Sage Journals: Discover world-class research

Abstract

Background

The primary hyperoxalurias are inherited disorders of glyoxylate metabolism, which cause over-production of oxalate leading to urolithiasis and subsequent renal failure. Other metabolites may be produced in excess in the different forms of PH: glycolate in PH1, glycerate in PH2 and 4-hydroxy-2-oxoglutarate and 2,4-dihydroxyglutarate in PH3. The aim of this study was to set up and validate a method for the simultaneous analysis of these metabolites in urine and to evaluate its use for preliminary identification of primary hyperoxaluria prior to definitive diagnosis by genetic testing.

Methods

Urine samples were derivitized by methoximation and silylation and extracted into organic solvent prior to analysis by gas chromatography mass spectrometry.

Results

Recovery of the analytes spiked into urine ranged from 91 to 103% and total analytical imprecision ranged from 3.0 to 13.6%. 4-Hydroxy-2-oxoglutarate was unstable in urine at room temperature, and preservation by acidification was required. Mean urinary glycolate, glycerate and 4-hydroxy-2-oxoglutarate or 2,4-dihydroxyglutarate (expressed as a ratio to creatinine) were significantly higher in patients with PH1, PH2 and PH3, respectively. Low 4-hydroxy-2-oxoglutarate was observed in some patients with PH3, probably due to the instability of this analyte, but all PH3 patients had elevated 2,4-dihydroxyglutarate. During five months of routine service, seven cases of PH were identified by this method and subsequently confirmed by gene sequencing including two with novel mutations in HOGA1.

Conclusions

This study confirms that the method is useful in aiding the diagnosis of primary hyperoxaluria and can direct genetic testing.

Keywords

Renal disease inborn errors of metabolism evaluation of new methods

Introduction

The primary hyperoxalurias (PH) are rare inherited metabolic diseases of glyoxylate metabolism, which cause excess production of oxalate and thus lead to increased oxalate excretion in the urine.¹ Hyperoxaluria is a well-recognized risk factor for calculus formation² because of the very low solubility of calcium oxalate, leading to urine supersaturation and crystallization. Patients with PH suffer renal damage directly via oxalate toxicity to renal tubular cells and indirectly as a result of the urolithiasis.³ The disease is progressive and can lead to end-stage renal failure (ESRF), after which the excess oxalate starts to be deposited around the body (systemic oxalosis). The first-line test for PH is urine oxalate, but this test is not diagnostic because hyperoxaluria is more commonly secondary to exogenous oxalate, either from increased intake (oxalate is found in many foodstuffs) or due to fat malabsorption.⁴ Even when secondary causes have been excluded, definitive diagnosis of PH requires gene sequencing (for all three types) or enzyme activity analysis in a liver biopsy (for PH1 and PH2). Genetic testing employing Sanger sequencing methodology has a relatively high cost per test, and liver biopsy is an invasive procedure with an associated, albeit low, burden of morbidity. Therefore, a ‘diagnostic testing gap’ exists between the initial screening test and the definitive tests. Other urine metabolites have been shown to be present in excess in these diseases including glycolate in PH1,⁵ glycerate in PH2,⁶ and 4-hydroxy-2-oxoglutarate (HOG) and 2,4-dihydroxyglutarate (DHG) in PH3.⁷

PH type 1 (Online Mendelian Inheritance in Man [OMIM] #259900) is caused by a deficiency of the peroxisomal enzyme alanine glyoxylate aminotransferase (AGT), encoded by AGXT, which catalyses the conversion of glyoxylate to glycine. Deficiency of AGT leads to a build-up of glyoxylate, which is converted to oxalate by lactate dehydrogenase (LDH) and to glycolate by glyoxylate reductase (GR). Elevated urine glycolate has been described in approximately three-quarters of cases of PH1,⁵ but it can also be derived from dietary sources.⁸ It has been quantified colourimetrically⁹ and enzymatically using glycolate oxidase (GO).¹⁰ Glycolate can also be detected by organic acid analysis using gas chromatography (GC);¹¹ however, it is poorly extracted into the solvent systems normally used.¹²

PH type 2 (OMIM 260000) is caused by mutations in GRHPR, which encodes GR. GR has both glyoxylate and hydroxypyruvate reductase activities and deficiency thus leads to accumulation of both substrates. LDH converts the excess substrates to oxalate and l-glycerate, respectively. Glycerate can be detected on organic acid analysis by gas chromatography mass spectrometry (GCMS); however, like glycolate, it is very polar and poorly extracted into organic solvents.¹² It is specifically l-glycerate and not d-glycerate that is produced in PH2, the latter being produced in excess in the unrelated condition of glycerate kinase deficiency. In the original description of PH2, glycerate was measured using a colourimetric method and the l-enantiomer determined by its optical rotation.⁶ Enantio-specific oxidation using LDH, followed by derivitization of the produced hydroxypyruvate and analysis by high-pressure liquid chromatography (HPLC) has been described.¹³ Analysis is also possible by liquid chromatography tandem mass spectrometry using a chiral column, which allows simultaneous determination of both enantiomers.¹⁴ However, PH2 and glycerate kinase deficiency have very different phenotypes and should not be mistaken for one another clinically, and it is not necessary to have measurement of specifically the l-enantiomer of glycerate in this setting.

