• Open Access

The Cardiac Biomarker Sodium-Calcium Exchanger (NCX-1) Can Differentiate between Heart Failure and Renal Failure: A Comparative Study of NCX-1 Expression in Dogs with Chronic Mitral Valvular Insufficiency and Azotemia


Corresponding author: Dr Changbaig Hyun, The Section of Small Animal Internal Medicine, School of Veterinary Medicine, Institute of Veterinary Medicine, Kangwon National University, Chuncheon 201-100, South Korea; e-mail: hyun5188@kangwon.ac.kr.


Background: The sodium-calcium exchanger (NCX-1), an established cardiac biomarker, was postulated previously as differentiating between heart failure (HF) and renal failure (RF) in dogs. The effect of azotemia on NCX-1 expression has not been studied.

Hypothesis: In contrast to other cardiac biomarkers (eg, N-terminal-proBNP), we hypothesized that the expression level of NCX-1 is not influenced by either azotemia or decreased renal clearance.

Animals: Fifteen client-owned healthy control dogs, 14 dogs with chronic mitral valvular insufficiency (CMVI), classified based on severity of the disease by the established International Small Animal Cardiac Health Council classification system, and 15 dogs with RF, grouped according to the International Renal Interest Society stage classification.

Methods: A comparative study of the expression levels of NCX-1, evaluated in peripheral blood samples from dogs with HF, RF, and healthy controls by quantitative PCR.

Results: NCX-1 expression was significantly increased in moderate (2.99 ± 0.61 [fold changes relative to normal group]) to severe (4.35 ± 1.44) CMVI dogs (P < .01). In contrast, NCX-1 expression was not increased in the azotemic dogs. Furthermore, there was also no correlation between increased concentrations of creatinine and urea nitrogen in serum and NCX-1 expression in the RF group.

Conclusions and Clinical Importance: Azotemia likely does not affect NCX-1 expression.


chronic mitral valve insufficiency


International Renal Interest Society


International Small Animal Cardiac Health Council


sodium-calcium exchanger


N-terminal proANP


N-terminal proBNP

The discovery of reliable cardiac biomarkers has led to an increase in the popularity of molecular diagnostics as a mainstream tool for detecting heart diseases in modern medicine. These cardiac biomarker assays are routinely performed on a very small amount of peripheral blood. The assays are used for a broad number of applications such as to screen for asymptomatic heart disease, to monitor therapeutic response (including cardiotoxicity of some anticancer drugs), and for rapid differential diagnosis of respiratory diseases in humans.1

Thus far, assay development has focused on circulating biochemical markers related to pathological processes of heart failure (HF), specifically, their presence and level of gene expression.2–6 Circulating concentrations of the cardiac biomarkers N-terminal proANP (NT-proANP) and N-terminal proBNP (NT-proBNP) were found to be significantly higher in dogs with severe HF.7 Furthermore, these biomarkers also were used to differentiate between cardiac and respiratory disease.7 However, the biomarkers NT-proANP or NT-proBNP are not suitable for discrimination in patients with renal disease because their concentration can be affected by renal failure (RF).8–9

Recently, more biomarkers targeting different aspects of the pathophysiological processes of heart disease have been discovered. These include the sodium-calcium exchanger (NCX-1), phospholamban, and HS-1-associated protein X-1, genes associated with intracellular calcium homeostasis.10NCX-1 plays a critical role in cardiac excitation-contraction coupling and intracellular calcium concentration in cardiac myocytes.4,11–15 The expression of NCX-1 has been found to be significantly increased in mice with cardiac hypertrophy and HF.16–18 In addition, NCX-1 also was found to be upregulated in dogs with moderate to severe chronic mitral valvular insufficiency (CMVI) but not in the mild groups.14 These findings indicated that NCX-1 has potential as a marker for monitoring CMVI progression and severity.

