Sepsis Impairs Thyroid Hormone Signaling and Mitochondrial Function in the Mouse Diaphragm

Abstract

Background:

Sepsis can cause the nonthyroidal illness syndrome (NTIS), resulting in perturbed thyroid hormone (TH) signaling and reduced thyroxine (T4) levels. TH is a major regulator of muscle function, via its influence on mitochondria. This study aimed at evaluating the relationship between TH signaling, mitochondrial function, and the antioxidant defense system in the diaphragms of septic mice.

Methods:

Male C57Bl/6 mice were divided into two groups: cecal ligation and puncture (CLP) and sham. Twenty-four hours after surgery, plasma, diaphragms, and livers were collected. TH metabolism and responses were analyzed by measuring messenger RNA (mRNA) expression of Dio1 in the liver, and Thra, Thrb, Dio2, Slc16a10, and Slc16a2 (encodes MCT 10 and 8), in the diaphragm. T4 plasma levels were measured by radioimmunoassay. Damage to diaphragm mitochondria was assessed by electron microscopy and real-time polymerase chain reaction (qPCR), and function with oxygraphy. The diaphragm antioxidative defense system was examined by qPCR, analyzing superoxide dismutase (SOD) 1 (Sod1), mitochondrial superoxide dismutase (SOD 2; Sod2), extracellular superoxide dismutase (SOD 3; Sod3), glutathione peroxidase 1 (Gpx1), and catalase (Cat) expression. The effect of TH replacement was tested by treating the mice with T4 and triiodothyronine (T3) (CLP+TH) after surgery.

Results:

CLP mice presented reduced total plasma T4 concentrations, downregulated Dio1, and upregulated Il1b mRNA expression in the liver. CLP mice also displayed downregulated Thra, Thrb, Slc16a10, and Slc16a2 expression in the diaphragm, suggesting that TH signaling was compromised. The expression of Ppargc1a (encoding PGC1a) was downregulated, which correlated with the decrease in the number of total mitochondria, increase in the percentage of injured mitochondria, downregulation of respiratory chain complex 2 and 3 mRNA expression, and reduced maximal respiration. In addition, septic animals presented a three-fold increase in Ucp3 and G6pdh expression; downregulated Sod3, Gpx1, and Cat expression; and upregulated Sod2 expression, potentially due to elevated reactive oxygen species levels. The mitochondrial number and the percentage of injured mitochondrial were similar between sham and CLP+TH mice.

Conclusions:

Sepsis induced responses consistent with NTIS, resulted in mitochondrial damage and functional impairment, and modulated the expression of key antioxidant enzymes in the diaphragm. Thus, impaired diaphragm function during sepsis seems to involve altered local TH signaling, mitochondrial dysfunction, and oxidative stress defense.

Color images are available online.

Introduction

Sepsis is a life-threatening condition associated with an unbalanced host response to infection (1). Septic patients frequently require expensive and intensive treatments, thus making sepsis a major public health concern (2 –4). It is also known that sepsis induces the nonthyroidal illness syndrome (NTIS) (5), and the severity of NTIS is associated with the mortality rate due to sepsis (6,7).

NTIS is characterized by a decrease in peripheral thyroid hormone (TH) serum levels while thyrotropin levels remain unchanged (8). NTIS can be viewed as part of the inflammatory response, and the changes in serum TH concentrations have been shown to modulate central and peripheral TH signaling and metabolism (8,9). Septic patients with lower triiodothyronine (T3) and thyroxine (T4) serum levels have a poor prognosis for a favorable outcome (5,10). In our previous work, it was found that sepsis, induced by cecal ligation and puncture (CLP) surgery, led to changes in TH serum levels and TH metabolism in the liver, which are also commonly observed during NTIS (11).

Further, in some cases, septic patients require mechanical ventilation, which can result in injury of the diaphragm. Further, sepsis can induce diaphragm weakness independent of mechanical ventilation (12). The mortality rate is greater in patients with increased diaphragm weakness, when compared with patients with a similar disease severity and without this condition (12). Currently, the mechanisms involved in septic-mediated diaphragm impairment are not known, and they require clarification.

Skeletal muscle function is directly and indirectly related to mitochondrial biogenesis and energy production (13). Further, TH is crucial to normal skeletal muscle physiology, since it is required for tissue regeneration and differentiation and is involved in the regulation of mitochondrial function. In addition, TH has been shown to be associated with the regulation of sarcolemmal Ca⁺² concentrations (14). Interestingly, stimulation of a muscle cell line with lipopolysaccharide, a bacterial endotoxin, decreased mitochondrial function in a TH-dependent manner (15). Moreover, our group previously demonstrated that sepsis, induced by S. pneumoniae infection, decreased the responsiveness of the diaphragm to TH (15).

We, therefore, hypothesized that altered TH signaling induces changes in mitochondrial function and, consequently, impairs the diaphragm function of septic animals. The impact of sepsis on TH signaling, mitochondrial function, and antioxidant defense in the diaphragm were investigated by using real-time polymerase chain reaction (qPCR), electron microscopy, and oxygraphy in a CLP-induced sepsis model.

Materials and Methods

Ethics statement and animal experimentation

The project was approved by the local Animal Care Committee at the Federal University of Rio de Janeiro (CEUA-088/15), and it was registered with the Brazilian National Council for Animal Experimentation Control. Animal care was in compliance with the “Guide for the Care and Use of Laboratory Animals” established by the U.S. National Academy of Sciences (16).

Sepsis was induced by CLP and control C57Bl6 male mice that were two to three months old, weighing 24–26 g, and that were subjected to sham surgery, as previously described (11). Briefly, mice were anesthetized with ketamine (100 mg/kg; Cristália, Itapira, Brazil) and xylazine (10 mg/kg; Syntec) i.p. After anesthesia, the lower abdomen area was disinfected. Longitudinal incisions were performed in the lower quadrant of the abdomen and peritoneal cavity with scissors. Later, the cecum was located and exteriorized. The cecal ligation was performed below the ileocecal valve by using a cotton suture line. The cecum was perforated once between the ligation and the end of the cecum by using a 18G needle. A small amount of fecal material was exposed before reinsertion of the cecum in the abdominal cavity. The fascia, abdominal musculature, peritoneum, and skin were closed via simple running sutures by using a nylon suture line. Subsequently, 1 mL of sterile saline was administrated subcutaneously for volume substitution. The sham-operated mice (control) underwent the same procedure of CLP mice, excluding cecal ligation and perforation. Animals were housed in a temperature control environment (22°C) with a 12-h light/dark cycle and food and water ad libitum.

