ncRNAs have been reported to participate in pathophysiology and repair in the case of AKI, and are additionally promising biomarkers and therapeutic targets [46]. miRNAs are the greatest studied ncRNAs in the AKI by distinct pathways. It was identified that mitochondrial quality control processes in the AKI may be controlled by miRNAs. Recent research revealed that miR-140-5p expression was repressed by hypoxia inducible factor 1a (HIF-1$\alpha$) in ischemic AKI, leading to an increase in PARKIN and consequently mitophagy [47]. miR-214 is capable of targeting mitofusin-2 (Mfn2), and leading to massive mitochondrial fragmentation, ATP synthesis depletion, and apoptosis in renal proximal tubular cells in the ischemic kidney [48]. Exosomal miR-19b-3p and miR-155 are exchanged between M1 macrophages and tubular cells, resulting in the impairment of ischemic kidney injury by suppressing cytokine signaling‑1 (SOCS1), a negative regulatory factor of nuclear factor kappa beta (NF‑$\kappa$B) [49, 50]. miR-182 is studied in AKI as a consequence of promoting apoptosis by different pathways [51, 52]. Differentially, some miRNAs are characterized by protecting against AKI. miR-21 is often reported as a renoprotective miRNA and highly expressed during AKI, targeting key mechanisms of inflammation, autophagy, apoptosis, and fibrosis [53, 54, 55]. HIF-1$\alpha$ was indicated as an upstream effector of miR-21 [40]. More recently, the same overexpression of miR-21 was shown by Pushpakumar et al. [56], but in aged-AKI associated with renal ischemic and reperfusion models. Like miR-21, other miRNAs were suggested to be protectors against AKI. miR-688 has 124 gene targets, among them mitochondria protein 18 (MTP18), which is involved in mitochondrial fragmentation. In the AKI MTP18 is suppressed, leading to mitochondria dynamic preservation, and avoiding tubular cell apoptosis [57]. miR-205 is down-regulated in AKI, but the administration of miR-205 mimic provides an anti-apoptotic effect, but differently from miR-688, its effect is mediated by PTEN/Akt pathway [58].

On the other hand, lncRNAs are lesser studied than miRNAs; however, there is an increasing quantity of studies shedding light on the contribution of these ncRNAs in AKI. Metastasis-associated-related lung adenocarcinoma transcript 1 (MALAT1) is highly conserved and one of the first lncRNAs linked to diseases, and was detected overexpressed in the plasma and kidney samples in humans, animals, and cells. However, knockdown of MALAT1 only inhibited apoptosis of tubular epithelial in septic AKI, promoting the overexpression of miR-146a, a negative regulator of NF-$\kappa$B [59, 60]. Geng et al. [61] reported that lncRNA Growth Arrest Specific 5 (GAS5) exerts an apoptotic effect in ischemic AKI, acting as a competing endogenous RNA for miR-21 and leading to overexpression of the antiangiogenic, pro-inflammatory, and pro-apoptotic thrombospondin 1 (TSP1). Currently, both in vivo and in vitro models indicated that the lncRNA maternally expressed 3 (MEG3) was upregulated during ischemic AKI, targeting miR-129-5p, and elevating the expression of high-mobility group box-1 (HMGB1), a protein intensively released by necrotic tubular epithelial cells in the AKI, conducting inflammation and apoptosis as well [62].