PH type 3 (OMIM 613616) is caused by mutations in HOGA1, encoding the mitochondrial enzyme 4-hydroxy-2-oxoglutarate aldolase (HOGA). This enzyme catalyses the cleavage of HOG, a metabolite of collagen degradation, into pyruvate and glyoxylate. It has been demonstrated that the PH3 phenotype is associated with loss-of-function mutations in HOGA1.¹⁵ The mechanism by which oxalate over-production occurs in PH3 is not definitively known. The accumulation of HOG may be involved in the pathogenesis; hypotheses include extra-mitochondrial degradation of HOG to oxalate⁷ and inhibition of GR by HOG, which would cause a PH2-like state.¹⁶ Outcomes in PH3 are thought to be more favourable,¹⁷ and only one case of PH3 has been reported to have ESRF.¹⁸ However, the level of hyperoxaluria in PH3 overlaps with PH1 and PH2,¹⁹ so the natural history may not yet be clearly defined for this relatively newly described disease. Prevalence estimates from public whole exome sequencing in the USA give a figure of ∼7 per million for PH3,¹⁸ which is much higher than the reported prevalence of PH1 (∼1 per million in Europe²⁰) or PH2. It is possible that there are a significant number of undiagnosed cases of PH3 amongst renal stone patients, who perhaps have not been worked up for PH because of a lack of renal damage. Measurement of HOG and its reduced metabolite DHG in urine is possible using GCMS.⁷ An alternative method for HOG by ion-exchange chromatography coupled to mass spectrometry that does not require derivitization has been published.¹⁶ At present these investigations are not in routine use in the UK, and gene sequencing is the only available test for diagnosing PH3.

The aims of this study were to set up a method for the measurement of the urinary PH metabolites (PHM) glycolate, glycerate, HOG and DHG for use in the preliminary investigation of patients with hyperoxaluria of unknown cause, to evaluate its ability to differentiate between the three forms of PH and to demonstrate that it can be used for case detection.

Methods

Calibrator preparation

Stock solutions of glycolate (glycolic acid >99%, Sigma-Aldrich), glycerate (l-(−)-glyceric acid hemicalcium salt monohydrate ≥97.0%, Aldrich), HOG (dl-4-hydroxy-2-ketoglutarate disodium salt 85.5%, Santa Cruz) and DHG (synthesized by the Department of Chemistry, University College London from HOG as previously described)⁷ were made in deionized H₂O. Calibrators were prepared in urine which had been shown, by the method of standard additions, to contain 103.1 ± 0.1 µmol/L-glycolate, and no glycerate, HOG or DHG. The five-point calibration ranged up to 761 µmol/L for glycolate, 699 µmol/L for glycerate, 205 µmol/L for HOG and 3047 µmol/L for DHG. The working internal standard solution contained 0.1 g/L dimethylglutarate (DMG, 3,3-dimethylglutaric acid 98%, Aldrich) and 0.8 g/L d2-glycolate (glycolic acid 2,2-D2 98%, Cambridge Isotope Laboratories) in D₂O (deuterium oxide 99.9 atom % D, Aldrich).