Expression of some cardiac biomarkers is known to be influenced by RF, a finding that limits their reliability as cardiac biomarkers in patients that also have renal disease. NCX-1 is expressed in kidneys19 and thus also may be increased in the peripheral blood of animals with RF or azotemia. Consequently, this may significantly decrease the diagnostic value of NCX-1 in dogs with CMVI, but there has been no evidence demonstrating such is the case. Therefore, the aim of this study was to evaluate the effect of azotemia on NCX-1 expression in dogs with CMVI.

Materials and Methods

Study Population

Before the study, we obtained approval of the animal ethics committee of Kangwon National University to obtain blood samples from healthy dogs for RNA extraction. Informed written consent for sample collection including information pertinent to our investigation was obtained from the dog owners before beginning the study. This animal testing program, including animal care, euthanasia, and disposal of dead animals, was performed in strict adherence with the guidelines of the National Research Council of Korea.

Dogs selected for the study had either CMVI without other systemic diseases, had RF without heart disease or any other complications, or were healthy (control group). Fourteen dogs with CMVI, 15 dogs with RF, and 15 normal healthy dogs were enrolled in this study. CMVI was diagnosed based on echocardiographic evidence of regurgitant flow into the left atrium (LA) during the systolic period, enlarged LA, myxomatous degeneration of mitral valve leaflets, and the presence of mitral valve prolapse. The animals also were physically examined, including CBC, serum biochemistry, and urinalysis to screen for the presence of other concurrent diseases such as chronic inflammatory diseases, systemic diseases, and other types of heart diseases. RF was diagnosed based on clinical signs indicative of renal disease without other concurrent organ failure or heart disease in conjunction with biochemical test results (eg, serum creatinine and BUN concentrations) and urinalysis (eg, proteinuria and low urine specific gravity). All dogs considered to have chronic renal failure (CRF) had been diagnosed with CRF for at least 3 months or longer. All RF dogs had no detectable cardiac murmurs. Healthy control dogs all were normotensive and had no radiographic or echocardiographic evidence of structural heart disease and had no laboratory evidence of RF (ie, reference range of urine specific gravity, no proteinuria, normal serum creatinine, and BUN concentrations). The dogs with CMVI were classified according to the International Small Animal Cardiac Health Council (ISACHC) classification system, selecting animals in classes II–III (moderate to severe HF) for this study. All CMVI dogs were medicated with furosemidea (1–4 mg/kg depending on severity) and enalaprilb (0.5–1.0 mg/kg). Some dogs also were medicated with pimobendanc (0.3–0.5 mg/kg) and fed a salt-restricted diet.d The dogs with CRF were grouped based on the International Renal Interest Society (IRIS) stage classification and were selected at stages III–IV. Presence of proteinuria was determined by a urine strip test.e Some RF dogs were medicated with azodylf (1–4 capsules based on body weight) and enalaprilb (0.5–1.0 mg/kg) and fed a salt-restricted diet.g

Quantification of NCX-1 Expression by Real-Time PCR

The methods for total RNA extraction, reverse transcription, and cDNA synthesis are described in a previous report.14 The primer sequences of the target (NCX-1; forward: 5′-ctataaaaccatcgaagggact-3′ reverse: 5′-ctttcttctcactcatctccac-3′) and reference genes (HPRT; forward: 5′-gctggattatatcaaagcactg-3′ reverse: 5′-tacttttatgtcccctgttgac-3′) also were determined and described previously.14 The NCX-1 primer set spanned a 240 kb intron, resulting in the exclusive amplification of cDNA.

