Kidney urinary biomarkers in patients with branched‐chain amino acid and cobalamin metabolism defects

There is a clinical need for early detection of chronic kidney disease (CKD) in patients with organic acidurias. We measured kidney markers in a longitudinal study over 5 years in 40 patients with methylmalonic aciduria (Mut0), propionic aciduria (PA), cobalamin A (CblA), and cobalamin C (CblC) deficiencies. Neutrophil gelatinase‐associated lipocalin (NGAL), calprotectin (CLP), kidney injury molecule‐1 (KIM‐1), dickkopf‐3 (DKK‐3), albumin and beta‐2‐microglobulin (B2MG) in urine, as well as cystatin C (CysC) in serum were quantified. In Mut0 patients, mean concentrations of B2MG, KIM‐1, and DKK‐3 were elevated compared with healthy controls, all markers indicative of proximal tubule damage. In PA patients, mean B2MG, albumin, and CLP were elevated, indicating signs of proximal tubule and glomerulus damage and inflammation. In CblC patients, mean B2MG, NGAL, and CLP were increased, and considered as markers for proximal and distal tubule damage and inflammation. B2MG, was elevated in all three diseases, and correlated with DKK‐3 in Mut0/CblA and with eGFR(CysC) and KIM‐1 in PA patients, respectively. None of the markers were elevated in CblA patients. Significant deterioration of kidney function, as determined by steady increase in CysC concentrations was noted in seven patients within the observation period. None of the investigated biomarker profiles showed a clear increase or added value for early detection. In conclusion, we identified disease‐specific biomarker profiles for inflammation, tubular, and proximal damage in the urine of Mut0, PA, and CblC patients. Whether these biomarkers can be used for early detection of CKD requires further investigation, as significant kidney function deterioration was observed in only a few patients.


| INTRODUCTION
Methylmalonic acidemias (MMA) and propionic acidemia (PA) are caused by a disturbance in the catabolism of branched-chain amino acids.MMAs are due to a complete (Mut 0 subtype) or partial (Mut À subtype) deficiency of the mitochondrial methylmalonyl-CoA mutase (EC 5.4.99.2;OMIM #251000) or due to defects in the uptake, transport, or synthesis of its cofactor 5 0 -deoxyadenosylcobalamin (cblA, cblB or cblD variant 2 type; OMIM #607481, #251110, #251100, #277410).Deficiency of the mitochondrial propionyl-CoA carboxylase (EC 6.4.1.3;OMIM #606054) results in PA.CblC deficiency is the most common inborn error of intracellular vitamin B 12 metabolism (Figure 1; OMIM #277400).In patients with Mut 0 and PA, and to lesser extend also with CblA and CblC, 2-methylcitric acid (MCA) accumulates in tissues and body fluids. 1,2Higher MCA levels are seen in patients with severe phenotype and significant long-term complications, and MCA levels are generally higher in patients with PA than in Mut 0 . 3MCA increases during metabolic decompensation 4 and plasma MCA correlates with disease burden. 57][8] Chronic renal failure (CKD) is the most important organ system long-term complication in MMA, Mut 0 , and CblB patients, 9 while CKD is a rare complication in PA, CblC, and CblA in childhood.In an adult American cohort of PA patients, CKD has been manifested in half of the patients, 10 in CblC, kidney involvement occurs preferentially in the lateinfantile or adult-onset form. 11A cross-sectional data from the European Registry revealed that CKD was less frequent in CblA 12 occurring primarily in adult patients. 13In clinical practice, glomerular filtration rate (GFR) estimation and albuminuria are widely used for CKD diagnosis and prognosis.GFR is often approximated by measuring serum CysC concentrations, 14 and urinary B2MG excretion is considered as a marker of tubular damage.However, these traditional biomarkers only increase when kidney damage is already advanced. 15Early identification of CKD is important to predict and prevent or slow CKD progression. 16There is an urgent clinical need of additional diagnostic non-invasive tools that allow us to determine kidney disease. 3Several new biomarkers for early detection of deterioration of kidney function have been identified and studied in several patient cohorts with kidney involvement. 17,18e, therefore, investigated three novel markers of tubular damage, i.e. neutrophil-gelatinase-associated lipocalin (NGAL), kidney injury molecule-1 (KIM-1), dickkopf 3 (DKK-3), [19][20][21][22] albumin as a well-established marker of glomerular injury and beta-2-microglobulin (B2MG), as a well-established marker of proximal tubule F I G U R E 1 Metabolic overview.Metabolic pathways affected by methylmalonic acidemia (Mut 0 ), propionic acidemia (PA), and cobalamin A (CblA) and C (CblC) deficiency.Adapted from Morath et al. 60 damage.In addition, we investigated calprotectin (CLP) as an immunological marker released during a variety of inflammatory processes. 23Furthermore, MCA was included as a marker for severe phenotype and disease burden.These urine markers were then related to serum CysC and eGFR (CysC).This approach should provide information on whether the new markers offer an advantage in early diagnostics of CKD.