At 8 and 23 hours after CLP or sham operation, mice were scored for severity of sepsis (17). The following variable were evaluated: piloerection, alterations in gait, lethargy, alterations in respiratory rate, lacrimation, grip strength, body tone, respiration after manipulation, lack of exploratory behavior, and body temperature alterations; if present, the animal receives one point. The sum of total score reflected the severity of sepsis. Score 2–3 points mild sepsis, 4–7 moderated sepsis, and 8–10 severe sepsis. In our experimental condition, most of the CLP animals were ranked as moderate sepsis.

A total of 15 control (sham), 26 CLP mice were used; 8 CLP mice were treated with TH (CLP+TH); and six CLP and two CLP+TH mice died 6 and 18 hours after the surgery and were excluded from the experiments. TH (250 ng T3/g body weight plus 7.5 ng T4/g body weight in a volume of 100 μL; both from Sigma Aldrich) was administrated subcutaneously immediately, four and eight hours after the surgery, and control mice received saline. For each analysis, the number of animals used are indicated in the legends.

Animals were sacrificed under anesthesia with isoflurane 1.5–2.0% (Isoforine; Cristália), 24 hours after surgery between 9 and 11 am, to avoid changes in hormonal and gene expression due to circadian rhythms. Blood, liver, and diaphragm were collected, immediately frozen in liquid nitrogen, and stored at −70°C, until use. Freshly dissected diaphragms were used for the mitochondrial studies. Hormone determinations were performed with blood serum that was stored at −20°C.

Plasma total T4 measurement

Plasma total T4 concentration was measured by using a commercial radioimmunoassay kit (MP Biomedicals, CA), according to the manufacturer's recommendations. The limit of detection for T4 was 13 μM.

Total RNA extraction and qPCR

Total RNA from liver samples was extracted by using the TRIzol method, according to the manufacturer's protocol (TRIzol Reagent; Life Technologies, CA). Total RNA from diaphragm samples was extracted by using a combination of the TRIzol Reagent (Invitrogen, Carlsbad) and the Nucleospin RNA II kit (Macherey Nagel, Düren, DE). For cDNA synthesis, 800 ng of total RNA was reverse transcribed by using the High-Capacity cDNA Reverse Transcription Kit (Applied Biosystems, CA), according to the manufacturer's protocol. After cDNA synthesis, messenger RNA (mRNA) expression was evaluated by qPCR, using the HOT FIREPol Evagreen qPCR Supermix (Solis, Denmark) and the Master Cycler Realplex system (Eppendorf, Germany).

The sequences for each primer pair are shown in Table 1, in summary for analyses TH responsiveness and metabolism: Dio1, Dio2, Thra1, Thrb1, Slc16a2, Slc16a10, Tnni2, and Hr; for systemic inflammation: Il1b; to investigate mitochondrial physiology: Ppargc1a, Opa1, and Dnm1l; Ucp3; G6pdh to analyze the respiratory chain: Ndufb8, Sdhb, Uqcrc2, and Cox4i1; to investigate reactive oxygen defenses: superoxide dismutase (SOD) 1 (Sod1), mitochondrial SOD2 (Sod2), extracellular SOD3 (Sod3), glutathione peroxidase 1 (Gpx1), and catalase (Cat). The reference genes are listed next. Quantification of mRNA expression was calculated by using the standard curve method and corrected by the geometric mean of the following reference genes: Phctr and Pib for the diaphragm (control 358.9 ± 26.7 and septic 388.8 ± 71.6, p = 0.6), Gapdh and Rpl0 for the liver (control 1032 ± 220 and septic 1115 ± 131, p = 0.6), and Vdac for mitochondrially encoded genes (control 1633 ± 94 and septic 1239 ± 262, p = 0.2).

Table 1.

