Functions of the C/EBPβ isoforms in breast cancer
Sterken, Britt
DOI:
10.33612/diss.172465560
IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite from it. Please check the document version below.
Document Version
Publisher's PDF, also known as Version of record
Publication date: 2021
Link to publication in University of Groningen/UMCG research database
Citation for published version (APA):
Sterken, B. (2021). Functions of the C/EBPβ isoforms in breast cancer. University of Groningen. https://doi.org/10.33612/diss.172465560
Copyright
Other than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).
Take-down policy
If you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediately and investigate your claim.
Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons the number of authors shown on this cover page is limited to 10 maximum.
Chapter IV
Conditional CAG-promoter driven expression of
C/EBPβ-LIP or C/EBPβ-LAP from the Col1a1-locus is epigenetically
silenced in mice, and Trp53-deficient breast cancer latency
and growth is not correlated to LIP/LAP isoform ratio
Britt A. Sterken1, Peter Bouwman2, Anne Paulien Drenth2, Eline van der Burg2, Colin Pritchard3, Ivo Huijbers3, Jos Jonkers2, Cornelis F. Calkhoven1
1 European Research Institute for the Biology of Ageing (ERIBA), University Medical Center Groningen, University of Groningen, 9700 AD Groningen, The Netherlands. 2 Division of Molecular Pathology, The Netherlands Cancer Institute, Plesmanlaan 121, 1066CX Amsterdam, The Netherlands
3 Mouse Clinic for Cancer and Aging – Transgenic facility, The Netherlands Cancer Institute, Plesmanlaan 121, Amsterdam, The Netherlands
Abstract
Breast cancer is the most commonly diagnosed cancer type in women, with its survival outcome being strongly dependent on the stage the breast cancer at the time of diagnosis. Among the different molecular subtypes of breast cancer, triple negative breast cancer (TNBC) is the most aggressive subtype, representing 15-20% of all breast cancers. Previous studies showed that expression of the truncated isoform of C/EBPβ called C/EBPβ-LIP is inversely correlated with the expression of estrogen (ER) and progesterone (PR) receptors in human breast cancer samples. The CEBPB-mRNA is translated into three distinct protein isoforms named C/EBPβ-LAP*, C/EBPβ-LAP and C/EBPβ-LIP that are functionally different, as the shorter isoform LIP lacks the N-terminal transactivation domains and acts as a competitive inhibitor of LAP* and LAP. Systemic ablation of LIP in mice results in reduced overall tumourigenesis, whereas overexpression of LIP systemically induces tumourigenesis and selective LIP overexpression in the mammary epithelium promotes hyperproliferation. However, so far no genetically engineered LIP and LAP breast cancer mouse models have been generated to study their functions in breast cancer development and metastasis. Here, we introduced exogenous expression of LIP and LAP in an existing p53-deficient mouse model for breast cancer to analyse the functions of LIP and LAP in breast cancer development and metastasis. In
vitro Cre-mediated recombination in Mouse Mammary Epithelial Cells
(MMECs) isolated from mice from the LIP and LAP cohorts resulted in a strongly upregulated expression of LIP and LAP. However, the LIP and LAP exogenous expression was reduced in the MMECs with increased passage numbers, and in the large majority of tumours the exogenous expression of LIP and LAP were reduced or lost. Analysis of mammary tumours revealed a consistent endogenous expression of LIP and LAP in all mammary tumours at variable LIP/LAP ratios, which however, did not correlate to tumour onset and overall survival. Given the high endogenous expression of LIP and LAP in p53-deficient mouse mammary tumours, isoform-specific loss of function in breast cancer models might be more
suitable to model the role of LIP and LAP in breast cancer development in future studies.
Introduction
Breast cancer is the most commonly diagnosed cancer type in women1, with its survival outcome being strongly dependent on the stage of the breast cancer at the time of diagnosis. Breast cancers are classified into different molecular subtypes that are characterised by distinct gene expression profiles 2–4. Among these different subtypes of breast cancer, basal-like breast cancer is a particularly aggressive breast cancer subtype and represents 15-20% of all breast cancers 5. Most basal-like tumours lack the expression of hormone receptors (estrogen (ER) and progesterone (PR) receptors) and human epidermal growth factor receptor 2 (HER2), which are therefore referred to as Triple-Negative Breast Cancer (TNBC). Due to the lack of expression of these druggable receptors, this subclass is particularly difficult to treat due to the lack of targeted therapies 6. Previous studies show that high expression of the N-terminally truncated protein isoform of the CCAAT/enhancer binding protein beta (C/EBPβ) transcription factor, called LIP (Liver-enriched inhibitory factor), is correlated with absence of the ER and PR receptors, a high proliferation index and aneuploidy in human breast cancer tumours 7. C/EBPβ comes in different protein isoforms (LAP* or Liver-enriched activating factor *, LAP or Liver-Liver-enriched activating factor, and LIP), which are translated from a single mRNA8–11. All isoforms share a highly conserved C-terminal domain containing the leucine zipper dimerization domain and the DNA binding domain. However, the N-terminal domain, containing the transactivation domains is present in the long isoform LAP and the extended isoform LAP*, but absent in the short isoform LIP 10,12. The transcriptional activators LAP* and LAP have been demonstrated to induce differentiation and inhibit proliferation, whereas LIP induces proliferation and decreases differentiation by competing for the same binding sites as LAP/LAP* and inhibits related gene transcription 10,11.