Sample analysis

Where samples had been preserved by acidification, they were adjusted to pH 7–9 prior to analysis using 1 mol/L NaOH, added drop wise. The analysis method was adapted from Belostotsky et al.⁷ To 200 µL of sample or calibrator was added 20 µL of methoximation reagent (a 1 mol/L solution of methoxyamine hydrochloride, Aldrich, adjusted to pH 9 with 25% ammonia, VWR) and 10 µL of internal standard solution. Tubes were incubated at 60℃ for 1 h and then allowed to cool to room temperature. To each tube was added 10 µL of 6 mol/L HCl, enough sodium chloride (≥99.0% Sigma Aldrich) to saturate and 1 mL of a 5:1 mixture of ethyl acetate and propan-2-ol (both AnalaR Normapur, VWR). After vortex mixing for 1 min to extract the compounds and centrifugation at 3000 rpm for 5 min to aid phase separation, the organic fraction was removed to another glass tube and evaporated to dryness. To the dried extracts were added 75 µL of silylation reagent (BSTFA + TMCS 99:1, Supelco) and 20 µL of pyridine (Emsure ACS reagent grade, Merck). Tubes were incubated at 60℃ for 40 min and then allowed to cool to room temperature. Each tube was diluted with 100 µL of cyclohexane (AnalaR Normapur, VWR), centrifuged to remove any precipitate and the supernatant transferred to a glass GC vial. A Shimadzu GCMS-QP2010S instrument with an HP-5 ms capillary column was used with helium carrier gas at a total flow rate of 5.4 mL/min (76.5 kPa). The sample (1 µL) was injected in splitless mode. The column temperature was held at 66℃ for 1 min, and then ramped at 4℃/min to 100℃, followed by 12℃/min to 220℃ and 20℃/min to 300℃. Total run time was 30 min per injection including equilibration between samples. The detector was operated in single-ion monitoring mode. Urine creatinine was measured on a Cobas 8000 autoanalyser (Roche diagnostics GmbH, Mannheim, Germany) using a kinetic Jaffé reaction according to the manufacturer’s instructions.

Data analysis

Chromatograms (an example is shown in Figure 1) were analysed by the Shimadzu post run analysis software and peak integration carried out automatically with a minimum slope of 50 min⁻¹ and minimum width of 1 s. The analytes of interest were identified by the parameters in Table 1. DHG appears as two peaks: one for the mesoisomer (2S,4R-DHG) and one for the mixture of stereoisomers (2R,4R- and 2S,4S-DHG). The latter peak was chosen for quantitation because it was found to consistently have much larger peak areas in urine from patients with PH3. Results were analysed using Excel (Microsoft Corporation, Redmond, USA). Imprecision was estimated by ANOVA using the method validation statistics templates available from the Association for Clinical Biochemistry and Laboratory Medicine website,²¹ and expressed as a coefficient of variation (CV). Means of groups were compared using Student’s t test, and a P value < 0.05 was considered to be statistically significant. Diagnostic sensitivity and specificity were calculated using gene sequencing as the gold standard.

Figure 1.

Example chromatogram.

Table 1.

Analyte retention times and ions.

Analyte	Retention time (min)	Molecular ion mass	Quantitation ion (m/z)	Qualification ion (m/z)
d2-Glycolate	8.75	222	207.10	179.10
Glycolate	8.80	220	205.10	177.10
Glycerate	14.05	322	307.25	205.10
DMG	15.35	304	289.25	186.10
DHG (mesoisomer)	18.17	452	Not quantified
DHG (stereoisomers)	18.28	452	319.20	245.20
HOG	18.32	407	290.25	392.20

HOG: 4-hydroxy-2-oxoglutarate; DHG: 2,4-dihydroxyglutarate; DMG: dimethylglutarate.

Genetic testing

Gene sequencing of AGXT, GRHPR and HOGA1 was carried out on DNA extracted from leucocytes as previously described.^15,22,23 Numbering was based on cDNA sequence NM_000030.2 (AGXT), NM_012203.1 (GRHPR) and NM_138413.3 (HOGA1) with nucleotide #1 denoting the first coding base. Sequence variants were described using Human Genome Variation Society nomenclature (www.hgvs.org/). Novel sequence variants were analysed in silico using Mutation Taster²⁴ (http://www.mutationtaster.org/), SIFT²⁵ (http://sift.jcvi.org/www/SIFT_BLink_submit.html) and Polyphen-2²⁶ (http://genetics.bwh.harvard.edu/pph2/).

Results

Accuracy and assay linearity

Accuracy was estimated by spiking two urine samples (with low and high concentrations of each analyte) using 10 g/L aqueous standard solutions and calculating the recovery from quadruplicate analysis. Recovery was: 103 ± 2.6% (368 and 597 µmol/L) for glycolate, 101 ± 7.4% (215 and 370 µmol/L) for glycerate, 93 ± 3.2% (85 and 144 µmol/L) for HOG and 91 ± 4.6% (1696 and 2519 µmol/L) for DHG. Linearity above the calibration range was estimated by spiking to a final concentration of three times the top calibrator value for each analyte and making doubling dilutions with blank urine. Defining acceptable recovery as 100 ± 10%, glycolate and glycerate were linear to 1000 µmol/L, HOG was linear to 600 µmol/L and DHG was linear to 3000 µmol/L.

Imprecision and lower limit of quantitation

Within- and between-batch imprecision was calculated by repeat analysis (four batches of five replicates) of four urine pools (Table 2). Two were produced by mixing surplus samples from patients with PH1, PH2 and PH3 to achieve low and high concentrations of the analytes, and two were produced by spiking known amounts of the analytes into blank urine so that bias could also be calculated. The lower limit of quantitation (LLOQ) was defined from these data as the lowest analyte concentration with a total analytical imprecision of <20% CV and bias of <20%: 20 µmol/L for glycolate and glycerate, 10 µmol/L for HOG and 200 µmol/L for DHG.