Primer Pair's Sequences

Gene	Primer sequences	GenBank accession no.	Annealing temperature
Phactr4	ACTTTACACGCTACCATCGCCCAT	NM_175306	65°C
Phactr4	AAGGGGAGCACAAGGACACG	NM_175306	65°C
Ppib	GAGACTTCACCAGGGG	NM_011149	60°C
Ppib	CTGTCTGTCTTGGTGCTCTCC	NM_011149	60°C
Vdac	GGGCTGACGTTTACAGAGAAG	NM_001362693	60°C
Vdac	CTCATAGCCAAGCACCAGAGC	NM_001362693	60°C
Gapdh	GTAACCAGGCGTCCGATAC	NM_008084	60°C
Gapdh	TCTCTGCTCCTCCCTGTTC	NM_008084	60°C
Rpl0	GGCCCTGCACTCTCGCTTTC	NM_007475.5	60°C
Rpl0	TGCCAGGACGCGCTTGT	NM_007475.5	60°C
G6pdx	CATGAGTCAGACAGGCTGGA	NM_008062.2	60°C
G6pdx	GATCTGGTCCTCACGGAAAA	NM_008062.2	60°C
Ppargc1a	CAATGAATGCAGCGGTCTA	NM_008904.2	60°C
Ppargc1a	GTGTGAGGAGGGTCATCGTT	NM_008904.2	60°C
Ucp3	GAGATGGTGACCTACGACATCA	NM_009464.3	60°C
Ucp3	GCGTTCATGTATCGGGTCTTTA	NM_009464.3	60°C
Opa1	TGGAAAATGGTTCGAGAGTCAG	NM_001199177.1	60°C
Opa1	CATTCCGTCTCTAGGTTAAAGCG	NM_133752.3	60°C
Dnm1l	TAAGCCCTGAGCCAATCCATC	NM_001360010.1	60°C
Dnm1l	CATTCCCGGTAAATCCACAAGT	NM_001360010.1	60°C
Ndufb8	TGTTGCCGGGGTCATATCCTA	NM_026061	60°C
Ndufb8	AGCATCGGGTAGTCGCCATA	NM_026061	60°C
Sdhb	AATTTGCCATTTACCGATGGGA	NM_001355515	60°C
Sdhb	AGCATCCAACACCATAGGTCC	NM_001355515	60°C
Uqcrc2	TCTCTGGAAAACTATGCTCCTCT	NM_025899.2	60°C
Uqcrc2	AAATGTGAGGTTCCCAAGTTGT	NM_025899.2	60°C
Cox4i1	TCCCCACTTACGCTGATCG	NM_009941	60°C
Cox4i1	GATGCGGTACAACTGAACTTTCT	NM_009941	60°C
Sod1	TATGGGGACAATACACAAGGCT	NM_011434	60°C
Sod1	CGGGCCACCATGTTTCTTAGA	NM_011434	60°C
Sod2	CCAAGGGAGATGTTACAACTCAG GGGCTCAGGTTTGTCCAGAA	NM_013671	60°C
Sod3	CCTTCTTGTTCTACGGCTTGC	NM_011435	60°C
Sod3	GCGTGTCGCCTATCTTCTCAA	NM_011435	60°C
Gpx1	AATGTCGCGTCTCTCTGAGG	NM_008160	60°C
Gpx1	TCCGAACTGATTGCACGGG	NM_008160	60°C
Cat	CCCCTATTGCCGTTCGATTCT	NM_009804	60°C
Cat	TTCAGGTGAGTCTGTGGGTTT	NM_009804	60°C
Il1b	TTGACGGACCCCAAAAGATG	NM_008361	60°C
Il1b	AGAAGGTGCTCATGTCCTCA	NM_008361	60°C
Dio1	AGCCAGCTCTACGCGGC	NM_007860.4	60°C
Dio1	CCCTTGTAGCAGATCCTGCC	NM_007860.4	60°C
Dio2	GCTTCCTCCTAGATGCCTACAA	NM_010050.3	60°C
Dio2	CCGAGGCATAATTGTTACCTG	NM_010050.3	60°C
Thra1	CATCTTTGAACTGGGCAAGT	NM_001313983	60°C
Thra1	CTGAGGCTTTAGACTTCCTGATC	NM_001313983	60°C
Slc16a2	GAGCAGAGAGATTCCAGCAAGG	NM_009197	60°C
Slc16a2	GTAGGATGAGGGGTGGAGTGG	NM_009197	60°C
Thrb	AACCAGTGCCAGGAATGTCG	NM_001113417.1	60°C
Thrb	CTCTTCTCACGGTTCTCCTC	NM_009380.3	60°C
Slc16a10	TGTGGGATTCATTGGACTCATGTAGTGACGGCTGGTAGGCAAAG	NM_001114332.1	60°C
Slc16a10	TGTGGGATTCATTGGACTCATGTAGTGACGGCTGGTAGGCAAAG	NM_028247.4	60°C
Tnni2	CGGAGGGTGCGTATGTCTG	NM_009405.3	60°C
Tnni2	CAGGTCCCGTTCCTTCTCA	NM_009405.3	60°C
Hr	GCAGGTAGTAGAACGGAAGATCC	NM_021877.3	60°C
Hr	CTGGTGACAATGGAAGGCTCA	NM_021877.3	60°C
Klf9	TTA TTG CAC GCT GGT CAC TAT C	NM_010638.5	60°C
Klf9	CTC ATC GGG ACT CTC CAG AC	NM_010638.5	60°C

The best reference genes were selected according to their Cq, amount values, low variance, and no significant differences between groups. The qPCR efficiency was between 90% and 100%. PCR cycles were as follows: denaturation for 12 minutes at 95°C, 40 cycles of 15 seconds at 95°C, 30 seconds of annealing temperature (65°C or 60°C), 30 seconds at 72°C, and the following melting program. Genomic DNA contamination and qPCR quality were evaluated by using intron-spanning primers, reverse transcriptase-negative samples, and melting curve analyses from each reaction.

Skeletal muscle high-resolution respirometry

Three milligrams of diaphragm was washed in chilled BIOPS buffer (2.77 mM CaK₂EGTA, 7.23 mM K₂EGTA, 20 mM imidazole, 20 mM taurine, 6.56 mM MgCl₂, 5.77 mM ATP, 15 mM phosphocreatine, 0.5 mM dithiothreitol, and 50 mM K-MES, pH 7.1); the tissues were permeabilized with cold BIOPS buffer supplemented with 50 μg/mL saponin, for 40 minutes with gentle agitation. The samples were then transferred to Mir05 buffer (110 mM sucrose, 60 mM K-MES, 0.5 mM EGTA, 3 mM MgCl₂, 20 mM taurine, 10 mM KH₂PO₄, 20 mM K-HEPES, and 1 g/L bovine serum albumin, pH 7.1) and maintained in this buffer for at least 10 minutes under mild agitation.

Oxygen consumption was measured with an O2K (Oroboros Instruments GmbH, Innsbruck, Austria) in Mir05 buffer. Basal O₂ consumption was recorded for at least five minutes followed by drug administration in the following order: 5 mM pyruvate and 5 mM malate, 3 mM ADP, 10 μM cytochrome c, 10 mM succinate, two pulses of 1 μg/mL oligomycin, 600 nM Carbonyl cyanide-4-(trifluoromethoxy) phenylhydrazone, and 10 mM KCN. All data were recorded after O₂ consumption stabilized.

Transmission electron microscopy and mitochondrial quantification

Fragments of the diaphragm were processed and analyzed qualitatively and quantitatively by using a JEM1011 microscope (JEOL). Briefly, the fragments were washed in PBS, fixed in 0.1 M sodium cacodylate buffer (pH 7.2) containing 2.5% glutaraldehyde for 24 hours, and washed three times, 10 minutes each, in the same buffer. The tissue was postfixed in 0.1 M sodium cacodylate buffer (pH 7.2) containing 1% osmium tetroxide (OsO₄), dehydrated with increasing acetone concentrations (30%, 50%, 70%, 90%, and 100%), and embedded in Poly/Bed812 resin (Ted Pella, Inc., Redding, CA).