Previous studies have focused on identifying the roles of the C/EBPβ-isoforms in
vivo. We showed that LIP-deficient (CebpbΔuORF) mice display an increased
lifespan and reduced overall tumour incidence in females 13,14. Moreover, mono-or biallelic replacement of the wt Cebpb gene with a cDNA only expressing LIP increases tumour incidence in mice 15. Genetic ablation of Cebpb in the mouse mammary gland results in defective mammary ductal morphogenesis and lobuloalveolar development and differentiation of mammary epithelial cells 16. Furthermore, selective overexpression of LIP in the mammary epithelium leads to an increase in hyperplasias and less frequently in neoplasias 17, and a tight regulation of the LIP and LAP isoforms is essential for normal breast development and tissue maintenance, with LIP being induced during pregnancy when epithelial proliferation is high 18,19. Previously, we demonstrated that C/EBPβ-LIP induces cellular metabolic reprogramming, similar to oncogenic metabolic reprogramming, through regulation of the let-7/Lin28b circuit 20. Further analyses in our lab reveal a high LIP/LAP ratio in cell lines derived from TNBC, and overexpression of LAP in cell lines derived from TNBC results in reduced migration and invasion rates in vitro (Sterken et al, Chapter III). Moreover, RNA-seq analysis using BT20 CEBPB-knockout cells revealed a regulation of genes involved in migration, extracellular matrix components and immunity through C/EBPβ. Previously, several studies have provided evidence that C/EBPβ can mediate cancer growth by regulating immune cells and interaction with the microenvironment 21,22,23. Even though previous models have described C/EBPβ-isoform specific effects in vivo, Genetically engineered mouse models (GEMMs) to study the role of LIP and LAP in mammary tumour development are lacking.
GEMMs have greatly contributed to the field of cancer research, as they develop spontaneous tumours in an immune-proficient microenvironment 24. By genetic modification of specific tumour drivers and tumour suppressors in target cells in the mouse mammary gland, models for invasive lobular carcinoma (ILC) and basal-like Invasive ductal carcinoma (IDC) have been established 25. TP53 is found to be mutant in ~80% of all basal like cancers 26. Therefore, we introduced
mammary-specific exogenous expression of LIP and LAP in the mammary gland in mice with a mammary-specific p53-deficient background to study the role of LIP and LAP in mammary tumour development. Here, we show that tumour latency is not altered in the models with LIP and LAP expression cassettes. We found that although exogenous expression of both C/EBPβ isoforms is present in mouse mammary epithelial cells (MMECs) it is silenced in the developed tumours and during passaging of the MMECs in culture. In addition, we show that the LIP/LAP ratio does not correlate with tumour latency and survival in this model for breast cancer.
Results
Conditional expression of C/EBPβ-LIP and C/EBPβ-LAP in a
Wap-Cre;Trp53F/F mammary tumour model
In order to generate LIP and LAP specific mammary tumour genetically engineered mouse models (GEMMs), we introduced Cre-inducible knockin alleles of mouse LIP and mouse LAP into an existing p53-deficient mammary tumour GEMM Wap-Cre;Trp53F/F (WP) 27,28. We designed Cre-inducible LIP and LAP expression constructs linked with IRES-driven firefly luciferase expression for monitoring of tumour development, which we integrated into the Col1a1 locus of mouse Embryonic Stem Cells (ESCs) derived from the existing WP model 29 (figure 1a). Chimeric mice were generated by injecting the ESCs containing LIP-IRES-luciferase and LAP-IRES-luciferase into blastocysts of WP mice, and consequently crossed back into WP mice to generate Wap-Cre;Trp53F/F;Col1a1inv-CAG-LIP-Luc (WP-LIP) and Wap-Cre;Trp53F/F;Col1a1inv-CAG-LAP-Luc (WP-LAP) cohorts, with a mammary gland-specific deletion of p53 and a overexpression of LIP- or LAP-IRES-luciferase. To confirm whether Cre-recombinase mediated induction of exogenous LIP and LAP was obtained, Mouse Mammary Epithelial Cells (MMECs) were isolated from Trp53F/F, Trp53F/F;Col1a1inv-CAG-LIP-Luc and Trp53F/F;Col1a1inv-CAG-LAP-Luc mice. Transduction of MMECs with an adenovirus containing Cre-recombinase (AdCre) resulted in an efficient Cre-recombinase
mediated deletion of p53 and inversion of the CAG promoter (figure 1b, c). Moreover, to test whether the inversion of the CAG promoter results in exogenous LIP and LAP expression, MMECs transduced with AdCre were immunoblotted for C/EBPβ and revealed a strongly upregulated expression of LIP and LAP (figure 1d). These results demonstrate successful targeting of the
Col1a1 locus with LIP and LAP, resulting in an exogenous expression of the
mouse LIP and LAP isoforms upon Cre-mediated recombination.