Table 2.

Imprecision data.

Analyte	Concentration (µmol/L)	Bias	Within-batch %CV	Between-batch %CV	Total analytical %CV
Glycolate	11	+9.6%	10.3	36.5	37.9
	21	+5.8%	2.2	15.8	16.0
	539	N/A	2.5	1.7	3.0
	765	N/A	2.2	3.0	3.7
Glycerate	17	+16.6%	5.6	8.4	10.1
	25	+6.8%	5.1	5.6	7.6
	37	N/A	5.9	9.9	11.5
	323	N/A	4.0	11.8	12.5
HOG	4	−24.9%	4.8	11.1	12.1
	8	−22.9%	1.7	12.5	12.6
	14	N/A	5.6	7.4	9.3
	81	N/A	3.3	6.8	7.6
DHG	161	+7.0%	14.4	9.5	17.2
	217	+8.0%	8.3	10.7	13.6
	164	N/A	5.8	21.2	22.0
	1737	N/A	5.1	10.7	11.9

HOG: 4-hydroxy-2-oxoglutarate; DHG: 2,4-dihydroxyglutarate.

Note: Bias could not be calculated for patient pools, denoted by N/A.

Comparison of glycolate results with an enzymatic method

Twenty-nine urine samples previously analysed by the GO method¹⁰ were also analysed by the present method. There was a mean bias of +15% compared to the enzymatic method, which was higher at lower concentrations (+18% up to 200 µmol/L, +11.5% from 200 to 600 µmol/L and +6.5% above 600 µmol/L).

Stability of HOG and DHG in urine

Urine samples from two patients with known PH3 were each divided into two portions and following acidification of one portion to pH 1, aliquots of these samples were stored at 2–8℃ or room temperature. After 3, 7 and 14 days, an aliquot from both storage conditions was removed and frozen prior to analysis in one batch. Results were expressed as a percentage of the baseline, an average of the two urines taken and the results plotted in Figure 2. Up to 90% of measurable HOG was lost over six days in non-acidified samples. Defining ‘stable’ as levels within ±10% of the baseline, HOG was stable for two weeks when preserved by acidification and refrigeration. DHG was stable in all conditions over two weeks. We initially hypothesized that HOG might be reduced to DHG in vitro. However, this hypothesis was refuted because there was no significant change in DHG seen in these samples over time (Figure 2) nor in blank urine artificially spiked with HOG (data not shown).

Figure 2.

Stability of (a) HOG and (b) DHG in urine. Solid line: acid urine at 2–8℃, dotted line: acid urine at room temperature, hollow line: plain urine at 2–8℃, dashed line: plain urine at room temperature. Results expressed as a % of the baseline.

Biological variation

Four healthy adults (two male, two female) with no renal disease or history of stone formation provided an early morning urine sample on four consecutive days. The samples were analysed in duplicate in one batch and the intra-individual variation (CV_I) estimated. The average CV_I was 18.9% for glycolate; for glycerate, HOG and DHG the results were below the LLOQ so the biological variation could not be calculated.

Urine samples from PH cases and controls

Urine samples from patients with confirmed PH (16 PH1, 15 PH2 and 19 PH3) and from 14 healthy control individuals (six male, eight female) were analysed by the present method. Where results were below the LLOQ, an arbitrary value of half the LLOQ was assigned to them to allow inclusion in the analysis. Where peaks were not detected, a value of 0 was given. Results were then expressed as a ratio to creatinine to correct for urine concentration (Figure 3).

Figure 3.

Concentration of PH metabolites expressed as µmol/mmol creatinine in patients with PH1 (n = 16), PH2 (n = 15), PH3 (n = 19) and controls (n = 14). Boxes are 25th–95th percentiles and whiskers range. Median is denoted by line and mean by dot.

Glycolate

Mean glycolate:creatinine ratio was higher in urine from patients with PH1 (248 µmol/mmol) than PH2, PH3 or controls (52, 39 and 43 µmol/mmol respectively, P < 0.005). Glycolate:creatinine ratios ranged from normal to grossly elevated (30–637 µmol/mmol) in the PH1 group and there was a clear overlap between the groups visible on the box and whisker plot (Figure 3). The highest values were only seen in urine from patients with PH1, which suggested that a higher cut-off can be used to diagnose PH1; a cut-off of 100 µmol/mmol gave 100% specificity for PH1 at a sensitivity of 71%. The range of glycolate:creatinine ratios in the control group was 24–93 µmol/mmol.