Real-time PCR analyses was performed with a commercial real-time PCR Kith with 4 μL cDNA in a Rotorgene 4000.i, 14 A melting curve analysis and agarose gel electrophoresis were performed after real-time PCR to confirm the amplicon size and absence of primer dimers. The melting peak and Ct values were determined with the Rotorgene data analysis software.j The relative expression level of NCX-1, normalized to the reference gene (HPRT), was analyzed by the inline image method.19 The amplification efficiencies of the target (NCX-1) and reference (HPRT) genes were used to validate the −ΔΔCt method by plotting the log cDNA dilution versus Ct and ΔCt. We also determined the intra- and interassay coefficients of variance, using 3 replicates per template concentration and 3 different experimental runs to ensure statistical significance, accuracy, and reproducibility. The data were expressed as the fold change in gene expression normalized to the reference gene and relative to the control group by the inline image analysis method.20

Statistical Analysis

The data were analyzed by the following method: ÄÄCt (mean ÄCt value of each affected group − mean ÄCt value of control group) values and ÄCt values áΔΔCt (mean value of ΔCt value of each affected group − mean ΔCt value of control group) values and ΔCt values.19 All values were expressed as mean ± SD. The statistical methods used were a 1-way analysis of variance and a Pearson correlation. A difference was considered significant at P-value of <.05. All analyses were performed with commercially available statistical software.k Equal variances between groups were tested by the F-test (variance ratio test). Hence, the statistical analysis of fold change for NCX-1 expression was performed using log values. The data were normally distributed according to a Kolmogorov-Smirnov normality test. Any observed statistical significance among groups was further analyzed by multiple comparisons (Dunnett's test and Tukey-Kramer test) to corroborate the observed significance. The correlation of serum creatinine, BUN, age, and body weight to the expression levels of NCX-1 was investigated by a univariate analysis.


Forty-four dogs were enrolled in this study in the following distribution: 15 healthy normal dogs, 14 CMVI dogs grouped as ISACHC II (n = 7), and III (n = 7), and 15 dogs with CRF classified as IRIS stage 3 (n = 10) and stage 4 (n = 5). The clinical features of all the dogs used in this study are summarized in Table 1. The dog breeds used consisted of Maltese (n = 12), Yorkshire Terrier (n = 8), Toy Poodle (n = 6), Shih Tzu (n = 6), Beagle (n = 4), miniature Schnauzer (n = 3), mixed breed (n = 2), and Chihuahua (n = 1). The average ages of the dogs in years were as follows: CMVI (II; 9.71 ± 2.21 and III; 11.67 ± 3.08), RF (3; 9.30 ± 4.52 and 4; 7.00 ± 3.16), and normal healthy group (3.46 ± 2.89). All RF dogs were proteinuric by a urine strip test. Systolic arterial pressure results of RF and CMVI dogs were summarized in Table 1.

Table 1.   Summary of the study population used in this study.
 ControlISACHC ClassIRIS Stage
  • BW, body weight; Rad, thoracic radiography; Echo, echocardiography; VHS, vertebral heart scale; FS, fraction shortening; LVIDs, left ventricular internal dimension in systole; LVIDd, left ventricular internal dimension in diastole; LA/Ao, ratio of dimensions of left atrium to aortic root; SAP, systolic arterial pressure; IRIS, International Renal Interest Society; ISACHC, International Small Animal Cardiac Health Council.

  • a

    Blood pressure measured by Doppler method.

Age3.46 ± 2.899.71 ± 2.2111.67 ± 3.089.30 ± 4.527.00 ± 3.16
SexF (8) M (7)F (4) M (3)F (5) M (2)F (6) M (4)F (4) M (1)
BW4.39 ± 2.053.94 ± 1.553.45 ± 1.724.57 ± 2.743.97 ± 2.81
BreedPoodle (3)Maltese (4)Maltese (4)Shih Tzu (2)Beagle (1)
Maltese (2)YT (1)YT (2)Beagle (1)Mixed (2)
Shih Tzu (3)Poodle (2)Shih Tzu (1)Maltese (2)YT (1)
YT (3)  Mixed (2)Chihuahua (1)
Beagle (2)  Schnauzer (1) 
Schnauzer (2)  Poodle (1) 
   YT (1) 
Blood work
 BUN (mg/dL)17 ± 320 ± 827 ± 11108 ± 57146 ± 31
 Creatinine (mg/dL)0.8 ± 0.10.9 ± 0.20.9 ± 0.23.5 ± 0.89.5 ± 3.6
 VHS9.1 ± 0.411.8 ± 0.812.3 ± 0.9  
 FS (%)43 ± 753 ± 751 ± 10  
 LVIDs (cm)1.1 ± 0.41.2 ± 0.41.3 ± 0.5  
 LVIDd (cm)1.9 ± 0.52.6 ± 0.52.7 ± 0.6  
 LA/Ao1.2 ± 0.11.8 ± 0.12.2 ± 0.4  
 SAP (mmHg)a131 ± 13145 ± 8143 ± 22138 ± 15147 ± 13