In relation to cardiac complications, a wide range of ncRNAs also play in the development of cardiac hypertrophy and HF as demonstrated by a large number of in vivo or in vitro hypertrophy models. miR-17-5p, miR-29a, miR-100-5p, miR-128, miR-199a, and miR-302-367 clusters were reported to regulate pathological autophagy during cardiac hypertrophy, abrogating the expression of Mfn2, regulating and stimulating the PTEN/Akt/mTOR pathway, directly targeting mTOR, and activating the hypertrophic GSK3$\beta$/mTOR pathway [63, 64, 65, 66, 67, 68]. In addition to autophagy, miRNAs regulate other cellular processes involved in cardiac remodeling, such as metabolism and mitochondrial integrity. miR-24, miR 27b-3p and miR-214 were discovered to be overexpressed and promote mitochondrial dysfunction and oxidative stress during distinct cardiac hypertrophy models by targeting glycose-6-phosphate dehydrogenase (G6PD), PCG1$\alpha$/PCG1$\beta$ and sirtuin 3 (SIRT 3) [69, 70, 71]. miR-223 also contributes to cardiac hypertrophy by targeting the apoptosis repressor with caspase recruitment domain (ARC), which inhibits the activation of mitochondrial permeability transition (MPT) [72].

Other miRNAs may be called pro-hypertrophic miRNAs because of regulating key cardiac components during hypertrophy. miR-339-5p contributes to cardiomyocyte hypertrophy by depleting valosin-containing protein (VCP), in turn promoting activation of the mTOR/S6K pathway, an important sarcomeric protein synthesis pathway [73, 74]. miR-208a and miR-208b are only expressed in the heart together with MCH and $\beta$-myosin heavy chain ($\beta$-MHC). It was demonstrated that miR-208 contributes to cardiac remodeling by suppressing two negative regulators of hypertrophy, thyroid hormone–associated protein 1 (Thrap1) and myostatin [75]. The overexpression of miR-208a and its pro-hypertrophic properties was reported in diabetic cardiomyopathy [76]. miR-195 was revealed to be up-regulated during cardiac hypertrophy [77], and a study proposed that high mobility group A1 (HMGA1) was the target with a prominent elevation of atrial natriuretic factor (ANF) expression [78]. Moreover, miR-195 also may induce cardiac arrhythmia during hypertrophy [79]. It was proposed that miR-195-5p mediated hypertrophic response in H9c2 cells treated with angiotensin II (Ang II), depleting Mfn2 and F-box and WD-40 domain protein 7 (FBXW7). FBXW7 has anti-hypertrophic properties [77, 80]. Different studies have indicated the overexpression of miR-21 in cardiac hypertrophy and HF [81], however it has shown dual mechanisms, being either detrimental or beneficial according to cell type [82].

Over the few years, CRS studies demonstrated that ncRNAs are promising mechanisms to be deeply studied. miR-21 has been demonstrated to be a potential target. Wang et al. [83] conducted a clinical study to assess whether miR-21 could be a diagnostic and prognostic biomarker in CRS 2 patients compared with traditional biomarkers (e.g., C-cystatin, KIM-1, and N-terminal proBNP (NT-proBNP)). The results showed that miR-21 and the well-known kidney and heart biomarkers were increased in plasma samples, but miR-21 had a medium diagnosis value per se, while miR-21 and C-cystatin together had a significant diagnostic value [83]. Di et al. [84] reported that renal tubular epithelial cells stimulated with transforming growth factor beta (TGF$\beta$) released a large number of EVs containing miR-21, and cardiomyocytes treated with these vesicles presented a hypertrophic response. miR-21 was also upregulated in a cardiac hypertrophy model induced by uremia, in which both miR-21 and miR-29 were downstream of angiotensin-converting enzyme (ACE) and angiotensin receptor 1a (ATR1) [85]. Despite all of these findings, studies clarifying whether ncRNAs may be involved in CRS 3 do not yet exist. Considering the wide range of ncRNAs promoting renal and cardiac diseases and the probability of finding them in circulation, it is possible to presume that these nucleic acids likely participate in kidney-heart crosstalk.