Glycerate

Mean glycerate:creatinine ratio was higher in urine from patients with PH2 (879 µmol/mmol) than PH1, PH3 or controls (1, 8, and 3 µmol/mmol, respectively, P < 0.005). The lowest level in the PH2 group (146 µmol/mmol) was almost three times higher than the highest level in the PH1, PH3 or control groups (49 µmol/mmol). Therefore, 100% specificity and sensitivity for diagnosis of PH2 was possible using a cut-off of between 50 and 145 µmol/mmol. The range of glycerate:creatinine ratios in the control group was 0–23 µmol/mmol.

HOG

Mean HOG:creatinine ratio was higher in urine from patients with PH3 (20 µmol/mmol) than PH1, PH2 or controls (all 0 µmol/mmol, P < 0.005). Two PH3 urines had HOG concentration below the LLOQ, and the range of HOG:creatinine ratios in PH3 (1–57 µmol/mmol) overlapped with the other groups. A cut-off of 4 µmol/mmol gave 100% specificity for PH3 at a sensitivity of 84%.

DHG

Mean DHG:creatinine ratio was higher in urine from patients with PH3 (651 µmol/mmol) than PH1, PH2 or controls (2, 20 and 7 µmol/mmol, respectively, P < 0.001). The lowest level in the PH3 group (282 µmol/mmol) was more than five times higher than the highest level in the PH1, PH2 or control groups (88 µmol/mmol). Therefore, 100% specificity and sensitivity for diagnosis of PH3 was possible using a cut-off of between 89 and 281 µmol/mmol.

Identification of new cases of PH

Over five months of routine clinical service, 60 samples were analysed for PHM, of which 16 had elevated levels suggestive of PH (glycolate or glycerate ≥100 µmol/mmol, HOG ≥4 µmol/mmol or DHG ≥150 µmol/mmol). Five of these were from patients with known PH and three samples with high glycolate had a normal oxalate, indicating a likely dietary origin of the glycolate. This left eight samples with possible undiagnosed PH and genetic testing for definitive diagnosis was recommended. In seven cases from five families DNA was analysed, confirming the diagnosis of PH (Table 3). In two cases, novel mutations in HOGA1 were identified, both of which were predicted to be disease causing in silico. The first, c.134C>T (p.Pro45Leu), changes a highly conserved residue and is predicted to cause a change in helical structure; no single-nucleotide polymorphisms have been reported at this location. The second, c.337G>A (p.Glu113Lys), is another missense change. Another rare variant (c.337G>C) has been reported at this position in the Exome Sequencing Project²⁷ (found in 1 out of 13,005 alleles sequenced) and was also predicted to be disease causing in silico (variant 10:99358657).

Table 3.

Confirmed new cases of PH diagnosed by this method.

Case	PHM results (abnormal in bold)				Genetic testing		Diagnosis
Case	(µmol/mmol creatinine)				Genetic testing		Diagnosis
	Glycolate	Glycerate	HOG	DHG	Gene	Sequencing results
1	19	2	40	1077	HOGA1	c.134C>T (p.Pro45 Leu) homozygous/hemizygous	PH3
2^a	72	1546	0	26	GRHPR	c.494G>A (p.Gly165Asp) homozygous	PH2
3^a	95	1703	1	50	GRHPR	c.494G>A (p.Gly165Asp) homozygous	PH2
4^a	48	628	0	0	GRHPR	c.494G>A (p.Gly165Asp) homozygous	PH2
5	154	5	0	0	AGXT	c.364C>T (p.Arg122^*) homozygous/hemizygous	PH1
6	43	49	7	589	HOGA1	c.700+5G>T (missplicing) heterozygous	PH3
						c.944_946delAGG (p.Glu315del)
						heterozygous
7	62	5	57	797	HOGA1	c.700+5G>T (missplicing) heterozygous	PH3
7	62	5	57	797	HOGA1	c.337G>A (p.Glu113 Lys) heterozygous	PH3
Controls	<100	<100	<4	<150

Members of same family.

Discussion

We have described a method for the quantitation of glycolate, glycerate, HOG and DHG in urine, which successfully provides an intermediary step in the investigation of PH. In the absence of certified reference materials or an external quality assurance scheme, accuracy was assessed by recovery, which was between 91 and 103%. The accuracy of glycolate measurement was further investigated by data comparison with the GO enzymatic method.¹⁰ There was a positive bias, but because the comparator assay is not a gold standard we cannot state that the results from it are any more accurate than the results from the present method. The data presented here meet the accepted criteria for within-batch imprecision for a mass spectrometry assay, i.e. <10% CV and <20% at the LLOQ.²⁸ Between-batch imprecision of <10% was not met for glycerate and DHG. However, because of the large difference in expected values between patients with and without PH2 and PH3, respectively, this magnitude of imprecision is unlikely to have an adverse clinical impact.

