Qualitative CysC MSIA assay in CSF and serum
Cystatin C is a serine proteinase inhibitor belonging to the type 2 cystatin gene family (Jarvinen, Rinne et al. 1987; Mussap and Plebani 2004). It inhibits both endogenous proteases, such as liposomal cathepsins, and proteases of parasites and microorganisms. It is a non-glycosilated single chain protein with a molecular weight of 13,343. Due to the important function, cystatin C is expressed at the stable levels in most nuclear cells. Its amino acid sequence consists of 120 amino acid residues encoded by a 7.3 kb gene located in chromosome 20 (Schnittger, Rao et al. 1993). Cystatin C has been indicated in numerous pathological states (Henskens, Veerman et al. 1996; Grubb 2000; Reed 2000), most notably in renal failure (Randers, Kristensen et al. 1998; Randers and Erlandsen 1999). There are a growing number of reports demonstrating that cystatin C is more preferable than creatinine for measurement of GFR (Naruse, Ishii et al. 2009). Also, a variant of human cystatin C (L68Q) is an amyloidogenic protein deposited in the cerebral vasculature of patients with hereditary cerebral hemorrhage with amyloidosis in which patients suffer from repeated cerebral hemorrhages (Ghiso, Jensson et al. 1986; Olafsson and Grubb 2000; Calero, Pawlik et al. 2001). In the clinical practice, cysC is well-desribed serum marker of renal failure that is not dependent of age, sex or lean muscle mass (Seronie-Vivien, Delanaye et al. 2008; Naruse, Ishii et al. 2009). At the same time, cystatin C is becoming acknowledged as a marker for elevated risk of death from myocardial infraction and stroke(Naruse, Ishii et al.2009). The concentration of cystatin C in healthy individuals range from 0.8 to 1.2 mg/L, depending on the measurement method (Roos, Doust et al. 2007). Increased serum levels are almost always associated with reduction in GFR. The role of cystatin C as potential biomarker for multiple scleroses was introduced in the work of Irani et al. They have reported a correlation between the ratio of one cysC truncated form des-SSPGKPPR and native cysC peak, and the occurrence of multiple sclerosis(Irani, Anderson et al. 2006). Other research groups do not support this concept, claiming that the cleaved peptide occurs as a result of sample storage, not exclusively MS existence (Del Boccio, Pieragostino et al. 2006; Hansson, Hviid Simonsen et al. 2006; Nakashima, Fujinoki et al. 2006).
Fig. 7. Cystatin C mass spectra showing signal from native form and additional "wild type" protein variants
In previous work MSIA assays have been developed for qualitative identification of cystatin C from human plasma (Kiernan, Nedelkov et al. 2006) and urine samples (Kiernan, Tubbs et al. 2003). After several assay optimization we were able to develop MSIA assay for determination of cysC in cerebrospinal fluid both from control group and a couple of matched serum/ CSF samples obtained from patients with multiple sclerosis(Trenchevska, Aleksovski et al. 2009).
MSIA provides excellent results for this type of single protein analysis, because it preserves the protein in its native form during the sample preparation procedure, and using the MALDI-TOF MS, allows an insight into the detailed protein structure, presence and distribution of isoforms. When completing MSIA data from analyzed cysC in human plasma sample, besides the native peak (CysC native MW=13 343), several protein variants can be noted: oxidation peak (MW=13 359), des-S variant (MW=13 260) and des-SSP variant (MW= 13 076) (Fig.7)
These isoforms were noted in all samples analyzed, and also in a large population proteomics study where 1000 samples were screened using the developed MSIA cystatin C assay (Nedelkov, Tubbs et al. 2004; Nelson, Nedelkov et al. 2004). The goal of the further research was to develop MSIA method for cysC identification in CSF. For utilization of this assay, a critical step was sample preparation in terms of optimizing the dilution. In this analysis, both control CSF samples and samples from patients diagnosed with MS were analyzed. Moreover, paired serum/ CSF samples origin from same individuals was screened. As presented in the previous studies, using the optimized MSIA qualitative assay, cysC and its wild type isoforms were detected in the analyzed CSF samples (Nedelkov, Shaik et al. 2008). However, several correlations were observed when matched serum/ CSF samples were analyzed. The sample that consisted most extensive truncations in the serum cysC, exploits most variants in the CSF sample as well (Fig.8/c,d). Also, additional peaks not common in serum cysC analysis were noted both in the control CSF samples and the samples from patients with MS. These are mainly truncated cysC isoforms missing 3, 4, 7, 8, 9, 10, 11, 14 and 17 N-termina! amino acids; some of which have been reported for the first time by our group (Fig. 8/a,b).
Fig. 8. MALDI-TOF MS spectra obtained after cystatin C MSIA in a) serum sample (control); b) CSF sample (control); c) paired serum/CSF sample (control) and d) paired serum/CSF sample (patients with diagnosed multiple sclerosis).
