Francesco Elia Marino, 2017


This section is in progress. Thanks for your patience. 














Francesco Elia Marino, Gail Risbridger and Elspeth Gold. 

Inhibins and activins are essential for normal reproductive and endocrine function. Previous literature showed that loss of INHA encoding for inhibin-α subunit in mice results in development of gonadal (sex-cord stromal) and adrenal tumorigenesis. However, a strong paradox has emerged regarding the function of INHA in human tumors primarily due to conflicting studies. A correlation has been found between the over-expression of activin-A by the tumors in the inhibin-deficient mice and severe cancer-associated cachexia, which has a profound impact on survival. Therefore, activin antagonism has been used to modulate the sex-cord stromal tumor growth rate both in vivo and in vitro. Recent work has shown that activin-C is an additional activin-A antagonist, despite being considered biologically redundant for several years. A preliminary study conducted in the inhibin-deficient mouse model, showed that over-expression of activin-βC has a beneficial effect, modulating sex-cord stromal tumors development and attenuating the cancer-associated cachexia phenotype.

The overall objectives of my research were to clarify the incongruity regarding the inhibin-α subunit in human versus mouse and to further explore the biological function of activin-βC in vivo and in vitro defining its effect on gonadal, adrenal tumorigenesis and cancer-cachexia. Finally, my work wanted to clarify if activin-C is a tumor suppressor in the inhibin knockout mice.

Initially my work re-evaluated the inhibin/activin pathway in human gonadal and adrenal cancers using contemporary protein and mRNA expression data for pathway components rather than inhibin-α alone. Clarification of the inhibin/activin expression profile in human cancers represented an essential part of the study in order to validate the in vivo findings described in the mouse model for potential translational applications. The study also described, for the first time, a comprehensive protein expression profile of activin-C in reproductive and adrenal cancers and its effect on a human Granulosa cell line (COV434).

The study then proceeded to determine whether over-expression of activin-βC modulated adrenal tumors in gonadectomized inhibin deficient mice. Male and female WT, ActC++, α-KO (INHAKO), and α-KO/ActC++ mice were gonadectomized and monitored up to 30 weeks of age. Markers of apoptosis and proliferation were assessed, and survival analyses were conducted in the castrated animals versus the sham operation control groups.

Finally my work identified the potential molecular mechanism by which activin-C increased survival and modulated cancer-associated cachexia in the α-KO mice. Western blot analysis for the specific E3 ubiquitin ligase, atrogin-1 and MuRF1, effectors Smad-2/Smad-3 and myostatin was performed in the gastrocnemius muscle. Histopathology and survival analysis was conducted in animals from the same breeding cohort. The study also explored the impact of over-expression of activin-βC on the serum levels of activin-A, inflammatory cytokines (TNF-α, IL-6 and IF-γ), hormonal profile and bone density.

Results showed a reduced inhibin-α expression at both protein and mRNA levels in human testicular and ovarian cancers, and increased activin signalling in human testicular and ovarian cancers. Increased expression of inhibin-α was found in benign versus malignant forms of adrenal cancers and no significant changes were noted in the inhibin-αmRNA levels between adrenocortical carcinoma and adrenocortical adenomas versus normal controls. Thus, showing a similar expression profile as the inhibin deficient mouse model. Comparing the immunoreactivity from two antibodies raised against different regions of the inhibin-α protein, my reseearch suggested that difference in antibodies previously used in the literature contributed to the biological paradox regarding the activity of INHA in cancer biology. The study also found that activin-C acts as an activin-A antagonist by binding to the activin receptor IIA and IIB and modulating the canonical Smad pathway.

The study showed that over-expression of activin-βC modulated the progression of Granulosa and Sertoli cell tumors in the α-KO/ActC++ mice versus the α-KO counterpart. Over-expression of activin-βC also increased survival in the same animals. However, despite a clear effect on modulating sex-cord stromal tumors and cancer-associated cachexia my research showed that over-expression of activin-βC had no effect on adrenal tumorigenesis. In fact, proliferation, apoptosis and survival evaluated in gonadectomized animals, were not different in the α-KO versus the α-KO/ActC++ mice.

Finally, the study showed that over-expression of activin-βC in the α-KO mice had an effect on muscle wasting reducing the activation of transcription factors such as atrogin-1, MurF-1, and Smad-2 involved in the activation of the muscle protein degradation pathway. Results confirmed that over-expression of activin-βC reduces serum levels of activin-A, and showed it also reduces inflammatory cytokines, factors recognized to have an essential role in the progression of cancer-associated cachexia and directly involved in triggering the mechanism of wasting. The study further validated the importance of the activin signaling pathway in cancer-associated cachexia and suggested activin-C as a new anti-activin strategy to combat cancer associated weight loss and prolong survival.

In conclusion, the research conducted in my Ph.D. thesis provided new evidence to explain part of the INHA dilemma by demonstrating similar inhibin-α expression in human and mouse tumors. Findings aim to have a significant impact on the way INHA is considered. The identification of activin-C as an important regulator of gonadal tumorigenesis and muscle wasting might have implications for the activin field and should provide the basis for a completely new body of work relevant to reproductive and cancer biology. Finally, data presented in my thesis showed that activin-C cannot be considered a tumour suppressor in the inhibin-knockout mice. However, its overexpression significantly modulated gonadal tumorigenesis, survival and muscle wasting in the inhibin-knockout mice.


Using an antagonist of activin receptors: ActRII (M108A), follistatin, inhibin, CRIPTO, BAMBI, a soluble receptor ActRII ECD, Activin-C or small-molecule inhibitors of ActRI (SB-431542 and SB-505124) induce a blockade of activin signalling pathway (Harrison et al., 2004, DaCosta Byfield et al., 2004, Inman, 2002, Cheng et al., 2003).


Weighing up evidence for inhibin-α as a tumor suppressor or tumor promoter.

Weighing up evidence for inhibin-α as a tumor suppressor or tumor promoter.

Shared mechanism of signal transduction mediated by activin-A (ActA) and Myostatin (GDF-8).

Effect of activin-βC on the activin-signalling pathway.

Fold difference (shown in table) calculated on the mean IRS or percentage of positive cells (Smad-2 and Smad-3) of the disease ovary (A), testis (B) and adrenal (C) was compared with the normal controls: ovary (n . 8), testis (n . 13) and adrenal (n . 16). ActRIIA: activin receptor IIA, ActRIIB: activin receptor IIB. The IRS for each tissue was treated as ranked data and analyzed with Mann–Whitney U-test for two independent groups.*P value , 0.05, **,0.01, ***,0.001. From Marino et al. 2014. Mol. Hum. Reproduction

Stereological assessment of testis (left) and ovary (right) shows a reduction in the percent tumor when activin-βC is over-expressed. From Marino et al. 2015, Molecular and Cellular Endocrinology