We investigated the use of this method to distinguish between patients with the three known forms of PH and controls. Glycolate:creatinine ratios in PH1 were similar to previously reported levels. A review of data from the Rare Kidney Stone Consortium (RKSC) PH registry in the USA gave a median level of 152 µmol/mmol from 96 PH1 cases (analysis method not stated).¹⁸ Another published GCMS method gave a mean of 239 µmol/mmol (range 79–418) in seven PH1 cases,¹² and an ion-chromatographic method gave a mean of 318 µmol/mmol (range 112–555) in six PH1 cases.²⁹ However, glycolate was not a completely sensitive marker for PH1 because not all patients had elevated levels; this corroborates earlier findings that up to a quarter of patients with proven PH1 have normal urine glycolate excretion.⁵ The differential excretion in oxalate and glycolate may reflect inter-individual variation in the relative activities of LDH and GR for glyoxylate metabolism. Nor was glycolate a specific marker for PH1, as mildly elevated levels were found in the other forms of PH and in the control group (Figure 3). Factors other than hepatic AGT activity influence glycolate production in vivo. Diet can have an impact and some fruit and vegetables have high glycolate concentrations⁸ to the extent that appreciable glycolic aciduria has been shown to occur in healthy individuals. In addition, there has been a recent report of elevated urine glycolate as the result of GO deficiency, although this disorder appears to be an incidental finding without a clinical phenotype.³⁰ The range of glycolate:creatinine ratios in the control group in this dataset was similar to previously reported ranges by a variety of methods: 20–107 µmol/mmol by GCMS,¹² 13–80 and 13–90 µmol/mmol by ion-chromatography,^29,31 22–125 µmol/mmol by chromotropic acid³² but was somewhat higher than the levels reported using GO (4–41 µmol/mmol).³³ This suggests that the positive bias we found with respect to the GO method may be attributed to under-recovery of glycolate by GO. The biological variation data show that the intra-individual contribution to glycolate imprecision is larger than the analytical contribution. This has an important consequence for use of the assay in a diagnostic setting, because a single measurement cannot therefore be relied upon to reflect the true baseline glycolate excretion. The cause of the high intra-individual variation is likely to be diet, which was uncontrolled in our studies.

Unambiguous identification of all PH2 cases was possible based on the glycerate results because there was no overlap in glycerate:creatinine ratios between PH2 and other groups. The RKSC PH registry also shows no overlap in glycerate levels between PH2 and other groups and has comparable results (median 837, IQR 289–1564 µmol/mmol).¹⁸ Urinary l-glycerate excretion has previously been shown to be negligible (<5 µmol/L) in healthy individuals using an HPLC method,¹³ whereas some controls in our study had measurable, albeit low, levels. This may be attributable to non-selective measurement of both l- and d-enantiomers of glycerate in the present method; however, the fact that levels in PH2 were at least three times higher bears out the assumption that this lack of selectivity is unimportant. A case of proven PH2 has been reported without elevated glycerate³⁴; however, this finding may be related to the organic acids method used, with inadequate extraction of polar glycerate into conventional solvent systems. The finding of elevated glycerate in the presence of hyperoxaluria makes a diagnosis of PH2 highly likely and genetic testing is indicated.

HOG:creatinine ratios were highest in patients with PH3, and the data compare well with a previously reported range of 1.7–118 µmol/mmol.⁷ HOG levels in PH3 overlapped with the PH1, PH2 and control groups in this dataset, and levels in healthy controls have also been reported to overlap with levels seen in PH3.⁷ It is reasonable to state that HOG is neither a sensitive nor specific marker for PH3. Results from the stability study offer an insight into why this may be. HOG was only stable in urine when acidified and refrigerated. Therefore, the apparent poor sensitivity of HOG to detect PH3 may be due to in vitro loss of the metabolite, rather than genuinely normal urinary excretion. We have also demonstrated that the loss of HOG in urine samples is not due to in vitro reduction to DHG. The metabolic route taken by HOG in PH3 is not known for certain. It has been suggested that some must be catabolized to glyoxylate by an extra-mitochondrial aldolase, and thus lead to hyperoxaluria.⁷ Another destination for HOG may be in vivo reduction by cytosolic enzymes to DHG. We hypothesize that metabolism by LDH may be one possibility given its broad substrate specificity and this is a subject for further study. In contrast to HOG, DHG was elevated in all PH3 urines and by a considerable margin over the PH1, PH2 and control groups. The results of this study suggest that DHG is the better marker and can identify PH3 with 100% diagnostic efficiency, in agreement with published data.⁷ Furthermore, DHG was essentially stable under all the conditions trialled, so the measured levels are likely to reflect physiological excretion. Glycerate levels were not elevated in the PH3 group. This finding provides evidence against the hypothesis that hyperoxaluria in PH3 is caused by inhibition of GRHPR by HOG;¹⁶ if this were the case, one might expect elevated glycerate in the urine similar to PH2.