Developing quantitative MSIA assay for CysC and its variants
Developing quantitative assays for protein determination can contribute to the diagnosis and prognosis in clinical proteomics studies due to the fact that oftentimes, not only the occurrence, but also the concentration distinguishes between "normal" and "diseased" condition. Using conventional ELISA and other immunologically based quantitative assays, we can obtain information about the total protein concentration. All these techniques, however, lack the ability to quantify specific protein variants and isoforms, which in turn, are more probable biomarker candidates.
Using MSIA, we have been able to develop and validate several fully quantitative assays, both for quantitation of intact proteins, as well as protein isoforms.
There are several possible approaches regarding mass spectrometry quantitation of proteins and protein variants. One requires usage of isotope labels, whereas, second uses different protein, termed as internal reference standard for quantitation. The goal of developing platforms to be implemented in every-day clinical practice requires simplified sample preparation procedure and handling. In our work we present the developed quantitative assay for determination and quantitation of cystatin C and its variants in biological samples. Results from this quantitative assay have recently been published (Trenchevska and Nedelkov 2011). In this topic we summarize the basic concepts and "hot spots" that need to be considered when developing this type of assay in clinical practice. The procedure includes several steps:
Choice of internal reference standard - There are several approaches in the process of choosing an adequate internal reference standard for a quantitative assay. An important prerequisite that an internal standard must possess is not to be present in human plasma or serum, so that its spiked concentration in the analytical sample remains constant in each analysis. Also, the signals that the IRS produces in the mass spectra should be in close proximity to the signal of the targeted protein, so that the same MS acquisition parameters can be used for analysis of both proteins. For developing quantitative assay for cysC we have used beta-lactoglobulin (BL), the major whey protein in cow’s milk, which has molecular weight of 18 281 (it is close to the MW of cysC = 13 343), and is present in many mammalian species, but not in human. In Figure 9, signals from cystatin C and the internal reference standard BL can be noted.
Optimization of standard curve concentration range – In this step first it is necessary to determine the ratios of immobilized cysC and BL antibodies immobilized on the affinity pipettes. In general, concentration of cysC antibody should be higher in order to accurately quantify the various amounts of cysC in the examined analytical samples. Since we can control the concentration of BL in the samples, the antibody concentration can be lower in the affinity pipettes. For the analysis we have used the following antibody ratio -cysC:BL=8.5:1 (m/m). Next, we determine the optimal concentration and volume of spiked BL in the samples, and last, we determine the concentration range for the standard curve. It is important to construct the curve in its linear range concentration and then dilute the samples if necessary, because of the greater accuracy. In this assay we have constructed a six point cysC standard curve, spanning the range from 0.0312 to 1.0 mg/L cystatin C (Fig.10).
Fig. 9. Mass spectra from cystatin C and Beta lactoglobulin used as internal reference standard for quantitation
Fig. 10. Presented is a typical standard curve used for cysC quantitation (left); representative cysC mass spectra used for standard curve generation (right).
Sample preparation – Prior to the analysis, the analytical samples were diluted twice in the assay buffer and spiked with 5 μL BL (c (BL) =10 mg/L).
Assay parameters - For this quantitative assay several characteristic parameters were analyzed in order to check the reproducibility and quality of assessment. Intra-assay precision was done by analyzing three plasma samples in replicates, each with a single standard curve. Inter-assay precision was determined by analyzing one plasma sample three times on different days with separate standard curve. Linearity of the assay was determined by analyzing serial dilutions of a sample with known cysC concentration and comparing the observed with the expected concentrations. Also, spiking-recovery experiments were performed by adding cysC standard in known concentration into the analytical sample, again with a known cysC concentration. By comparing the expected with the observed results we were able to recover the spiked cysC concentration.
Assay validation - In order to further validate the newly developed assay, a comparison with a well-established method was performed. Several samples were analyzed both with commercially available ELISA and with the developed MSIA method. The good correlation between both assays validated the results obtained with the new cystatin C assay (ref. statija cysC).
Assay application - The next phase, and practically the last step between the method and its application in the full potential is actually screening many samples and quantifying not only the native protein, but the isoforms present. Using this assay, we were able to quantify cystatin C and its variants in a total of 500 plasma samples. For all the samples, concentrations of cystatin C and its variants was determined, and averaged (average cysC concentration = 0.94 mg/L) which corresponds to the average CysC concentration previously established using ELISA (Erlandsen, Randers et al. 1998). In all the samples the "wild type" variants were noted; CysC 3Pro-OH and two truncated forms lacking one (des-S) and three (des-SSP) N-terminal amino acids respectably (Fig.11).
