Tristetraprolin promotes survival of mammary progenitor cells by restraining TNFα levels

Tristetraprolin (TTP) is an RNA binding protein that destabilizes mRNAs of factors involved in proliferation, invasiveness, and inflammation. Disruption of the gene that codes for TTP (Zfp36) led to severe arthritis, autoimmunity, cachexia and dermatitis in mice. It has been shown that these phenotypes were mostly due to excessive TNFα levels in the affected tissues. We have previously reported that TTP expression is required for lactation maintenance. Our results indicated that conditional MG TTP-KO female mice displayed early involution due to the untimely induction of pro-inflammatory pathways led mostly by TNFα overexpression. Here we show that reducing TTP levels not only affects the fully differentiated mammary gland, but also harms morphogenesis of this tissue by impairing the progenitor cell population. We found that Zfp36 expression is linked to mammary stemness in human and mice. In addition, diminishing TTP expression and activity induced apoptosis of stem-like mouse mammary cells, reduced its ability to form mammospheres in culture and to develop into complete glands when implanted into cleared mammary fat pads in vivo. Our results show that survival of the stem-like cells is compromised by increased levels of inflammatory cytokines and stimulation of signaling cascades involving NFκB, STAT3 and MAPK-p38 activation. Moreover, TNFα overexpression and the consequent p38 phosphorylation would be the leading cause of progenitor cell death upon TTP expression restriction. Taken together, our results reveal the relevance of TTP for the maintenance of the mammary progenitor cell compartment by maintaining local TNFα levels at bay.


Introduction
Tristetraprolin (TTP) is a protein, encoded by the Zfp36 gene, that negatively regulates expression of multiple targets involved in inflammation, cell proliferation, angiogenesis, epithelial to mesenchymal transition and invasiveness.Typically, TTP binding to the 3′untranslated regions (3′UTR) of specific mRNAs leads to their destabilization, although this protein can also exerts transcriptional modulation of various genes (Guo et al., 2017;Kovarik et al., 2021).Importantly, it has been reported that TTP-KO mice showed a dramatic systemic inflammatory syndrome due, at least partly, to the overexpression of TNFα.On the other hand, TTP overexpression protected against the development of inflammatory diseases in various experimental models (Patial and Blackshear, 2016).
Mastitis, an inflammation of the mammary gland caused by infection or autoimmunity, has a relevant impact in women of fertile age (Goulabchand et al., 2020).It has been shown that TNF-α inhibitors may be a potential treatment for at least some types of this disease (Chiu et al., 2022).We have previously reported that TTP is expressed in the mammary differentiated epithelium (Goddio et al., 2012) and that its presence is required for lactation maintenance in the female mice.We found that lactating TTP fl/fl x Wap-Cre bitransgenic females (MG TTP-KO) showed premature involution due, at least partly, to excessive levels of TNFα in the mammary tissue (Goddio et al., 2018).
The mammary epithelium entails two main cellular lineages: luminal cells that border the ductal and alveolar lumen, and myoepithelial or basal cells, which are located next to the basement membrane (Watson and Khaled, 2008).The proliferative changes within this epithelium are fueled by embryonic pluripotent mammary stem cells (MaSCs), capable of originating both basal and luminal lineages, and by unipotent progenitor cells (either luminal or basal) that amplify and maintain each of these compartments in the post-natal gland throughout the successive reproductive cycles of the adult female mouse (Visvader and Stingl, 2014).Starting at puberty, a tree of branching ducts driven by the terminal end buds (TEBs) located at their ends, advance until they reach the limits of the mammary fat pad, when these structures disappear (Williams and Daniel, 1983).In each pregnancy, progenitor cells promote the development of the secretory lobuloalveolar compartment, which is the main source of milk production during lactation.After weaning, these differentiated structures regress due to massive cell death and tissue remodeling.This involution process is associated with inflammatory pathways that involve activation of transcription factors, such as STAT3 and NFκB (Watson and Kreuzaler, 2011;Inman et al., 2015).However, a pregnancy-induced mammary epithelial cell (PI-MEC) population that originates in the first pregnancy and lactation cycle do not dieout during involution and persists throughout the rest of the female mouse life.These cells possess stem cell-like features, are at least partially responsible for subsequent lobular development and may behave as cancer stem cells (Wagner and Smith, 2005).
Here we show that TTP downregulation affects not only the fully differentiated mouse mammary epithelium, but also the progenitor cell population.We found that this deleterious effect was mostly due to the dramatic overexpression of TNFα that led to p38-dependent cell death.Therefore, these results suggest that inflammatory diseases of the mammary gland may have an impact on lactation success in following reproductive cycles.