Results from use of this method in routine clinical practice identified seven cases of PH which have been confirmed by gene sequencing in this laboratory. This represents a pick-up rate of 12%, which is high for such rare diseases, though there is likely to be a high index of suspicion in all cases referred for this service. In the seven cases confirmed by gene sequencing, the elevated PHM directed testing to the gene of interest, rather than stepwise analysis of all three genes known to cause PH based on the prevalence of the disorders. It should be noted that a normal glycolate does not rule out PH1, so in these cases, if glycerate, HOG and DHG are also low, then sequencing should be directed to AGXT. Even anticipating the introduction of next-generation sequencing methods to facilitate genetic testing for these diseases,³⁵ establishing the biochemical phenotype is important, particularly to support the finding of novel variants. We identified two novel changes in HOGA1, which were predicted to be disease causing in silico. Confirmation of the effect of both these changes ideally requires in vitro expression of the mutant protein but in the absence of these studies, the PHM profile and in silico analysis make a strong case for pathogenicity.

Patients with PH can undergo a diagnostic odyssey prior to receiving the correct treatment, and delays in this can be costly in terms of preserving renal function.³⁶ This study demonstrates that measurement of PHM by this method allows focussing of genetic testing to those in whom PH is likely and also to the gene responsible.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Ethical approval

Not applicable. The patient samples analysed in this study were referred to our department for investigation of primary hyperoxaluria.

Guarantor

GR.

Contributorship

OCM validated the method and wrote the manuscript. LH did initial method development work. GR conceived the study and contributed to the manuscript.

References

Danpure

Primary hyperoxaluria. In: Scriver

Beaudet

Valle

(eds). The metabolic and molecular bases of inherited disease, 8th ed. New York: McGraw-Hill, 2001, pp. 3323–3367.

Robertson

Peacock

Heyburn

. Risk factors in calcium stone disease of the urinary tract. Br J Urol 1978; 50: 449–454.

Cochat

Rumsby

. Primary hyperoxaluria. N Engl J Med 2013; 369: 649–58.

Robijn

Hoppe

Vervaet

. Hyperoxaluria: a gut-kidney axis? Kidney Int 2011; 80: 1146–1158.

Danpure

. Molecular and clinical heterogeneity in primary hyperoxaluria type 1. Am J Kidney Dis 1991; 17: 366–369.

Williams

Smith

. L-Glyceric Aciduria. N Engl J Med 1968; 278: 233–239.

Belostotsky

Pitt

Frishberg

. Primary hyperoxaluria type III – a model for studying perturbations in glyoxylate metabolism. J Mol Med 2012; 90: 1497–504.

Harris

Richardson

. Glycolate in the diet and its conversion to urinary oxalate in the rat. Invest Urol 1980; 18: 106–109.

Niederwieser

Matasović

Leumann

. Glycolic acid in urine. A colorimetric method with values in normal adult controls and in patients with primary hyperoxaluria. Clin Chim Acta 1978; 89: 13–23.

10.

Kasidas

Rose

. A new enzymatic method for the determination of glycollate in urine and plasma. Clin Chim Acta 1979; 96: 25–36.

11.

Chalmers

Watts

RWE

. The quantitative extraction and gas-liquid chromatographic determination of organic acids in urine. Analyst 1972; 97: 958–967.

12.

Dietzen

Wilhite

Kenagy

. Extraction of glyceric and glycolic acids from urine with tetrahydrofuran: utility in detection of primary hyperoxaluria. Clin Chem 1997; 43(8 Pt 1): 1315–1320.

13.

Petrarulo

Marangella

Cosseddu

. High-performance liquid chromatographic assay for L-glyceric acid in body fluids. Application in primary hyperoxaluria type 2. Clin Chim Acta 1992; 211: 143–153.

14.

Rashed

Aboul-Enein

Al-Amoudi

. Chiral liquid chromatography tandem mass spectrometry in the determination of the configuration of glyceric acid in urine of patients with D-glyceric and L-glyceric acidurias. Biomed Chromatogr 2002; 16: 191–198.

15.

Williams

Bockenhauer

van’t Hoff

. The enzyme 4-hydroxy-2-oxoglutarate aldolase is deficient in primary hyperoxaluria type 3. Nephrol Dial Transplant 2012; 27: 3191–3195.

16.