Histone acetylation is the most evaluated epigenetic mechanism in AKI. Both histone acetylation and deacetylation may negatively or positively contribute to AKI pathophysiology according to the stimuli. In a model of 30 minutes of unilateral renal ischemia and reperfusion (IR) injury, Zage et al. [86] reported a time-dependent increase in H3 acetylation associated with overexpression of pro-inflammatory cytokines and pro-fibrotic genes. Additionally, a gender-specific IR injury model developed by Kim et al. [87] showed increased H3 acetylation mediated by the reduced negative regulation of plasminogen activation inhibitor 1 (PAI-1) conferred by HDAC11. An opposite outcome was observed by Ruiz-Andres et al. [88] when the H3 deacetylation of PCG-1$\alpha$ was mediated by the activation of NF-$\kappa$B through TWEAK (a member of the TNF superfamily), leading to the attraction of HDAC. Moreover, PCG-1$\alpha$ depletion worsened AKI. The same deacetylation was reported by Li et al. [89] through the attraction of HDAC1 to the IL6 and IL12b promoter region mediated by activating transcription factor 3 (ATF3) during IR injury. Class III deacetylase SIRT3 was associated with enhanced AKI by reducing superoxide dismutase 2 (SOD2) and p53, and autophagy [90, 91].

Although DNA methylation data in AKI is still limited, it has been demonstrated that this epigenetic mechanism is altered or aberrant in kidney injury. In an ex vivo experiment of cold kidney ischemia for 24 hours without reperfusion and followed by 2 hours of reperfusion, Pratt et al. [92] reported for the first time that CpG sites of IFN-$\gamma$ response element in the C3 gene promotor were remarkably demethylated in both groups compared to the control group. Two years later, the same research group reported that demethylation remained 6 months after the renal IR in the C3 gene promoter corresponding to NF-$\kappa$B binding sites and IL-1/IL-6 response element sites [93]. Another study conducted by Zhao et al. [94] reinforced that the global, gene promotor, exons, and introns 5hmC level was significantly reduced in ischemic kidney disease, along the 1 and 7 days of reperfusion. Two other studies with AKI and transplanted patients pointed out that methylation status may be a biomarker in AKI patients through the presence of hypermethylation of kallikrein1 (KLK1) and calcitonin [95, 96].

Conversely demethylation, another study showed the presence of hypermethylation in genes responsible for avoiding AKI worsening and progression to CKD. Chou et al. [97] induced AKI in C57Bl/6 mice by right kidney nephrectomy and then ischemia in the left kidney after 2 weeks. The results demonstrated activation of pericytes, or myofibroblast, from 1 day to 14 days of renal reperfusion, which led to elevated expression of $\alpha$-smooth muscle actin ($\alpha$-SMA), a specific myofibroblast marker, by the hypermethylation of $\alpha$-actin 2 repressor Ybx2. Myofibroblasts are key contributors to kidney fibrosis after AKI [97]. A different fibrotic kidney injury model induced by unilateral ureteral occlusion in mice was carried out by Yin et al. [98]. They observed that the fibrotic kidneys presented hypermethylation of Klotho due to overexpression of DNTM1 and DNTM3 promoted by TGF-$\beta$ [98]. All these reports indicated that altered DNA epigenetics is crucial to the development of AKI, inducing the expression of genes that contribute to the aggrievement of injury and lead to chronic kidney disease.

Histone acetylation was first highlighted regarding cardiac hypertrophy and HF, and HAT and HDAC activities were indicated to favor or inhibit cardiac hypertrophy, depending on the targets. Zhang et al. [99] demonstrated that the cardiac-specific class II HDACs, HDAC9 or MIRT, and HDAC 5 had anti-hypertrophy effects by inhibiting the expression of myocyte enhancer factor-2 (MEF2), a transcription factor of pro-hypertrophy genes like ANF and B-type natriuretic peptide (BNP). CREB-binding protein (CBP) and p300 co-activators were reported to have HAT activity and induce hypertrophic responses in cardiac muscle cells through treatment with phenylephrine (PE) [100]. After these reports, the contribution of core histone modifications to cardiac remodeling and also the target genes remained an extensive research field, considering that fetal gene expression is activated during this non-adaptive response.