Exogenous expression of C/EBPβ-LIP and C/EBPβ-LAP is silenced in mammary epithelium by DNA-methylation
To determine whether LIP and LAP were overexpressed in tumours from WP-LIP and WP-LAP mice, end-stage tumours from WP-WP-LIP and WP-LAP mice were analysed and compared to WP tumours. Locus-specific PCRs for the p53 deletion and recombination of the Col1a1 locus showed a successful p53 deletion and Cre-mediated recombination of the promoter in vivo (figure 2a). However, immunoblot analysis revealed that the exogenous overexpression of LIP or LAP proteins were absent in the majority of WP-LIP and WP-LAP mouse tumours (figure 2b). Additionally, most WP-LIP and WP-LAP tumours showed only background luminescence signals in vivo, which indicates that similarly to the C/EBPβ isoforms also the luciferase is not expressed (figure 2c). Kaplan-Meier analysis of WP, WP-LIP and WP-LAP cohorts revealed no differences in tumour-free survival and mammary tumour specific survival (figure S1a, b). While an analysis of the numbers of tumours per mouse shows a trend for higher numbers of tumours in WP-LIP and WP-LAP mice compared to WP mice (figure S1c), a comparison of tumour growth curves from WP, WP-LIP and WP-LAP tumours revealed large variability and no clear trend for altered tumour growth in WP-LIP and WP-LAP tumours (figure S1d, e, f). In vitro AdCre-mediated recombination resulted in recombination of the Col1a1 locus and consequent overexpression of LIP and LAP in MMECs. Therefore, we tested whether the LIP and LAP expression decreases over passaging. AdCre-treated Trp53F/F, Trp53F/F;Col1a1inv-CAG-LIP-Luc and Trp53F/F;Col1a1inv-CAG-LAP-Luc MMECs were
passaged and LIP and LAP immunoblot analysis revealed a loss of LAP upon early passages and a loss of LIP upon higher passage numbers (figure 2d). Previous analysis of tumour DNA (figure 2a) revealed presence of recombined promoter DNA in end-stage tumours, meaning that the LIP and LAP expressing cells were not outcompeted by cells with p53 deletion only. These results, together with the data from the MMECs, suggest that there is an initial overexpression of LIP and LAP in the murine mammary gland, which is silenced after the recombination in
vivo. Therefore, we sought out to investigate whether expression of LIP and LAP
is epigenetically silenced. MMECs transduced with AdCre-virus, which were passaged and had an established loss of LIP and LAP overexpression (figure 2d) were treated for 48 hours with DNA-methyltransferase inhibitor 5-Aza-2-deoxycytidine to test whether DNA methylation is involved in the silencing of the LIP and LAP expression. Immunoblot analysis revealed clear upregulation of LIP and of LAP upon 5-Aza-2-deoxycytidine treatment (figure 2e). Moreover, parallel luminescence imaging revealed an upregulation of luciferase signal upon treatment with 5-Aza-2-deoxycytidine (figure 2f). Therefore, we propose that the
in vivo loss of expression of LIP and LAP and the coupled luciferase is
epigenetically silenced by methylation.
P53-deficient mammary tumours display endogenous variation in LIP/LAP ratio that does not correlate with tumour onset
Even though the exogenous expression of LIP and LAP is repressed in the majority of tumours, we observed expression of endogenous C/EBPβ-LIP and C/EBPβ-LAP at varying levels and different LIP/LAP ratios. Quantification of the LIP/LAP expression ratio from tumour immunoblot analyses showed large endogenous variation in LIP/LAP ratio for all tumours isolated from all cohorts (WP, WP-LIP and WP-LAP) (figure 3a + S2). Previous studies have suggested that a high LIP/LAP ratio stimulates cell proliferation and tumour development, and therefore we hypothesised that tumours with a high endogenous LIP/LAP ratio may have shorter tumour latency and accelerated tumour growth compared to tumours with a low LIP/LAP ratio. Therefore, we performed a retrospective analysis plotting tumour latency and growth data from selected mice with
Figure 1: a) Schematic representation of the cloning of mouse LIP and LAP into the Frt-
invCAG-IRES-Luc vector. b) Schematic representation of the Trp53F locus (upper) and Col1a1 locus (lower) containing Frt-invCAG-LIP-IRES-LAP and Frt-invCAG-LIP-IRES-LIP. Upon cre-mediated recombination, exon 2-10 of Trp53 are excised and the CAG promoter is inverted. Primers detecting the deletion of Trp53 (FwT1, RvT1, FwT10, RvT10) and recombination of the Col1a1 locus (RvC2, RvC1, FwC1) are indicated (Adapted from Cornelissen et al 2019). c) PCR analysis of MMECs derived from Trp53F/F, Trp53F/F;Col1a1inv-CAG-LAP-Luc and Trp53F/F;Col1a1inv-CAG-LIP-Luc mice
β-globin mouse LIP or LAP
a + Cre-recombinase + Cre-recombinase c d #1 #1 #1 #3 #2 #2 #4 #3 #3 b
to detect AdCre-induced recombination of the Trp53 locus (upper band P53∆ and lowest band P53F) and recombination of the Col1a1 locus (upper band recombined, lower band not-recombined), with a spleen isolated from a WCreH;Trp53F/F;Col1a1inv-CAG-LIP-Luc mouse used as a
non-cre-recombined control. Corresponding primer locations are indicated in figure 1b. d) Immunoblot analysis of LIP and LAP expression in MMECs isolated from Trp53F/F,
Trp53F/F;Col1a1inv-CAG-LAP-Luc and Trp53F/F;Col1a1inv-CAG-LIP-Luc mice after in vitro AdCre-induced
recombination. β-actin is used for loading control.
tumours with the 25% highest and the 25% lowest LIP/LAP ratios (figure 3b). Survival analysis revealed no differences in tumour-free survival and mammary tumour specific survival (figure 3c, d). Next, we compared the growth curves from five tumours with the highest LIP/LAP ratios (4.55, 3.68, 3.53, 3.03 and 3) and the lowest LIP/LAP ratios (0.16, 0.21, 0.27, 0.34 and 0.36) (figure 3e), displaying no clear trends for altered growth patterns. Kaplan Meier curves reveal no clear differences in tumour onset between mice with tumours with high and low LIP/LAP ratios (figure 3f). In summary, these data suggest that endogenous variation of the LIP/LAP ratio in this p53-deficient model for mammary tumour development is not correlated to tumour onset and survival.