Riedel

Knight

Murray

. 4-Hydroxy-2-oxoglutarate aldolase inactivity in primary hyperoxaluria type 3 and glyoxylate reductase inhibition. Biochim Biophys Acta 2012; 1822: 1544–1552.

17.

Beck

Baasner

Buescher

. Novel findings in patients with primary hyperoxaluria type III and implications for advanced molecular testing strategies. Eur J Hum Genet 2013; 21: 162–172.

18.

Hopp

Cogal

Bergstralh

. Phenotype-genotype correlations and estimated carrier frequencies of primary hyperoxaluria. J Am Soc Nephrol. Epub ahead of print 2 Febuary 2015. doi:10.1681/ASN.2014070698.

19.

Clifford-Mobley

Tims

Rumsby

. The comparability of oxalate excretion and oxalate:creatinine ratio in the investigation of primary hyperoxaluria: review of data from a referral centre. Ann Clin Biochem 2015; 52: 113–121.

20.

Cochat

Deloraine

Rotily

. Epidemiology of primary hyperoxaluria type 1. Societe de Nephrologie and the Societe de Nephrologie Pediatrique. Nephrol Dial Transplant 1995; 10(Suppl 8): 3–7.

21.

Kallner A. Measurement verification, www.acb.org.uk/whatwedo/science/best_practice/measurement_verification.aspx. (accessed 14 May 2014).

22.

Cregeen

Williams

Hulton

. Molecular analysis of the glyoxylate reductase (GRHPR) gene and description of mutations underlying primary hyperoxaluria type 2. Hum Mutat 2003; 22: 497–497.

23.

Williams

Rumsby

. Selected exonic sequencing of the AGXT gene provides a genetic diagnosis in 50% of patients with primary hyperoxaluria type 1. Clin Chem 2007; 53: 1216–1221.

24.

Schwarz

Cooper

Schuelke

. MutationTaster2: mutation prediction for the deep-sequencing age. Nat Meth 2014; 11: 361–362.

25.

Henikoff

. SIFT: predicting amino acid changes that affect protein function. Nucleic Acids Res 2003; 31: 3812–3814.

26.

Adzhubei

Schmidt

Peshkin

. A method and server for predicting damaging missense mutations. Nat Meth 2010; 7: 248–249.

27.

NHLBI. Exome Sequencing Project, http://evs.gs.washington.edu/EVS. (accessed 25 March 2014).

28.

Honour

. Development and validation of a quantitative assay based on tandem mass spectrometry. Ann Clin Biochem 2011; 48: 97–111.

29.

Wandzilak

Hagen

Hughes

. Quantitation of glycolate in urine by ion-chromatography. Kidney Int 1991; 39: 765–770.

30.

Frishberg

Zeharia

Lyakhovetsky

. Mutations in HAO1 encoding glycolate oxidase cause isolated glycolic aciduria. J Med Genet 2014; 51: 526–529.

31.

Hagen

Walker

Sutton

. Plasma and urinary oxalate and glycolate in healthy subjects. Clin Chem 1993; 39: 134–138.

32.

Leumann

Dietl

Matasovic

. Urinary oxalate and glycolate excretion in healthy infants and children. Pediatr Nephrol 1990; 4: 493–497.

33.

Barratt

Kasidas

Murdoch

. Urinary oxalate and glycolate excretion and plasma oxalate concentration. Arch Dis Child 1991; 66: 501–503.

34.

Rumsby

Sharma

Cregeen

. Primary hyperoxaluria type 2 without L-glycericaciduria: is the disease under-diagnosed? Nephrol Dial Transplant 2001; 16: 1697–1699.

35.

Williams

Bagg

EAL

Mueller

. Performance evaluation of Sanger sequencing for the diagnosis of primary hyperoxaluria and comparison with targeted next generation sequencing. Mol Genet Genomic Med 2014; 3: 69–78.

36.

Van der Hoeven

van Woerden

Groothoff

. Primary hyperoxaluria Type 1, a too often missed diagnosis and potentially treatable cause of end-stage renal disease in adults: results of the Dutch cohort. Nephrol Dial Transplant 2012; 27: 3855–3862.

Simultaneous analysis of urinary metabolites for preliminary identification of primary hyperoxaluria

Abstract

Background

Methods

Results

Conclusions

Keywords

Introduction

Methods

Calibrator preparation

Sample analysis

Data analysis

Genetic testing

Results

Accuracy and assay linearity

Imprecision and lower limit of quantitation

Comparison of glycolate results with an enzymatic method

Stability of HOG and DHG in urine

Biological variation

Urine samples from PH cases and controls

Glycolate

Glycerate

HOG

DHG

Identification of new cases of PH

Discussion

Footnotes

Declaration of conflicting interests

Funding

Ethical approval

Guarantor

Contributorship

References