After polymerization, ultrathin sections were obtained and contrasted with uranyl acetate-lead citrate. For ultrastructural evaluations, sections were collected in three grids with a distance of 7 μm between them. A total of 10 images (70 nm) were acquired from random fields in each grid. The total number of mitochondria, as well as the number of healthy and injured mitochondria were quantified to obtain the percentage of injured mitochondria (18). Digital micrographs were captured by using an AMT XR80 CCD digital camera (Advanced Microscopy Techniques, Woburn, MA, EUA) at 12,000 × magnification

Statistical analysis

The Kolmogorov–Smirnov test was used to determine data normality. If the data were not normally distributed, then differences between groups were analyzed by using the Mann–Whitney U test. Alternatively, if the data were normally distributed, then the Student's t-test was used to analyze differences between groups. Data outliers were identified by using Grubbs' test, which only identifies one outlier per group using alpha 0.05. Chi-square tests were performed between the CPL and CLP+TH groups comparing score 0–1 and 4–7. Results are shown as the mean values ± SD of at least four animals per group. Statistical analyses were performed by using the GraphPad prism 5 software (GraphPad Software, Inc., San Diego, CA). Differences were considered to be significant at p < 0.05.

Results

Sepsis due to CLP induces NTIS

The CLP procedure decreased total serum T4 levels by 33% (p < 0.01) when compared with sham-operated control animals (Fig. 1A). In addition, liver Dio1 mRNA expression was decreased by 42% (Fig. 1B, p < 0.05), while liver ll1b mRNA expression increased eight-fold (Fig. 1C, p < 0.01), when compared with controls. Liver Il1b is an acute-phase response marker and a T4 decrease associated with a decrease in liver Dio1 are hallmarks of NTIS. Thus, CLP induced systemic inflammation and NTIS.

FIG. 1.

Serum T4 levels and gene expression confirm that acute CLP is a nonthyroidal illness syndrome model. Total serum T4 (A, sham n = 15, and CLP n = 11). Liver Dio1 (B, sham n = 5, and CLP n = 5) and Il1b (C, sham n = 5, and CLP n = 5) mRNA expression. Expression of Dio2 (D, sham n = 7, and CLP n = 6); Thra1 (E, sham n = 6, and CLP n = 5), Thrb (F, sham n = 12, and CLP n = 9), Slc16a2 (G, control n = 8, and CLP n = 7), and Slc16a10 (H, control n = 11, and CLP n = 9) in the diaphragm. Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. Values are expressed as mean ± SD. CLP, cecal ligation and puncture; mRNA, messenger RNA; T4, thyroxine.

Expression of Thra1 and Slc16a2 is reduced during sepsis in the diaphragm

Based on a previous study, demonstrating the involvement of Thra1, Thrb, Dio2, Slc16a10, and Slc16a2 in TH signaling (19), the expression of these genes was investigated in the diaphragms of septic and control mice. As shown in Figure 1D–H, the expression of Thra1, Thrb, Slc16a10, and Slc16a2 was reduced by 78% (p < 0.05), 71% (p < 0.0001), 70% (p < 0.05), and 55% (p < 0.0001), respectively, in the septic diaphragm, while Dio2 expression was unaffected. In summary, skeletal muscle TH receptors (Thra1 and Thrb) and transporters (Slc16a10 and Slc16a2) were decreased, suggesting a decreased responsiveness to TH action in the diaphragm. In addition, liver Dio1 is highly sensitive to TH and can be used to evaluate TH signaling. Thus, the investigation of TH responsive genes could help to elucidate this hypothesis.

Sepsis affects mitochondrial function

Ppargc1a is a T3 responsive gene of interest. Interestingly, the protein encoded by the Ppargc1a gene, PGC1α, is important for regulating transcription and stimulating mitochondrial biogenesis and function (14). In addition to other tissues, Ppargc1a is also expressed in skeletal muscle and has been used to confirm reduced TH signaling (20). Ppargc1a expression was reduced by 90% in the septic mice, when compared with controls (Fig. 2A, p < 0.01).

FIG. 2.

Mitochondrial dynamics in the diaphragm. Expression of Ppargc1a (A, sham n = 6, and CLP n = 5), Opa1 (B, sham n = 6, and CLP n = 5), and Dnm1l (C, sham n = 6, and CLP n = 5) in the diaphragm. Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. Values are expressed as mean ± SD.

Given the reduced Ppargc1a mRNA expression levels in the diaphragm, markers of mitochondrial fusion (Opa1) and fission (Dnm1l) processes were also investigated, so as to characterize mitochondrial physiology in more detail. No significant differences were observed for the expression of Opa1 (Fig. 2B) or Dnm1l (Fig. 2C) in the diaphragms of septic or control animals.

On the other hand, the number of mitochondria was significantly reduced in the septic group, when compared with controls (Fig. 3A–E, p = 0.01). In addition, there was a detectable increase in the size of the mitochondria from septic animals; the increased organelle size was caused by mitochondrial swelling (Fig. 3A–D, F, p = 0.01). In fact, mitochondrial swelling is associated with reduced organelle function and increased permeability (21). Indeed, when quantifying mitochondrial damage in the diaphragm of septic and control mice, it was found that a greater percentage of injured organelles were detected in the septic animals (Fig. 3G).

FIG. 3.

Mitochondrial quantification in the diaphragm. TEM photomicrographs of sham (A, C) and CLP (B, D) diaphragm fragments. White arrows indicate healthy mitochondria, and black arrows indicate injured mitochondria. Quantification of total mitochondria number (E), area (F), and percentage of injured mitochondria (G). Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. n = 3. Values are expressed as mean ± SD. TEM, transmission electron microscopy.

As shown in Figure 4, the physiology of mitochondria from the permeabilized diaphragms of septic and control mice was measured by HRR (Fig. 4A). Basal rates of respiration, considering complex I substrate and oxygen consumption coupled to ADP phosphorylation after succinate addition, did not differ between septic and control mice (Fig. 4B, C). On the other hand, the maximal respiration (in the presence of FCCP) was significantly reduced by 31% in the septic group (Fig. 4D, p = 0.04). In line with these results, mRNA expression levels of genes encoding for proteins related to complex II and III (Sdhb and Uqcrc2, respectively), in the respiratory chain, were clearly downregulated (Fig. 5).

FIG. 4.