Papait et al. [101] assessed histone modifications aiming to describe the epigenetic changes in adult cardiomyocytes in cardiac remodeling induced by transverse aortic constriction (TAC) by chromatin immunoprecipitation assay (ChIP), namely: three associated with activation of H3K9ac, H3K27ac, and H3K4me3 regulatory regions, a marker of the transcribed genes (H3K79me2), and three markers related to repression regions (H3K9me2, H3K9me3, and H3K27me3). Even though this full epigenetic screening demonstrated that genes involved in cardiac function were affected by TAC, a clearer understanding was necessary. Thus, Palomer et al. [102] carried out in vivo and in vitromodels reinforcing the protective role of SIRT3 against cardiac fibrosis and inflammation. Treatment of AC16 linage (human cardiac cells) with TNF$\alpha$ increased FOS protein expression, a subunit of AP-1 (activation protein 1), which is associated with fibrosis during inflammation in the heart. Further assessment revealed that SIRT3 deacetylated H3K27 in the promoter region of FOS, leading to the repression of the gene expression [102].

Unlike the previous SIRT3 study, Gu et al. [103] reported that the polycomb protein PH19 reduced H3K36m3 and increased H3K27m3 of gene promoter of SIRT2 in cardiomyocytes and heart tissue treated with AngII, providing an exponential expression of ANF and BNP, and consequently cardiac hypertrophy. More recently, Funamoto et al. [104] evaluated acetylation in the isolated left ventricle primary cardiomyocyte cells stimulated by PE, and the results indicated hyperacetylation of H3K9 and H3K122 provided by p-300 in the promoter region of hypertrophic genes ANF, BNP and $\beta$-myosin heavy chain ($\beta$-MHC). Interestingly, H3K9 acetylation predominated in the hypertrophic stage, whereas H3K122 acetylation was enhanced in the HF stage [104].

Many publications regarding histone and DNA methylation have also highlighted the signature and outcomes in the development of cardiac hypertrophy and HF. For example, Kaneda et al. [105] published a study in which H3K4me3 and H3K9m3 were shown as epigenetic markers of HF in a heart hypertrophic model with cardiomyocytes of left ventricles from Dahl salt-sensitive rats. Modifications in H3 and H4 in mice hearts were also observed 8 weeks after inducing HF by TAC. ChIP indicated that the H3K4me2 histone marking in the Atp2a2 gene promoter, the SERCA2 gene, was significantly reduced in the TAC group, whereas there was an increase in the Myh7 gene promotor, the p$\beta$-MHC corresponding gene. Additionally, ChIP analysis of the Atp2a2 gene promoter presented an elevated H3K36me2 level, resulting in recruiting DNMT1 and DNMT3b and the methyl CpG binding protein 2 (MeCp2). However, the opposite event was observed in the Myh7 gene promotor [106]. These reports indicated that specific epigenetic modifications are key contributors to the pathophysiology of cardiac hypertrophy and subsequent HF.

In relation to epigenetic modifications, CRS shows a restricted quantity of reports. As previously mentioned, uremia and inflammation participate in CRS pathophysiology. The increased homocysteine and S-adenosylhomocysteine levels in plasma are capable of inducing DNA hypomethylation and atherosclerosis. This evidence indicates that uremia provides cardiac epigenetic modification after kidney injury [107]. Inflammation was also associated with aberrant DNA methylation, however more robust data are lacking [108]. Gaikwad et al. [109] reported that acetylation at H3K9 and H3K23, dimethylation at H3K4 and H3K9, and phosphorylation at H3K10 were exacerbated in hearts during unilateral nephrectomy in mice with diabetic cardiomyopathy. As consequence, several genes involved in the cardiac remodeling were overexpressed in the heart of renal failure and diabetic mice, such as myosin heavy chains 3, 6, and 7, myosin light chain 3, metalloproteinase 1, laminin-$\beta$2, tubulin-$\alpha$, plasminogen, etc. More recently, Huang et al. [110] executed CRS type 4 in vivo and in vitro models, when it was reported that hyperphosphatemia leads to hyperacetylation of H3K9 by histones acetyltransferases p300, CREB, and HAT1 in the promoter region of the transcription factor interferon regulatory factor 1 (IRF1). IRF1 inhibited the expression of PCG-1$\alpha$, resulting in cardiac mitochondria dysfunction and metabolic changes [110]. Even though it is clear that kidney injury may yield histone modifications in the heart, specific reports in the CRS type 3 do not exist so far, being an open field to be examined.

