**4. Discussion**

The main result of this study is that the effect of pravastatin and gemfibrozil is organ specific and dose dependent. Both drugs seem to have a deteriorating effect on hepatic mitochondria but rather positive influence on colonic mitochondrial respiration.

The chosen experimental setting is based on our previous study [24]. The drug concentrations correspond to the literature describing similar in vitro experiments [13,15,19,27]. The measurements are performed at 30 ◦C which is a methodological standard [13,15,20], but not a physiological condition. Thus, the data may not reflect full effects of the drugs in vivo. While liver is mainly composed of hepatocytes, colon consists of different cell lines like epithelial cells, smooth muscle cells, adipocytes and many others. Therefore, our results cannot relate to a special cell line.

In hepatic mitochondria, pravastatin dose dependently reduced ADP-induced mitochondrial respiration-state 3, and coupling between electron transport chain (ETS) and oxidative phosphorylation (OXPHOS)-RCI, without changing the efficacy of oxidative phosphorylation-ADP/O.

In colonic mitochondria, pravastatin also reduced state 3 and RCI, however, to a minor extent, mainly at the higher concentration and preferably through complex II. In contrast to hepatic mitochondria, pravastatin increased the efficacy of OXPHOS in the colon for both complexes.

Our results concerning pravastatin and mitochondrial respiration are new findings compared to the results of other authors who mainly could not show any effects of this drug on mitochondrial respiration. Marques et al. examined the effect of pravastatin on hepatic mitochondrial in LDL knockout mice after oral pretreatment with 40 mg/kg pravastatin and did not observe any changes either [16]. Sugiyama et al. tested the effects of pravastatin on age-related changes in mitochondrial function in rats after long-term therapy. They could show that pravastatin significantly accelerated the age-related decline in the activity of complex I of diaphragm mitochondria, whereas the aging effect on mitochondrial respiratory function was not observed on heart muscle and liver. Pravastatin did not significantly a ffect cardiac and hepatic respiratory function [28]. Kaufmann et al. [13] did not observe any e ffects of pravastatin on mitochondrial oxygen consumption in isolated muscle mitochondria in vitro using similar pravastatin concentrations (50–400 μM). Godoy et al. [1] tested the influence of atorvastatin and pravastatin (10 μM) on HL-1 cardiomyocyte mitochondrial function and could show, that atorvastatin altered mitochondrial function compared to cardiomyocytes treated with pravastatin. The di fference between our results (declined mitochondrial respiration in liver) and the other findings (unchanged mitochondrial respiration) could be caused by many factors like di fferent experimental conditions (in vitro vs. ex vivo, di fferent drug concentrations in in vitro experiments and oral pretreatment), long-term therapy vs. single dose and di fferent tissues (liver, muscle, cardiomyocytes). It is well known, that mitochondrial function varies between organs [29].

The e ffect of pravastatin on colonic mitochondria was di fferent from that in the liver. The drug moderately reduced state 3 and RCI, mainly in high concentration and preferably through complex II but increased the e fficacy of OXPHOS for both complexes. It seems to be a rather positive e ffect reflected in a higher e fficacy of OXPHOS with reduced mitochondrial respiration. In general, an increase in mitochondrial respiration is considered as an improvement and vice versa. However, mitochondria may also dynamically respond to specific conditions, and changes in oxphos capacity or RCI may simply reflect a response rather than an improvement or impairment. Under physiological conditions with su fficient oxygen supply, the impact on cell metabolism is probably of minor relevance. However, this e ffect might gain a major importance as an adaptive response under compromised conditions associated with e.g., cellular hypoxia. To clarify, whether this observation is favorable for the cell metabolism, the underlying processes like activity of the single complexes of the respiratory chain, mitochondrial membrane potential or tissue ATP-concentration need to be further analyzed.

To our best knowledge, there are no data about influence of pravastatin on mitochondrial function in the colon so we cannot refer our results to those of other authors.

