Hepatic hydrogen sulfide levels are reduced in mouse model of Hutchinson-Gilford progeria syndrome

Hutchinson-Gilford progeria syndrome (HGPS) is a rare human disease characterised by accelerated biological ageing. Current treatments are limited, and most patients die before 15 years of age. Hydrogen sulfide (H2S) is an important gaseous signalling molecule that it central to multiple cellular homeostasis mechanisms. Dysregulation of tissue H2S levels is thought to contribute to an ageing phenotype in many tissues across animal models. Whether H2S is altered in HGPS is unknown. We investigated hepatic H2S production capacity and transcript, protein and enzymatic activity of proteins that regulate hepatic H2S production and disposal in a mouse model of HGPS (G609G mice, mutated Lmna gene equivalent to a causative mutation in HGPS patients). G609G mice were maintained on either regular chow (RC) or high fat diet (HFD), as HFD has been previously shown to significantly extend lifespan of G609G mice, and compared to wild type (WT) mice maintained on RC. RC fed G609G mice had significantly reduced hepatic H2S production capacity relative to WT mice, with a compensatory elevation in mRNA transcripts associated with several H2S production enzymes, including cystathionine-γ-lyase (CSE). H2S levels and CSE protein were partially rescued in HFD fed G609G mice. As current treatments for patients with HGPS have failed to confer significant improvements to symptoms or longevity, the need for novel therapeutic targets is acute and the regulation of H2S through dietary or pharmacological means may be a promising new avenue for research.


INTRODUCTION
Progeroid syndromes are a set of genetic disorders characterized by a shortened lifespan and early onset phenotypic alterations typically associated with advanced age such as hair loss, frailty, and atherosclerosis [1]. Consequently, studies of progeroid syndromes in both humans and in model organisms have provided valuable insight into the ageing process [2][3][4]. Progeria syndromes are extremely rare in humans, the most common is Hutchinson-Gilford Progeria Syndrome (HGPS), which has a prevalence of 1 in 20 million [5]. HGPS is a laminopathy, one of a set of diseases caused by mutations in the lamin A AGING (LMNA) gene or genes that encode for proteins involved in lamin processing [6,7]. Specific mutations activate a cryptic splice site, resulting in the production of a permanently farnesylated and methylated lamin-A isoform named progerin. Progerin disrupts the nuclear membrane by binding to lamin-A, stimulating senescence, and is associated with dramatically reduced lifespan [8]. Progerin also accumulates in tissues during normative ageing due to spontaneous activation of the cryptic splice site in HGPS, which further suggests that normative ageing and progeroid syndromes may share some common molecular processes [9].
The most common mutations in HGPS is the point mutation c.1824C>T;p.Gly608Gly in the human LMNA gene. A mouse model carrying the equivalent 1827C>T;p.Gly609Gly mutation in the Lmna gene has been developed and is known as Lmna G609G+/G609G+ or simply G609G [10]. G609G mice capture much of the human HGPS disease phenotype [11][12][13]10], although some aspects of the human disease are not present in G609G mice. For instance, the leading cause of death (>90%) in HGPS patients is myocardial infarction or other cardiovascular pathologies such as stroke [14]. In the G609G mouse model the major cause of death is cachexia and starvation [15]. Diet has recently emerged as an intervention capable of modulating the progeroid phenotype in mice, with methionine restriction extending mean and median lifespan in G609G mice by approximately 20% [16,17]. Methionine restriction is also reported to elevate tissue hydrogen sulfide (H2S) levels in various model systems [18], although H2S levels in G609G mice under methionine restriction has not yet been measured. G609G mice fed a high-fat diet (HFD) had an average lifespan of 193 days compared to 110 days on regular chow (RC) [15]. Surprisingly, G609G HFD mice had significantly lower serum glucose and plasma insulin levels compared to WT HFD mice [15], and improved glucose tolerance compared to G609G RC fed mice [15]. HFD feeding in G609G mice also elicited the onset of previously unseen pathologies in mice that closer replicate the phenotype of HGPS patients, including full body alopecia, skeletal dysplasia and aortic wall stiffening [15]. As such, the application of HFD in G609G mice may be considered as therapeutic (as is greatly extends lifespan compared to RC controls) but also may prove useful if employed as standard animal care conditions for G609G to more fully recapitulate the disease phenotype seen in patients with HGPS.
Therapeutic treatments for patients with progeroid diseases are lacking, with an average life expectancy typically less than 15 years [19]. Current treatments include farnesyltransferase inhibitors, rapamycin analogues, sulforaphane and vitamin D analogues, all of which all have clear impacts on cellular processes associated with HGPS but have yet to translate to substantial improvements to patient lifespan or incidence of comorbidities [19]. H2S has recently emerged as a potential effector of various longevity boosting interventions, and elevated tissue H2S levels are associated with long lived mutant mouse models [20][21][22][23][24].
While no studies to date have investigated the role of H2S in HGPS, there is overlap between H2S signalling and the mechanisms that underpin rapamycin, sulforaphane, and vitamin D treatment, as discussed in depth in our recent review [25]. Studies in mice have shown that ex vivo H2S production is negatively affected across tissues under HFD feeding, with a reduction in the protein levels of H2S-generating enzyme cystathionine--lyase (CSE) [26]. Additionally, intraperitoneal injection of NaHS, a H2S donor compound, mitigated against markers of renal injury caused by HFD-induced obesity in mice [27]. These anti-obesity effects appear to be, in part, modulated through interaction between H2S and the mammalian target of rapamycin pathway [28], but also via the maintenance of mitochondrial import homeostasis and metabolic activity [29]. These data suggest that the deleterious effects of HFD in mice are associated with a reduced H2S production. However, given the unexpected benefits of HFD to lifespan in G609G, a greater understanding of H2S metabolism within the unique context of G609G mice on HFD may help further our understanding of the role of diet and H2S in progeria.

