1.

Introduction

Left ventricular remodeling (LVR) as a consequence of acute myocardial infarction (AMI) appears on both macroscopic and microscopic levels. Changes with regard to the macroscopic level refer to the modification of shape, size, and wall thickness of the myocardium. The disproportionate thinning and dilatation of acutely infarcted myocardium is caused by extensive deposition of collagen as well as an increase in the tensile strength. In 1978, Hutchins and Bulkley¹ described this process and it was termed “infarct expansion.”^2,3 It led to microscopic changes related to a reduction in the number of cardiomyocytes and alterations in connective tissue involving the degradation of connective compartments and proliferation of fibroblasts in the cardiac muscle. However, the myocardium is more than myocytes; it is also an extensive network of extracellular matrix (ECM) which contains collagens and fibrillar proteins, the structural support to resistant cells. The ECM provides a viscoelastic scaffold consisting of type I and III collagen.^4,5 The remodeling of the ECM plays a major role in the changes that take place in the LVR.^6–8 Whittaker et al.⁹ demonstrated the significant damage of the myocardial collagen network as early as 1 day after infarction. Furthermore, the authors reported that fibroblasts are responsible for creating the microstructure because of the thickness and crosslinking of collagen fiber bundles, which explains the properties of infarct scar tissue.^10,11 Therefore, cardiac wound repair after myocardial infarction involves phases which include an inflammatory phase and tissue remodeling phase. The first phase starts early after coronary artery occlusion and involves degradation of normal ECM, invasion of inflammatory cells at the site of the injury, and the induction of bioactive peptides and cytokines. Degradation of the ECM during the acute phase is an essential event which is related to the appearance of inflammatory cells and the proliferation and maturation of macrophages and fibroblasts. Furthermore, it also provides the necessary substructure for scar formation.¹²

The concentration of collagen degradation products (CDPs) may reflect the process of LVR. The specific structure of collagen provides its resistance to proteolytic degradation except by matrix metalloproteinases (MMPs). The activated MMPs require the proteolytic cleavage of a propeptide sequence. The classes of MMPs that may have particular relevance to myocardial remodeling possessing high substrate specificity for fibrillar collagens are MMP1, MMP8, and MMP13 and other ECM proteins such as proteoglycans, versican, perlican, and aggrecan.^5,13 MMPs are capable of hydrolysing type I, II, and III collagens at the Gly775–Leu776 or Gly775–Ile776 peptide bond of the corresponding α chains.¹⁴ Type I and type III collagens are the main fibrillary constituents in developing tissue.¹⁵ They are characterized by their ability to assemble into highly orientated supramolecular aggregates with a characteristic suprastructure and the specific quarter-staggered fibril-array with diameters between 25 and 400 nm.^16,17 Type III collagen is a homotrimer of three $\alpha 1(\mathrm{III})$-chains and it is widely distributed in collagen type I containing soft tissues.¹⁶ It is a significant component of reticular fibers in the interstitial tissue, including vessels.¹⁸ Type III collagen has slightly different functions compared to type I collagen. Type I collagen confers tensile strength and resistance to stretch and deformation, whereas type III collagen fibers confer resilience.¹⁹ The cross-linked N-terminal telopeptide (IIINTP) is one of the products from collagen catabolism which is assessed in tissue degradation products as a marker of mature type III collagen.²⁰

Optical methods provide a possibility to study biological samples quickly and often noninvasively. Time-resolved fluorescence spectroscopy (TRFS) is a powerful tool in physics and chemistry²¹ which can be successfully applied in medical diagnosis. Herein, the fluorescence of a sample is measured as an approximately exponential (often multiexponential) decay function of time as a result of the radiative relaxation of the molecules after a remarkably short pulse of excitation light. Many biological macromolecules, including CDPs, which exhibit complex intensity decay patterns, could be distinguished by different fluorescence lifetimes (FLTs). The solution of this complexity is the picosecond time resolution obtained with time-correlated single-photon counting (TCSPC). It is possible to extend the data collection by multiple excitation/emission cycles, even if a sample consists of just a few molecules.^22,23

The aim of this study was to evaluate the potential diagnostic usefulness of TRFS in the assessment of type III CDPs. It has been done using the in vitro model of CDPs and human plasma by checking the influence of different amounts of CDPs on the FLT of plasma. This preliminary experiment was designed to establish if CDPs’ characteristics by FLT might be visible under determined conditions.