| Study design and population
A longitudinal study was conducted at Heidelberg University Children's Hospital, Germany, from 2017 to 2022.Urine and serum samples from patients with methylmalonic aciduria (Mut 0 ; n = 11), Propionic aciduria (PA; n = 15), Cobalamin A (CblA; n = 5), and Cobalamin C (CblC; n = 9) deficiencies were collected during routine metabolic follow-up every 6-12 months resulting in 128 urine samples and 106 serum samples from 40 patients.Samples were collected in well-state and never during metabolic crisis.Diagnosis for all patients was genetically confirmed.
The control group (n = 49) consisted of apparently healthy children and adolescents aged 0-18 years.Urine from healthy children and urine and serum samples from patients were collected, centrifuged and the supernatant was stored frozen (À20 C).This study was approved by the ethic committee of the University Heidelberg (S-436/2016).

| Measurement of urinary and serum markers
Calprotectin was measured using an enzyme-linked immunosorbent assay (ELISA) kit (Biovendor R&D; Czech Republic; Catalog No. RD191217100R), KIM-1 was measured with the KIM-1 (human) ELISA kit (Enzo Life Sciences GmbH, Lörrach, Germany; Catalog No. ADI-900-226-0001) and NGAL with the NGAL Rapid ELISA kit (BioPorto Diagnostik, Gentofte, Denmark; Catalog No. 036CE).DKK-3 was measured using an enzyme-linked immunosorbent assay (ELISA) kit (Dia Ren UG; Homburg).2-methylcitric acid (MCA) was measured by organic acid analysis carried out by gas chromatography-mass spectrometry. 24Urinary B2MG and albumin were measured immunoturbidimetrically as described by Martin et al. (2011).All urinary markers were related to creatinine.Serum CysC was measured immunoturbidimetrically as described. 25Reference values were defined by the Central Laboratory of University Hospital Heidelberg.

| Statistical analysis
All results are presented as mean ± SD for continuous variables, however median for ordinal scaled data.For showing differences in biomarker levels between groups oneway ANOVA with Tukey's test was performed using GraphPad Prism version 9.4.1 for macOS, GraphPad Software, San Diego, California USA.Age-dependent effects of urinary biomarker concentrations were analyzed by multiple regression with biomarker as dependent variable, and age in years and disease as independent variables.