RESULTS
Hepatic tissue protein homogenates were analysed by the lead acetate assay ( Figure 1A). A greater than 50% reduction in H2S production capacity was found between regular-chow (RC) fed WT and G609G RC mice (p = 0.041), while a non-significant reduction of approximately 45% was found between WT and G609G HFD mice (p = 0.210). There was no significant difference in H2S production capacity between G609G mice on either diet (p > 0.999). Given the significant reduction in hepatic H2S production capacity in G609G RC mice compared to WT controls, we then investigated whether major components of the enzymatic H2S production and disposal processed were altered ( Figure  1B and Figure 1C respectively). Transcript levels of Cse (also known as Cth) were found to be increased by almost 8-fold in G609G mice irrespective of diet ( Figure  1B, p = 0.005 for RC and p = 0.007 for HFD). Cbs expression was almost doubled in G609G RC-fed mice relative to WT controls (p = 0.028), but comparable after HFD-feeding between genotypes (p = 0.558). Mpst transcript levels were not different between WT and AGING G609G RC mice (p = 0.216) but were reduced by approximately 30% in G609G HFD mice relative to WT controls (p = 0.050). Cse, Cbs, and Mpst expression were unaffected by diet in G609G mice. We then examined expression levels of enzymes that regulate the oxidation and disposal of H2S. Ethylmalonic encephalopathy 1 Figure 1. Gene expression of selected H2S production and removal enzymes. (A) Hepatic H2S production capacity assay assessed in protein lysates from WT, G609G RC and G609G HFD mice. Data quantified by densitometry analysis of lead acetate assay results. A representative image of the lead sulfide precipitate that form as the output of the lead acetate assay is shown beneath the plot. Darker precipitates indicate higher hepatic H2S producing enzymes Cse, Cbs, Mpst (B) and H2S disposal enzymes Ethe1, Tst and Suox (C) as measured by RT-qPCR. Relative expression values were calculated using the 2 −ΔΔCt method. WT data shown in black, G609G RC in pink, G609G in Green. Statistical significance determined by one-sample t-test comparing the fold changes to a theoretical mean of one. Grubbs outlier test with alpha = 0.05 was performed, no outliers removed. Bars show mean values with error bars representing standard error of the mean. Abbreviations: Cse: Cystathionine-beta-lyase; Cbs: Cystathionine-Beta-synthase; Mpst: 3-Mercaptopyruvate Sulfurtransferase; Ethe1: Ethylmalonic encephalopathy 1 protein; Tst: Thiosulfate Sulfurtransferase; Suox: Sulfite Oxidase. * p < 0.05, ** p < 0.01. AGING protein (Ethe1) expression was significantly reduced in G609G HFD mice compared to WT mice (p = 0.003) but not different in G609G RC mice relative to WT mice (p = 0.525). Thiosulfate sulfurtransferase (Tst) expression was reduced in G609G mice relative to WT mice, irrespective of diet (p = 0.047 for RC and p = 0.027 for HFD). Finally, Suox expression was comparable between G609G and WT mice irrespective of diet (p = 0.128 for RC and p = 0.340 for HFD).
The protein levels of major H2S producing and disposal enzymes were also measured ( Figure 2, representative blots in Supplementary Figure 1). A significant difference between groups in CSE protein levels was found (H = 8.115, p = 0.003), Figure 2A. Dunn's multiple comparisons determined a significant reduction of approximately 60% in mean CSE protein levels between G609G RC and G609G HFD samples (p = 0.013) but no significant difference between any other comparison. Neither CBS (H = 3.598, p = 0.178) nor MPST (H = 2.192, p = 0.370) differed between groups as shown in Figure 2B and Figure 2C. As with CSE, protein levels of TST ( Figure 2D) were found to be significantly different between groups (H = 8.115, p = 0.003), with post hoc testing determining a significant difference between G609G RC and G609G HFD mice (p = 0.013) only. ETHE1 levels ( Figure 2E) did not differ between groups (H = 0.731, p = 0.746).
Gene expression analysis showed that TST mRNA transcripts were significantly reduced in G609G mice fed either diet compared to WT controls ( Figure 1C). However, at the protein level there was a divergence in levels of TST protein between G609G groups, with TST significantly elevated by approximately 400% in G609G HFD compared to G609G RC ( Figure 2D). We then employed a TST activity assay to interrogate the effect of genotype and diet on the rhodanese activity of TST ( Figure 2F). Statistical analysis determined no significant difference between groups (H = 5.115, p = 0.074).