Discussion
Breast cancer is the most commonly diagnosed type of cancer in women and even though many advances have been made, still a frequent cause of cancer related death in women. Therefore, the identification of regulators contributing to breast cancer development and metastasis are of great interest. High expression of C/EBPβ-LIP has been found in hormone receptor negative breast cancer patient samples and is correlated to a poor prognosis 7. Several mouse models have been generated to study the isoform-specific tumour promoting functions of LIP, revealing that its overexpression predisposes to tumour development 14,15,17. Moreover, translation into LIP is often favoured by growth signalling in breast cancer and the translational regulation of C/EBPβ into its distinct isoforms might influence therapy resistance and metastasis 31–33. Recent analyses in our lab revealed high LIP/LAP ratios in cell lines derived from TNBC and that overexpression of LAP in TNBC cells in decreased cell migration and invasion
Figure 2: a) PCR analysis of tumours derived from WP, WP-LAP and WP-LIP mice to detect
AdCre-induced recombination of the Trp53 locus (upper band P53∆ and lowest band P53F) and the Col1a1 locus (upper band recombined, lower band not-recombined). Corresponding primer locations are indicated in Figure 1b. b) Immunoblot analysis of LIP and LAP expression in tumours derived from WP, WP-LAP and WP-LIP mice. β-actin is used for loading control. c) IVIS scans from tumours derived from WP-LAP mice. Shown are representative images of in vivo bioluminescence imaging of luciferase expression (corresponding to immunoblot 2b). d) Immunoblot analysis of LIP and LAP expression in AdCre-treated and passaged MMECs isolated from Trp53F/F, Trp53F/F;Col1a1inv-CAG-LAP-Luc and Trp53F/F;Col1a1inv-CAG-LIP-Luc mice. β-actin is used
for loading control. e) Immunoblot analysis of LIP and LAP expression and f) corresponding luminescence measurements of AdCre-treated and passaged MMECs from Trp53F/F,
Trp53F/F;Col1a1inv-CAG-LAP-Luc and Trp53F/F;Col1a1inv-CAG-LIP-Luc mice with loss of LIP and LAP
expression, treated with or without 5-Aza-2-deoxycytidine (a tumour-derived cell line from WB1P-Myc mice is taken along as control for luminescence due to successful exogenous WB1P-Myc expression and linked luciferase expression from the Col1a1 locus 34).
f e
d
MECC line Passage number a * c Cell line Cell line Cell line mouse # mouse #
Figure 3 a) Quantification of the LIP/LAP ratio in all tumours from the WP, WP-LIP and WP-LAP
cohorts (Geometric mean with geometric SD), One-way ANOVA, n.s. (non-significant). b) Generation of new cohorts based on the 25% highest and 25% lowest LIP/LAP ratios in tumours (Geometric mean with geometric SD). c) Kaplan-Meier curve of tumour-free survival of cohorts with the 25% highest LIP/LAP ratio vs the 25% lowest LIP/LAP ratio. Mantel-Cox: 25% highest LIP/LAP (n = 11, median survival 185 days) versus 25% lowest LIP/LAP (n = 11, median survival 168 days) n.s. d) Kaplan-Meier curve of mammary tumour specific survival in cohorts with the 25% highest LIP/LAP ratio vs the 25% lowest LIP/LAP ratio. Mantel-Cox: 25% highest LIP/LAP (n = 11, median survival 199 days) versus 25% lowest LIP/LAP (n = 11, median survival 182 days) n.s. e) Tumour growth curves of 5 tumours with the highest LIP/LAP ratio vs tumour growth curves of the 5 tumours with the lowest LIP/LAP ratio. f) Kaplan-Meier curve of mammary-specific tumour latency of mice corresponding to tumour growth curves from 2e. Mantel-Cox: 5 highest LIP/LAP (n = 4, median survival 186 days) versus 5 lowest LIP/LAP (n = 4, median survival 175 days) n.s. (for each cohort 1 mouse contains 2 tumours).
a d b c e Time (days) Tu m ou rs iz e (mm ^2 ) 0 10 20 30 40 0 50 100 150 200 250 5 highest LIP/LAP 5 lowest LIP/LAP Time (days) M amm ar y tu m ou r sp ec ifi c su rv iva l( %) 0 100 200 300 400 0 50 100 5 highest LIP/LAP 5 lowest LIP/LAP 25% highe st 25% lowes t 0.0625 0.125 0.25 0.5 1 2 4 8 LI P/ LA P ra tio WP WP-LI P WP-LA P 0.0625 0.125 0.25 0.5 1 2 4 8 16 LI P/ LA P ra tio Time (days) M amm ar y tu m ou r sp ec ifi c su rv iva l( %) 0 100 200 300 400 0 50 100 Time (days) Tu m ou r-f ree su rv iva l( %) 0 100 200 300 400 0 50 100 25% highest LIP/LAP (n=11) 25% lowest LIP/LAP (n=11) f 25% highest LIP/LAP (n=11) 25% lowest LIP/LAP (n=11) n.s. n.s. n.s. n.s. 3,68 3,53 3 3,03 4,55 0,36 0,27 0,34 0,21 0,16 LIP/LA P LIP/LA P
(Sterken et al, Chapter III). In this study, we sought out to generate a breast specific C/EBPβ-LIP and C/EBPβ-LAP overexpression in a p53-deficient mammary tumour model to investigate the functions of LIP and LAP in breast cancer development. Even though LIP and LAP were successfully integrated and overexpressed after Cre-mediated recombination in mammary epithelial cells, immunoblots of tumour lysates and in vivo luminescence imaging analysis revealed deficient exogenous expression of LIP and LAP in the tumours isolated from the experimental cohorts. Furthermore, retrospective analysis showed that endogenous variation in LIP/LAP ratio is not correlated with tumour onset. Data obtained from cell culture experiments, revealed that epigenetic silencing is causing the loss of ectopic luciferase, LIP and LAP expression in MMECs isolated from the WP-LIP and WP-LAP cohorts. Therefore, we propose epigenetic silencing as the mechanism for insufficient exogenous expression in the tumours
in vivo. Previous studies have shown successful integration and consequent