Mouse models
Mice were maintained in a pathogen-free facility at constant temperature of 22 °C ± 2 °C and 40%-70% humidity in a 12-h light cycle with food and water ad libitum.Wap-Cre/TTP fl/fl (MG TTP-KO) strain was obtained by crossing TTP fl/fl with Wap-Cre mice, both in C57BL/6 genetic background (Goddio et al., 2018).TTP fl/fl mice, generated and provided by Dr. Perry Blackshear´s lab (NIEHS, NIH, USA), have exon 2 of Zfp36 surrounded by LoxP sequences (Qiu et al., 2012).The Wap-Cre mice were obtained from the NCI-NIH Mouse Repository, United States.Genomic DNA was obtained from mouse tails by "HotSHOT," as previously described (Truett et al., 2000).Wap-Cre/TTP fl/fl mice were identified by PCR.Zfp36 floxed and Zfp36 WT alleles were amplified with P1_Fw and P2_Rv primers.Wap-Cre allele was amplified with W003_Fw and C031_Rv primers (see Supplementary Table S1).Removal of Zfp36 exon 2 in MG-TTP KO mammary glands was tested by multiplex PCR with two forward primers (P1_Fw, P3_Fw) and reverse primer P4_Rv.Product size of each PCR is shown in Supplementary  S5.
MMTV-Cre mice (from the NCI-NIH mouse repository) were crossed with TTP fl/fl mice.Bitransgenic female mice were identified by PCR as described above.MMTV-Cre allele was amplified using MMTV-Cre_Fw and CRE_C031_RV primers (Supplementary Table S1).Product size is shown in Supplementary Table S2.KO of TTP was checked by multiplex PCR as described above (Supplementary Figure S5).Eight-week-old females were euthanized and whole mount analysis of their mammary glands were performed.
Inguinal mammary fat pads of 21-day-old C57BL/6 (for TTP-KO tissue implants) or BALB/c (for HC11 cell injection) female mice were cleared of endogenous epithelium to perform implantation studies as described by Blair and DeOme, 1961.For TTP-KD or Sh-Ctrl HC11 cells, 1.0 × 10 5 and 2.0 × 10 5 cells were implanted in 100 µl.After 10 weeks, recipient mice were euthanized and mammary whole mounts were performed.Tertiary ducts, which originate at second branching events (Macias and Hinck, 2012;Huebner and Ewald, 2014), were analyzed in quantity and length.

Ethics statement
Mouse experiments were approved by local IACUC authorities (Comisión Institucional para el Cuidado y Uso de Animales de Laboratorio-CICUAL FCEN-UBA) and complied with regulatory standards of animal ethics (ARRIVE guidelines, Rs principles and AVMA guidelines).

Cell viability and apoptosis assays
We used the CellTiter 96 ® AQueous One Solution Cell Proliferation Assay (Promega, #G3582) for determining number of viable cells.Calibration curves were made by adding MTS 24 h post-plating and measuring the absorbance at 490 nm between 2.5 and 3.5 h later.Apoptotic cells were detected using the DeadEnd ™ Colorimetric TUNEL System (Promega, #G7360).Diaminobenzidine (DAB) was used as chromogen and hematoxylin as counterstaining.At least four fields per condition were photographed and quantified.For positive control, cells were treated with DNase I (4u/mL) for 10 min.