DISCUSSION
Elevated endogenous hydrogen sulfide (H2S) appears to be essential for dietary restriction benefits in mice [20,21,30], is a conserved phenotype in long-lived mouse models [20,21], while tissue H2S levels decrease with age across model organisms [31,32]. Further, a reduction in persulfidation (a post-translational modification of cysteine residues by H2S) with age is a conserved phenotype across E. coli, C. elegans, D. melanogaster, and mice [33]. To date no studies have directly measured H2S production in Hutchinson-Gilford Progeria Syndrome (HGPS) despite the emerging understanding that H2S may be central to longevity and healthy ageing, and the commonalities seen in the signalling modalities of H2S and several established HGPS treatments. In this study, we employed the HGPS mouse model G609G [34] to test the hypothesis that, in contrast to anti-ageing increases in H2S production, the accelerated ageing typical of progeroid mice is associated with reduced hepatic H2S production. Given that G609G mice show lifespan extension on high-fat diet (HFD) compared to regular chow (RC) [15], we also examined the H2S pathway in these mice under HFD. HFD in WT mice typically shortens lifespan, and has been shown to reduce both plasma H2S levels and hepatic activity of CSE, CBS and MPST [26].
We report a significant reduction in H2S production capacity in G609G mice on RC compared to WT controls and a non-significant reduction in G609G HFD compared to WT controls, indicative of a partial rescue of H2S levels on HFD compared to RC in G609G mice. Cystathionine--lyase (CSE) is the major producer of H2S in hepatic tissues [35], and transcripts for Cse were significantly elevated in G609G mice on both RC and HFD compared to WT controls. These data may point to a futile compensatory elevation of H2S producing enzymes in response to reduced H2S levels, as previously reported in a rat model of human myocardial infarction [36]. One potential explanation for these findings is that transcriptional regulation of Cse expression can be altered by reactive oxygen species (ROS). ROS, and hydrogen peroxide in particular, are capable of elevating expression of Cse [37]. Crucially, G609G mice on RC show chronic elevated ROS production [38]. Cse and Cbs comprise the cytosolic endogenous H2S production pathway in mammals, although there is also a distinct pathway for H2S generation within mitochondria mediated by 3mercaptopyruvate sulfurtransferase (Mpst). However, Mpst transcripts were not found to be elevated in G609G RC and were in fact reduced in G609G on HFD. In order to provide additional insight, future studies could include additional dietary interventions, for example increasing or decreasing the abundance of dietary cysteine or methionine, to better understand which specific dietary components result in modification of hepatic H2S production in the context of the G609G mutation.
To gain better understanding of the regulation of H2S production we then quantified protein levels of H2S producing and disposal enzymes. CSE protein levels may be controlled at the transcriptional level by repression of Cse gene expression [21]. However, such an explanation does not account for the data presented here, where Cse mRNA transcripts were significantly upregulated in G609G mice on both diets but protein abundance of CSE appeared diet-dependent, being elevated in G609G mice only under HFD feeding. One potential explanation is that the extent of posttranslational degradation of CSE differed between groups. Ubiquitination is an established mechanism by which CSE stability is controlled and CSE ubiquitination and degradation is promoted by superoxide anions [39]. Indeed, the difference between RC and HFD-fed G609G mice in CSE protein levels could also be explained by the observation that, unlike most oxidative damage compounds, superoxide anion generation in rodent liver is reduced by HFD [40]. As such, future studies should measure ROS production (e.g., superoxide anion), SOD activity, and/or markers of oxidative damage as these may help understand both the transcriptional and post-translational regulation of CSE. While HFD in WT mice has deleterious effects on H2S metabolism and nutrient metabolism [26], in G609G mice it appears to be beneficial to lifespan [15]. This has implications for the importance of nutrient rich diets in human patients with HGPS, where "nutritious and high calorie foods and supplements" are recommended [41]. We also found that TST protein was elevated in HFD-fed G609G mice compared to RC fed animals, but the H2S-oxidising activity of this enzyme was decreased under HFD. We previously identified TST as a putative antidiabetic enzyme and demonstrated that homozygous TST knock out confers a pro-diabetogenic HFD-like phenotype in C57BL/6J mice [42]. These findings indicate an antagonistic relationship between HFD and TST activity, which might explain our observation of reduced TST activity under HFD feeding in G609G mice.