overexpression of c-Myc and Cas9 in this GEMM model 34. However, integration of human and mouse ERα constructs appeared to be silenced in a similar way to the LIP and LAP expression 35. Inarguably, c-Myc is a strong driver of mammary tumourigenesis and its exogenous expression is favourable and not subjected to silencing. Most likely, temporary overexpression of Cas9 is enough to result in the KO of target genes of interest. The development of ERα positive tumours in mice has proven to be difficult, with most GEMMs developing ERα-negative tumours and the few cases that do develop ERα-positive tumours do not respond to estrogen or endocrine therapy 36. Whereas integration of an ERα-IRES-luc allele in the Col1a1 locus of the WP model results in initial exogenous expression of ERα in the mammary gland at young age, it is silenced during mammary gland development and tumourigenesis 35. Similar to this study, we observe a loss of exogenous LIP and LAP expression in the developed tumour. Even though we have not analysed mammary glands of young WP-LIP and WP-LAP mice, we do observe trends for higher tumour numbers in WP-LIP and WP-LAP mice. However, this trend does not appear to be directional for LIP and LAP, where both WP-LIP and WP-LAP appear to have higher tumour numbers than the WP model, which contains an unmodified Col1a1 locus and therefore might not be
the optimal control. Therefore, we presume that only integrations of (strong) oncogenic drivers into the Col1a1 locus result in successful overexpression of genes of interest. Our results indicate that the exogenous expression of LIP and LAP is not as favourable in tumourigenesis as the expression of strong oncogenic drivers such as c-Myc. However, previous studies have shown that the overexpression of LIP both in vitro and in vivo promotes proliferation of mammary epithelial cells, but our in vitro MMEC data show that LIP expression is lost with increased passage numbers and in in vivo mammary tumour development. This suggests that silencing of the LIP allele still is promoted by selection against LIP-expressing cells, even though previous studies have pointed out that LIP predisposes to tumourigenesis.
Even though exogenous expression of LIP and LAP is repressed in the majority of tumours, we observed expression of endogenous LIP and C/EBPβ-LAP at varying levels and LIP/C/EBPβ-LAP ratios. Previously, we have observed high expression of C/EBPβ in cell lines derived from TNBC, with particularly high LIP/LAP ratios (Sterken et al, Chapter III), whereas differentiated tissues typically display higher expression of LAP than LIP 37. In this basal like tumour model, we observe an average LIP/LAP ratio of around 1, which is higher than reported in other differentiated tissues, but lower than in the TNBCs. These differences in expression between human and mouse basal-like tumours potentially reflect different roles and importance of the C/EBPβ isoforms in mouse and human tumours. However, our analysis has limitations. Firstly, tumours were dissected into small compartments for protein and DNA analysis, and contents may vary per region. Secondly, with immunoblotting of tumour pieces we are unable to distinguish between proteins from cancer cells and proteins from the (cells in the) microenvironment. Lastly, only correlative conclusions can be drawn from these analyses, since the LIP/LAP ratio plotted is the LIP/LAP ratio as detected in the end-stage tumour and LIP/LAP ratios might alter during tumour development. Even though this dataset is too small to draw conclusions, we propose that in the future mammary tumours with more strongly changed LIP/LAP isoform have to be analysed to study the functions of the isoform ratio in breast cancer development and metastasis. Given the consistent expression of LIP and LAP in
p53-deficient mammary tumours, a loss of function isoform specific model might be appropriate to study the functions of the isoforms in tumour development. Our previously published CebpbΔuORF model drastically reduces expression of LIP
and could be crossed with a p53-deficient model to induce mammary tumourigenesis, to test whether LIP is required for mammary tumourigenesis. Conversely, to model the effects of a high LIP/LAP ratio, the previously published LIP-overexpressing transgenic R26LIP mice could be crossed with the p53-deficient model, as this model has been demonstrated not to be susceptible to silencing of LIP expression.
In conclusion, we show that the ectopic expression of LIP and LAP are epigenetically silenced and that therefore their contribution to p53-deficient tumour development cannot be addressed. However, we did observe consistent endogenous expression of C/EBPβ in the p53-deficient mammary tumours. Even though endogenous variation in LIP/LAP ratio does not seem to correlate with tumour latency, we hypothesise that more extreme alterations of the LIP/LAP ratio will be required to see effects on mammary tumour development. We therefore propose the generation of mouse models with more drastically changed LIP/LAP isoform ratios to study its effects on mammary-specific tumour development. By crossing models with established modified C/EBPβ isoform expression we hope to circumvent epigenetic events repressing the expression and study the effects of LIP and LAP in mammary tumour development.
Acknowledgements
We thank the Lona Kroese and Colin Pritchard for generation of the mouse models and caretakers in the Mouse Clinic for Cancer and Aging for the help with maintaining the mice. We thank Koen Schipper for the help with the isolation of MMECs and Lisette Cornelissen for the scientific discussions about the project. Experiments performed in this manuscript are funded by the Dutch Cancer Society (KWF, #10080) through a grant to C.F.C. and by Stichting de Cock-Hadders through a grant to B.A.S.