Primary cultures
Thoracic, abdominal and inguinal mammary glands were obtained with sterilized surgery tools and mechanically dissected in RPMI supplemented with HEPES, penicillin/streptomycin 1% v/ v, 0.15% collagenase type 4, 0.2% trypsin, DNase 10 μg/mL and 2% FBS (Fetal Bovine Serum).Then, they were incubated in agitation for 1 h at 37 °C and centrifuged several times to remove adipocytes from mammary epithelial cell pellets, which were resuspended in RPMI supplemented with HEPES, penicillin/streptomycin 1% v/v and 10% FBS.Cells were plated and monitored for up to 4 days, when they were trypsinized and used for mammosphere assays.

Western blot (WB) analysis
Cells were harvested in RIPA buffer with 1:100 protease inhibitor cocktail set I (Calbiochem, #539131) and phosphatase inhibitors (1:100 Na 2 Ov 4 ; 1:20 NaF).Most membranes were blocked for 1 h in 5% milk, but those for detection of phosphorylated proteins were blocked overnight in 5% gel fish (Sigma, #G7765).Blots were tested with antibodies listed in Supplementary Table S4.For MAPK analysis, fluorescent secondary antibodies were used and detected with the Odyssey imaging system.For studying other proteins, horseradish peroxidase (HRP) conjugated secondary antibodies were utilized and detection was performed with Amersham ImageQuant 800 equipment.

Immunofluorescence (IF)
For mammary gland and cell culture immunofluorescence, fixation was carried out in 4% paraformaldehyde (PFA), tissue was then included in OCT compound, frozen, and sectioned in 20 µm slides by cryostat.For DNA staining, slides were incubated with RNase A (10 μg/mL) for 3 h at 37 °C and 1 μg/mL propidium iodide (PI) was added for 5 min.Primary and secondary antibodies are listed in Supplementary Table S5.

Immunohistochemistry (IHC)
Mouse tissues were formalin-fixed, paraffin-embedded, and sectioned.After sample dewaxing and rehydration, antigen retrieval was performed with 10 mM Citrate buffer pH 6.0 for 10 min at sub-boiling tempeture.Sections were exposed for 2 min to 2% H2O2 and, after washing, blocked for 1 h in 2.5% bovine serum albumin (BSA).The samples were incubated overnight at 4 °C with the following antibodies diluted in 2.5% BSA: anti-CC3 (#9661, Cell Signaling Technology, 1:200), and anti-ZFP36/Tristetraprolin (#C431045, LSBio, 1:200).The next day, after washing, the samples were incubated for 1 h with an anti-rabbit biotinylated antibody, diluted 1:400 in 2.5% BSA.Next, samples were washed and incubated with the ABC reagent (PK-6101, Vector Laboratories) for 30 min.Finally, the sections were washed and stained with the DakoCytomation LSAB + System-HRP (Dako).Sections were counter-stained with hematoxylin.For CC3 and ZFP36/ Tristetraprolin-staining, pictures (200X and 400X, respectively) covering the whole area of the gland were taken.For CC3 quantification was done using ImageJ software and expressed as the number of positive CC3 cells/1,000 cells ±S.D.In the figure legends, the magnification of the pictures is indicated by scale bars.

Mammary whole mounts (WM)
Whole mammary glands were extended on slides and fixed for 2 h in Carnoy's solution (6:3:1 Ethanol: Chlorophorm: Glacial acetic acid).Tissue was hydrated to be stained O.N. with carmine-alum, dehydrated and mounted.

In silico analysis
Analysis of gene expression throughout different molecular subtypes of breast cancer were performed using The Cancer Genome Atlas (TCGA) -Pan-Cancer database and the METABRIC dataset.To evaluate expression levels differences, Kruskal Wallis and Dunn tests were carried out.To determine correlations, Spearman's ranked correlation test was used.For scRNA-seq analysis, matrices from Bach et al., 2017, Giraddi et al., 2018;Pal et al., 2017 were downloaded and loaded into R, version 3.6.3.The data corresponding to adult mice mammary samples were integrated into a single matrix and analyzed using the Seurat R package, version 3.4.4.A UMAP was generated using the top 3,000 variable genes selected via the default Variance Stabilizing Transformation (VST) method.Cluster/cell type labels were preserved from the original manuscripts for visualization and downstream analysis.Specific details in García Solá et al., 2021 and https://github.com/martings/JOMG-Garcia_Sola_et_al.