CONCLUSIONS
This study was designed and undertaken due to the lack of understanding in the mechanistic targets of known treatments against HGPS and considering the positive association between H2S and longevity in model organisms. The data acquired here confirmed some aspects of the relevance of H2S in HGPS but is of course descriptive in nature and raises more questions about the specific mechanisms at play. Studies in primary cell culture from patients and progeria animal models that knock down/out or overexpress CSE would assist in proving a mechanistic role for H2S in the first instance. Lentiviral constructs for the overexpression of progerin in mammalian cells are already established (see [43]) and could help clarify any direct relationship between progerin protein levels and H2S production. Subsequent application of H2S donating compounds (such as NaHS and GYY4137 [44]) could then be used to investigate if exogenous H2S treatment can rescue the progeroid phenotype, establish causation and a potential therapeutic window. Supplementation or depletion of cysteine in the diets fed to G609G mice would also be a key experiment to demonstrate a nutritional basis for modulating tissue levels of various sulfide species, including H2S, in the context of progeria. Additionally, sex differences play an important role in the regulation of Cse expression and in mammalian lifespan, with oestrogen, for example, acting as a positive regulator of Cse expression [45]. Future studies could include sexspecific control groups to untangle the role of sex on H2S metabolism in progeria. Finally, additional limitations in this study were a lack of statistical power due to the small sample size reported and that we were limited to a single tissue, the liver. This was due to the difficulty in breeding and maintenance of mice with the G609G mutation [34]. These mice are infertile, have an abundance of comorbidities, and due to their small size have limited tissue available. Regardless, the work presented here addresses an area of research that remains critically understudied and provides new evidence that the accelerated ageing phenotype observed in HGPS may be partially explained by a reduction in hepatic H2S levels.