Authorship contributions
B.A.S., A.P.D. and E.B. performed the experiments, C.P. and I.H. generated the mouse model and C.F.C. and P.B. supervised the project. B.A.S. wrote the manuscript.
Materials and methods
Mouse models
LIP and LAP constructs preceeded by a β-globin leader sequence were subcloned into pcDNA3 vector. Constructs were sequenced and globin-mLIP and β-globin-mLAP were inserted as BamHI-BamHI fragments into the BglII site in the Frt- invCAG-IRES-Luc vector 29. Next, Flp-mediated integration of the shuttle vectors into the Wap-Cre;Trp53F/F;Col1a1frt/+ ESC clones (FVB) and subsequent blastocyst injections of the modified ESCs were performed as described previously 29 to obtain chimeric animals. Chimeric animals were crossed with Wap-Cre;Trp53F/F FVB animals to generate Wap-Cre;Trp53F/F;Col1a1inv-CAG-LIP-Luc (WP-LIP) and Wap-Cre;Trp53F/F;Col1a1inv-CAG-LAP-Luc (WP-LAP) cohorts. Genotyping of the Wap-Cre, Trp53F/F, and Col1a1 LIP-IRES-Luc and Col1a1 inv-cag-LAP-IRES-Luc were confirmed by multiplex PCR using Mytaq HS red mix (BIO-25048, Bioline, Waddinxveen, The Netherlands) with an annealing temperature of 60 °C (genotyping primers listed in table S1). Monitoring of tumour growth was performed twice weekly by palpation and mice were sacrificed when total mammary tumour burden reached a size of 1500 mm3. All mouse experiments were approved by the Animal Ethics Committee of the Netherlands Cancer Institute and performed in accordance with institutional, national and European guidelines for animal care and use.
Isolation of mouse mammary epithelial cells (MMECs)
Primary MMECs were isolated as previously described 35,38,39. In summary, mammary glands were isolated from 8-12 week old Trp53F/F, Trp53F/F;Col1a1inv-CAG-LIP-Luc and Trp53F/F;Col1a1inv-CAG-LAP-Luc mice (different #s correspond to MMECs isolated from different mice). Briefly, mammary glands were minced, and incubated for 30 min at 37 degrees in collagenase/trypsin
solution in DMEM/F12 glutamax (31331–093, ThermoFisher Scientific) with 1 mg/ml collagenase A (11088793001, Sigma-Aldrich), 3 mg/ml trypsin (#215250, BD Biosciences, Breda, The Netherlands) and 5 μg/ml insulin (I0516, Sigma- Aldrich). Enzyme activity was neutralized by addition of DMEM/F12-Glutamax, containing 2% fetal bovine serum (FBS; F0926, Sigma-Aldrich), and the suspension was dispersed through a 40 μm cell strainer. Suspensions were centrifuged at 1500 rpm, and subsequently resuspended in DMEM/F12-Glutamax 3x. Cells were seeded in DMEM/F12- DMEM/F12-Glutamax, containing 10% FBS, 50 IU/ml penicillin, 50 μg/ml streptomycin (15070–63, ThermoFisher Scientific), 5 μg/ml insulin, 5 ng/ml EGF (E4127, Sigma- Aldrich) and 5 ng/ml cholera-toxin (Inaba 569B, Gentaur, Kampenhout, Belgium). Isolated MMECs were treated with 1-10 × 107 IU/ml AdCre (Gene Transfer Vector Core, University of Iowa, USA) 24 h after isolation. MMECs were harvested for experiments 48 h after AdCre transduction.
In vitro and in vivo bioluminescence imaging
For in vitro bioluminescence imaging of cultured MMECs, cells were treated with 5-Aza-2-deoxycytidine 24 hrs prior to bioluminescence imaging.
Just
before the measurement the medium was removed and phenol red free DMEM was added. The measurement of firefly-luciferase activity was performed according to the manufactures protocol (Promega, E1910). For detection, a GloMax-Multi Detection System (Promega) was used (1 sec measure).Bioluminescence imaging of mice was performed as previously described 40. Signal intensity was measured over the region of interest and quantified as flux (photons/s/square-centimeter/sr).
DNA isolation and analysis
DNA was isolated from tumours and from cell lines using lysis buffer containing 10 mM Tris pH7.8, 10 mM EDTA, 10 mM NaCl and 0,5% SDS, containing proteinase K (0.5 mg/ml) and incubated at 55 degrees C overnight. The next day, DNA was precipitated with EtOH and diluted in DNA hydration buffer (Qiagen 158914). For PCR analysis of the recombination of the Col1a1 locus, Phusion Flash High-Fidelity PCR Master Mix (# F548S) was used in combination with
locus specific recombination primers listed in supplementary table S1) as recommended by manufacturers. Detection of the Trp53F2–10 allele was performed by PCR of the loxP site in intron 1 or 10 as described in 28. For each reaction 100 ng of DNA was used for amplification.
Protein analysis
Tumour samples and cell lines were lysed in RIPA buffer as previously described35. Protein lysates were quantified using the Bradford assay. 50 ug of protein were separated on precast gels, 4-20%, 15 wells (BioRad Laboratories cat# 456-1096) and transferred onto PVDF membrane in transfer buffer from BioRad Laboratories 170-4273 according to manufacturer’s recommendations. Membranes were blocked in 5% milk in TBS-T (pH7.6, 20 mMTris, 138mM NaCl, 0.05% Tween-20 in demineralized water) and afterwards incubated in primary antibodies C/EBPβ (E299) from Abcam (ab32358, 1:1000) and β-actin (clone C4) (#691001) from MP Biomedicals (1:10.000). For detection,
HRP-conjugated secondary antibodies (Amersham Life Technologies) were used. The
signals were visualised by chemiluminescence (ECL, Amersham Life Technologies) using ImageQuant LAS 4000 mini imaging machine (GE Healthcare Bioscience AB) and the supplied software was used for the quantification of the bands. Due to unequal levels of loading control β-actin between different tumours, membranes were stained with Ponceau for 10 mins to check for unequal loading.