| Urinary biomarker concentrations in the patient cohort
Urine concentrations of albumin, B2MG, NGAL, CLP, KIM-1, and DKK-3 were determined in 49 urine samples from healthy children.128 urine samples from 40 patients were collected over a period of 5 years and analyzed.Multiple regression analysis of CLP, KIM-1, and DKK-3 concentrations revealed no age-dependent effects in disease groups and respective controls.Borderline, opposing effects were observed for NGAL in Mut 0 and CblC patients, probably caused by the small sample size (Figure S1).For 22 patients, multiple urine samples (up to 8 samples per patient) were available.The data were analyzed both in the entire collective (as shown in Figure 2) and compared to the averaging of the values per patient (Table S1).The biomarker concentrations were both normalized to creatinine (crea), the nonnormalized data can also be found in Figure S2.
Compared to healthy subjects, mean albumin was increased by twofold in Mut 0 and PA and B2MG up to fourfold in those patients (Figure 2A).None of the biomarkers were elevated in CblA patients.Mean urinary excretion for NGAL was sevenfold higher in CblC patients, mean CLP was sevenfold higher in PA and CblC patients, mean KIM-1 was twofold higher in Mut 0 and PA patients and mean DKK-3 up to 30-fold higher in Mut 0 patients (Figure 2B, Table 1).Comparing the patient groups with each other, mean urinary NGAL excretion was two-, three-and eightfold higher in CblC compared with samples from Mut 0 , PA, and CblA patients, respectively (Table 1).Three of the CblC patients, including an 18-year-old patient with late-onset presentation of hemolytic uremic syndrome history at 2 and 5 years of age, had elevated CysC levels since study enrollment.The NGAL and CLP concentrations in these three patients were in the same range as for the patients with normal CysC concentrations (119 ± 53 vs. 312 ± 203 NGAL pg/mg creatinine and 68 ± 73 vs. 72 ± 65 CLP ng/mg creatinine; both p = ns).CLP excretion was approximately 3-, and 20-fold higher in PA and CblC compared with Mut 0 and CblA patients.KIM-1 excretion was approximately twofold higher in Mut 0 and PA patients compared with CblC patients (Figure 2B, Table 1).The mean value for DKK-3 in Mut 0 patients was highly variable, due to individual patients with generally high and patients with low DKK-3 concentrations.Two or more urine samples from the same patient were available from six patients with Mut 0 .Half of the patients had elevated mean DKK-3 concentrations (1193 ± 1061 mg/mg crea, n = 7; 35 014 ± 22 474 mg/mg crea, n = 2; 7655 ± 8946 mg/mg crea, n = 3), and the mean eGFR (CysC) was 26, 26, and 8 mL/min/1.73m 2 in the patients, respectively.In the other three patients, mean DKK-3 levels were within the normal range (6.3 ± 1.7 mg/ mg crea, n = 4; and 7.3 ± 6.1 mg/mg crea, n = 5; 2.7 ± 1.2 mg/mg crea, n = 3), and mean eGFR (CysC) was 64, 49 and 26 mL/min/1.73m 2 , respectively.None of the Mut 0 patients showed a clear increase in DKK-3 concentration over time.Neither DKK-3, KIM-1, and B2MG concentrations correlated with urinary albumin, only DKK-3 correlated with B2MG in those patients (Table S3).
The highest mean urinary MCA concentrations were found in PA patients, which were almost fourfold higher compared to Mut 0 patients.MCA concentration did not correlate with biomarker concentrations (Figure S3).Combining the distribution of conventional and new kidney biomarkers revealed marker-specific patterns (Table 2).

| Serum CysC levels as a marker of kidney function
The patients' serum samples were divided into two groups: in patients with eGFR(CysC) above and below  60 mL/min/1.73m 2 , indicating mild to severe loss of kidney function.In the patient cohort with eGFR(CysC) below 60 mL/min/1.73m 2 , albumin, B2MG, NGAL, and DKK-3 were elevated compared to the control group.Mean NGAL and CLP concentrations were also higher compared to controls for patients with eGFR(CysC) above 60 mL/min/1.73m 2 (Figure 3).Correlation of biomarkers with eGFR(CysC) revealed a significant result for B2MG in PA patients and for KIM-1 in CblA patients (Figure 4).

| Urinary biomarkers in progression of CKD
During the observation period, only 2 of 11 Mut 0 , 1 of 15 PA, and 2 of 9 CblC patients developed for the firsttime kidney function deterioration based on CysC elevation above the age-related normal range (Table S4).In those five patients, biomarker concentrations fluctuated and showed no clear increase over time (Figure S4).Two further patients, one Mut 0 and one PA, with increased CysC concentrations from the beginning of the study, developed significant disease progression during the observation time, measured by steadily increasing CysC concentration (from 2.3 to 3.8 over 3.11 years and from 1.8 to 2.7 over 4.5 years, respectively).Based on the small sample size, no statement can be made on prediction of kidney disease development or even progression in these patient cohorts.