Statistical tests
Statistical significance of differences was evaluated using GraphPad Prism8.Data was presented as mean ± SEM unless otherwise noted.In every case, at least three independent experiments were evaluated.

TTP/Zfp36 downregulation induces mammary progenitor cell impairment in vivo
WAP gene promoter allows the expression of a transgene specifically in the mammary gland during lactation (Andres et al., 1987).Interestingly, throughout that period, in addition to the secretory cells, the WAP promoter is active in a progenitor compartment, the parity-induced mammary epithelial cells (PI-MEC), which are able to proliferate and produce new secretory acini in successive pregnancies (Boulanger et al., 2005).Therefore, to determine the impact of TTP expression impairment in progenitor cells using TTP fl/fl x Wap-Cre bitransgenic females (MG TTP-KO), we analyzed the mammary tissue at the second lactation period.As previously observed during first lactation (Goddio et al., 2018), MG TTP-KO mammary glands displayed signs of premature involution, like significant increase of cleaved caspase 3 (CC3), STAT-3 phosphorylation (p-STAT3, Y705) (Figures 1A, B) and lighter litters than those from control mice (Supplementary Figure S1A) at day 15 post-delivery.In addition, we observed that developmental deficiency of MG TTP-KO mammary glands was greater in the second reproductive cycle compared with the first one (Figure 1C).Apoptosis increase in mammary glands of MG TTP-KO was also detected during the second pregnancy of these mice (Supplementary Figure S1B).
To confirm the impairment of the progenitor mammary cell compartment, primary cultures of post-involuting glands were carried out.We found that MG TTP-KO cells took longer to attach and grow than cells from control mice (Supplementary Figure S1C), and that when plated on ultra-low attachment plates, they formed less mammospheres than their control counterparts (Figure 1D).These data suggested that survival of WAP-expresser PI-MECs has been negatively affected by TTP loss.In addition, flow cytometry assay of post involuting MG-TTP KO and control mammary glands revealed that the former displayed an important deficit of mammary epithelial cells, including in the basal/ progenitor compartment (Supplementary Figure S1D).
To verify whether mammary progenitor compartments may be altered in other mouse models with TTP deficiency, we implanted mammary tissue fragments from total TTP-KO or C57BL/6 wild type (control) mice into inguinal cleared fat pads of 3-week-old syngeneic females.After 10-12 weeks, whole mount analysis revealed that mammary glands from TTP-KO implants developed significantly lower number and length of tertiary ducts (Figure 1E).Similarly, mammary glands of 8-10 weeks-old MMTV-Cre x TTP fl/fl females displayed fewer major and branching ducts as well as a higher number of remaining TEBs, associated to incomplete invasion of cleared fat pads, than TTP fl/fl control mice (Supplementary Figure S1E).