Statistical analysis
Statistical analyses were performed using Graphpad Prism version 8. Statistical tests used were one-way ANOVA and Log-rank (Mantel-Cox) test. P-values of <0.05 were considered to be significant.
References
1. 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. 0, 1– 41 (2021).
2. Perou, C. M. et al. Molecular portraits of human breast tumours. letters to nature 748. Nature 533, 747–752 (2000).
3. Sørlie, 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).
4. Hu, Z. et al. The molecular portraits of breast tumors are conserved across microarray platforms. BMC Genomics 7, 1–12 (2006).
5. Fragomeni, S. M., Sciallis, A. & Jeruss, J. S. Molecular Subtypes and Local-Regional Control of Breast Cancer. Surg. Oncol. Clin. N. Am. 27, 95–120 (2018). 6. Wahba, H. A. & El-Hadaad, H. A. Current approaches in treatment of
triple-negative breast cancer. Cancer Biol. Med. 12, 106–116 (2015).
7. Zahnow, C. a, Younes, P., Laucirica, R. & Rosen, J. M. Overexpression of C/EBPbeta-LIP, a naturally occurring, dominant-negative transcription factor, in human breast cancer. J. Natl. Cancer Inst. 89, 1887–1891 (1997).
8. Akira, S. et al. A nuclear factor for IL-6 expression (NF-IL6) is a member of a C/EBP family. EMBO J. 9, 1897–1906 (1990).
9. Descombes, P., Chojkier, M., Lichtsteiner, S., Falvey, E. & Schibler, U. LAP, a novel member of the C/EBP gene family, encodes a liver-enriched transcriptional activator protein. Genes Dev. 4, 1541–1551 (1990).
10. Descombes, P. & Schibler, U. A liver-enriched transcriptional activator protein, LAP, and a transcriptional inhibitory protein, LIP, are translated from the sam mRNA. Cell 67, 569–579 (1991).
11. Calkhoven, C. F., Müller, C. & Leutz, A. Translational control of C/EBPalpha and C/EBPbeta isoform expression. Genes Dev. 14, 1920–32 (2000).
12. Nerlov, C. The C/EBP family of transcription factors: a paradigm for interaction between gene expression and proliferation control. Trends Cell Biol. 17, 318–324 (2007).
13. Zidek, L. M. et al. Deficiency in mTORC1-controlled C/EBPβ-mRNA translation improves metabolic health in mice. EMBO Rep. 16, 1022–36 (2015).
14. Müller, C. et al. Reduced expression of C/EBPβ-LIP extends health and lifespan in mice. Elife 7, e34985 (2018).
15. Begay, V. et al. Deregulation of the endogenous C/EBPβ LIP isoform predisposes to tumorigenesis. J. Mol. Med. 93, 39–49 (2014).
16. Seagroves, T. N. et al. C/EBPbeta, but not C/EBPalpha, is essential for ductal morphogenesis, lobuloalveolar proliferation, and functional differentiation in the
mouse mammary gland. Genes Dev. 12, 1917–28 (1998).
17. Zahnow, C. A., Cardiff, R. D., Laucirica, R., Medina, D. & Rosen, J. M. A role for CCAAT/enhancer binding protein beta-liver-enriched inhibitory protein in mammary epithelial cell proliferation. Cancer Res. 61, 261–9 (2001).
18. Raught, B., Liao, W. S. & Rosen, J. M. Developmentally and hormonally regulated CCAAT/enhancer-binding protein isoforms influence beta-casein gene expression. Mol. Endocrinol. 9, 1223–1232 (1995).
19. Hirai, Y. et al. Epimorphin mediates mammary luminal morphogenesis through control of C/EBPbeta. J. Cell Biol. 153, 785–794 (2001).
20. Ackermann, T. et al. C/EBPβ-LIP induces cancer-type metabolic reprogramming by regulating the let-7/LIN28B circuit in mice. Commun. Biol. 2, 208 (2019). 21. Kurzejamska, E. et al. C/EBPβ expression is an independent predictor of overall
survival in breast cancer patients by MHCII/CD4-dependent mechanism of metastasis formation. Oncogenesis 3, (2014).
22. Li, W. et al. Aerobic Glycolysis Controls Myeloid-Derived Suppressor Cells and Tumor Immunity via a Specific CEBPB Isoform in Triple-Negative Breast Cancer. Cell Metab. 28, 87-103.e6 (2018).
23. Damiano, L. et al. Oncogenic targeting of BRM drives malignancy through C/EBPβ-dependent induction of ’5 integrin. Oncogene 33, 2441–2453 (2014). 24. Jonkers, J. & Berns, A. Conditional mouse models of sporadic cancer. Nat. Rev.
Cancer 2, 251–265 (2002).
25. Annunziato, S., Barazas, M., Rottenberg, S. & Jonkers, J. Genetic dissection of cancer development, therapy response, and resistance in mouse models of breast cancer. Cold Spring Harb. Symp. Quant. Biol. 81, 141–150 (2016).