2.

Materials and Methods

3.

Results

3.1.

Hydrolyzed Collagen III

To confirm the hydrolysis, the colorimetric reaction with ninhydrin color reagent was performed and the results were shown in Table 1. According to the ninhydrin reaction, after 7, 11, and 23 h from the beginning of the hydrolysis, similar values of absorbance were received ($\lambda =570\mathrm{nm}$). Apparently, in less than 11 h, the activity of the enzyme significantly decreased and it is possible that the enzymatic reaction had finished.

Table 1

Results from ninhydrin color reagent reactions.

	Time point (h)	Absorbance
1.	3	0.026
2.	7	0.439
3.	11	0.421
4.	23	0.398

The fluorescence decay curves of the samples of hydrolyzed collagen III significantly deviated from a single-exponential function; therefore, in order to obtain the correct fluorescence decay characterization, it was necessary to use the multiexponential fluorescence decay model. The model of three components was used during the analysis of hydrolyzed collagen without the addition of plasma (Fig. 1). However, this model is not efficient if more plasma is used. The real distribution of FLTs in such a system is probably much more complex, especially when one realizes that different conformations of the same specimen may exhibit different deactivations of the excited state. In this work, the authors analyzed only a mean FLT (mFLT) weighted by the fractional contribution of each component to the steady-state intensity calculated from four-exponential models.

Fig. 1

The decomposition of fluorescence lifetime (FLT) decay on particular components: (a) the signal intensity of hydrolyzed collagen III, (b) the model of three exponential fluorescence decay model, (c) the residuals of the model.

The averaged values of hydrolyzed collagen FLTs after the addition of different amounts of plasma were presented in Fig. 2. When the concentration of plasma was substantial, mFLT increased. Furthermore, when the proportion of hydrolyzed collagen and plasma was $5:6$, the value of mFLT was highly approximate to the $5:6$ dilution of plasma with physiological saline. The influence of the amount of added plasma was significantly different between hydrolyzed collagen after 4 and 8 h from the beginning of the reaction and after 12 and 24 h. The tendency of the curve was still the same, but the increase of mFLT after a further addition of plasma was relatively slower. The comparison of different amounts of collagen (with respect to the plasma concentration) in samples allowed us to assess the limit of traceability using TRFS in order to detect CDPs. Thus, it is possible to detect CDPs when the CDP concentration in plasma is higher than $100\mu \mathrm{g}/\mathrm{mL}$. When the amount of hydrolyzed collagen in the sample was less than $100\mu \mathrm{g}/\mathrm{mL}$, the increase of mFLT was not observed. CDPs in the presence of the highest amount of plasma did not dominate the signal originating from the fluorescence emission of other proteins.

Fig. 2

The averaged values of hydrolyzed collagen FLT after addition of plasma; collagen = purified type III collagen from human placenta (Sigma-Aldrich), plasma = human plasma from healthy subject diluted in NaCl: (a) mFLT of samples after 4 h of hydrolysis, (b) mFLT of samples after 8 h of hydrolysis, (c) mFLT of samples after 12 h of hydrolysis, (d) mFLT of samples after 24 h of hydrolysis.

It is necessary to point out that the function of the curves changes in a particular way. At the first time point ($t=4\mathrm{h}$), it is the concave-convex function from below, whereas in the next two time points ($t=8\mathrm{h}$ and $t=12\mathrm{h}$), the sigmoidal function was observed. In the last time point ($t=24\mathrm{h}$), it is the convex function from the bottom. The obtained results look like a function of a dose-response relationship. The changes appear to be regular and depend on the hydrolysis time.

