br TP INP overexpression resulted in
TP53INP2 overexpression resulted in a trend toward increased cancer-induced muscle loss also in female mice, at both early and late stages of cachexia (Fig. S3).
To understand the molecular alterations associat-ed with muscle wasting in WT and Tg TB mice, the Haloperidol of relevant genes was assayed (Fig. 5).
Several atrogenes were induced in WT TB mice, atrogin-1 and MuRF1 being further increased in TP53INP2 TB mice, suggesting that both the proteasomal and the autophagic degradation con-tribute to the increased muscle protein catabolism. Autophagy was reported to be strongly induced in cachectic muscles  and in keeping with those data an enhanced expression of autophagy genes (LC3, GABARAPL1, p62, LAMP2A, BNIP3) was detected in the muscle of WT C26 hosts. In TP53INP2 TB mice, transcript levels for all the assayed genes were comparable to WT C26 hosts, despite the role played by TP53INP2 in autophagic degradation, implying that TP53INP2-induced catabolism is not mediated by transcriptional regulation. Beyond bulk or selective autophagy, mitochondrial dynamics modulates muscle wasting as well [12,18]. In the present study, BNIP3 mRNA was similarly induced in WT and Tg TB mice, whereas both mitochondrial
Autophagy and mitochondria in cancer cachexia 2679
Fig. 4. TP53INP2 overexpression worsens tumor-induced muscle, though not adipose loss. Lean (a) and fat (c) % mass time course of controls (C) and C26-bearing mice (C26) either WT (n = 5 for C and 9 for C26) or overexpressing TP53INP2 specifically in the skeletal muscle (Tg; n = 7 for C and 9 for C26). Gastrocnemius (GSN; b) and gonadal white adipose tissue (WAT; d) mass expressed in mg × 100 g of initial body weight. Different letters indicate statistically different results. Statistical significance is set at p b 0.05.
fission- (DRP1) and fusion- (OPA1 and MFN2) related gene expression was suppressed in WT and even more severely in TP53INP2 TB mice. On the contrary, the mitochondrial biogenesis regulator PGC-1α transcript levels were unchanged.
Depression of muscle mitochondrial function in C26-bearing mice
The association between autophagy and the repression of mitochondrial dynamics genes prompted us to analyze mitochondrial function. Ex vivo mitochondrial respiration in permeabilized EDL fiber bundles was severely impaired in TB mice irrespective of TP53INP2 overexpression (Fig. 6a). Respiratory complexes I and II showed a similar decrease in muscles from TB mice (Fig. 6a). Such a reduction in mitochondrial respiration might be explained by mitophagy, as suggested by the increased BNIP3-mitochondria (SDHa) co-localization in both WT and TP53INP2 TB mice compared to control conditions (Fig. 6b and S4).
Consistently with increased bulk autophagy, LC3B-II levels were increased in the cytosol from the muscle of TB mice (Fig. 6c), the more so in the mitochondrial fraction, suggesting a marked activa-tion of mitophagy (Fig. 6C). SDHa protein content remained unaltered in WT TB mice with respect to control values, while a mild reduction was detected in the muscles of TP53INP2 TB mice, suggesting that TP53INP2 could promote an increased autoph-agy flux, allowing to speed mitochondrial clearance (Fig. 6d).
The present study was aimed to clarify the role of autophagy in cancer-induced muscle wasting, in the attempt to spare muscle mass through the modulation of muscle proteolysis. The results show that, despite autophagy is indubitably induced in the skeletal muscle of tumor hosts in a short experimental setting, its partial blockade by genetic
2680 Autophagy and mitochondria in cancer cachexia
Fig. 5. TP53INP2 overexpression exacerbates atrogene induction and mitochondrial dynamic suppression. Gene expression in the TA of controls (C) and C26-bearing mice (C26) either WT (n = 5 for C and 9 for C26) or overexpressing TP53INP2 specifically in the skeletal muscle (Tg; n = 7 for C and 9 for C26). The expressed values are relative to C WT mice (set to 1 and gray color). Red denotes gene induction and green gene repression. Different letters indicate statistically different results. Statistical significance is set at p b 0.05.
manipulation does not improve tissue wasting. The explanations for such inefficacy might be various. First, given that beclin-1 knockdown was not complete, the tendency to increased TA mass in TB mouse suggests that a complete beclin-1 silencing/inhibition might turn in a consistent muscle sparing. Second, autophagy could be relevant in the short-term atrophy induction observed in the C26 model, followed by differential adaptations in the long term, vanishing the effects of blocking autophagy. However, we recently set up a more “chronic” model by treating C26-bearing mice with chemotherapy . In that condition, although a flux experiment was not performed, the alterations of autophagy-related proteins are exacerbated, sug-gesting that the acute boost progresses into a chronic state of autophagy impairment. Third, autophagy is activated in the skeletal muscle upon wasting stimuli in parallel with other proteolytic systems, namely, the calpains and the proteasome . As for the latter, a tight crosstalk with autophagy is becoming evident in the last years  and a compensatory proteasome-mediated protein hyper-catabolism potentially occurs as a consequence of