Search

While interplay between BRCA1 and AURKA-RHAMM-TPX2-TUBG1 regulates mammary epithelial polarization, common genetic variation in HMMR (gene product RHAMM) may be associated with risk of breast cancer in BRCA1 mutation carriers. Following on these observations, we further assessed the link between the AURKA-HMMR-TPX2-TUBG1 functional module and risk of breast cancer in BRCA1 or BRCA2 mutation carriers. Forty-one single nucleotide polymorphisms (SNPs) were genotyped in 15,252 BRCA1 and 8,211 BRCA2 mutation carriers and subsequently analyzed using a retrospective likelihood approach. The association of HMMR rs299290 with breast cancer risk in BRCA1 mutation carriers was confirmed: per-allele hazard ratio (HR) = 1.10, 95% confidence interval (CI) 1.04 – 1.15, p = 1.9 x 10−4 (false discovery rate (FDR)-adjusted p = 0.043). Variation in CSTF1, located next to AURKA, was also found to be associated with breast cancer risk in BRCA2 mutation carriers: rs2426618 per-allele HR = 1.10, 95% CI 1.03 – 1.16, p = 0.005 (FDR-adjusted p = 0.045). Assessment of pairwise interactions provided suggestions (FDR-adjusted pinteraction values > 0.05) for deviations from the multiplicative model for rs299290 and CSTF1 rs6064391, and rs299290 and TUBG1 rs11649877 in both BRCA1 and BRCA2 mutation carriers. Following these suggestions, the expression of HMMR and AURKA or TUBG1 in sporadic breast tumors was found to potentially interact, influencing patients’ survival. Together, the results of this study support the hypothesis of a causative link between altered function of AURKA-HMMR-TPX2-TUBG1 and breast carcinogenesis in BRCA1/2 mutation carriers.