| DISCUSSION
Early diagnosis and treatment of chronic kidney disease (CKD) in patients with organic acidurias and treatment are two important measures to prevent further deterioration of renal function and to delay adverse consequences.CKD is T A B L E 2 Disease-specific pattern of the conventional and new urinary markers when compared to a healthy control group.defined as a gradual loss of kidney function due to structural or functional abnormalities with or without a decreased glomerular filtration rate. 29Measured glomerular filtration rate (mGFR) is considered the gold standard for assessing kidney function, but is rarely used due to its invasive nature, high cost and time. 30An overestimation of mGFR could be demonstrated for creatinine-based calculations in adolescent and adult MMA patients. 9][33][34] Serum creatinine is subject to individual variation and is an unreliable marker for estimating renal function in pediatrics due to dependence on muscle mass and age 35,36 Especially in CblC deficiency, borderline reduced serum creatinine concentrations have been observed. 37ormalization of biomarker concentrations to urinary creatinine levels is a common practice, but it should be noted that in CKD and in disorders of creatinine metabolism, such as in CblC deficiency, may lead to lower urinary creatinine levels.For this purpose, we also performed the analysis without creatinine normalization, which revealed similar but less significant results.We further provided the calculation of biomarker concentrations by averaging all samples and averaging values per patient in case of multiple samples per patient.Averaging of all samples carries the risk of skewing the data toward the findings of a smaller subset of patients who happen to have more frequent sample collection.Both types of calculation revealed a comparable disease-specific biomarker profile, especially for NGAL, CLP, KIM-1, and DKK-3.
In addition, the two urinary markers, albumin, and B2MG, were used to detect glomerular and tubular renal damage.Despite its prognostic significance, tubular disease cannot be reliably detected with standard clinical measures of glomerular health (eGFR and UACR) and is invisible to the clinician except in the rare instances when biopsies are obtained. 38The pathomechanisms of CKD development in the studied disease groups are only partly understood, but seem to be based on different mechanisms.Patients with branched-chain organic acidurias, that is, Mut 0 , CblA, and PA, develop chronictubulo-interstitial nephritis (cTIN), demonstrated in human 13,39 and murine 40 kidney biopsies.A severe disturbance of mitochondrial network especially in renal proximal tubule cells has been suggested as a causative pathomechanism of kidney damage. 41,42In CblC deficiency, kidney involvement occurs preferentially in the late-infantile or adult-onset form, in most cases as acute or chronic thrombotic microangiopathy (TMA), sometimes resulting in hemolytic uremic syndrome or even chronic renal failure. 43The underlying pathophysiological mechanism remains fairly unknown, hyperhomocysteinemia and elevated MMA appear not to be the responsible nephrotoxins.KIM-1 is expressed at low levels in both normal kidney and other organs, but its expression is upregulated in proximal tubule cells in kidney after post-ischemia/ reperfusion injury.It has been used as a sensitive indicator for early detection of renal tubular injury in patients with acute kidney injury. 20Urinary KIM-1 is closely related to tissue KIM-1 levels and correlates with tissue damage. 20,45In CKD, the role of KIM-1 is controversially discussed.Studies in patients with diabetes and hypertension found no correlation of urinary KIM-1 with lowering eGFR, 46,47 whereas in the systolic blood pressure intervention study, urinary KIM-1 was associated with an increased risk for CKD progression. 48,49After acute kidney injury urinary KIM-1 has been associated with a significantly increased risk for CKD development 50 Mean urinary KIM-1 concentrations were higher in Mut 0 patients but in no other patient cohort (Figure 2).A novel and promising urinary biomarker, particularly for the detection of ongoing tubular injury and early stages of CKD, is the glycoprotein DKK-3, a stress-induced, renal tubular epithelial cell-derived, pro-fibrotic molecule. 51It promotes tubulointerstitial fibrosis, a histopathological feature of Mut 0 and CblA. 13,39,40rinary DKK-3 is able to predict acute kidney injury 21,52 and identifies patients at high risk for shortterm progression of CKD, regardless of the cause of kidney disease, baseline kidney function, and albuminuria 51 and indicates a short-term risk of severe kidney function impairment in children with chronic kidney disease. 53KK-3 was significantly elevated in some, but not all, Mut 0 patients and correlated with B2MG but not with albumin excretion (Table S3).In six Mut 0 patients who already had impaired renal function (eGFR 8-64 mL/min/m 2 ) since study inclusion, there tended to be a relationship between the severity of renal function impairment and the increase of urinary DKK-3 concentration.However, in the two Mut 0 patients who showed an increase in CysC for the first time during the study period, no increase in urinary DKK-3 concentrations was observed (Table S4).