26. Koboldt, D. C. et al. Comprehensive molecular portraits of human breast tumours. Nature 490, 61–70 (2012).
27. Derksen, P. W. B. et al. Mammary-specific inactivation of E-cadherin and p53 impairs functional gland development and leads to pleomorphic invasive lobular carcinoma in mice. Dis. Model. Mech. 4, 347–58 (2011).
28. Jonkers, J. et al. Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer. Nat. Genet. 29, 418–425 (2001). 29. Huijbers, I. J. et al. Using the GEMM-ESC strategy to study gene function in
mouse models. Nat. Protoc. 10, 1755–1785 (2015).
30. Begay, V., Smink, J. & Leutz, A. Essential Requirement of CCAAT / Enhancer Binding Proteins in Embryogenesis. Mol. Cell. Biol. 24, 9744–9751 (2004). 31. Li, H., Baldwin, B. R. & Zahnow, C. a. LIP expression is regulated by IGF-1R
signaling and participates in suppression of anoikis. Mol. Cancer 10, 100 (2011). 32. Arnal-Estapé, A. et al. HER2 silences tumor suppression in breast cancer cells by
switching expression of C/EBPβ isoforms. Cancer Res. 70, 9927–9936 (2010). 33. Baselga, J., Norton, L., Albanell, J., Kim, Y. M. & Mendelsohn, J. Recombinant
humanized anti-HER2 antibody (herceptin(TM)) enhances the antitumor activity of paclitaxel and doxorubicin against HER2/neu overexpressing human breast cancer xenografts. Cancer Res. 58, 2825–2831 (1998).
34. Annunziato, S. et al. Comparative oncogenomics identifies combinations of driver genes and drug targets in BRCA1-mutated breast cancer. Nat. Commun. 10, (2019).
35. Cornelissen, L. M. et al. Exogenous ERα Expression in the Mammary Epithelium Decreases Over Time and Does Not Contribute to p53-Deficient Mammary Tumor Formation in Mice. J. Mammary Gland Biol. Neoplasia 24, 305–321 (2019).
36. Özdemir, B. C., Sflomos, G. & Brisken, C. The challenges of modeling hormone receptor-positive breast cancer in mice. Endocr. Relat. Cancer 25, R319–R330 (2018).
37. Müller, C. et al. Reduced expression of C/EBPβ-LIP extends health and lifespan in mice. Elife 7, (2018).
38. Boelens, M. C. et al. PTEN Loss in E-Cadherin-Deficient Mouse Mammary Epithelial Cells Rescues Apoptosis and Results in Development of Classical Invasive Lobular Carcinoma. Cell Rep. 16, 2087–2101 (2016).
39. Kas, S. M. et al. Insertional mutagenesis identifies drivers of a novel oncogenic pathway in invasive lobular breast carcinoma. Nat. Genet. 49, 1219–1230 (2017). 40. Henneman, L. et al. Selective resistance to the PARP inhibitor olaparib in a mouse model for BRCA1-deficient metaplastic breast cancer. Proc. Natl. Acad. Sci. 112, 8409–8414 (2015).
Table S1: primer sequences
Allele 5’ -> 3’ sequence
Wap-Cre Fw ACAGCCATCAGTCACTTGCC
Rv CATCACTCGTTGCATCGACC
Trp53floxed Intron 10 FwT10 AAGGGGTATGAGGGACAAGG
RvT10 GAAGACAGAAAAGGGGAGGG
Trp53floxed Intron 1 FwT1 CACAAAAACAGGTTAAACCCAG
RvT1 AGCACATAGGAGGCAGAGAC Col1a1WT Fw CTCGCACGTACTTCATTC Rv CCCAAGAAATTCTCATCCTC Fw 2 LIP/LAP Fw AACCTGGAGACGCAGCACAAGG GENO-ires-REV / Shuttle geno RV Rv ACACCGGCCTTATTCCAAGC Col1a1 (Recombined/Non-recombined) FwC1 GGCCGGCCATAACTTCGTATAATG RvC1 CTGCGTTATCCCCTGATTCTGTGG RvC2 CCTACATCGAAGCTGAAAGCACGAG
Figure S1
Figure S1 a) Kaplan-Meier curve of tumour-free survival from WP, WP-LIP and WP-LAP mice.
Mantel-Cox: WP (n = 17, median survival 168 days), WP-LIP (n = 17, median survival 179 days) and WP-LAP (n = 23, median survival 167 days) n.s. b) Kaplan-Meier curve of mammary tumour specific survival from WP, WP-LIP and WP-LAP mice. Mantel-Cox: WP (n = 17, median survival 175 days), WP-LIP (n = 17, median survival 203 days) and WP-LAP (n = 23, median survival 184 days) n.s. c) Number of tumours per mouse WP (n=16), WP-LIP (n=15), WP-LAP (n=17). One-way ANOVA, n.s. d-f) Tumour growth curves of WP, WP vs WP-LAP, and WP vs WP-LIP.
Tumour growth Tum our s iz e (m m ^2 ) 0 1 0 2 0 3 0 4 0 0 1 00 2 00 3 00 WP Tumour growth Time (days) 0 1 0 2 0 3 0 4 0 0 1 00 2 00 3 00 Tumour growth 0 1 0 2 0 3 0 4 0 0 1 00 2 00 3 00 WP WP-LIP
# tumours per mouse
N um be r of tum our s WP WP LIP WPLAP 0 5 1 0 a b c d n.s. n.s. n.s. e f Tum our s iz e (m m ^2 ) WP WP-LAP Tum our s iz e (m m ^2 ) Time (days) Time (days)
Figure S2. Immunoblots, ponceau stainings, and quantification of LIP/LAP ratio in all tumours