Els models de malalties són una eina bàsica per la comprensió de les malalties humanes. Actualment, la majoria de la informació de la que disposem de malalties humanes es basa en models animals. Tot i això, els models animals difereixen molecular i fenotípicament dels humans, i no sempre reprodueixen fidelment la malaltia humana. En les últimes dècades, les cèl·lules mare humanes s’han establert com una opció molt interessant en el camp de la modelització cel·lular. En aquest treball hem volgut caracteritzar les cèl·lules mare embrionàries com a models per a l’estudi de la inestabilitat de la repetició de trinucleotids a la distròfia miotonica tipus 1 (DM1) i la malaltia de Huntington (HD). Així mateix, hem volgut estudiar la inactivació del cromosoma X amb la intenció de fer servir linees cel·lulars com a models per l’estudi del desenvolupament embrionàri humà. A la primera part d’aquest treball, hem observant una inestabilitat de repeticions de trinucleotids significativa al locus de la malaltia DM1 de les cèl·lules mare estudiades. La diferenciació d’aquestes cèl·lules va estabilitzar el número de repeticions. L’estabilització de les repeticions va ser concomitant amb la regulació a la baixa de l’expressió de gens involucrats en els mecanismes de reparació cel·lular. Posteriorment a la publicació del nostre article, altres grups varen reproduir els nostres resultats, però en aquest cas utilitzant cèl·lules mare induïdes. Els estudis recolzen la reproductibilitat dels nostres resultats, suggerint que poden ser extrapol·lats a altres linees de cèl·lules mare arreu del mon. Referent a la mutació de HD, varem trobar que era estable en totes les condicions estudiades, en cèl·lules indiferenciades, diferenciades a progenitors d’os, teratomes i progenitors neurals. Aquests resultats estan en concordancia amb els resultats obtinguts per altres grups que descriuen un baix nombre de repeticions al locus de HD. Per altra banda, varis grups han descrit la presencia de inestabilitat de les repeticions en cèl·lules diferenciades a la linea neural. La discrepància entre els nostres resultats i aquests últims podria ser deguda a la obtenció de cèl·lules neurals menys madures en el moment del nostre estudi. A la segona part d’aquesta tesis hem estudiat la inactivació del cromosoma X en 23 línies femenines de cèl·lules mare pluripotents. Vàrem observar una ràpida progressió de les cèl·lules de dependència de XIST en la inactivació del cromosoma X cap a un estat d’adaptació al cultiu que es caracteritza per un estadi de inactivació independent de l’expressió de XIST i amb una erosió de la metilació. També describim un patró d’inactivació esbiaixat en la majoria de les línies estudiades, contrari al patró aleatori observat en cèl·lules femenines adultes. A més a més, aquest patró és independent de XIST, de l’origen del cromosoma X i d’aberracions cromosòmiques. Aquests resultats suggereixen que el patró esbiaixat observant esta dirigit provablement per l’activació o repressió d’al·lels específics que es troben en el cromosoma X i que li confereixen a la cèl·lula un avantatge respecte a les altres cèl·lules. En conclusió, les cèl·lules mare pluripotents semblen ser un bon model in vitro per a l’estudi d’ambdues malalties, DM1 i HD, ja que presenten el mateix patró d’inestabilitat de la repetició del trinucleotid que s’observa in vivo. Cal remarcar també la depencia Overall, hPSC appear to be a good in vitro model for the study of both DM1 and HD TNR instability, as the repeat follows in vitro the same patterns as found in vivo, including its dependency of the MMR machinery, particularly in the case of DM1. However, our results on the study of the X chromosome inactivation (XCI) state suggest caution when using hPSC as early human developmental research models. The eroded state of XCI found in many of the hPSC lines, and the frequency of skewed XCI patterns suggests that these cells are not a good proxy to early embryonic cells, at least what XCI is concerned. Conversely, they may still provide an interesting model to study gene function and mechanisms implicated., Disease modelling is an essential tool for the understanding of human disease. Currently, much of the information we have on human diseases is based on animal models. However, animal models differ molecularly and phenotypically from humans, and are not always suitable to reproduce with fidelity human diseases. In the past decades, human pluripotent stem cells (hPSC) have emerged as an interesting option in the field of cellular modelling, this development recently having taken up much momentum. In this work, we aimed at characterizing hPSC as models for the study of Myotonic dystrophy type 1 (DM1) and Huntington’s disease (HD) trinucleotide repeat (TNR) instability and to investigate the status of the X-chromosome inactivation with an eye on using these cells as models for early human development. In the first part of our work, we observed a significant TNR instability for the DM1 locus in hESC, and that differentiation resulted in a stabilization of the repeat. This stabilization was concommitant with a downregulation of the mismatch repair (MMR). Our results were later replicated in hiPSC by other researchers, showing their reproducibility and suggesting they may be extrapolated to other hPSC lines worldwide. Regarding the HD repeat, we found it was very stable in all conditions studied, both in undifferentiated hESC and cells differentiated into osteogenic progenitor-like cells, teratoma cells and neural progenitors. This is in line with other studies showing that hESC show very limited TNR in the HD locus. On the other hand, some groups have now reported some instability of this locus in cells differentiated into the neuronal lineage. The instability seen in neuronal lineage in later studies, not in our study, is probably explained by the use of hPSC derived neurons more similar to the cells showing in vivo instability than the ones we were able to generate at the time of the study. In the second part of the thesis we studied the X-chromosome inactivation in 23 female hPSC lines. We found that hPSC rapidly progress from a XIST-dependent XCI state to a culture-adapted, XIST-independent XCI state with loss of repressive histone modifications and erosion of methylation. We also report a remarkably high incidence of non-random XCI patterns, and that this skewing of the methylation patterns is independent from the transition to the XIST-independent XCI state, the origin of the X chromosome or chromosomal aberrations. These results suggest that XCI skewing is possibly driven by the activation or repression of a specific allele on the X chromosome, conferring a growth or survival advantage to the cells. Overall, hPSC appear to be a good in vitro model for the study of both DM1 and HD TNR instability, as the repeat follows in vitro the same patterns as found in vivo, including its dependency of the MMR machinery, particularly in the case of DM1. However, our results on the study of the X chromosome inactivation (XCI) state suggest caution when using hPSC as early human developmental research models. The eroded state of XCI found in many of the hPSC lines, and the frequency of skewed XCI patterns suggests that these cells are not a good proxy to early embryonic cells, at least what XCI is concerned. Conversely, they may still provide an interesting model to study gene function and mechanisms implicated.