For pravastatin, plasma concentrations after oral administration of 20–40 mg/day are in the range of 0.1–0.229 μM [30] which is considerably lower than those concentrations applied in our in vitro study. The oral bioavailability of pravastatin is low (17%) because of incomplete absorption and a first-pass e ffect. The drug is rapidly absorbed from the upper part of the small intestine, probably via proton-coupled carrier-mediated transport, and then taken up by the liver by a sodium-independent bile acid transporter [31,32]. The uptake of the drugs into di fferent tissues di ffers substantially and leads to wide ranges of drug concentrations in the target organs. Yamazaki et al. [33] examined the pharmacokinetic properties of pravastatin after intravenous and portal vein application in rats and could show that the largest clearance was observed for the liver, followed by the kidney whereas other tissues like small intestine exhibited only a minor uptake. Hatanaka et al. showed that pravastatin accumulated in the liver and reached even higher concentrations compared to plasma levels. Interestingly, the pravastatin plasma concentration increased with increasing dose, whereas the contrary was the case in liver and small intestine [34]. We chose relatively high drug concentration according to the similarin vitro experiments described in the literature [13,17]. Nevertheless, our results may be relevant for the in vivo situation since the most adverse events with statins have been described in patients having a drug–drug interaction leading to a higher drug concentration or underlying mitochondrial disease rendering them more sensitive to statins [35,36]. Moreover, many patients su ffer from multimorbidity. Renal and/or liver insu fficiency can also result in supraclinical or even toxic drug plasma concentrations.

The e ffects of gemfibrozil on hepatic mitochondrial respiration were similar to pravastatin but seem to be more pronounced. Gemfibrozil reduced dose dependently state 3 and RCI for both complexes and ADP/O-ratio for complex I. However, the increase in ADP/O ratio after treatment with gemfibrozil 1000 μM most likely reflects rather a sign of terminal uncoupling than improvement of e fficacy of OXPHOS. To test this hypothesis, we treated hepatic mitochondria with an uncoupler—2-[2-(3-Chlorophenyl)hydrazinylyidene]propanedinitrile (CCCP)—and there was no further enhancement in mitochondrial respiration confirming the terminal uncoupling.

Our results are consistent with those of other authors, who also observed a decrease in mitochondrial respiration and uncoupling e ffect of gemfibrozil. Nadanaciva et al. [15] showed that gemfibrozil in concentration of 500 nmol/mg mitochondrial proteins (which would correspond to 2000 μM in our experimental setting) lowered state 3 in isolated hepatic rat mitochondria about more than 50%. Zhou et al. [20] examined the e ffects of di fferent fibrates on mitochondrial bioenergetics and showed that gemfibrozil at 75 μM uncouples isolated hepatic rat mitochondria and reduces the efficacy of the OXPHOS. Zhou et al. sugges<sup>t</sup> that the underlying mechanism of uncoupling could be the induction of mitochondrial permeability transition pores.

In the colon, gemfibrozil decreased slightly ADP-dependent mitochondrial respiration, but did not a ffect the coupling between ETS and OXPHOS and improved the e fficacy of OXPHOS for complex I. Similar to pravastatin, we presume that gemfibrozil could also have a protective e ffect on colonic mitochondria, allowing an e fficient ATP-production by reduced mitochondrial oxygen consumption. Also in this case, this mechanism may be relevant under pathologic conditions like sepsis or hemorrhagic shock, but probably does not play a pivotal role when the oxygen supply is su fficient. Nevertheless, mechanisms compensating a lack of oxygen are substantially important in organs like the colon. When circulation becomes unstable, e.g., in septic shock, blood flow is redistributed to maintain the oxygenation of vital organs as heart or brain, while microcirculation in less essential organs like the splanchnic region, kidney and liver is critically reduced [37]. We are the first to analyze the e ffects of gemfibrozil on colonic mitochondrial function, so no comparison with other results can be made.

As described above, we used in our experiment higher gemfibrozil concentrations than clinically occur. After the standard oral administration of 1200 mg/day, the plasma concentration of gemfibrozil reaches 10–20 mg/<sup>L</sup> (40–80 μM) [38,39]. Also in this case we consider our result as clinically relevant, because fibrates are often combined with other lipid-lowering drugs like statins or thiazolidinediones and the combination can lead to higher plasma levels. Moreover, the clinically used drug combinations, like statins plus fibrates, show di fferent e ffects to the single application. In our experiments, pravastatin and gemfibrozil showed similar negative e ffects on mitochondrial respiration in liver and positive influence on the colonic mitochondria. So it is conceivable, that additive e ffects are observed with co-incubation. Nadanaciva et al. showed that gemfibrozil at 62 nmol/mg protein (which would correspond to 248 μM in our experimental setting) did not a ffect mitochondrial function, but in combination with cerivastatin depressed the mitochondrial respiration significantly [15].