Bio-informatic analysis of TTP/Zfp36
expression profile in mouse mammary gland and breast cancer data sets TTP/Zfp36 expression in mammary progenitor cell populations was evaluated in previously reported integrated single-cell RNAseq data sets of post-natal mouse mammary glands (García Solá et al., 2021).This analysis showed that relatively high levels of TTP/Zfp36 mRNA were detected in basal cell clusters, Procr+ mammary stem cells and Aldh1a3+ luminal progenitor cells (Figure 2A).A similar distribution in the assembled UMAP was determined for all members of the Zfp36 gene network that has been previously reported (Canzoneri et al., 2020) (Supplementary Figure S2).Furthermore, focusing specifically on the luminal alveolar populations, we found relevant TTP/Zfp36 expression levels mostly in progenitor cell clusters (Figure 2B).
In the human breast, most cancer subtypes have shown reduced TTP/ZFP36 expression compared with normal tissue or "Normal-Like" mammary tumors (Goddio et al., 2012;Canzoneri et al., 2020).However, "Claudin-Low" breast cancers, which present stem-like expression profiles (Fougner et al., 2020), displayed significantly higher ZFP36 expression levels than Luminal, Her2 and Basal subtypes (Figure 2C).Furthermore, using METABRIC data, we found positive correlations between expression of ZFP36 and genes associated with the stem-like phenotype, as TWIST1, TWIST2, SNAI1, ZEB1, ZEB2, ALDH1A and YAP1 (Fougner et al., 2020) in all breast cancer subtypes (Figure 2D).Similar observations were made analyzing TCGA-Pan Cancer data (Supplementary Figures S3A, B) and it was determined that positive correlations of TTP/ZFP36 expression with those stemlike markers were more significant in primary tumors (all subtypes considered) than in normal tissue (Figure 2E).Besides, analysis of this dataset also revealed that, in spite of clear TTP/ZFP36 expression downregulation in mammary tumors, high expression of this gene was not associated with better prognosis, considering all breast cancer subtypes, and that overall survival (OS) time was even shorter if only patients with basal-like subtype were considered (Figure 2F).Therefore, we consider that in the human mammary gland, low expression of TTP/ZFP36 may favor tumor development as several oncogenic and inflammatory pathways would be upregulated and therefore may contribute to progression of the disease.On the other hand, breast cancer cells with relatively high levels of this protein would display a stem-like phenotype, which is associated with breast cancer recurrence (Mao et al., 2022).

TTP/ZFP36 knock-down decreases survival, self-renewal, and repopulation capacity of HC11 mammary stem-like cells
To prove the impact of diminishing TTP levels in mammary progenitor cells, we proceeded to transfect specific Zfp36 shRNAs complementary to different protein-coding regions in HC11 mammary cells, which display stem-like features (Ferrari et al., 2015;Sornapudi et al., 2018).Although most clones in which TTP/Zfp36 expression was significantly decreased did not survive more than a few passages, we were able to establish a cell subline, transfected with shRNA 325, named TTP-KD.These cells showed significantly less Zfp36 mRNA and TTP protein expression than cells transfected with scrambled shRNAs (Sh-Ctrl cells) (Figure 3A and Supplementary Figures S4A, B).By immunofluorescence, we have also found weaker nucleic and cytoplasmic TTP staining in TTP-KD compared with ShCtrl cells (Supplementary Figure S4C).Similarly, TTP immunostaining performed in mammary glands of pregnant MG-TTP KO mice displayed less positive cells, and almost complete loss of nuclear staining (Supplementary Figure S4D).
To determine TTP activity in Sh-Ctrl and TTP-KD cells, they were transfected with the pTRE-LUC-3′UTR-Tnfα plasmid in which the TTP binding region of Tnfα 3′UTR is located downstream Luciferase cDNA (Carballo et al., 1998;Beisang and Bohjanen, 2012).Figure 3B shows that luciferase activity increased about 12-fold, although reduction of TTP/Zfp36 expression was only around 40%, in TTP-KD compared with Sh-Ctrl cells.
TTP-KD cell line showed reduced viability in different culture conditions (Figure 3C and Supplementary Figure S4E), increased apoptosis (Figure 3D) as well as higher levels of pro-apoptotic BAX and CC3 (Figure 3E).In addition, these cells displayed impaired selfrenewal capacity, since they produced smaller and fewer mammospheres than Sh-Ctrl cells in low-attachment plates (Figure 3F).Furthermore, when 1 × 10 5 TTP-KD cells were implanted into BALB/c inguinal cleared fat pads, no mammary development was observed after 10 weeks, although Sh-Ctrl cells gave rise to fully developed glands.When the number of implanted cells was doubled, both sublines originated mammary ductal networks.However, ducts from 2 × 10 5 TTP-KD cells did not reach mammary fat pad limits and displayed remaining TEBs, structures that had already disappeared in glands from Sh-Ctrl cells (Figure 3G).Therefore, TTP downregulation resulted in incomplete mammary ductal development in vivo, probably due to the impairment of the progenitor cell population.Interestingly, whole mount analysis of 1 × 10 5 TTP-KD and Sh-Ctrol cell implants performed during lactation revealed that pregnancy allowed development, although incomplete in most cases, and differentiation of the experimental cells (Supplementary Figure S4F).In addition, the ability of TTP-KD cells to differentiate when stimulated with lactogenic hormones in culture has been demonstrated when HC11 TTP-KD cells treated with prolactin and glucocorticoids displayed a relevant increase in STAT5 phosphorylation (Supplementary Figure S4G).