e12112 Background: Pathological complete response (pCR) after neoadjuvant chemotherapy for BC has been related with survival, and the sequence of anthracyclines followed by taxanes (AT) has been the strategy to achieve this response. Some studies suggest that the reverse sequence (TA) might improve cancer outcomes. The aim of our study was to compare clinical response, toxicity and efficacy in terms of pCR rate (ypT0/pTis, ypN0) between both sequences Methods: A retrospective cohort study was designed. TA cohort received Docetaxel (100 mg/m2) x 4, followed by AC (60/600 mg/m2) x 4. AT cohort received the reverse sequence at the same doses. To be included, it was necessary to have received at least one cycle in the planned sequence. Logistic regression analysis was performed to obtain a model adjusted by age, menopausal status, tumor size, nodal status, grade, estrogen receptor (ER), progesterone receptor, and Ki67 Results: From June 2008 to December 2015, 135 consecutive patients with stage II – III HER2 negative BC were treated: TA (n = 48) and AT (n = 87). No significant differences in patient characteristics between groups were found, except for histological grade 3 (63.2% to AT vs. 35.4% to TA), and ER negative (41.4% to AT vs. 22.9% to TA). The pCR rate for TA was 16.7% (8 of 40) vs. 14.9% (13 of 74) for AT. Multivariate logistic regression analysis found OR = 3.65 to achieve pCR for TA (95%CI 1.03 – 12.94; p = .045). We also performed the same analysis for two clinical parameters of response: Neoadjuvant response index (NRI) with cut-off > 0.5 (OR 3.96, 95%CI 1.42-11.08; p = .009) and Clinical response (CR) > 50% (OR 3.65, 95%CI 1.03 – 12.94; p = .045). These parameters were correlated in our series with DFS (p = .011 to NRI and p = .00023 to CR) and OS (p = .034 to NRI and p = .0014 to CR) using the Kaplan-Meier method. No significant differences between rates of either, hospitalizations, neutropenia or dose intensity were found Conclusions: Sequence of T followed by A was slightly significantly more effective to achieve pCR as well as better clinical outcomes in our series of HER2 negative BC than classical sequence of A followed by T, and support the design of large prospective studies to confirm these results

Introduction Trinucleotide repeat (TNR) instability is the mechanism responsible for numerous inherited human disorders, including Huntington's disease (HD) and myotonic dystrophy (DM1). The TNRs have different patterns of instability according to the disorder as well as the cell type investigated. It is known that the mismatch repair (MMR) machinery is involved in TNR instability, and particularly MSH2 and MSH3 seem to be required for the intergenerational and somatic CTG and CAG expansions. Most of these results have been obtained from studies on transgenic mice, due to the lack of suitable human models. The aim of this work is to study the behaviour of the TNRs causing HD and DM1 in human embryonic stem cells (hESCs), both in the undifferentiated and differentiated state, and its relationship to the MMR machinery, in order to evaluate the potential of these cells as a model in the human. Materials and methods HESC lines were established in our laboratory using spare IVF/PGD embryos. VUB03_DM1 carries a CTG expansion in the DMPK1 gene and VUB05_HD carries a CAG expansion in the huntingtin gene. VUB02 was used as a control line. The lines were kept in standard culture conditions, on mouse embryonic fibroblasts, and samples were taken at different passages. A well-established protocol was used to differentiate hESCs into osteoprogenitor-like cells (OPL). Whenever possible, samples for RNA and DNA extraction were taken after each passage. TNR stability was studied by PCR followed by southern blotting and/or analysis on a DNA sequencer. The MMR machinery was studied by quantitative real-time PCR (qRT-PCR) of the MSH2, MSH3, MLH3 and PMS2 genes. Results While the HD repeat showed a stable pattern in undifferentiated hESCs over more than 80 passages, the DM1 repeat was unstable from the first passages onwards, the TNR mostly increasing in size, leading to a mosaic culture. In the OPLs, the HD TNR remained stable, whereas the DM1 repeat showed a tendency towards a clonal growth (only a few alleles remained detectable), and a stabilization in the behaviour of the repeat. The results of the qRT-PCR showed a downregulation of all the studied genes in the OPLs, in comparison to the undifferentiated hESCs. Conclusions The data on VUB05_HD show that the behaviour of this TNR in hESC follows the same pattern as its somatic behaviour in vivo, in which only striatal neurons show TNR instability. Further experiments on in-vitro differentiated neuronal tissue from hESC are planned. In the case of DM1, the results suggest a correlation between the downregulation of the MMR machinnery and the stabilization of the TNR, which will be further studied by siRNA experiments.