Diaphragm mitochondrial function. HRR in the diaphragm. (A) Solid lines represent oxygen flux (Black—Sham, n = 7/gray—CLP, n = 5), and dotted lines correspond to oxygen concentration (Black—Sham/gray—CLP). Numbers indicates drug addition: [1] pyruvate and malate (5 mM/5 mM), which make it possible to analyze basal oxygen consumption, [2] ADP (3 mM) used to investigate ATP synthase function, [3] cytochrome c (10 μM) as control of mitochondrial integrity and permeability, [4] succinate (20 mM) to stimulate mitochondria complex II activity, [5] oligomycin (2 mg/mL) used to inhibit ATP synthase activity, [6] FCCP (600 nM) mitochondrial uncoupler for measuring maximal oxygen consumption capacity, and [7] KCN (20 mM) inhibits respiration. (B) Basal respiration: O₂ consumption after pyruvate and malate administration. (C) oxygen consumption coupled to ADP phosphorylation: It was calculated by subtracting O₂ consumption after succinate administration values from O₂ consumption after oligomycin administration values; the data are presented related to pyruvate and malate O₂ consumption. (D) maximal respiration (in presence of FCCP): O₂ consumption after FCCP administration minus O₂ consumption after oligomycin administration related to pyruvate and malate O₂ consumption values. Expression of Ucp3 (E, sham n = 7, and CLP n = 5) and G6pdh (F, sham n = 6, and CLP n = 5) mRNA in the diaphragm. Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. Values are expressed as mean ± SD.

FIG. 5.

Mitochondrial respiratory chain. Expression of complex I (Ndufb8—A, sham n = 5, and CLP n = 5); II (Sdhb—B, sham n = 7, and CLP n = 8); III (Uqcrc2—C, sham n = 6, and CLP n = 6); and IV (Cox4i1—D, sham n = 6, and CLP n = 5) in the diaphragm. Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. Values are expressed as mean ± SD.

The decrease in maximal respiration was not associated with a significant difference in oxygen consumption coupled to ADP phosphorylation. This could be related to a change in the proton gradient. As proton leaking is commonly associated to uncoupling protein activity, we investigated the expression of the premier uncoupling protein in skeletal muscle UCP3 (Ucp3). In addition, Ucp3 mRNA expression in the diaphragm was increased three-fold in the septic mice when compared with controls (Fig. 4E, p = 0.006).

The increase in UCP3 would suggest increased uncoupling, not observed in the HHR. UCP3 is also related to reactive oxygen species (ROS) defense (22). Thus, we suggest that the impact on UCP3 could be linked to ROS. In view of this possibility, we analyzed another gene related to oxidative phosphorylation and ROS, G6pdh, and found a significant increase in G6pdh mRNA expression in the septic diaphragm (Fig. 4F, p = 0.007).

Sepsis modulates ROS defenses in the diaphragm

To determine whether sepsis alters the expression of genes involved in defending the cell against ROS, the mRNA expression of soluble Sod1, Sod2, Sod3, Gpx1, and Cat was analyzed. Although Sod1 expression was unchanged (Fig. 6A), the expression levels of Sod3, Gpx1, and Cat were significantly decreased in the diaphragm of septic mice (Fig. 6C–E, p = 0.02). In contrast, Sod2 expression was three-fold higher in septic animals (Fig. 6B, p = 0.0015).

FIG. 6.

Diaphragm reactive oxygen species defense and generation. Diaphragm Sod1 (A, sham n = 5, and CLP n = 7); Sod2 (B, sham n = 6, and CLP n = 7); Sod3 (C, sham n = 6, and CLP n = 6); Gpx1 (D, sham n = 9, and CLP n = 7); and Cat (E, sham n = 5, and CLP n = 5) mRNA expression. Black bar: CLP—cecal ligation puncture technique, white bar: sham—caused by surgical stress. Values are expressed as mean ± SD. Cat, catalase; Gpx1, glutathione peroxidase 1; Sod1, superoxide dismutase (SOD) 1; Sod2, mitochondrial superoxide dismutase SOD2; Sod3, extracellular superoxide dismutase SOD3.

TH replacement therapy improves mitochondrial quality

To evaluate whether TH administration could affect sepsis outcome and improve mitochondrial physiology, we investigate the survival and the sepsis severity (Table 2). After TH treatment, the percentage of animals with no signs of distress was 33.3% in the CLP+TH group compared with 8.7% in the CLP group. The comparison between the animals with no signs of distress (score 0–1) with moderate sepsis (score 4–7) in the CLP and CLP+TH groups did not reach significance (p-value of 0.1). TH treatment did not change the percentage of animal survival (CLP 79% and CLP+TH 75%, Table 2).

Table 2.

Sepsis Survival Rate and Severity

	Mice (total/died)	Score 0–1	Score 2–3 (mild sepsis)	Score 4–7 (moderate sepsis)	Score 8–10 (severe sepsis)
Sham	15/0	15/15 (100%)	0/15	0/15	0/15
CLP	29/6	2/23 (8.7%)	3/23 (13%)	17/23 (73.9%)	1/23 (4.3%)
CLP+TH	8/2	2/6 (33.3%)	0/6	3/6 (50%)	1/6 (16.6%)

CLP, cecal ligation and puncture; TH, thyroid hormone.

T3 responsiveness was investigated by determining the expression of Tnni2, Klf9, and Hr, a positively regulated gene. CLP significantly decreased Tnni2 and Hr mRNA expression in the diaphragm compared with sham-treated mice. However, Tnni2 expression in the CLP+TH was not significantly different compared with sham or CLP animals. It is important to notice that Klf9 and Hr expression increased in the CLP+TH group compared with controls (Supplementary Fig. S1).

Next, we investigated the Ppargc1a mRNA expression levels in the diaphragm, the total mitochondrial number, and the percentage of injured mitochondria (Fig. 7). The CLP-induced decrease of Ppargc1a expression is not different in TH-treated mice versus untreated mice (Fig. 7D). However, the impact of sepsis on the total number of mitochondria and percentage of healthy organelles improved after TH treatment (Fig. 7A–C, E, F). In addition, we investigated whether TH treatment would improve complex II and III expression, which was downregulated by sepsis. TH treatment in CLP mice did not change Sdhb (complex II) and Uqcrc2 (complex III) mRNA expression compared with CLP mice and the sham group (Fig. 7G, H).

FIG. 7.

TH replacement improves mitochondrial parameters. TEM photomicrographs of sham (A), CLP (B) and CLP+TH (C) diaphragm fragments, white arrows indicate healthy mitochondria, and black arrows indicate injured mitochondria. Expression of Ppargc1a (D) in the diaphragm. Quantification of total mitochondria number (E) and percentage of injured mitochondria (F). Expression of complex II (Sdhb—G) and III (Uqcrc2—H). Black bar: CLP—cecal ligation puncture technique, white bar: sham due to surgical stress, gray bar: CLP mice treated with T3 250 ng/g body weight plus T4 7.5 ng/g body weight sc immediately, four and eight hours after the surgery. For Ppargc1a expression sham n = 9, CLP n = 8, and CLP+TH = 6. n = 5 for TEM in all groups. Values are expressed as mean ± SD. T3, triiodothyronine; TH, thyroid hormone.