TTP/ZFP36 downregulation leads to inflammatory cytokine over-expression and activation of stress-associated signaling pathways
TTP-KD cells showed increased II-6 and Tnfα mRNA levels and phosphorylation of transcription factors commonly activated by  these cytokines, as STAT3 and NFkB p65/RelA, respectively (Berishaj et al., 2007;Kalliolias and Ivashkiv, 2016) (Figures 3H,  I).In fact, p65/RelA phosphorylation on S536 was mostly dependent on TNFα overexpression, since this effect was blocked by inhibiting that cytokine with Etanercept (Figure 3J).Importantly, an increase of p-p65 was also observed in the mammary glands of lactating MG TTP-KO compared with control mice (Figure 3K).
Figure 4A shows a drastic increase of constitutive MAPK p38 phosphorylation (p-p38, T183/Y182) while, ERK1/2 and JNK1/2 phosphorylation (Y204 and T183/Y185 respectively) was lower in TTP-KD compared with Sh-Ctrl cells.Interestingly, inhibition of p38, using SB203580, caused a rise of ERK1/2 and JNK1/2 phosphorylation, indicating that p38 overactivation may be responsible for downregulation of ERK1/2 and JNK1/2 activity in TTP-KD cells.On the other hand, ERK and JNK inhibition (by PD98059 and SP600125 treatment, respectively) decreased p-p38 levels in this cell line.Whether that was due to unspecific effects of these pharmacological inhibitors on p38 phosphorylation, or to ERK1/2 and JNK1/2 positive contribution to p38 phosphorylation, remains to be elucidated.
Over-activation of p38 as a consequence of TTP downregulation was also evident in mammary glands of lactating MG-TTP KO females, in which nuclear and cytoplasmic p-p38 labeling was observed.Nuclear localization was prioritized for quantification since it may be related with the pro-apoptotic activity of this enzyme (Wood et al., 2009;Cuadrado and Nebreda, 2010) (Figure 4B).
We determined that secreted TNFα was required for p65/Rel A and p38 phosphorylation as well as caspase 3 activation in TTP-KD cells, as these effects were blocked by Etanercept treatment.Besides, inhibition of p38 phosphorylation reduced p65/RelA phosphorylation and caspase 3 cleavage in these cells (Figure 4C).These results indicated that NFκB activation and apoptosis induction were controlled by p38 activation, which, in turn, depended on TNFα extracellular levels.In addition, Figure 4D shows that Etanercept and SB203580 treatment also prevented the dramatic Tnfα-3′UTR stabilization induced by TTP downregulation.Furthermore, inhibiting p38 and TNFα activity not only reversed, but greatly increased mammosphere formation capacity of TTP-KD cells (Figure 4E).Therefore, blocking the TNFα-p38 signaling pathway restores survival of TTP-deficient mammary progenitor cells.However, by analyzing mammosphere sizes, we observed that Etanercept treatment did not reverse proliferative impairment of TTP-KD cells, and that SB203580 inhibited proliferation of Sh-Ctrl cells (Supplementary Figure S4H).These data suggest that mechanisms regulating progenitor cell survival are different from those controlling proliferative capacity of the daughter cells.