Fig. 11. Typical MALDI-TOF spectra from cystatin C showing the detected variants
Presented in Table 3 are partial results from the quantitative analysis of cystatin C, where we were able to calculate the concentration of each variant present in the analytical sample.
|
sample |
Native cysC [mg/L] |
Hydroxyl cysC [mg/L] |
cysC des-S [mg/L] |
cysC des-SSP [mg/L] |
Total cysC [mg/L] |
|
1 |
0.25 |
0.33 |
0.07 |
0.07 |
0.71 |
|
2 |
0.32 |
0.41 |
0.08 |
0.03 |
0.84 |
|
3 |
0.39 |
0.48 |
0.09 |
0.08 |
1.04 |
|
4 |
0.42 |
0.54 |
0.08 |
0.08 |
1.11 |
|
5 |
0.35 |
0.46 |
0.11 |
0.11 |
1.02 |
|
6 |
0.27 |
0.35 |
0.07 |
0.09 |
0.79 |
|
7 |
0.24 |
0.34 |
0.06 |
0.04 |
0.68 |
|
8 |
0.69 |
0.83 |
0.15 |
0.18 |
1.85 |
|
9 |
0.45 |
0.58 |
0.12 |
0.11 |
1.27 |
|
10 |
0.29 |
0.38 |
0.08 |
0.09 |
0.85 |
Table 3. Example of quantitative data obtained for cystatin C protein variants
Further strategies in clinical proteomics
From a diagnostic point of view, it is of great advantage to analyze and assess biological markers of disease from several biological fluids from the same individual. Albeit on a small scale, the results shown here indicate that such studies are possible if the right assays are utilized. The developed MSIA method provides a unique way of delineating protein isoforms and their abundance in serum and CSF. This way, additional population proteomics studies can be done, that will provide with further insight into the physiology of biological processes and diseases. Group protein profiling still remains an irreplaceable method in clinical practice, due to the amount of information that can be provided in a single analysis. When combined with the advanced methods for statistical classification and analysis and using the tools of bioinformatics, analyses at a large scale can be performed. Obtained data can further be used in creating software programs and data bases for recognition and identification of specific protein profiles.
Further analyses regarding our group will progress in several directions: (1) develop quantitative assays for other proteins that are considered to be potential diagnostic markers for specific neurological diseases; (2) implement these assays to fully quantify proteins and protein isoforms in serum and cerebrospinal fluid samples; (3) introduce the concept of immunoaffinity separation techniques into clinical practice (4) develop algorithms for protein profile data base creation and pattern recognition and (5) investigate the possibility to implement the novel methods to analyze proteins using other media such as tears and saliva.
Conclusions
Analyzing proteins is a complicated task since many approaches can be followed (Andersson, Alvarez-Cermeno et al. 1994; Anderson and Anderson 1998; Bakhtiar and Nelson 2001; Diamandis 2003; Conrads, Hood et al. 2004; Bons, Wodzig et al. 2005; Anderson 2010). There are two majorly differentiated approaches that can be applied in routine clinical practice, analyzing protein profile (protein profiling) and choosing single protein (biomarker discovery) and optimizing methods for identification and determination. Regarding these remarks, there is a quite divergence between the "group" protein screening and "individual" protein determination. However, by combining different novel techniques we are able to bridge the gap in between. In addition, we can discuss the necessary development both in the methodology and techniques in order to have the highest impact on the study of the molecular composition of CSF proteins.
The application of these novel techniques to the study of neurological disorders is providing an insight into the pathogenesis of neurodegeneration, and is fueling major efforts in biomarker discovery.In the protein profiling concept we are able to analyze patterns obtained from the proteome of a patient with a certain disease, and compare it with a sample from a control group. This approach is still widely used in screening patients with neurological diseases, using electrophoresis techniques as a basic technique. By comparing these profiles, a typical prototype pattern is created for separate conditions, and by simply comparing these patterns (nowadays with the use of informatics tools and statistics programs) differential diagnosis is possible.
This approach, however, only presents the big picture. It is the single protein biomarker assays that are more and more abundant in clinical laboratories. The quest for novel biomarkers results in developing techniques and mechanisms of investigating protein intrinsic characteristics at a molecular level.
From the obtained results we can conclude that even though in routine clinical practice enzymatic immunoassays are still dominant, especially for quantitative protein analysis, they have one crucial disadvantage. These assays are oblivious to protein modifications, unless there is an antibody toward that exact modification. The mass spectrometry immunoassay, on the other hand, is designed to detect protein modifications and intact proteins in a single assay. This qualitative determination is already a way to delineate between control subjects and diseased samples. With the added quantitative feature, this and similar assays are poised to change the way we look proteins and protein modifications and their role in health and disease.
An integrated evaluation of the data that are obtained by all the holistic approaches, in combination with clinical and epidemiological data will eventually not only increase our understanding of disease mechanisms, but subsequently also enable us to develop more specific and individualized medicines and therapies.
All our efforts are oriented toward the main goal – to aid in the process of revealing the protein complexity of neurological disorders and contribute to its better understanding.