Discussion

This study demonstrates that CLP-induced sepsis promotes the development of NTIS since we observed a decrease in T4 serum levels and liver Dio1 expression, hallmarks of NTIS, in our murine model (Fig. 1). In addition, CLP-induced sepsis impairs TH signaling in the diaphragm (Fig. 2 and Supplementary Fig. S1). The results also demonstrate a decrease in mitochondrial function and an imbalance in the ROS defense system in the diaphragms of septic mice (Figs. 3–6). TH treatment in CLP mice increased the expression of Hr and Klf9, T3 responsive genes that are positively regulated (Supplementary Fig. S1), and improved mitochondrial parameters (Fig. 7).

Taken together, the results suggest that the decrease in TH responsiveness in the diaphragm impairs mitochondrial function, and it reduces oxidative phosphorylation capacity and ROS defense. It is plausible that both factors are related to diaphragm weakness, which is often present in critically ill septic patients.

Normal muscle TH levels are necessary for stimulating the function and activity of skeletal muscle mitochondria (14). For example, TH has been implicated in the stimulation of oxidative phosphorylation, mitochondrial biogenesis fission, and mitophagy (23). The reduction in total plasma T4 levels and the observed decreases in Thra1, Thrb, Slc16a10 (MCT10), and Slc16a2 (MCT8) expression levels suggest that the TH availability of the diaphragm is reduced. Further, it is tempting to speculate that the observed decrease in TH availability could perturb and/or impair mitochondrial function in the diaphragm.

T3 treatment did not improve mechanical ventilation parameters in patients admitted to the ICU with NTIS (24). However, our data demonstrate improvement of mitochondrial parameters after treatment with a combination of T4 and T3 starting at the beginning of sepsis (Fig. 7). This is in contrast to the study of Bello et al. (24), who started treatment when respiratory dysfunction was already established. However, it is important to point out that our survival data do not suggest an improvement in sepsis outcomes (CLP 75%, CLP+TH 79%; Table 2). Thus, TH treatment can improve parameters associated with diaphragm physiology, but it may not improve survival in septic patients.

The oxidation of glucose to obtain energy leads to pyruvate formation in the cytoplasm, which is completely oxidized inside the mitochondria via the citric acid cycle. Concomitantly, nicotinamide adenine dinucleotide and flavin adenine dinucleotide are reduced but the electron transfer chain promotes their re-oxidation. As a consequence, these redox processes provide energy, in the form of a proton gradient across the inner mitochondrial membrane, which defines the mitochondrial membrane potential. Thus, ATP synthesis is coupled to the return of protons to the mitochondrial matrix through the FoF1-ATPase enzyme, which then uses the proton motive force to synthesize ATP (25).

The observed decrease in the number of mitochondria (Fig. 3A–E) is probably a consequence of increased mitochondrial damage (Fig. 3G) and both phenomena may lead to decreased maximal respiration (Fig. 4D). To better understand the mechanism involved in the decreased maximal phosphorylation, we investigated the genes associated with the four complexes of the electron transfer chain. There was a downregulation in Sdhb (complex II) and Uqcrc2 (complex III) expression in the diaphragms of septic mice. Complexes II and III are critical for maintaining electron flux in the respiratory chain and oxygen consumption, and the observed decrease in gene expression suggests that sepsis impairs diaphragm mitochondria physiology. TH treatment improved Sdhb and Uqcrc2 expression, suggesting that sepsis may affect mitochondrial respiration via reduction of TH signaling.

Mitochondrial uncoupling also decreases the proton gradient across the inner mitochondrial membrane. The premier uncoupling protein in skeletal muscle is UCP3. The observed increase in UCP3 expression in the skeletal muscle of septic animals is in agreement with previous studies (26,27). The role of the UCP3 uncoupling protein is controversial in the literature (28 –30). Most of the authors consider UCP3 to be a mitochondria uncoupling protein with low activity (22). However, it is possible that increased UCP3 levels are reducing mitochondrial coupling, which can also influence the oxidative phosphorylation capacity. The decrease in maximal phosphorylation observed in the CLP group is not likely to be only associated with the change in skeletal muscle UCP3 expression. Previous studies with septic Ucp3(−/−) mice showed that these mice did not present any differences in the mitochondrial membrane potential, when compared with septic wild-type mice, and the data presented here do not show differences in oxygen consumption coupled to ADP phosphorylation between groups (Fig. 4C) (22,31). Previous studies have also suggested that the upregulation of UCP3 in muscle is part of a defense mechanism employed to protect the cell against oxidative damage (32,33). To further investigate mitochondrial ROS generation and defense recruitment, we measured G6pdh expression, which is linked to increased mitochondrial ROS (34), and the mitochondrial superoxide dismutase (Sod2). In addition to the increase in Ucp3 expression, there was also an increase in G6pdh and Sod2 expression, and a decrease in Sod1, Sod3, Gpx1, and Cat expression in the diaphragms of septic mice. The upregulation of Ucp3, G6pdh, and Sod2 suggests that there is an increase in mitochondrial ROS production. It should also be noted that the downregulation of genes encoding for cytoplasmic antioxidant enzymes (Sod3, catalase, and glutathione peroxidase) could be related to the observed increase in cellular damage susceptibility.

It has been demonstrated that TH augments the expression of Ppargc1a and that the inflammatory response inhibits Ppargc1a expression (35 –37). The gene product, PGC1α, is primarily responsible for mitochondrial biogenesis and ROS defense (38 –40). The reduction in Ppargc1a expression in the diaphragm of septic animals could be due to increased proinflammatory cytokine expression or reduced TH action. We previously showed that stimulation of a muscle cell line with a bacterial endotoxin decreased mitochondrial function, which was associated with reduced Dio2 expression (15). Unfortunately, we could not measure diaphragm TH levels and, therefore, could not demonstrate a causal role for decreased TH concentrations in the Ppargc1a decrease in the diaphragm. Since PGC1α is capable of stimulating Sod1, Sod2, Sod3, Cat, and Gpx1 expression (37,41), the observed decrease in Sod3, Gpx1, and Cat expression is consistent with the observed decrease in Ppargc1a expression, and might be an indirect effect of TH. The increase in Sod2 expression could also be a compensatory mechanism used to protect cells from mitochondrial ROS-mediated damage.