Discussion
Here we show that TTP/Zfp36 expression is indispensable for controlling TNFα levels in mammary cells, even in the absence of infection, mutations, or autoimmune diseases.This suggests that there is constitutive transcription of the gene encoding that cytokine and its mRNA levels has to be regulated post-transcriptionally to avoid tissue damage.The relevance of post-transcriptional control for TNFα levels in different tissues and embryonic development has been previously pointed out by other authors (Clayer et al., 2020).
Our results indicate that in the mammary gland this mechanism is particularly important for stem cells, in which relatively high TTP levels preserve the integrity of the progenitor compartment allowing successive developmental rounds for feeding multiple litters.
Bipotent mammary stem cells (MaSCs) as well as long-lived unipotent stem cells in addition to diverse luminal progenitor subtypes have been identified in mouse and human mammary tissue (Visvader and Stingl, 2014).Particularly, the MG TTP-KO model allowed us to determine the importance of TTP in PI-MECs, which behave as multipotent mammary epithelial progenitors upon transplantation, and as cancer-initiating cells in MMTV-Her2/neu multiparous transgenic mice (reviewed by Wagner and Smith, 2005).These cells are able to form mammospheres in culture and express markers associated with mammary stem cells (Matulka et al., 2007).We have previously shown that TTP/ Zfp36 expression is highly increased during lactation, mainly due to prolactin induction, through Stat5A activation.In addition, we have demonstrated that during lactation, TTP wards off early involution by preventing the untimely increase of local inflammatory factors (Goddio et al., 2018).It has not been determined whether PI-MECs originate from cells expressing WAP that are not terminally differentiated, or whether they arise from differentiated cells that bypass apoptosis during involution or both (Wagner and Smith, 2005).Nevertheless, it can be speculated that TTP/Zfp36 expression decrease observed in involuting mammary glands (Goddio et al., 2012) would correspond to downregulation of this gene in the differentiated cells, making them susceptible for dying, while PI-MECs would maintain relatively high expression levels, preventing the impairment of this stem-like compartment.
Our analysis of scRNA-seq data sets confirm previously reported observations about conserved co-expression of TTP/Zfp36 with other early-responsive genes (Canzoneri et al., 2020;Gutierrez et al., 2022).Here we show that coincidence occurs, at least partly, in Procr+ MaSCs and Aldh1a3+ luminal progenitor cells.We have also found expression of this set of genes in some populations of the basal compartment, which might correspond to an "obligatory transitional transcriptional state" described by Gutierrez et al., 2022.We postulate that TTP/Zfp36 expression may the right of each graph, color scale (red: positive, blue: negative) indicates correlation level.Spearman's ranked correlation test was applied (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).(F) Overall survival (OS) of breast cancer patients (all subtypes, on the left; basal-like subtype, on the right) from TCGA database categorized according to ZFP36 expression levels.Number of patients (n) is indicated in each graph.To test significant differences, Log rank<0.05 was used.be relevant to maintain these mammary cell populations throughout different periods of the postnatal female mouse life.
Numerous reports have shown that TTP/ZFP36 downregulation is associated with breast cancer progression and/or treatment resistance (Marderosian et al., 2006;Brennan et al., 2009;Al-Souhibani et al., 2010;Griseri et al., 2011;Goddio et al., 2012;Upadhyay et al., 2013;Canzoneri et al., 2020).However, as pointed out previously (Griseri et al., 2011;Goddio et al., 2012), the relevance of TTP/ZFP36 mRNA level for breast cancer prognosis is debatable.In fact, using TCGA data, we did not find a negative correlation between high expression of this gene and breast cancer patient survival.Besides, we show that the association between the mammary stem-like phenotype and TTP/Zfp36 expression can be observed not only in mouse mammary cells, but also in human breast cancer tissue.Interestingly, positive correlation of this gene with different stem-like gene markers was stronger in primary tumors than in breast samples from healthy women, which might be due to a higher proportion of cells that express these markers in neoplastic tissue than in normal mammary epithelium.Therefore, we propose that high TTP levels might be relevant for maintaining the cancer stem cell population in some breast tumors.Furthermore, Wang et al. (2017) have demonstrated that CD8 + cytotoxic T lymphocytes (CTLs) are increased in Zfp36 −/− mice compared with WT mice.This was due to enhanced production of IL-27 by macrophages in the tumor microenvironment.In agreement with this observation, mammary tumors showed retarded growth in the TTP-KO mice, Therefore, high TTP levels may impair anti-tumor immunity contributing to lower the survival rate of basal breast cancer patients, as displayed in Figure 2F.