Additional file 4: Figure S4. RANK but not RANKL expression increased after dual anti-HER2 therapy in HR+ and HR- patient samples (n = 151) from the PAMELA trial. A and B. Ladder plots (left panels) showing RANK and RANKL gene expression in PAMELA HER2-positive HR+ (A) and HR- (B) tumors before (baseline) and after (surgery) dual anti-HER2 treatment. An increase in gene expression is represented in red and a decrease in blue. Each line represents a tumor sample from one patient. P values in A were calculated by comparing mean values between both groups and in B were determined by paired two-tailed t-tests. Density plots (right panels) showing RANK and RANKL gene expression in PAMELA HER2-positive HR+ (A) and HR- (B) tumors before (baseline) and after (surgery) treatment.

While interplay between BRCA1 and AURKA-RHAMM-TPX2-TUBG1 regulates mammary epithelial polarization, common genetic variation in HMMR (gene product RHAMM) may be associated with risk of breast cancer in BRCA1 mutation carriers. Following on these observations, we further assessed the link between the AURKA-HMMR-TPX2-TUBG1 functional module and risk of breast cancer in BRCA1 or BRCA2 mutation carriers. Forty-one single nucleotide polymorphisms (SNPs) were genotyped in 15,252 BRCA1 and 8,211 BRCA2 mutation carriers and subsequently analyzed using a retrospective likelihood approach. The association of HMMR rs299290 with breast cancer risk in BRCA1 mutation carriers was confirmed: per-allele hazard ratio (HR) = 1.10, 95% confidence interval (CI) 1.04 – 1.15, p = 1.9 x 10−4 (false discovery rate (FDR)-adjusted p = 0.043). Variation in CSTF1, located next to AURKA, was also found to be associated with breast cancer risk in BRCA2 mutation carriers: rs2426618 per-allele HR = 1.10, 95% CI 1.03 – 1.16, p = 0.005 (FDR-adjusted p = 0.045). Assessment of pairwise interactions provided suggestions (FDR-adjusted pinteraction values > 0.05) for deviations from the multiplicative model for rs299290 and CSTF1 rs6064391, and rs299290 and TUBG1 rs11649877 in both BRCA1 and BRCA2 mutation carriers. Following these suggestions, the expression of HMMR and AURKA or TUBG1 in sporadic breast tumors was found to potentially interact, influencing patients’ survival. Together, the results of this study support the hypothesis of a causative link between altered function of AURKA-HMMR-TPX2-TUBG1 and breast carcinogenesis in BRCA1/2 mutation carriers.

Additional file 2: Figure S2. RANK and RANKL staining in TMAs. A. Pictures of RANK and RANKL protein expression analyzed by IHC in the TMA cores from the treatment-naïve cohort. B. Pictures of RANK and RANKL protein expression analyzed by IHC in the TMA cores from the anti-HER2 resistant cohort.

Additional file 2: Figure S2. RANK and RANKL staining in TMAs. A. Pictures of RANK and RANKL protein expression analyzed by IHC in the TMA cores from the treatment-naïve cohort. B. Pictures of RANK and RANKL protein expression analyzed by IHC in the TMA cores from the anti-HER2 resistant cohort.

BackgroundAround 15–20% of primary breast cancers are characterized by HER2 protein overexpression and/orHER2gene amplification. Despite the successful development of anti-HER2 drugs, intrinsic and acquired resistance represents a major hurdle. This study was performed to analyze the RANK pathway contribution in HER2-positive breast cancer and anti-HER2 therapy resistance.MethodsRANK and RANKL protein expression was assessed in samples from HER2-positive breast cancer patients resistant to anti-HER2 therapy and treatment-naive patients. RANK and RANKL gene expression was analyzed in paired samples from patients treated with neoadjuvant dual HER2-blockade (lapatinib and trastuzumab) from the SOLTI-1114 PAMELA trial. Additionally, HER2-positive breast cancer cell lines were used to modulate RANK expression and analyze in vitro the contribution of RANK signaling to anti-HER2 resistance and downstream signaling.ResultsRANK and RANKL proteins are more frequently detected in HER2-positive tumors that have acquired resistance to anti-HER2 therapies than in treatment-naive ones.RANK(but notRANKL) gene expression increased after dual anti-HER2 neoadjuvant therapy in the cohort from the SOLTI-1114 PAMELA trial. Results in HER2-positive breast cancer cell lines recapitulate the clinical observations, with increased RANK expression observed after short-term treatment with the HER2 inhibitor lapatinib or dual anti-HER2 therapy and in lapatinib-resistant cells. After RANKL stimulation, lapatinib-resistant cells show increased NF-κB activation compared to their sensitive counterparts, confirming the enhanced functionality of the RANK pathway in anti-HER2-resistant breast cancer. Overactivation of the RANK signaling pathway enhances ERK and NF-κB signaling and increases lapatinib resistance in different HER2-positive breast cancer cell lines, whereas RANK loss sensitizes lapatinib-resistant cells to the drug. Our results indicate that ErbB signaling is required for RANK/RANKL-driven activation of ERK in several HER2-positive cell lines. In contrast, lapatinib is not able to counteract the NF-κB activation elicited after RANKL treatment in RANK-overexpressing cells. Finally, we show that RANK binds to HER2 in breast cancer cells and that enhanced RANK pathway activation alters HER2 phosphorylation status.ConclusionsOur data support a physical and functional link between RANK and HER2 signaling in breast cancer and demonstrate that increased RANK signaling may contribute to the development of lapatinib resistance through NF-κB activation. Whether HER2-positive breast cancer patients with tumoral RANK expression might benefit from dual HER2 and RANK inhibition therapy remains to be elucidated.

