2011).
50. Shen, F. et al. Novel small-molecule CX3CR1 antagonist impairs metastatic seeding and colonization of breast cancer cells. Mol.
Cancer Res. 14, 518–527 (2016).
51. Qian, C. et al. Impeding circulating tumor cell reseeding decelerates metastatic progression and potentiates chemotherapy. Mol.
Cancer Res. 16, 1844–1854 (2018).
52. Langfelder, P. & Horvath, S. WGCNA: An R package for weighted correlation network analysis. Bmc Bioinform. 9, 1–13 (2008).
53. Yang, Y. et al. Gene co-expression network analysis reveals common system-level properties of prognostic genes across cancer
types. Nat. Commun. 5, 1–9 (2014).
54. Stewart, P. A., Luks, J., Roycik, M. D., Sang, Q. X. A. & Zhang, J. F. Dierentially expressed transcripts and dysregulated signaling
pathways and networks in African American breast cancer. PLoS ONE 8, e82460 (2013).
55. Love, M. I., Huber, W. & Anders, S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome
Biol. 15, 550 (2014).
56. Oliveros, J. C. Venny. An interactive tool for comparing lists with Venn’s diagrams. https:// bioin f ogp. cnb. csic. es/ tools/ venny/ index.
html (2007).
57. Győry, B. Survival analysis across the entire transcriptome identies biomarkers with the highest prognostic power in breast
cancer. Comput. Struct. Biotechnol. J. 19, 4101–4109 (2021).
Acknowledgements
e authors thank Dr. Laura Fejerman for helpful discussions and advice on issues examined in this paper. E.D.V.
was supported by a GEN-PINC scholarship. X.F. was supported by a CSUPERB Doris A. Howell Award. C.V.
and N.L. were supported by NIH-Bridges to Baccalaureate program 3R25GM050078-19S1. is study was sup-
ported by the National Institute of General Medical Sciences of the National Institutes of Health under Award
Number SC2GM135135.
Author contributions
E.D.V., X.F., L.A. L., M.H., C.V., and N.L. conducted experiments and analyzed results. N.S. planned the experi-
ments, analyzed results, and wrote the manuscript. All authors reviewed the manuscript.
Competing interests
e authors declare no competing interests.
Additional information
Supplementary Information e online version contains supplementary material available at https:// doi. org/
10. 1038/ s41598- 022- 14734-5.
Correspondence and requests for materials should be addressed to N.S.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access is article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
© e Author(s) 2022
Content courtesy of Springer Nature, terms of use apply. Rights reserved

1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com