This study demonstrates that acute sepsis promotes the development of NTIS and decreases the TH availability in the mouse diaphragm. This was accompanied by an imbalance in the antioxidant defense system, which appears to have severely impacted the mitochondrial physiology in the diaphragm. In conclusion, the results provide evidence that maintenance of the mitochondrial physiology in the diaphragm during sepsis is associated with TH response.

Footnotes

Acknowledgments

The authors thank the Biology Institute from the Fluminense Federal University for allowing them to use the transmission electron microscope (JEM 1011), and they acknowledge technical support from the biologist Alan Moraes, during the image acquisition. They are also very grateful to Juliana Pena Gonçalvez and Hannaily Ribeiro, for basic research support.

Author Disclosure Statement

The authors declare that there are no conflicts of interest.

Funding Information

This research was supported by the Brazilian Society of Endocrine and Metabolism—Thyroid department (SBEM), Fundação Carlos Chagas Filho de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ, CNE 2015 E26/203.190/2015 and CNE E26/203.095/2016), Conselho Nacional de Pesquisa e Desenvolvimento (CNPq, 304667/2016-1), and Coordenação de Aperfeiçoamento Pessoal de Nível Superior (CAPES, 01). F.F.B. was a recipient of The Domestic Post Doctorate Program (PNPD, CAPES). A.T.S. was a recipient of the PhD program (CNPq).

Supplementary Material

Supplementary Figure S1

References

Singer

, Deutschman

, Seymour

, Shankar-Hari

, Annane

, Bauer

, Bellomo

, Bernard

, Chiche

, Coopersmith

, Hotchkiss

, Levy

, Marshall

, Martin

, Opal

, Rubenfeld

, van der Poll

, Vincent

, Angus

. 2016. The third international consensus definitions for sepsis and septic shock (Sepsis-3). JAMA, 315:801–810.

Hall

, Williams

, DeFrances

, Golosinskiy A 2011 Inpatient care for septicemia or sepsis: a challenge for patients and hospitals. NCHS Data Brief 1–8.

Fleischmann

, Thomas-Rueddel

, Hartmann

, Hartog

, Welte

, Heublein

, Dennler

, Reinhart

. 2016. Hospital incidence and mortality rates of sepsis. Dtsch Arztebl Int, 113:159–166.

Torio

, Andrews RM 2006 National Inpatient Hospital Costs: The Most Expensive Conditions by

Payer

, 2011: Statistical Brief #160 Healthcare Cost and Utilization Project (HCUP) Statistical

Briefs

. Rockville, MD.

Luo

, Yu

, Li

. 2017. Thyroid hormone disorders and sepsis. Biomed Mater Eng, 28:S237–S241.

Fliers

, Kalsbeek

, Boelen

. 2014. Beyond the fixed setpoint of the hypothalamus-pituitary-thyroid axis. Eur J Endocrinol, 171:R197–R208.

Miro

, Ambrosio

, De Stefano

, Di Girolamo

, Di Cicco

, Cicatiello

, Mancino

, Porcelli

, Raia

, Del Vecchio

, Salvatore

, Dentice

. 2017. The concerted action of type 2 and type 3 deiodinases regulates the cell cycle and survival of basal cell carcinoma cells. Thyroid, 27:567–576.

Boelen

, van der Spek

, Bloise

, de Vries

, Surovtseva

, van Beeren

, Ackermans

, Kwakkel

, Fliers

. 2017. Tissue thyroid hormone metabolism is differentially regulated during illness in mice. J Endocrinol, 233:25–36.

de Vries

, Fliers

, Boelen

. 2015. The molecular basis of the non-thyroidal illness syndrome. J Endocrinol, 225:R67–R81.

10.

Fliers

, Bianco

, Langouche

, Boelen

. 2015. Thyroid function in critically ill patients. Lancet Diabetes Endocrinol, 3:816–825.

11.

Fontes

, Cabanelas

, Bloise

, de Andrade

CBV

, Souza

, Wilieman

, Trevenzoli

, Agra

, Silva

, Bandeira-Melo

, Silva

, Rocco

PRM

, Ortiga-Carvalho

. 2017. Differential regulation of thyroid hormone metabolism target genes during non-thyoidal illness syndrome triggered by fasting or sepsis in adult mice. Front Physiol, 8:828.

12.

Supinski

, Callahan

. 2013. Diaphragm weakness in mechanically ventilated critically ill patients. Crit Care, 17:R120.

13.

Ciciliot

, Rossi

, Dyar

, Blaauw

, Schiaffino

. 2013. Muscle type and fiber type specificity in muscle wasting. Int J Biochem Cell Biol, 45:2191–2199.

14.

Bloise

, Cordeiro

, Ortiga-Carvalho

. 2018. Role of thyroid hormone in skeletal muscle physiology. J Endocrinol, 236:R57–R68.

15.

Bloise

, van der Spek

, Surovtseva

, Ortiga-Carvalho

, Fliers

, Boelen

. 2016. Differential effects of sepsis and chronic inflammation on diaphragm muscle fiber type, thyroid hormone metabolism, and mitochondrial function. Thyroid, 26:600–609.

16.

National Research Council (U.S.). Committee for the Update of the Guide for the Care and Use of Laboratory Animals; Institute for Laboratory Animal Research (U.S.); National Academies Press (U.S.).

17.

Goncalves-de-Albuquerque

, Rohwedder

, Silva

, Ferreira

, Kurz

ARM

, Cougoule

, Klapproth

, Eggersmann

, Silva

, de Oliveira

, Capelozzi

, Schlesinger

, Costa

, Estrela Marins

RCE

, Mocsai

, Maridonneau-Parini

, Walzog

, Macedo Rocco

, Sperandio

, de Castro-Faria-Neto

. 2018. The Yin and Yang of tyrosine kinase inhibition during experimental polymicrobial sepsis. Front Immunol, 9:901.

18.

Terman

, Dalen

, Eaton

, Neuzil

, Brunk

. 2004. Aging of cardiac myocytes in culture: oxidative stress, lipofuscin accumulation, and mitochondrial turnover. Ann N Y Acad Sci, 1019:70–77.