Additional file 4: Figure S4. RANK but not RANKL expression increased after dual anti-HER2 therapy in HR+ and HR- patient samples (n = 151) from the PAMELA trial. A and B. Ladder plots (left panels) showing RANK and RANKL gene expression in PAMELA HER2-positive HR+ (A) and HR- (B) tumors before (baseline) and after (surgery) dual anti-HER2 treatment. An increase in gene expression is represented in red and a decrease in blue. Each line represents a tumor sample from one patient. P values in A were calculated by comparing mean values between both groups and in B were determined by paired two-tailed t-tests. Density plots (right panels) showing RANK and RANKL gene expression in PAMELA HER2-positive HR+ (A) and HR- (B) tumors before (baseline) and after (surgery) treatment.

Additional file 1: Figure S1. TMA H-scores and controls. A. RANK and RANKL H-scores in HER2-positive breast cancer samples, treatment-naïve (left panels) or anti-HER2-resistant (right panels). In treatment-naïve TMAs, each number represents a “core” from a single patient. In anti-HER2-resistant TMAs, scored independent tumor cores are numbered for each patient (after the symbol #). B. Representative pictures of human breast tumors from patient-derived xenografts used as positive and negative controls for RANK and RANKL IHC.

Additional file 1: Figure S1. TMA H-scores and controls. A. RANK and RANKL H-scores in HER2-positive breast cancer samples, treatment-naïve (left panels) or anti-HER2-resistant (right panels). In treatment-naïve TMAs, each number represents a “core” from a single patient. In anti-HER2-resistant TMAs, scored independent tumor cores are numbered for each patient (after the symbol #). B. Representative pictures of human breast tumors from patient-derived xenografts used as positive and negative controls for RANK and RANKL IHC.