The basis for further analysis of the results was the mFLT of the solutions of different proportions of plasma with hydrolyzed collagen as a function of the duration of the hydrolysis (Fig. 3) Additionally, the mFLT of pure hydrolyzed collagen in each time point of the performed hydrolysis was presented. The pure hydrolyzed collagen showed the monotonic increase of the value of mFLT in further steps of hydrolysis. The addition of a small amount of plasma led to an increase of the mFLT of samples obtained at 4 and 8 h of hydrolysis. In further steps of this process, mFLT was substantially shorter, whereas the addition of plasma in the amount of $20\mu \mathrm{l}$ and more had decreased the correlation between the solutions and the time points of the hydrolysis.

Fig. 3

The mean FLT of the solutions with different proportions of plasma with hydrolyzed collagen as a function of the duration of the hydrolysis; collagen = purified type III collagen from human placenta (Sigma-Aldrich), plasma = human plasma from healthy subject diluted in NaCl: (a) mFLT of samples without addition of plasma, (b) mFLT of samples with 2,5 μl addition of plasma, (c) mFLT of samples with 5 μl addition of plasma, (d) mFLT of samples with 10 μl addition of plasma, (e) mFLT of samples with 20 μl addition of plasma, (f) mFLT of samples with 50 μl addition of plasma, (g) mFLT of samples with 100 μl addition of plasma, (h) mFLT of samples with 600 μl addition of plasma

4.

Discussion

We presented data of analysis of FLT due to the complexity of the blood plasma. The biological material contains multiple kinds of proteins and other components concerning the fluorescent properties, which may affect the mFLT. It is extremely difficult to characterize the sample because of the large proportion of albumin, the extensive range in abundance of other proteins as well as the heterogeneity of glycoproteins.²⁵ The endogenous fluorophores which are often used for biological material characterization are aromatic amino acids (tryptophan, tyrosine, phenylalanine), structural proteins (collagen cross-links, elastin), enzyme metabolic co-factors (nicotinamide adenine phosphate dinucleotide and flavin adenine dinucleotide), porphyrins, and lipids.²⁶ Thus, given the multiplicity of the factors determining the fluorescence, analysis of each component could not give reliable information regarding FLT.

Our in vitro model of CDPs only partly reflects the in vivo conditions. In the human body, the activity of the used enzyme is strictly regulated by many factors which maintain a balance between ECM deposition and degradation during different physiological processes.²⁷ The activity of MMPs is regulated by transcription, activation, and inhibition.²⁸ Transcription is stimulated by several factors including interleukin 1, platelet-derived growth factor, and tumor necrosis factor $\alpha$. However, it is inhibited by transforming growth factor $\beta$, heparin, retinoids, and corticosteroids. The activation of pro-MMPs to the MMPs is greatly stimulated by the urokinase plasminogen activator, which is expressed in monocytes and macrophages and is inhibited by tissue inhibitors of metalloproteinase (TIMPs). This system, with a specific receptor and inhibitors plasminogen activator inhibitor (PAI) 1 and 2, regulates the proteolytic activity. The inhibition of activated MMPs by TIMPs as well as drugs, such as tetracyclines and anthracyclines, regulates the proteolysis of ECM.²⁹ The myocardial MMPs and TIMPs are also secreted by several cell types including fibroblasts, inflammatory cells, and endothelial cells. The gene expression is firmly controlled at the transcription level. Several cytokines, polypeptide growth factors, hormones, steroids, and phorbol esters modulate the synthesis and secretion of pro-MMPs and TIMPs.³⁰

Liu et al.³¹ demonstrated that the collagenases are responsible only for the first degradation step of collagen, in which the fibers are cleaved into specific $1/4$ and $3/4$ fragments. However, gelatinases and cysteine proteases can further degrade the collagen fragments. The data confirmed our results from the ninhydrin reaction because the decreased activity of the enzyme and completion of the reaction was observed earlier than we expected. Kerkvliet et al.³² suggested that MMP1 is responsible for the first degradation step of collagen, and then MMP2 further digests the degraded collagen fibers, whereas Creemers et al.³³ demonstrated that MMP2 is responsible for collagen digestion in the soft connective tissues. The precise molecular mechanism of collagen degradation is still inexplicable; however, there are hypotheses which explain the structural changes during this process. Collagen degradation begins with the cleavage of collagen by MMPs at specific sites, which is a covalent bond between glycine and leucine or isoleucine residues, followed by alanine or leucine.³⁴ Therefore, from the analysis of the sequence of target proteins, which fragments will result from the hydrolysis with any of common proteases could be predicted. Due to the collagen triple helical structure, all side chains in collagen are exposed to solvent which differentiates them from globular proteins. Nevertheless, it shows that the covalent bond in the canonical triple helical structure is hidden from the solvent; thus, the collagen triple helix does not fit into the active site of collagenases. This confirms that the existence of locally unfolded states could probably play a major role in the covalent bonding to become available to collagenases.³⁵ Nonetheless, it is essential to explain the specific mechanism of the MMP1 cleavage to conduct further research.