This needs to be verified in a larger cohort.The partially large scatter of urinary DKK-3 values in Mut 0 patients is possibly due to a binding of plasma components preventing the crossing of the glomerular barrier. 51ean NGAL concentrations were only significantly increased in the CblC group.In Mut 0 and PA patients, NGAL concentrations were heterogenous with mean concentrations higher than in respective controls, albeit not significantly.NGAL is considered as a marker for distal tubule, is reabsorbed by the proximal tubules and released from damaged distal tubules in the setting of acute tubular injury.It can be detected within hours of tubular injury, even in the absence of functional acute kidney injury. 54lthough several studies support the use of NGAL as an early marker for acute kidney injury, its utility as a marker for CKD is uncertain and disease-dependent. 206][57][58] A previous study showed a strong correlation of plasma NGAL with creatinine and eGFR (CysC) in Mut 0 MMA subjects and Mut À/À;TgINS-Alb-Mut mouse model. 40The high concentrations of urinary NGAL and CLP in CblC patients might indicate distal tubular damage and inflammation processes in these patients but this needs further investigations.For NGAL, a correlation could be shown with hyperhomocysteinemia in kidney-transplanted adult patients as well as in association with vascular damage. 59This is another indicator that the biomarker profiles of the investigated disease groups are influenced by the underlying pathophysiology.Three of the CblC patients had elevated CysC concentrations; two of them had an early-onset and one had a late-onset presentation with two episodes of HUS and were all treated with parental hydroxocobalamin, betaine, and folinic acid after diagnosis.
Mean urinary CLP levels, a mediator protein of the innate immune system, were especially elevated in PA and CblC and to a lesser extend in Mut 0 patients.Urinary CLP values were six times higher in structural acute kidney injury than in functional acute kidney injury. 23In CblC patients, Lemoine et al demonstrated IgM deposits of the glomerular basement membrane, in addition to arteriolar and glomerular lesions of thrombotic microangiopathy, indicating immunological involvement in CblC patients. 11However, Seibert et al found no difference of CLP excretion between primary inflammatory and non-inflammatory etiologies of CKD, although the differentiation between "completely inflammatory and noninflammatory" renal disease is difficult. 58It remains unknown why CLP is less elevated in Mut 0 patients compared to PA patients, immunological tubulointerstitial involvement was detected in both diseases. 13,34n the patient cohort studied, only five patients experienced an increase in age-appropriate CysC via the upper cut-off during the observation period.No increase in urine concentration with the onset of CysC-based renal impairment could be identified for any of the biomarkers investigated, including the conventional markers albumin and B2MG (Table S4), whereas elevated CyS and eGFR(CysC) correlated with increased urinary excretion of albumin, B2MG, and DKK-3.Due to the very small number of patients with an early onset of CKD (5 patients) and progression of CKD (2 patients) during the observation period, no statement can be made on the suitability of the investigated novel biomarkers as early indicators of CKD or as a marker for disease progression on the basis of these data.Future investigations of serum biomarker concentrations, for example, serum NGAL concentrations, 40 in comparison to urinary biomarker concentrations in the described patient's cohort, would be of interest and may lead to a better understanding of the detected biomarker profiles.

| CONCLUSION
In conclusion, a disease-specific biomarker profile in patients with branched-chain amino acid and cobalamin metabolism defects with a risk of CKD was found, consisting of markers for proximal, distal tubule damage and/or an inflammatory component.More samples are needed to evaluate the diagnostic value of investigated kidney biomarkers for early detection of CKD and disease progression in patients with Mut 0 , PA, CblA, and CblC.

F
I G U R E 4 Correlation of urinary markers with eGFR(CysC).(A) Urinary concentration of B2MG and albumin and (B) urinary concentrations of NGAL, CLP, KIM-1, and DKK-3 were correlated with eGFR(CysC).In PA patients, B2MG correlated with eGFR(CysC) decline.No clear correlation could be demonstrated for the other investigated markers.