The expression levels of NCX-1 were significantly increased in the ISACHC II (2.99 ± 0.61) and III (4.35 ± 1.44; P < .01) groups, compared with the control group. There was no significant difference between the ISACHC II and III groups (Fig 1). NCX-1 expression levels in the IRIS stage 3 (0.95 ± 0.57) and stage 4 (1.57 ± 0.53) groups did not differ from the healthy control group. There was no significant difference between the IRIS stage 3 and stage 4 groups. NCX-1 expression in IRIS stage 3 and stage 4 dogs was lower than NCX-1 expression in ISACHC II and III dogs (P < .01). A univariate analysis of NCX-1 expression with respect to age (P= .684), body weight (P= .755), BUN (P= .191), and serum creatinine (P= .083) indicated no correlation (P < .05).

Figure 1.

 Altered expression levels of sodium-calcium exchanger (NCX-1) in the blood cells of each group. The expression levels of NCX-1 were plotted in box plots as fold changes relative to normal group. Asterisk (*), dagger (†), and circle (•) are utilized to indicate statistical significance. NCX-1 in International Small Animal Cardiac Health Council (ISACHC) classes II and III is significantly overexpressed compared with the control group (*), but not in the International Renal Interest Society (IRIS) groups 3 and 4. The significant difference in expression levels of NCX-1 between ISACHC (classes II, III) and IRIS (groups 3, 4) was documented (•, †). However, there was no difference in the expression levels between ISACHC classes II and III, and between IRIS groups 3 and 4. *Fold change of each group was calculated by inline image method. *A difference was considered significant at a value of P < .05.


Accurately determining disease stage and monitoring HF are critical factors for making a correct diagnosis and planning a therapeutic strategy for HF in small animals. Although advances in diagnostic imaging technologies have enabled the correct diagnosis of HF, this technology has limitations. This technique requires not only the use of expensive equipment, but also expert training in cardiology and radiology. Distinguishing dyspnea due to cardiac versus noncardiac causes (especially respiratory diseases) is a challenging task in general veterinary practice. Therefore, the use of cardiac biomarkers has become more prominent, not only for the early detection of heart disease, but also to differentiate between cardiac causes of dyspnea and respiratory causes without the aid of expensive diagnostic imaging.1

Thus far, several cardiac biomarkers, including cardiac troponins, natriuretic peptides, endothelin, and cytokines, have been evaluated in small animals with various heart diseases.12,21–26 The most widely used biomarker in veterinary medicine is NT-proBNP, used to differentiate cardiogenic dyspnea from noncardiogenic diseases and as indicator of the severity of cardiac disease.7,27–29 However, recent evidence has indicated that NT-proBNP expression can be influenced and increased in response to other systemic diseases such as RF.8,9 Therefore, recent studies have focused on finding biomarkers that are not influenced by other complications known also to be present in patients already diagnosed with HF.2–4

The increased expression patterns of NCX-1 in blood cells were first described in studies of humans with HF and in dogs with CMVI.4,14 In contrast, a microarray study of humans with HF did not find any increased expression.30 In this study, we investigated NCX-1 expression with respect to azotemia and CMVI. We found that expression levels of NCX-1 in IRIS stage 3 and stage 4 RF groups were not significantly different from the normal group, in contrast to other cardiac biomarkers, including NT-proBNP.