The significance of TTP/Zfp36 expression in the survival of stem-like cells was confirmed by the loss of cleared-fat pad repopulation and mammosphere formation capacity of TTP-KD cell line.These cells, which are still able to express significant amounts of TTP, showed a drastic decrease in their ability to reduce Tnfα mRNA probably due to the constitutive activation of p38, which phosphorylates the MAPK-activated protein kinase 2 (MK2) that, in turn, may completely block TTP activity (Tiedje et al., 2016).
Although TTP modulates multiple cytokines, it clearly exerts a dramatic effect on TNFα production by controlling its expression at transcriptional, post-transcriptional and translational level (reviewed by Guo et al., 2017).
Furthermore, it has been shown that cachexia, arthritis, and autoimmunity developed by TTP-KO mice were prevented if they were early treated with TNFα antibodies (Taylor et al., 1996).This indicated that the observed phenotype was mostly due to TNFα excessive levels.Here, we show that a pathway triggered by this cytokine would be mostly responsible for apoptosis induction in TTP-KD cells.Similarly, in lactating MG-TTP KO female mice, which also displayed IL-6 overexpression and STAT3 phosphorylation, excessive production, and secretion of TNFα played a key role in cell death induction (Goddio et al., 2018).Nevertheless, high levels of IL-6 and hyperactivation of STAT3 might be also relevant for inducing cell death, as previously reported in involuting mammary glands (Zhao et al., 2002;Hughes and Watson, 2018).TTP activity in modulating STAT3 activation by regulating levels of multiple factors have been reported in different cell types (e.g., Gaba et al., 2012;Jamal Uddin et al., 2013).Interestingly, loss of p63 in the mammary gland, a protein which has also been demonstrated to be essential for the survival and maintenance of PI-MECs, led to epithelial cell death, together with increased activation of the oncostatin M/Stat3 pro-apoptotic pathways (Yallowitz et al., 2014).In addition, TTP may modulate PI-MECs by regulating expression levels of different cytokines, as IL-2 or IL-10 in the mammary gland microenvironment (Ogilvie, R. et al., 2005;Stoecklin, G. et al., 2008).
Our results show that TNFα induces p38 phosphorylation in TTP-KD cells, as reported in other models (Sabio and Davis, 2014;Kalliolias and Ivashkiv, 2016), and that activation of this MAPK leads to mammary cell death, as it has also been previously observed (Wen et al., 2011).Besides, constitutive over activation of p38 may be, at least partially, also responsible for downregulation of ERK1/2 and JNK1/ 2 phosphorylation.This effect would be similar to what was reported in macrophages by Hall and Davis, 2002.In summary, we had previously demonstrated that TTP/ Zfp36 expression in the mammary gland is necessary for maintaining lactation and here we show a novel important role for TTP, which is to preserve the progenitor cell compartment.These cells are fundamental for fueling mammary development during puberty as well as throughout each reproductive cycle of female mice.In addition, our results point out the risk involved in long term exposure to high TNFα levels for physiological development, renewal and function of the mammary tissue.

FIGURE 2
FIGURE 2 Bioinformatic TTP/Zfp36 expression analysis in mouse mammary epithelial cells and human breast tissue.(A) Uniform Manifold Approximation and Projection (UMAP) plot of postnatal MEC populations displaying scRNAseq relative expression levels of Zfp36, Procr and Aldh1a3 in color intensity gradient.(B) UMAP plot of mammary luminal cell populations displaying relative expression levels of Zfp36, in color intensity gradient.Cell populations, as previously identified by García Solá et al., 2021, are indicated by color patterns.(C) Expression of ZFP36 mRNA (log2 values) in breast cancer molecular subtypes according to METABRIC database.Kruskal Wallis and Dunn's tests were performed.Significant differences are represented by asterisks (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001).Number of patients (n) is indicated below each plot.(D, E) Expression correlation between ZFP36 and genes related to stem/progenitor behavior comparing different breast cancer subtypes according to METABRIC database (D), and between normal tissue and primary tumors according to TCGA data (E).On (Continued )