Animal model, high-fat diet, and experimental design
Mice carrying the human HGPS single-base mutation (LMNA G609G−/+ ) were generated on a C57BL/6 background by Prof. Carlos López-Otin (Universidad de Oviedo, Spain) [10]. Heterozygous LMNA G609G−/+ female and male mice were obtained from the Otin lab and subsequently bred in the laboratory of Prof. Susana Gonzalo (Saint Louis University, MO, USA) to generate wild type (WT) control mice and homozygous LMNA G609G+/G609G+ mice, referred to simply as G609G mice hereafter [15]. All animal studies were approved (protocol #2299) and conducted in accordance with the Animal Studies Committee at Saint Louis University. Mice were housed at a constant temperature of 23°C with food and water provided ad libitum under a 12:12 light-dark cycle. Litter mates of G609G genotype were randomly assigned into one of two diet groups and fed either regular chow (RC) or high fat diet (HFD) immediately post-weaning (approximately 3 weeks of age)-diet composition for diets is provided in Table 1. Mice were culled by cervical dislocation at either 70 days (WT, males, n = 4), 100 days (G609G RC, females, n = 4) or 150 days (G609G HFD, females, n = 4). Immediately following death, the liver was collected, snap-frozen in liquid nitrogen and stored at −80°C. Tissue samples were subsequently shipped to University of Glasgow for analysis.

Hepatic hydrogen sulfide (H2S) production capacity
Measurement of H2S levels was performed in liver homogenates according to the previously described method [46]. Briefly, 100 mg of flash-frozen liver tissue was lysed in passive lysis buffer. Protein concentration was determined by bicinchoninic acid (BCA) assay (G Biosciences, MO, USA) and 100 ug of protein was loaded onto a 96-well plate. 150 uL of reaction solution containing 10 mM L-cysteine and 1 mM pyridoxal-5′phosphate was added to the protein. Filter paper that had previously been cut to the size of the plate, soaked in 20 mM lead (II)acetate trihydrate for 20 min, then dried under vacuum, was then securely attached to the plate. The assembled plate was incubated at 37°C for 1 hr. H2S sulfide gas produced during this time collects in the head space between the top of the solution in the well and the lead (II)acetate paper, forming a brown-black substrate on the paper. The amount of H2S generated in each sample was quantified by densitometry analysis of the brown-black substrate (ImageJ).

Reverse transcriptase quantitative-PCR
RNA was isolated from liver tissue by addition of 500

TST activity assay
Thiosulfate sulfurtransferase (TST) rhodanese activity was determined by measuring thiocyanate production capacity as previously described [48]. The assay was performed in a 96-well plate, where 20 µg of protein lysate was added. The reaction was initiated by the addition of 10 µL 500 mM thiosulfate and taken to 90 µL with 500 mM potassium phosphate pH 5.5 buffer. After a 2 min incubation at 37°C 10 µL 500 mM potassium cyanide was added to each sample. A standard curve of 0.1, 0.25, 0.5, 1, 2.5, 5, 10, 25, 50 mM potassium thiocyanate solutions was also prepared following the same conditions as above but without the addition of potassium cyanide. The reaction proceeded for 5 min at 37°C and was stopped by the addition of 11 µL of 38% formaldehyde. Lastly, 125 µL Fe(NO3)3/26% HNO3 was added which produced a colorimetric change. Results were quantified by spectrophotometry, measured at 460 nm absorbance. All samples were performed in duplicate, calculating the average absorbance.

Statistical analysis
All statistical analyses were performed using Prism 9 (GraphPad Inc., La Jolla, CA, USA) software. Data were first analysed by Grubbs outlier test with alpha set to 5%. Any outliers identified are noted in figure legends. Statistical significance was determined by performing non-parametric Kruskal-Wallis tests comparing WT, G609G RC and G609G HFD groups with Dunn's correction for post-hoc multiple comparison tests. * denotes a p value of < 0.05, ** denotes a p value of < 0.01, and *** denotes a p value < 0.001.

AUTHOR CONTRIBUTIONS
Generation of animal models and tissue samples by SG. Experiments designed by SG, CS, NM, RNC and SEW. Experiments performed by SW and DEM. Initial draft written by CS and SEW, with subsequent drafts commented on by all authors.

CONFLICTS OF INTEREST
The authors declare no conflicts of interest related to this study.

ETHICAL STATEMENT
All procedures were conducted in accordance with the U.S. Public Health Service Policy, approved by Saint Louis University Institutional Animal Care and Use Committee (IACUC).