The NCX-1 is involved in calcium regulation and homeostasis in cardiomyocytes strongly influencing myocardial contractibility.31NCX-1 is highly expressed in the heart and there is a rapid up-regulation of NCX-1 in response to pressure overload by HF.3–4 Consequently, many studies have found that the change in NCX-1 expression contributed to the pathophysiology of HF.3–4 Although molecular and pharmacological aspects of NCX genes have been well documented in literature,31 many parts of the signalling pathway (especially through peripheral blood cells) related to NCX-1 and HF are relatively unknown. In this study, it is still unclear why NCX-1 is also overexpressed in peripheral blood cells in dogs with CMVI. One possible explanation is that NCX-1 in peripheral blood cells may be coexpressed with cardiac NCX-1 because of increased concentrations of stimulants (eg, phenylephrine, endothelin 1, angiotensin II, and some growth factors) of NCX-1 expression in the bloodstream in conjunction with the progression of HF.

To date, no studies have found NCX-1 expression to be increased in CRF in mammals, although NCX-1 expression could be increased by salt-dependent hypertension and ischemic-reperfusion renal injuries.32,33 These studies found that NCX-1 was involved in salt-dependent hypertension because of increased concentrations of calcium via endogenous cardiac glycosides and Na+, K+-ATPases in vascular smooth muscle cells.34 In contrast, the RF groups (all CRF subjects) did not show NCX-1 overexpression in comparison to the normal groups. Furthermore, the dietary influence (salt content) was minimized in this study because our disease groups (CMVI and RF groups) were fed salt-restricted diets. Although high blood pressure is known to affect NCX-1 expression,32 mean systolic blood pressures in the disease groups were not significantly different compared with the normal group. Although the level of NCX-1 expression in the IRIS stage 4 group was higher than in the control and IRIS stage 3 groups, this difference was not statistically significant (P > .01). In addition, this difference could not be solely attributed to hypertension, because many aspects of NCX-1-mediated hypertension remain relatively unknown. Therefore, the influence of blood pressure on the level of NCX-1 expression in this study was minimal.

In conclusion, this study was conducted to evaluate the expression levels of NCX-1 in peripheral blood samples from dogs with HF, with RF, and in healthy controls by quantitative PCR. The results indicate that azotemia likely does not affect NCX-1 expression.

Although our results have highlighted the value of NCX-1 as a potential biomarker, it is prudent to note the limitations of this study. The sample size in each group was small. Although we used reliable methods for analyzing mRNA expression levels (gene expression) with relative quantification, evaluating protein expression of these markers via protein assays would be useful. Additionally, because of difficulties in sample collection, we did not assess NCX-1 expression in dogs with concurrent HF and RF. Doing so may provide further insight. Furthermore, we could not evaluate the influence on NCX-1 expression of acute renal injury. Lastly, the effect of diuretics on NCX-1 expression was not addressed in this study. Although the HF groups were treated with 1–4 mg/kg of furosemide, the RF groups were not. Therefore, the effect of diuretics on NCX-1 expression in a controlled study population should be investigated further.


a Lasix, Handok Pharmaceuticals, Seoul, Korea

b Encard, Merial, Duluth, GA

c Vetmedin, Boehringer-Ingelheim, Ingelheim am Rhein, Germany

d h/d; Hill's Pet Nutrition Inc, Topeka, KS

e Labstrip U11plus, Inter Medico, Markham, ON, Canada

f Vetoquinol, Lure cedex France

g k/d; Hill's Pet Nutrition Inc

h QuantiTect SYBR Green PCR kit, Qiagen, Valencia, CA

i Real-time cycler, Corbett Research, Sydney, Australia

j Rotor gene 6.0 software program, Corbett Research

k SPSS 12.0.0, SPSS Inc, Chicago, IL


This study was supported by research grants from the Korea Sanhak Foundation (120090588) and the Institute of Veterinary Science (KNU). Authors thank Dr Lopeti Lavulo (BIOLINE Australia, Sydney, Australia) for the advice on manuscript preparation.