All these findings over the last few years let to presume that both miRNAs and epigenetics modifications may be extensively studied in AKI patients in order to predict further heart complications and possibility interrupt CRS 3 development.

According to Kulvichit et al. [126], the continuum of AKI from initial kidney stress to an early injury, and then to dysfunction and long-term outcomes is the current conceptual model of the clinical course of AKI, wherein biomarkers at each point of the continuum may help define the mechanisms and predict the evolution of AKI.

Traditional biomarkers for assessing kidney function include serum creatinine, albuminuria, and cystatin C, as well as urine output and eGFR [127]. In any case, due to the limitations of serum creatinine and the complexity of AKI, a great effort has been made to develop biomarkers to improve the diagnosis and management of AKI. According to the AKI continuum, biomarkers can be broadly classified into two main categories: damage markers and dysfunction markers, such as human neutrophil gelatinase-associated lipocalin (NGAL), KIM-1, liver-type fatty acid–binding protein (L-FABP) and C-C motif chemokine ligand 14 (CCL14) as damage markers; proenkephalin amino acids 119 through 159 (penKid) as a function marker, and as stress tissue inhibitor of metalloproteinases-2 (TIMP2) and insulin-like growth factor–binding protein 7 (IGFBP7) [126].

Novel serum, plasma, and urinary AKI biomarkers have been discovered with the use of transcriptomic and proteomic technologies, and they have also garnered excitement; however, their adoption into routine clinical care has been slow due to the multi-factorial availability of testing platforms, cost, variability in assay techniques and results, as well as a lack of approval from national and international governance bodies [128].

Interestingly and similarly, the search for an ideal AKI biomarker was once viewed to be similar to the discovery of cardiac troponin for MI, but with a multifactorial cause [126]. HF biomarkers can be classified into several categories: biomarkers associated with myocardial and vascular stretching, such as BNP, NT-proBNP, troponins; biomarkers that reflect an alteration in the neurohormonal pathways and RAAS; biomarkers seen in inflammation and the oxidative process; and finally, biomarkers associated with myocardial and vascular fibrosis, such as suppression of tumorigenicity 2 (ST2) and galactin 3 [129].

As described by Biasucci et al. [130], novel biomarkers in HF may support the routinely used traditional ones by improving diagnosis and prognosis and thereby enhancing patient care. Thus, there is a growing interest in the multi-marker approaches because of their benefit over single biomarkers to increase diagnostic accuracy and to improve risk stratification in HF. miRNAs have greater stability and stronger targeting compared to traditional diagnosis and treatment, as presented by Luo et al. [131], which is a potential strategy for the diagnosis and treatment of HF.

It is known that cardiac dysfunction can determine injury to the kidney and kidney injury will reciprocally affect both the circulatory system and the heart, as the causal relationship between CKD, cardiovascular risk, and HF has been demonstrated by clinical and epidemiological studies [132]. In this sense, many biomarkers have been proposed such as cystatin C, IL-6, IL-18, KIM-1, NGAL, and Netrin-1 (NTN1) [133]. A metabolomic study reported a shift in energy production in both the heart and kidney in an animal model of CRS 3, leading to oxidative stress. Besides the potential therapeutic applicability, these finds suggest that oxidative stress systemic biomarkers may be plausible diagnostic tools [134].