The obtained results showed that the increase of added plasma to hydrolyzed collagen extends the mFLT. The visible fluorescence signal from hydrolyzed collagen was observed when a small amount of added plasma was used. This suggests that it is highly likely to detect CDPs in plasma if the interfering proteins were eliminated. According to our experiment, it is possible to detect CDPs when the concentration in plasma is higher than $100\mu \mathrm{g}/\mathrm{mL}$. There are a number of scientific papers,^36–39 which confirm the apparition of CDPs in plasma or serum during different disorders. Langberg et al.³⁶ reported that the concentration in plasma of type I collagen degradation product [cross-linked carboxyterminal telopeptide of type I collagen (ICTP)], which is the most common detectable CDPs from type I collagen, after intensive physical training was about $5\mu \mathrm{g}/\mathrm{dL}$. Nurmenniemi et al.³⁷ determined that in patients with IV stage head and neck squamous cell carcinomas, the amount of ICTP was $\sim 5.5\mathrm{ng}/\mathrm{mL}$. Rosenquist et al.³⁸ measured the C-terminal telopeptides of type I collagen in serum and detected a concentration of about $2000\mathrm{pmol}/\mathrm{L}$ for premenopausal women and about $4000\mathrm{pmol}/\mathrm{L}$ for healthy postmenopausal women. Perkins et al.³⁹ used Western blotting to check the concentration of circulating precursor forms of type XVIII collagen in acute lung injury patients. The densitometric analysis of gel electrophoresis confirmed increased plasma precursors, especially with the weights of 130 and 170 kDa.

Alterations in collagen deposition during scar development in the early phase of wound healing following AMI appear to be associated with the development of infarct expansion and the pathogenesis of myocardial rupture.¹⁹ Kasama et al.⁴⁰ revealed significant correlations between the degree of change in the concentration of PIIINP and echocardiographic exponents of LVR in patients with ST-segment elevation myocardial infarction. Markers of collagen turnover also identify patients with diastolic dysfunction.^41,42 Weber et al.⁴³ demonstrated that procollagen peptides as well as MMPs could be used for monitoring cardiac collagen turnover. Iraqi et al.⁴⁴ showed long-term kinetics of collagen biomarkers in patients with AMI with left ventricle dysfunction and congestive heart failure (HF). Jensen et al.⁴⁵ confirmed that serum PIIINP remained elevated for more than 4 months after AMI, suggesting ongoing repair processes. Moreover, markers of collagen metabolism (ICTP and PIIINP) reflected reverse left remodeling following cardiac resynchronization therapy in patients with HF.⁴⁶ These findings suggest that the collagen assessment may assist in identifying patients while they are developing heart remodeling that is still present in a diagnostically challenging condition. For better understanding, the microscopic level of the LVR process in vitro and in vivo models with cardiac tissue should be analyzed. Jugdutt⁴⁷ studied a canine model of the LVR after AMI. Different models for studying LVR and heart failure development were examined involving large animals (dog, pig, and sheep) and small animals (rabbit, rat, and mouse).^48,49 In most models, both types of mRNA (from types I and III collagens) increase in 3 days⁵⁰ and fall by 7 to 14 days.⁵¹

5.

Conclusion

The detection method of CDPs based on TRFS may have a significant diagnostic value. However, it is crucial to continue further investigation in this field. Differences between mean FLT from consecutive steps of hydrolysis suggest that it would be possible to determine the degree of collagen degradation. It is essential to further this field of research with the use of cardiac tissue from in vitro and in vivo models, as well as tissue from the human heart.