19.

Bloise

, van der Spek

, Surovtseva

, Ortiga-Carvalho

, Fliers

, Boelen

. 2016. Differential effects of sepsis and chronic inflammation on diaphragm muscle fiber type, thyroid hormone metabolism, and mitochondrial function. Thyroid, 26:600–609.

20.

Dentice

, Ambrosio

, Damiano

, Sibilio

, Luongo

, Guardiola

, Yennek

, Zordan

, Minchiotti

, Colao

, Marsili

, Brunelli

, Del Vecchio

, Larsen

, Tajbakhsh

, Salvatore

. 2014. Intracellular inactivation of thyroid hormone is a survival mechanism for muscle stem cell proliferation and lineage progression. Cell Metab, 20:1038–1048.

21.

Javadov

, Chapa-Dubocq

, Makarov

. 2018. Different approaches to modeling analysis of mitochondrial swelling. Mitochondrion, 38:58–70.

22.

Busiello

, Savarese

, Lombardi

. 2015. Mitochondrial uncoupling proteins and energy metabolism. Front Physiol, 6:36.

23.

Lanni

, Moreno

, Goglia

. 2016. Mitochondrial actions of thyroid hormone. Compr Physiol, 6:1591–1607.

24.

Bello

, Spinazzola

, Giammatteo

, Montini

, De Pascale

, Bisanti

, Annetta

, Troiani

, Bianchi

, Pontecorvi

, Pennisi

, Conti

, Antonelli

. 2019. Effects of thyroid hormone treatment on diaphragmatic efficiency in subjects with nonthyroidal illness syndrome and on ventilation. Respir Care, 64:1199–1207.

25.

Madeira

VMC

. 2018. Overview of mitochondrial bioenergetics. Methods Mol Biol, 1782:1–6.

26.

Sun

, Wray

, Tian

, Hasselgren

, Lu

. 2003. Expression of uncoupling protein 3 is upregulated in skeletal muscle during sepsis. Am J Physiol Endocrinol Metab, 285:E512–E520.

27.

, Barger

, Boyer

, Brand

, Pan

, Adams

. 2000. Impact of endotoxin on UCP homolog mRNA abundance, thermoregulation, and mitochondrial proton leak kinetics. Am J Physiol Endocrinol Metab, 279:E433–E446.

28.

Lombardi

, Busiello

, De Matteis

, Lionetti

, Savarese

, Moreno

, Gentile

, Silvestri

, Senese

, de Lange

, Cioffi

, Lanni

, Goglia

. 2019. Absence of uncoupling protein-3 at thermoneutrality impacts lipid handling and energy homeostasis in mice. Cells, 8:916.

29.

Pohl

, Rupprecht

, Macher

, Hilse

. 2019. Important trends in UCP3 investigation. Front Physiol, 10:470.

30.

Macher

, Koehler

, Rupprecht

, Kreiter

, Hinterdorfer

, Pohl

. 2018. Inhibition of mitochondrial UCP1 and UCP3 by purine nucleotides and phosphate. Biochim Biophys Acta Biomembr, 1860:664–672.

31.

Zolfaghari

, Carre

, Parker

, Curtin

, Duchen

, Singer

. 2015. Skeletal muscle dysfunction is associated with derangements in mitochondrial bioenergetics (but not UCP3) in a rodent model of sepsis. Am J Physiol Endocrinol Metab, 308:E713–E725.

32.

Costford

, Seifert

, Bezaire V

M F

G, Bevilacqua

, Gowing

, Harper

. 2007. The energetic implications of uncoupling protein-3 in skeletal muscle. Appl Physiol Nutr Metab, 32:884–894.

33.

Brand

, Esteves

. 2005. Physiological functions of the mitochondrial uncoupling proteins UCP2 and UCP3. Cell Metab, 2:85–93.

34.

Mailloux

, Harper

. 2010. Glucose regulates enzymatic sources of mitochondrial NADPH in skeletal muscle cells; a novel role for glucose-6-phosphate dehydrogenase. FASEB J, 24:2495–2506.

35.

Schilling

, Lai

, Sambandam

, Dey

, Leone

, Kelly

. 2011. Toll-like receptor-mediated inflammatory signaling reprograms cardiac energy metabolism by repressing peroxisome proliferator-activated receptor γ coactivator-1 signaling. Circ Heart Fail, 4:474–482.

36.

Olesen

, Larsson

, Iversen

, Yousafzai

, Hellsten

, Pilegaard

. 2012. Skeletal muscle PGC-1α is required for maintaining an acute LPS-induced TNFα response. PLoS One, 7:e32222.

37.

St-Pierre

, Drori

, Uldry

, Silvaggi

, Rhee

, Jäger

, Handschin

, Zheng

, Lin

, Yang

, Simon

, Bachoo

, Spiegelman

. 2006. Suppression of reactive oxygen species and neurodegeneration by the PGC-1 transcriptional coactivators. Cell, 127:397–408.

38.

Sinha

, Singh

, Zhou

, Wu

, Farah

, Ohba

, Lesmana

, Gooding

, Bay

, Yen

. 2015. Thyroid hormone induction of mitochondrial activity is coupled to mitophagy via ROS-AMPK-ULK1 signaling. Autophagy, 11:1341–1357.

39.

Lesmana

, Sinha

, Singh

, Zhou

, Ohba

, Wu

, Yau

, Bay

, Yen

. 2016. Thyroid hormone stimulation of autophagy is essential for mitochondrial biogenesis and activity in skeletal muscle. Endocrinology, 157:23–38.

40.

Singh

, Sinha

, Tripathi

, Mendoza

, Ohba

, Sy

JAC

, Xie

, Zhou

, Ho

, Chang

, Wu

, Giguere

, Bay

, Vanacker

, Ghosh

, Gauthier

, Hollenberg

, McDonnell

, Yen

. 2018. Thyroid hormone receptor and ERRalpha coordinately regulate mitochondrial fission, mitophagy, biogenesis, and function. Sci Signal, 11:eaam5855.

41.

Geng

, Li

, Yin

, Yan

. 2011. PGC-1alpha promotes nitric oxide antioxidant defenses and inhibits FOXO signaling against cardiac cachexia in mice. Am J Pathol, 178:1738–1748.