In the same way, mineral and bone disorder biomarkers such as fibroblast growth factor 23 (FGF-23) and vitamin D have been suggested to participate in CRS development; while mean platelet volume (MPV), hepcidin, soluble urokinase-type plasminogen activator receptor (suPAR), placental growth factor (PIGF), urinary cofilin-1, urinary adrenodoxin (ADX), eosinophil cationic protein (ECP), fetuin B (FETUB), growth differentiation factor 15 (GDF15), guanine deaminase (GUAD) and neurogenic locus notch homolog protein 1 (NOTCH1, urinary proteins) might be useful in prognostic CRS features [127]. In addition the participation in pathophysiology, miRNAs are also emerging biomarkers in the context of CRS. Ahmed et al. [135] carried out an in silico research of miRNAs in the CRS, in which hsa-miR-21-5p was found in 8 pathways out of principal 10, and more 5 miRNAs were prominent, hsa-miR-122-5p, hsa-miR-222-3p, hsa-miR-146a-5p, and hsa-miR-29b-3p. The research suggests that the found 6 miRNAs may be promising diagnostic biomarkers. Recently, Fan et al. [136] reported that three miRNAs were significantly downregulated in the serum of acute MI patients with subsequent AKI. miR-24, miR-23a, and miR-145 were associated with the regulation of TGF-$\beta$ and apoptosis pathways. Another study with cardiac surgery patients with AKI demonstrated that miR-21 in plasma, and specially in urine, may predict future renal complications, prolonged stay in hospital as well as high mortality [137].

Vibrational spectroscopy has become an important tool in biomedical analysis considering its specificity and sensibility using biological samples. Given the potential to provide essential information on the composition and structural conformation of specific molecular species, Fourier Transform (FT) Infrared and Raman spectroscopies have been explored in several diagnostic experiments [138, 139, 140]. FT-Raman spectroscopy has also been used to study CRS 3 induced by renal IR in an animal model by [140], in which bands related to tyrosine and tryptophan (which are the main raw materials of the protein-bounded uremic toxins) were found. Recent studies have also shown that infrared spectroscopy can be successfully adapted to experimental practice to analyze EVs [138].

Taking into account all the new mechanisms and biomarkers pointed out in this systematic review regarding both kidney and heart disease isolated studies, as well as CRS research, it is proposed that these molecules may be directly related to the AKI-acute HF axis, and even seem to be promising tools to diagnose and monitor the CRS 3 pathophysiological course (Fig. 2). Besides the discovery of promising biomarkers by high-throughput technologies, such as transcriptomic and proteomic, in animals and cellular models, the application in the clinical is still elusive and needs a better evaluation.

Fig. 2.

New pathophysiological mechanism and promising molecules in the CRS 3. Several types of molecules have been indicated to contribute to kidney and heart diseases, from nucleic acids, such as miRNAs, to proteins and vitamin D. There is increasing consent that these molecules are also associated with AKI-acute heat failure axis and confer promising biomarkers for diagnosing and monitoring CRS 3. ACF, Acute cardiac failure; ADX, urinary adrenodoxin; AKI, Acute kidney injury; CH, Cardiac hypertrophy; ECP, Eosinophil cationic protein; FETUB, Fetuin B; FGF23, Fibroblast growth factor 23; GAS5, Growth Arrest Specific 5; GUAD, Guanine deaminase; GDF15, Growth differentiation factor 15; IL-18, Interleukin 18; KIM-1, Kidney injury molecule 1; lncRNAs, long non-coding RNAs; MALAT1, Metastasis-associated-related lung adenocarcinoma transcript 1; MEG3, Maternally expressed 3; miRNAs, micro RNAs; NGAL, Neutrophil gelatinase-associated lipocalin; PIGF, Placental growth factor; suPAR, Soluble urokinase-type plasminogen activator receptor.