El subministrament d’aliments a les zones urbanes en creixement segueix una tendència lineal: les ciutats consumeixen una gran quantitat d’aliments importats alhora que generen residus i impactes ambientals en diferents fronts. L’agricultura urbana (UA) ha destacat com una pràctica per mitigar la quantitat d’aliments importats, generant beneficis en les tres dimensions de la sostenibilitat i una excel·lent oportunitat per a la restauració de fluxos. El concepte d’economia circular (CE) pot contribuir a aquesta mitigació minimitzant els fluxos de residus i aprofitant estratègies per recuperar recursos mentre s’exploten les sinergies entre sistemes urbans, contribuint a la millora del metabolisme urbà. No obstant això, l’aplicació dels principis de la CE als sistemes de la UA s’ha de controlar estrictament en termes de comportament ambiental per evitar un xoc entre els objectius de la CE i la sostenibilitat. Per evitar-ho, la present tesi pretén avaluar el comportament ambiental de l’aplicació d’estratègies circulars en sistemes de UA. Per arribar a aquest objectiu, utilitzem l’anàlisi de cicle de vida (ACV) combinada amb altres mètodes i eines complementàries: estudi dels fluxos de nutrients, anàlisi de variables climàtiques, sistemes d’informació geogràfica o avaluació de la circularitat mitjançant l’indicador de circularitat de materials (MCI). En primera instància analitzem el rendiment ambiental de 25 cicles de 7 cultius diferents en un hivernacle hidropònic en coberta per determinar quins haurien de ser els elements de l’inventari a optimitzar i definir les millors combinacions anuals de cultius. Els resultats mostren que els fertilitzants i les seves emissions a l’aigua són l’element amb més marge de millora. En aquest sentit, avaluem l’aplicació de tres possibles estratègies de recuperació de nutrients: filtració per membranes, precipitació química i recirculació directa de lixiviats, trobant que aquesta última és la que presenta el millor comportament ambiental. Tenint en compte això, avaluem mitjançant diferents enfocaments el rendiment de dues estratègies de recirculació diferents: la recirculació de lixiviats en un mateix cultiu i la recirculació de lixiviats en un cultiu paral·lel en el que es coneix com a sistema en cascada. Els resultats de l’anàlisi descriuen els potencials i les limitacions d’ambdós sistemes, concloent que la configuració definitiva inclouria un sistema en cascada en què el cultiu receptor reutilitzi els nutrients que lixivia. El fòsfor (P) és un recurs escàs i valuós a causa de la demanda antropogènica per produir fertilitzants. La sinergia entre sistemes urbans pot contribuir a augmentar la utilitat dels recursos de P a nivell urbà. En aquest sentit, analitzem la recuperació d’estruvita a les plantes de tractament d’aigües residuals urbanes i la seva aplicació en sistemes UA amb dos enfocaments diferents: (1) experimentalment per comprovar el seu potencial i limitacions en sistemes hidropònics, i (2) amb una perspectiva regional que tracta una àrea metropolitana com una entitat auto-suficient. Els resultats mostren com l’ús d’estruvita pot generar produccions més elevades que els fertilitzants minerals, a més de disminuir les pèrdues de P. La perspectiva regional a la recuperació i reutilització d’estruvita mostra com l’àrea estudiada és capaç de recuperar prou P per alimentar tota l’agricultura de la regió, tot i que alguns paràmetres com l’elecció de la planta de tractament d’aigües residuals o la tecnologia de recuperació implementada són importants per evitar impactes ambientals addicionals. Tenint en compte les conclusions de la tesi, es poden definir diferents línies de recerca futures: estandardització de mètriques i conceptes al voltant de l’economia circular o integració de diferents perspectives en l’avaluació de sistemes UA. El suministro de alimentos en zonas urbanas en crecimiento sigue una tendencia lineal: las ciudades consumen una gran cantidad de alimentos importados, generando residuos e impactos ambientales en distintos frentes. La agricultura urbana (UA) ha permitido disminuir la cantidad de alimentos importados, generando beneficios en las tres dimensiones de la sostenibilidad y una excelente oportunidad para la restauración de flujos. El concepto de economía circular (CE) puede contribuir a esta mitigación minimizando los flujos de residuos y aprovechando estrategias para recuperar recursos mientras se explotan las sinergias entre sistemas urbanos, contribuyendo a la mejora del metabolismo urbano. Sin embargo, la aplicación de los principios de la CE a los sistemas de la UA se debe controlar estrictamente en términos de comportamiento ambiental para evitar un choque entre los objetivos de la CE y la sostenibilidad. Para evitar esto, la presente tesis pretende evaluar el comportamiento ambiental de la aplicación de estrategias circulares en sistemas de UA. Para llegar a este objetivo, utilizamos el análisis de ciclo de vida (ACV) combinado con otros métodos y herramientas complementarias: estudio de los flujos de nutrientes, análisis de variables climáticas, sistemas de información geográfica o evaluación de la circularidad mediante el indicador de circularidad de materiales (MCI). En primera instancia analizamos el rendimiento ambiental de 25 ciclos de 7 cultivos diferentes en un invernadero hidropónico en cubierta para determinar cuáles deberían ser los elementos del inventario a optimizar y definir las mejores combinaciones anuales de cultivos. Los resultados muestran que los fertilizantes y sus emisiones al agua son el elemento con más margen de mejora. En este sentido, evaluamos la aplicación de tres posibles estrategias de recuperación de nutrientes: filtración por membranas, precipitación química y recirculación directa de lixiviados, concluyendo que esta última es la que presenta el mejor comportamiento ambiental. Teniendo en cuenta esto, evaluamos mediante distintos enfoques el rendimiento de dos estrategias de recirculación diferentes: la recirculación de lixiviados en un mismo cultivo y la recirculación de lixiviados en un cultivo paralelo en lo que se conoce como un sistema en cascada. Los resultados del análisis describen los potenciales y las limitaciones de ambos sistemas, concluyendo que la configuración definitiva incluiría un sistema en cascada en el que el cultivo receptor reutilice los nutrientes que lixivia. El fósforo (P) es un recurso escaso y valioso debido a la demanda antropogénica de P para producir fertilizantes. La sinergia entre sistemas urbanos puede contribuir a aumentar la utilidad de los recursos de P a nivel urbano. En este sentido, analizamos la recuperación de estruvita en las plantas de tratamiento de aguas residuales urbanas y su aplicación en sistemas UA con dos enfoques diferentes: (1) experimentalmente para cuantificar su potencial y limitaciones en sistemas hidropónicos, y (2) con una perspectiva regional que trata un área metropolitana como una entidad auto-suficiente. Los resultados muestran como el uso de estruvita puede generar producciones más elevadas que los fertilizantes minerales, además de disminuir las pérdidas de P. La perspectiva regional en la recuperación y reutilización de estruvita muestra como el área estudiada es capaz de recuperar suficiente P para alimentar toda la agricultura de la región, aunque algunos parámetros como la elección de la planta de tratamiento de aguas residuales o la tecnología de recuperación implementada son importantes para evitar impactos ambientales adicionales. Teniendo en cuenta las conclusiones de la tesis, se pueden definir diferentes líneas de investigación futuras: estandarización de métricas y conceptos alrededor de la economía circular o integración de diferentes perspectivas en la evaluación de sistemas UA. Food supply to ever-growing urban areas follows a linear tendency: cities consume a vast amount of imported food while generating waste and environmental impacts in different fronts. Urban agriculture (UA) has stood out as a practice to mitigate the volume of the imported flow, generating benefits in all three dimensions of sustainability and an excellent opportunity for the restoration of flows. The young concept of a circular economy (CE) can contribute to this mitigation by minimizing waste flows and take advantage of strategies to recover resources while exploiting synergies between urban systems, contributing to an improvement of the urban metabolism. However, the application of CE principles in UA systems should be strictly monitored in terms of environmental performance to avoid a clash between CE and sustainability goals. To avoid this, the present thesis aims to evaluate the environmental performance of applying circular strategies in UA systems. We use the Life Cycle Assessment (LCA) combined with other complementary methods and tools: nutrient balances, analysis of climatic variables, geographical information systems or a circularity assessment through the material circularity indicator (MCI). We first analyze the environmental performance of 25 cycles of 7 different crops in a hydroponic rooftop greenhouse to determine which should be the targets to optimize within the inventory and define the best year-round crop combinations. The results show that the fertilizers and their related emissions to water are the item with the biggest room for improvement. To avoid nutrient depletion, we evaluate the application of three possible nutrient recovery strategies: membrane filtration, chemical precipitation and direct leachate recirculation, finding that the latter had the best environmental performance. Considering this, we evaluate with different approaches the performance of two different recirculation strategies: the recirculation of leachates in the same crop and the recirculation of leachates in a parallel crop in what is known as a cascade system. The results of the analysis outline the potentials and limitations of both systems, concluding that the ultimate configuration would include a cascade system in which the receiving crop reuses the nutrients that leaches. Phosphorus (P) is a scarce and valuable resource due to the anthropogenic demand to produce fertilizers. The synergy between urban systems can contribute to enlarge the utility of P resources at the urban level. In this sense, we analyze the recovery of struvite in urban wastewater treatment plants and its application in UA systems with two different approaches: experimentally to test its potential and limitations in hydroponic systems, and with a regional perspective that treats a metropolitan area like a self-sufficient entity. The results show how the use of struvite can produce higher yields than mineral fertilizers while diminishing P losses. The regional perspective showed how the area under study is able to recover enough P to feed all the agriculture of the region, although parameters like the choice of the wastewater treatment plant or the recovery technology are important to avoid additional environmental impacts. Finally, we gather all the recovery strategies and evaluate their environmental performance and degree of circularity to prioritize circular strategies in UA systems. After solving detected limitations of the assessment through indicator development, we find that nutrient recirculation, the use of struvite or recycled materials are the best strategies to improve both the circularity and environmental performance of the system. Considering the findings of the thesis, different future research lines were defined: standardization of circular economy metrics and concepts or integration of different perspectives in the assessment of UA systems. Universitat Autònoma de Barcelona. Programa de Doctorat en Ciència i Tecnologia Ambientals